Digital SACD Ripping Guide
vf.su...@gmail.com
Because my view on music sharing is that it should be public, not private, I have decided to share, in full detail, how I rip SACDs digitally - with the intention that anyone can follow in my footsteps and learn how to make a Digital SACD Rip without having to go through the pain and learning/perfecting process that I had to.
I've finally gotten round to making the guide, so here it is. Jolson has expressed serious intentions to join me in the new effort, so here you are, too, my friend! Can't wait for you to join me!
If anyone has ideas to improve or would like to critique the process - I'd love to hear it. After all, our audiophile pursuit is about perfection in sound - for enjoyment in music.
I've attached a html file of the guide - in case you can't see the formatting that I've done for the guide. (makes it way easier to read). I very well might make some little edits to this guide over time, so I'll host it at some webpage eventually.
- - -
Digital
SACD Ripping Guide
Introduction
by HiResOrNothing (what a great guy, eh? :P):
"In the past we have
only been able to rip SACDs by recording the multichannel analog outs of an
SACD player and into our analog sound cards. Many people have used horrible
Creative Audigy cards, and quality has been disappointing at best. Other rips
that we have seen have been nicer, using the best that analog sound cards can
offer from the likes of M-audio, Lynx, etc. However, quality has still suffered
due to the multitude of problems that arise with analog capturing in general.
Unless the SACD player used for the transfer is a very expensive one, the
analog transfer quality will leave a lot to be desired.
"But we now can
celebrate in a new "standard" for SACD ripping. We can now capture a
pure 24-bit PCM digital signal which is converted directly from the pure DSD
stream from the disc - and this is achieved by using a special
"modifed" SACD player. These modified players been around ever since
the SACD format has, but they have always been extremely expensive (like, try a
$2000 Denon player which you have to order from Switzerland). But things have
progressed and a few companies are offering a modified version of the famous
Oppo DV-980H player, which is well known for its affordability yet
uncompromised quality - especially with regards to digital SACD output!
"I have purchased
such a player to make this SACD album available to you in the highest quality
currently possible. The Oppo offers pure DSD output over HDMI 1.2 but it also
gives the option for PCM output. The Oppo internally converts the DSD to PCM at
24-bit/88.2kHz and the "mod" captures this PCM signal and outputs it
through three stereo S/PDIF (coaxial) jacks. To capture this glorious digital
signal, I am using three M-Audio Delta 1010LT sound cards simultaneously. They
each have a single stereo coaxial input, so in order to record all six channels
simultanesously, three cards must be used. I'm using Adobe Audition 1.5 in
multitrack mode to simultaneously record the 6 channels in real time, then I
map the channels with Audition's "multichannel encoder", and export
the individual mono wav channels. After that I split the album into the tracks
and encode it to MLP with surcode, and author it to DVD-A with Discwelder
Chrome. No editing of the audio signal is made in Adobe Audition, this is a
straight digital capture of the signal and as close as you can get to copying
an SACD disc-to-disc at the present time."
- - -
The following is a detailed guide from HiResOrNothing, of how to digitally rip
SACD with the above method and equipment. Other equipment, software, or
variations in the editing/cutting/encoding process may be used, and get the
same results quality-wise. But this guide is for exactly how I do it.
Setting
up the Modified Oppo 980H
- It all starts with buying a modified Oppo player. You can get them from www.switch-box.com.
- First, it is important to understand that while we are ripping a digital
signal, this is still real-time recording and the signal we are capturing is
indeed a processed
signal. We are still not copying the information bit for bit, but instead we
are ripping a signal which has been fed through the Oppo player, processed by
its internal components, including the chip which converts the DSD signal to a
PCM one.
- Before you start recording, get to know the Oppo player very well, read the
manual and learn how to operate it and play back SACD discs before you
continue.
- These are the settings you need to set in the Oppo player, to achieve the best
processing and output quality of the PCM signal. Assuming the rest of the
settings are from Factory Default, do the following:
General Setup Page
- Resolution - set to "1080p"
Speaker Setup Page
- Downmix - set to "5.1 CH"
- Front, Center, Surround Speaker - set to "Large", and set Subwoofer
to "On"
- Channel Delay - set to "0ft"
- Make sure Channel Trim and Audio Delay are set to "0dB" and
"00".
Audio Setup Page
- LPCM Rate - set to "192K"
- Pro Logic 11 - turn off
- HDMI Audio - set to "Off"
- SACD over HDMI - set to "PCM"
- Make sure EQ Type, Sound Field, Audio tone are set to "None",
"Off" and "00".
Digital Volume Control
- On your Oppo remote you will see a "+" and a "-" button
to control the volume of the Oppo. Make sure you keep this digital volume at
100% (level 20). Lowering the output volume will actually lower the output
signal quality. The way digital volume control works, is that it reduces the
volume by taking out bits from the audio signal - it actually removes audio
information, and thus lowers the quality of the output signal! Volume control
should ALWAYS be done in the analog domain (your amplifier), not the digital
domain. So leave it at full volume, where no quality loss will occur.
- If you find a disc has dynamic clipping (with the amplitude touching the 0dB
mark), it is not because the Oppo output signal is too "hot" and
should be turned down, it is because the (stupid) mastering engineer/producer
applied dynamic compression to the original disc. Lowering the volume of the
Oppo will only output a softer, but still clipped signal, and reduce the
quality at the same time. All digital formats (and even vinyl!) have the
ability to suffer from the loudness war :/.
You will also need
three quality coaxial S/PDIF cords to connect the Oppo to the Delta Cards.
Installing/configuring
the Delta Cards
- You will need a motherboard with three PCI slots for the three Delta cards.
You will also need relatively high performance components (fast CPU, RAM, SATA
drive preferred) to ensure that no system lag occurs and gets in the way of
accurate and error-free capturing of the high bandwidth 6 channel digital
signal - and using three PCI cards at once with such intensity.
- Have plenty of HDD space, have a fresh system install, and only install
programs that you will absolutely need for the system.
- The OS that you must use when using the Delta cards is Windows XP SP2.
There is actually a bug whereby multiple M-audio cards sometimes have trouble
syncing together and pops/clicks in the audio can even occur. So do not use
Vista, do not use Windows 7, do not use XP SP3, use XP SP2. And you must use the
5.10.00.5057v3 version of the driver, as well! Also, READ THE USER MANUAL
before you even touch the cards with your bare hands - learn a bit about what
the cards can do, and how they generally work, and what you should and
shouldn't do with the cards - e.g. if you attach the breakout cables to the
cards while the computer is powered, you could damage the cards and void your
warranty.
- Once you're ready, install the three cards on the system. Once installed,
it's time to configure them in the Delta
Control Panel. You will see the three cards in the H/w installed area
on the right, which you simply select to individually configure each one.
In the Control Panel go to the "Hardware Settings" tab.
- Set the Master Clock to S/PDIF
In.
- Set the S/PDIF Sample Rate to 88200.
- Set the MultiTrack Driver Devices to Single
and In-Sync.
- Make sure these settings are configured for all three Delta Cards.
As for setting up playback with the M-Audios, consult the user manual and work
it out yerself ;).
Setting
up Adobe Audition
- I use Adobe Audition 1.5 to do the recording. Later versions are obviously
fine to use (as are many other programs), but the settings might be different
to what I outline in this guide.
- Go to Options -> Device Order. Click on Recording devices tab. Remove any
devices previously selected for use in Audition, and click on "Use >>"
to allow the three S/PDIF input devices. They will appear as:
M-Audio Delta 1010LT S/PDIF
M-Audio Delta 1010LT S/PDIF (2)
M-Audio Delta 1010LT S/PDIF (3)
- Click OK.
- Now we need to set the default recording settings to the right resolution. Go
to Edit View, and click on the record button. Select 88200 as the sample rate,
stereo as the channels, and 32-bit (float) as the Resolution. Click OK to start
recording, then stop the recording and restart Adobe Audition.
- Now go into Multitrack mode. Here we will do the recording, using three
stereo "tracks" for the three S/PDIF cards - totalling to 6 channel.
Link FL+FR to M-Audio Delta 1010LT S/PDIF, C+LFE to M-Audio Delta 1010LT S/PDIF
(2), and Rs+Ls to M-Audio Delta 1010LT S/PDIF (3). It might take some trial and
error to match them up, but it's better if you can make it logical. In most
multichannel mixes, the fronts are the loudest, C and LFE are pretty easy to
isolate and then the remaining two, usually softer than the fronts, are the
backs.
- So, right click in the Track 1 area, and choose Properties. Make sure M-Audio
Delta 1010LT S/PDIF is selected as the input device, and change 16-bit to
32-bit. It should be stereo by default.
- Then do the same for Track 2 and 3, as per the suggested signal mapping
above.
- Lastly, arm the tracks for recording by clicking on the red "R"
button for each track. Now Adobe Audition is set for recording!
Recording
- Connect the three stereo S/PDIF outputs from the modded Oppo (FL+FR, C+LFE,
Rs+Ls) to the three Delta S/PDIF inputs.
- When recording, disconnect any other audio connections from the back of the
Oppo player apart from the S/PDIF, to remove any possible extra unnecessary
processing/interference in the Oppo. This might include analog or
Toslink/Coaxial connections. HDMI can still be connected though, as we have
already turned HDMI
Audio off in the Audio Setup Page anyway.
- On the recording computer, do not run any other application while you record
with Audition, and make sure the system (RAM especially) is plenty free for the
recording process!
- Just before you record, set the Oppo player to Audio-Only Mode!
This turns off the video processing and output, reducing the inteference
between the video and audio signals - helping to optimise the signal output of
the Oppo. The screen will turn off after 5 seconds and you will see A.ONLY illuminated
in the Oppo front panel. When you stop a disc and put a new one in, it goes
back to normal mode so make sure you set it to A.ONLY before every recording! And give
it a few good few moments before you start recording too - let the player focus
all its processing power for the audio.
- With the three Tracks in Audition Multitrack mode configured and armed, press
record, then start playing back the SACD in the Oppo! Now the magic happens!
Always listen to your
capture all the way through to make sure there were no disc skips/glitches!
Repeat if anything went wrong. Be sure to clean the disc before recording to
prevent anything happening.
Other
Tips and Hints
- It's a good idea to get the length of silence at the beginning of the first
track exactly correct (when you edit your album). To make a good release, you
should get it right, and it might also help when cutting the album's track
points later on. To get the length of silence down to the milisecond, start
playback of the SACD by pausing the disc in track 1, and pressing the
"Previous" button on the Oppo player itself (instead of the remote).
This will create a little click in the audio signal which Adobe Audition will pick
up in the recording. You can then later use this click to remember where the
album starts, exactly.
But note that the click doesn't occur in exactly the same place in all the
channels (even in stereo). If you zoom right in you will see this. So make sure
to cut the album just after the last
click that occurs, wherever it is amongst the channels.
- Annoyingly, Audition seems to not remember the device settings very well.
Sometimes you have to keep going back to the Device Order settings and adding
the S/PDIF input devices in again. Also, remember to check EVERY setting in
Audition before every time you record - make sure all three tracks are set to
32-bit, (not 16-bit!) etc. You can also see down the bottom the current setting
for recording - it should say 88200
* 32-bit Mixing.
- For the process of setting everything up, or if you want to easily
monitor/play back your recordings in Audition (either during or after
recording), it's a good idea to connect the analog outs of the Deltas to your
multichannel receiver - to listen to the channels in your surround system.
Simply use the WavOut
1/2 device for each Delta card as the three stereo pairs,
matching them accordingly into your receiver/amp's multichannel analog input.
Then in Audition, in the same way that you set the S/PDIF inputs to be
recording devices and then set them in the track properties, do the same thing
for the analog playback devices and set them for each track's output.
Wrapping
Up
- Once you have your glorious digital capture on your computer, you can export
the tracks to mono wav channels. Do this with Audition's Multichannel Encoder
at the bottom of the View menu. Map the three tracks to their pertaining
channels, and select the correct export format (6 mono wav channels, 24-bit).
It will then export the channels for you. For the C+LFE track, move the Sub
Channel Level to "100", it is 0 by default.
- One you have the mono channel files, you can load them back in Audition (I
load them as six individual tracks in MultiTrack mode), and go on and edit the
tracks for your release. When you have them loaded as six individual tracks,
you can then pan them and monitor the recordings with any stereo setup that you
edit the tracks with. If it's just a stereo mix, you can simply save Track1 to
a stereo file in Wave Edit mode and then edit it as a stereo file as is.
- If your intention is to author it to DVD-A with DiscWelder, the easiest way
to cut the tracks is to set track points in Audition (right click at the tip of
the cursor bar where you want to cut a track and select "Insert CD Track
Master"), then write down the track points (right click at any track point
and "Go to Cue List", then copy and paste those cue values somewhere
and change them into Discwelder's time format), then in DiscWelder import the
full-length album and then insert the track points in there. This is also the
only way to do gapless albums properly in DiscWelder.
- So once you edit the beginning and end of the album and determine the track
points, you're ready to encode the mono channels to MLP, FLAC or whatever you
desire. There is NO need to edit the tracks in any other way, the aim of this
method is to rip the disc as close as possible to the original. If there's a
problem, re-transfer the disc, DO NOT edit anything in the audio.
- So there you are. If you have the courage to take on this new method of
ripping SACDs, now you know how to do it!
Yours,
HiResOrNothing
jolson
> 16KVisaHämta
george...@comcast.net
VF,
Your guide does not mention how to set the recording gain to prevent clipping. This topic is much more important than mentioning that one can control the output volume of the player. Recording at too high a gain will result in a clipped recording.
George
david n sanchez
Marek-F
Before long time i have a CD-Transport and some DACs. I test some
Digital Cables from 70€ to 1000€ range. A very musical and neutral
cable is: Apogee Wyde Eye, and it cost not much.
Marek
VF
Ok, I said I'd like ideas or critique to the process, which is true, but George, we've been through all this already. I can't believe you still don't understand this clipping thing.
Gain is something that pertains to analog audio. Gain pertains to amplitude, or the amount of power, given to the amplification of an ANALOG audio signal! NOT a digital signal!
You must think that the Oppo "adds" amplification to the digital signal? I already explained how this is NOT the case. If you want to learn about it, read this page: http://extra.benchmarkmedia.com/wiki/index.php/Digital_volume_controls. It explains how digital volume control works, and amongst the useful info, it states: "When the digital volume control is set to 100% (or 0.0dB, or Unity Gain, depending on nomenclature), no multiplication is performed, and this distortion is not present."
And to finally dispel any doubts, I took the liberty of emailing Oppo themselves today to ask how the volume control works and if the signal is too hot at +20 or if it's the right volume for digital output (I explained that I'm a audiophile listening to SACD and don't want to reduce the dynamic range by setting it too high). They responded with, "You should leave the volume at +20. If you reduce the volume, you may introduce digital truncation."
So can I provide any more info to prove to you that some SACDs simply COME clipped?
Now I move onto your claim that you can:
1. set the recording gain, and
2. prevent clipping by doing this.
OK - 1. There's no such thing as gain control with digital recording! Have you actually tried this, with digital S/PDIF recording equipment? Have you tried it yourself? Or have you only done analog recording? I searched high and low (once again, like when you brought this up months ago) to try and find somewhere where I can lower the "recording" gain in the Delta Control Panel or Adobe Audition, but I could not find it. Is this because gain doesn't "exist" for digital signals? Let me guess - yes!!!! The Delta Control Panel has gain controls for the analog inputs, that's for sure, but not digital! IF such a thing existed for digital recording I'm pretty sure it would be there in the Delta Control Panel. And if there's some way of lowering it with some recording programs that you might know, it wouldn’t be called "gain", and it would be disastrous as it would remove bits from the audio, just like the Oppo volume control!
2. Why can't you get this - if you lower the amplitude of an already clipped SOURCE (the SACD - and I've PROVEN above that some SACDs simply come clipped), you will only get a softer, STILL clipped, waveform! I already showed this explicitly by providing screenshots of recorded signals output by the Oppo at volumes less than +20. They were still clipped (you could see the truncating), just softer! Meaning, exactly the same dynamic range!
Much like distortion, you cannot just "fix" clipping and bring back audio information that was cut out when the engineer snipped it away. And as for trying to tamper with the audio and *marginally* improve the dynamic range with some tedious method (like some people do to "fix" loud redbook CDs and then re-upload them) - would you do that with a clipped DVD-A, or would you rather share a perfect DVDFab rip and let *others* tamper with the dynamic range if they want to? The latter is my philosophy with ripping SACDs – I want to share something as close as possible to the original disc. No tampering, no edits.
I'm sorry George, but you're completely wrong about this whole gain thing. If you see that one of my rips is clipped, it doesn’t mean it is a clipped recording! It means it’s a clipped disc!
Can someone else chime in and give their take on the matter, please? Am I mad/incorrect about all this?
george...@comcast.net
VF, You're a funny guy. Whenever I mention input gain, you respond about output volume. You don't seem to understand that your application has input gain controls. This has nothing to do with Oppo output volume.
Also there is no way you could determine an SACD was clipped unless you recorded it and looked at the files. If you recorded it clipped due to excessive input gain, you would see it clipped and think the disc came that way. LOL.
Digital recording does indeed have input gain control and is a lot more important than mentioning player ouput volume control which no one should ever attempt to adjust.
DSOTM was recorded about 3dB too hot by the way.
MIKEY
VF
Ok. I’m a patient guy. Let me think in terms of your view for a second and go through with what you would do if you were me.
“your application has input gain controls”
Where are these “input gain controls” in adobe audition? Tell me and I’ll make a recording and see if it lowers the gain for the digital input.
Also – you say there’s no way to know if an SACD is clipped unless you make a (proper, adjusted gain controlled) recording of it and looked at it. Well with DVD-A, we DO know for sure if a disc was clipped or not. DVDAExplorer can extract the original files from the disc. So what do you say to the situation of recording clipped DVD-As? I can make a “Digital DVD-A Rip” via the exact same method with my setup, right now. (I have just the one in mind – bryan ferry’s frantic. Clipped like heck (IIRC - I’ll have to double check or find a definite clipped disc), and it’s at 24-bit 88.2khz – not sure if the mod does 96khz or 48khz, maybe it’s stuck at 88.2khz – I think I have a crap Britney Spears DVD-A at 88.2khz which is sure to be clipped, as a backup :P).
If it comes to testing THAT out, I will do that to prove to you that I record discs such that the dynamic range of my recording is identical to the original disc. And NOT higher than the original disc. Or, if it’s higher than the original disc, I guess I’ll have to do something about it, wont’ I? :P
But first let’s first try out this “input gain control” thing which you say is there but I can’t find it. Please instruct me.
This is all pure science here, so honestly – we should be able to resolve this argument on pure scientific grounds.
elshagon
those "eating popcorn" watching the entertainment one's!
On Aug 4, 12:56 pm, "VF" <vf.surro...@gmail.com> wrote:
> Ok. I’m a patient guy. Let me think in terms of your view for a second and go through with what you would do if you were me.
>
> “your application has input gain controls”
>
> Where are these “input gain controls” in adobe audition? Tell me and I’ll make a recording and see if it lowers the gain for the digital input.
>
> Also – you say there’s no way to know if an SACD is clipped unless you make a (proper, adjusted gain controlled) recording of it and looked at it. Well with DVD-A, we DO know for sure if a disc was clipped or not. DVDAExplorer can extract the original files from the disc. So what do you say to the situation of recording clipped DVD-As? I can make a “Digital DVD-A Rip” via the exact same method with my setup, right now. (I have just the one in mind – bryan ferry’s frantic. Clipped like heck (IIRC - I’ll have to double check or find a definite clipped disc), and it’s at 24-bit 88.2khz – not sure if the mod does 96khz or 48khz, maybe it’s stuck at 88.2khz – I think I have a crap Britney Spears DVD-A at 88.2khz which is sure to be clipped, as a backup :P).
>
> If it comes to testing THAT out, I will do that to prove to you that I record discs such that the dynamic range of my recording is identical to the original disc. And NOT higher than the original disc. Or, if it’s higher than the original disc, I guess I’ll have to do something about it, wont’ I? :P
>
> But first let’s first try out this “input gain control” thing which you say is there but I can’t find it. Please instruct me.
>
> This is all pure science here, so honestly – we should be able to resolve this argument on pure scientific grounds.
>
> Sent: 04 August 2009 05:09
> To: Surrou...@googlegroups.com
> Subject: [SurroundSound] Re: Digital SACD Ripping Guide
>
> VF,
>
> Your guide does not mention how to set the recording gain to prevent clipping. This topic is much more important than mentioning that one can control the output volume of the player. Recording at too high a gain will result in a clipped recording.
>
> George
>
> ----- Original Message -----
> From: "vf surround" <vf.surro...@gmail.com>
> To: Surrou...@googlegroups.com
> Sent: Monday, August 3, 2009 6:50:20 AM GMT -06:00 US/Canada Central
> Subject: [SurroundSound] Digital SACD Ripping Guide
>
> Because my view on music sharing is that it should be public, not private, I have decided to share, in full detail, how I rip SACDs digitally - with the intention that anyone can follow in my footsteps and learn how to make a Digital SACD Rip without having to go through the pain and learning/perfecting process that I had to.
>
> I've finally gotten round to making the guide, so here it is. Jolson has expressed serious intentions to join me in the new effort, so here you are, too, my friend! Can't wait for you to join me!
>
> If anyone has ideas to improve or would like to critique the process - I'd love to hear it. After all, our audiophile pursuit is about perfection in sound - for enjoyment in music.
>
> I've attached a html file of the guide - in case you can't see the formatting that I've done for the guide. (makes it way easier to read). I very well might make some little edits to this guide over time, so I'll host it at some webpage eventually.
>
> - - -
>
> Digital SACD Ripping Guide
>
> Introduction by HiResOrNothing (what a great guy, eh? :P):
>
> "In the past we have only been able to rip SACDs by recording the multichannel analog outs of an SACD player and into our analog sound cards. Many people have used horrible Creative Audigy cards, and quality has been disappointing at best. Other rips that we have seen have been nicer, using the best that analog sound cards can offer from the likes of M-audio, Lynx, etc. However, quality has still suffered due to the multitude of problems that arise with analog capturing in general. Unless the SACD player used for the transfer is a very expensive one, the analog transfer quality will leave a lot to be desired.
>
> "But we now can celebrate in a new "standard" for SACD ripping. We can now capture a pure 24-bit PCM digital signal which is converted directly from the pure DSD stream from the disc - and this is achieved by using a special "modifed" SACD player. These modified players been around ever since the SACD format has, but they have always been extremely expensive (like, try a $2000 Denon player which you have to order from Switzerland). But things have progressed and a few companies are offering a modified version of the famous Oppo DV-980H player, which is well known for its affordability yet uncompromised quality - especially with regards to digital SACD output!
>
> "I have
>
>
> read more »- Hide quoted text -
>
> - Show quoted text -
elshagon
audio had been imported? So you'd have to have all the channels in
Adobe Audition and then intentionally add the gain or whatever other
effect you choose. Is this correct, or can you record the audio with
realtime gain applied as it's recorded? The fact that Adobe takes a
while to apply any effect makes me think that audio is imported "as
is" and gain/reverb/limiting and any other effect is applied by the
user afterwards.
ArnoldLayne
frequencies just become too much for 88,2/24 PCM? Or even more likely,
impossible to handle for the audio processor when converting on a 64fs
basis? The solution would be a more potent conversion like DXD but we
are not there just yet.
I do hear differences between pure DSD and converted. Nevertheless,
the small alterations I can sense in these digital transfers are no
worse than those I have experienced when it's done internally in my
SACD players. One transfer I find really silk smooth, the Toto IV. It
would be very interesting to learn about a "clipping analysis" of it.
I place my bets beforehand that it is negligible.
If my hypothesis mentioned before would be correct, there is
unfortunatelly nothing to do about it since the conversion is done by
hardware. It would have been interesting to try a reduction of dynamic
range at lower frequencies and see what happens with the alleged
clipping.
Heating up some butter in my popcorn pan now... :-)
>
> leer más »- Ocultar texto de la cita -
>
> - Mostrar texto de la cita -
elshagon
of time due to poor engineering. Some sacd's just sound hot, and we
can also look at the wavform's of dvd-audio discs to see that it's a
problem. Here's a screenshot of Roseanna wav's from the Toto digital
conversion http://www.pbase.com/dscott/image/115729811 As
you can see no clipping and the wav's look fine. You can go to the
Quad forum and see threads with similar screenshots which show that
many of these discs are clipped to begin with.
VF
mid range before, and to me this sounds only like a wild theory based on
those observations with not much science in it :P. I'm not sure it's
anything extreme like that.
Lots of (in fact probably all) DSD mixing consoles have a DSD to PCM
conversion chip or whatever. To be honest, none of us know all that much
about DSD conversion, we can only assume the Oppo's conversion is a decent,
and hopefully very good one at that.
But what I do know, is that it's really just mathematics, quantization (and
I think dithering), low pass filters and this and that - I'm sure DSD to PCM
conversion is pretty much worked out and there's probably a standard for the
best mathematical way to digitally convert DSD to PCM. Would be interesting
to research...
Add to that, the fact that many discs are NOT hot, but in fact very low -
discs that feature the mid range frequencies very intensely. Nothing's wrong
with the Oppo, or DSD to PCM conversion.
Anyway.....awaiting George's response....
GREG CARRIGAN
All too commonly over the past 10 years, mastering engineers have been forced by artists and labels to maximize the volume of their releases. Hard limiting boosts the volume of the entire mix by up to 10db's - essentially creating many peaks in the recordings - but then squashes the peaks down to -1db below the maximum volume (0 dB) that can be handled. The resulting waveforms look like a solid mess from start to finish - no dynamic variation.
I was asked to use hard limiting on a few recordings I was mastering for bands. Here's an interesting point - when in their native wav format, there were no peaks above the 0 dB level (because that's what the hard limiter does), but later when I converted those same wav files to other formats, peaks appeared in the wave forms.
It was as though the audio conversion process messed up the volume levels of the original file. I have no idea why this happened - but it happened in 100% of the cases.
Something to think about
G
From: elshagon <elsh...@msn.com>
To: SurroundSound <Surrou...@googlegroups.com>
Sent: Tuesday, August 4, 2009 6:37:54 PM
VF
it's not surprising that many of these discs are clipped as well.
It's when you get a rare pure dsd disc, which is dsd from start to finish of
the recording and mastering process, that you will definitely not have
clipping - I'm sure the Steve Reich is one such example.
BTW - I have three SACD albums coming up in relatively short succession - by
German prog rock ban Can - this band is awesome!:
Tago Mago
Saw Delight
Rite Time
I discovered Can at my local hifi shop when I found these three gems- the
last SACDs left in the shop :/. Glad I snapped em up. They're all
stereo-only remasters, but they're well worth getting. Sound quality of the
Tago Mago is not amazing (master tape must just be in that condition) but
the MUSIC is so good on that album! It's their best album and a very
influential one - they were headliners of the Krautrock movement.
The other two akbums are much better sonically. Saw Delight is great
(musically) too! Rite time is so-so IMO. But they're all great stereo mixes.
ArnoldLayne
even supposed to be a little wild. It is common to assess formats and
conversions by a) things that get lost, for example high frequencies,
and b) bad things added, like aliasing. The notion that c) correctly
maintained stuff could also cause problems is maybe a little more
abstract. But not automatically false just because of that. Anyway, it
is effectively difficult to discuss it from a theoretical point of
view. It would take quite some knowledge of the conversion itself.
Better with a practical approach, which in fact already has been
provided. If Toto IV doesn't clip, we can more or less discard a
general DSD->PCM problem, unless this album had some peculiarly soft
Low-/Midrange. And if many albums are deliberately mixed in a way that
introduces clipping, it is no surprise if many SACD suffer this
problem. So I myself find this latter explanation more plausible.
Haven't heard anything at all by Can, this will be interesting :-)
RW
simply engineered very "hot" and there isn't jack shit you can do
about it. I believe his methodology of taking the signal as-is and
simply getting it onto a disc we can all enjoy is the best way to
approach it. If it was recorded too hot, that's the way it is,
hombre, and you gotta live with that. Even getting the original
master tape wouldn't help you any, the signal is fucked right there
for all eternity. The fault lies in the original production, live with
it...
-RW-
jolson
people knowing much less...)
First, I think we all need to understand "What Is Gain?"
In the analogue world, whenever you connect two pieces of equipment
you need to match the output voltage level with the input sensitivity.
Usually it's the input sensitivity that is the static one, if you know
the maximum output voltage equipment A can give you can set the
sensitivity to - for example - 2V. If A can't output higher than 2V,
this connection can never give you any clipping.
If the actual output however voltage is often very much lower than 2V,
the inherit noise level may be too close to the signal level and you
won't be able not to have that noise heard. You solve that by
adjusting input gain. getting the signal level enough higher than
noise level. But now, when gain is higher than 0dB, you have the risk
of clipping if the signal suddenly in a passage is higher than you
expected. So you have to monitor the whole recording and watch out so
that the amplified input signal doesn't get higher than what can be
handled. If that's the case, the gain has to be lowered and the
recording has to be redone.
On SPDIF you may think you don't have voltage levels, it's one's and
zeroes. Of course, the "1" and "0" are different voltage levels
anyway. But SPDIF is a standard with specified levels, it's not so
that a little higher voltage makes a 1 into 1.1 or 1.2. If the
equipment doesn't follow SPDIF standards, a 0 can be taken for a 1.
But if the standard (as usual) is followed, an output 0 will always be
an input 0 in equipment B.
So, is there no digital gain? Of course there is. For instance, in
this example: You are creating an electronic masterpiece and are
recording 8 tracks, one by one, from a couple of synths. You are then
going to mix these 8 tracks together in your computer to get a killer
stereo melody. When you mix these tracks together the signals are
added, which may in the end result in clipping. Therefore, tracks in a
multi-track recording that should later be mixed and downmixed are
sampled with a negative gain (often -12dB) to get enough headroom.
If, however, you are not going to either mix, add or otherwise process
the audio you just use the standard 0dB "gain".
ArnoldLayne
chance to tweek and improve, It's in our nature, But at some certain
point we just have to let go. There is so much good music floating
around. I wake up happy every day because of the improved access to
high-quality audio. These digital transfers is a step in that
direction. So let's take them for what they are and use them for what
they were meant for.
I forgot to say before that I think it it beautifully done by VF
sharing the knowledge. I'm not likely to be able to use it in the
forseeable future. But it just feels so good to know that we are a
community where we help each other to advance and to enjoy.
RW
alter the input/output voltages for digital devices. The point that
VF and I are trying to make is that once the recording has been hard
limited by the original recording engineer and it is then transferred
to SACD, there is NOTHING you can do to un-do it. Period. You cannot
un-do what was cocked-up at the recording studio. You'd have to re-
assemble the musicians and do re-takes at the proper levels in order
to get what we all want, which is "properly" recorded music the way
God intended it to be. The hard limiting and compression being used
nowadays is a travesty and if I were a band member I'd be pretty
pissed-off that my work was ruined for all of eternity. However, if I
was an especially clever band member, I'd be sure that I had a copy
that was done properly and would vault that away for re-release at a
later date to an audiophile-friendly label. Or pull a Robin Hood and
release it to the Internet on the sly.
And don't even get me started on Auto-Tune and other such jiggery
pokery. If someone cannot sing on-pitch, then that's the way it ought
to be recorded. Or else the singer should just admit the truth and put
the mic away to be used by someone who can sing properly. Our music
is being destroyed by the suits in the tall shiny buildings who care
nothing for the art and/or the talent required to produce this art.
It's all about the money, honey...
-RW-
elshagon
compared them to VF's sacd conversion last night. For the first time
I was able to listen to them both side by side in any order I chose.
RW
>> On a lighter note, I finally loaded up tracks to Parson's DSOTM and compared them to VF's sacd conversion last night. For the first time I was able to listen to them both side by side in any order I chose. <<
-RW-
george...@comcast.net
Yes, I am here. Traveling through the midwest. Even the RV parks which claim to have internet service do not. VF, I appreciate the open-mindedness in your response. I haven't had a chance to read your initial response yet. It looks like you more or less vomited on the PC. I am away from my tools on my home computer for most of August. I have downloaded and installed audition and have found the digital input gain controls. When I return, we can resume this discussion. I am mildly surprised that I am the only one who has ever made a digital recording before. Input gain is the most basic of recording process variables.
Your friend,
George
PS - Friends don't let friends produce distorted recordings.
elshagon
Guthrie mix. They both sound great (and are quite different) and I'm
glad we have them both.
jolson
are sampled indivudally (as is usually done when someone creates a
record, even if it's in the end is downsampled to a stereo CD) you
will have to lower the input gain. This to take care of the fact that
mixing means adding the tracks together, which means adding to the
volume. Which means clipping, if you don't start out low enough (often
-12dB).
But here's it's six discrete channels, staying as six discrete
channels. Input gain should always be 0dB. Can't be so hard to
understand.
>
> läs mer »
george...@comcast.net
VF,
If there is no such thing as gain control with digital recording, then how do we explain this? This is a picture of my laptop desktop with a rudimentary digital recording gain control. Your application and other recording applications such as Sony Soundforge Acid have more sophisticated gain controls. Usually they have meters by which one can monitor the gain and watch out for clipping. When and if clipping occurs, the gain is then reduced and the recording monitored again until the gain is correctly set.
Anyway, here is the picture:
http://home.comcast.net/~georgeshannon/a.rtf
----- Original Message -----
From: "VF" <vf.su...@gmail.com>
To: Surrou...@googlegroups.com
Sent: Tuesday, August 4, 2009 12:09:11 PM GMT -07:00 US/Canada Mountain
Subject: [SurroundSound] Re: Digital SACD Ripping Guide
MIKEY
did you noticethat the gain controller for SPDIF is greyed out
I highly doubt that it will work or have an effect IF you CAN move it
MIKEY
(Sound engineer for almost 20 years now tells you: don’t touch digital controls …. Quality will get worse….. simple as that)
gshannon
soundcard or input attached to my little laptop. It really doesn't
seem reasonable for any other explanation. Microsoft wouldn't build
anything that wouldn't work would they? (Boy, is that a loaded
question). I mean Microsoft Volume Control has been around since 2007.
What would be the sense of keeping the SPDIF control current if the
sliders didn't move up and down?
Whether to use this control is a matter of opinion. Each of us has an
opinion on this and that is how it should be. But you can't dispute
the existance of digital inupt gain control. I mean there it is
staring me in the face. Its on everyone's PC who is using Windows XP.
I haven't checked Vista yet. Vista is on my wife's computer back home.
I'll do so when I return.
Maybe someone at home with SPDIF plugged into their soundcard can try
the control. That may answer this question and prevent RW from
believing that every SACD ever produced was clipped by the sound
engineers. LOL.
George
On Aug 11, 10:43 pm, "MIKEY" <MikeyWen...@gmx.de> wrote:
> Anyway, here is the picture:
>
> http://home.comcast.net/~georgeshannon/a.rtf
>
> did you noticethat the gain controller for SPDIF is greyed out
>
> I highly doubt that it will work or have an effect IF you CAN move it
>
> MIKEY
>
> (Sound engineer for almost 20 years now tells you: don't touch digital
jolson
and when to do it.
george...@comcast.net
That, of course, is your opinion.
----- Original Message -----
From: "jolson" <j.jo...@gmail.com>
To: "SurroundSound" <surrou...@googlegroups.com>
Sent: Friday, August 14, 2009 3:49:06 PM GMT -07:00 US/Canada Mountain
Subject: [SurroundSound] Re: Digital SACD Ripping Guide
VF
During the ripping process I learnt indeed how very "non-perfect"
digital ripping is, and that yes, it's not just 1s and 0s. Many things
affect and *change* those 1s and 0s - copying a file bit for bit is
one thing, but processing a signal is another (which is what is
happening in my ripping process). Anyway glad to learn some things and
I sure have a lot more to learn about audio in general!
RW
George, not only are you needlessly obstinate and obnoxious, you're a
schmuck for twisting my words into meaning something other than what I
>> Maybe someone at home with SPDIF plugged into their soundcard can try the control. That may answer this question and prevent RW from
believing that every SACD ever produced was clipped by the sound
engineers. LOL. <<
this? I wrote that *if* the engineeer laid the tracks down too hot
and then committed them to the final masters like that, there was
nothing you could do about it. SOMEHOW, you twisted this into my
saying that EVERY SACD ever produced was done too hot. Frankly, this
pisses me off quite a bit. OF COURSE not every SACD has been authored
this way. SOME SACDs have been done like this, but certainly not
EVERY one. In fact, a very small minority of SACDs have been done this
way, and most certainly not EVERY one. You need to take a reading
comprehension course so that you can begin to understand what others
on this forum are writing. It is obvious that have a problem with
this basic self-educational skill
Stop doing this, you dope. Go get some education before you post your
mis-guided nonsense here again.
-RW-
george...@comcast.net
RW,
I was just joking about that. Sorry you took offense.
George
----- Original Message -----
From: "RW" <rlwain...@gmail.com>
To: "SurroundSound" <surrou...@googlegroups.com>
Sent: Saturday, August 15, 2009 1:46:38 PM GMT -07:00 US/Canada Mountain
Subject: [SurroundSound] Re: Digital SACD Ripping Guide
grill
gain control in digital audio. It may or may not help, all of you can
decide.
http://books.google.com/books?id=gIFwda8iGSQC&pg=PA63&lpg=PA63&dq=digital+audio+gain+control&source=bl&ots=HJDewDcIg2&sig=ftvk8FlLsrVU9cUi18edTVFMvvw&hl=hu&ei=nCWJSpjJHIOG_Aai-OyOAg&sa=X&oi=book_result&ct=result&resnum=9#v=onepage&q=digital%20audio%20gain%20control&f=false
grill
>
> - Idézett szöveg megjelenítése -
RW
would recommend you sprinkle a few emoticons throughout your post(s)
just to make sure folks know that you are kidding. Adding a <g> to
the end of a post goes a long way towards letting people know you're
joking...
Best Wishes,
-RW-
On Aug 17, 6:34 am, grill <gr...@index.hu> wrote:
> I'm not an audio specialist at all. I found this book chapter about
> gain control in digital audio. It may or may not help, all of you can
> grill
>
> On aug. 12, 04:16, georgeshan...@comcast.net wrote:
>
>
>
> > VF,
>
> > If there is no such thing as gain control with digital recording, then how do we explain this? This is a picture of my laptop desktop with a rudimentary digital recording gain control. Your application and other recording applications such as Sony Soundforge Acid have more sophisticated gain controls. Usually they have meters by which one can monitor the gain and watch out for clipping. When and if clipping occurs, the gain is then reduced and the recording monitored again until the gain is correctly set.
>
> > Anyway, here is the picture:
>
> >http://home.comcast.net/~georgeshannon/a.rtf
>
> > ----- Original Message -----
> > From: "VF" <vf.surro...@gmail.com>
> > To: Surrou...@googlegroups.com
> > Sent: Tuesday, August 4, 2009 12:09:11 PM GMT -07:00 US/Canada Mountain
> > Subject: [SurroundSound] Re: Digital SACD Ripping Guide
>
> > OK - 1. There's no such thing as gain control with digital recording! Have you actually tried this, with digital S/PDIF recording equipment? Have you tried it yourself? Or have you only done analog recording? I searched high and low (once again, like when you brought this up months ago) to try and find somewhere where I can lower the "recording" gain in the Delta Control Panel or Adobe Audition, but I could not find it. Is this because gain doesn't "exist" for digital signals? Let me guess - yes!!!! The Delta Control Panel has gain controls for the analog inputs, that's for sure, but not digital! IF such a thing existed for digital recording I'm pretty sure it would be there in the Delta Control Panel. And if there's some way of lowering it with some recording programs that you might know, it wouldn’t be called "gain", and it would be disastrous as it would remove bits from the audio, just like the Oppo volume control!- Idézett szöveg elrejtése -
>
george...@comcast.net
I thought I did that with the "LOL" at the end of the sentence. Maybe you need some training in reading comprehension. <g>
Seriously, I think we all knew that was out of character for you. Apology accepted. Let's not take this hobby too seriously.
Between what I posted of my desktop and what grill gave a link to, I'm starting to think there is a digital gain control available. What do you think?
George
----- Original Message -----
From: "RW" <rlwain...@gmail.com>
To: "SurroundSound" <surrou...@googlegroups.com>
VF
What grill linked to, although useful info, was irrelevant to the main, important discussion (which I thought we resolved), which was, *should you lower the (digital) recording gain of an already-clipped signal*….to which the answer is, NO.
The answer to your question, George, is yes. There IS such a thing as digital input gain control and I was wrong before.
RW
lies in the fact that invoking it causes a loss of word length.
Someone else here posted that you lose 1 bit for every 6db of
attenuation. Therefore, if you (digitally) lower the input signal by
12 db, you've just tossed away 2 bits. On a 24bit file, this may not
be such a big deal, but doing so with a 16bit CD signal would be
significant because now you're down to only 14 bits and this would
probably be quite noticeable to those of us with good ears and good
systems. And it STILL will do nothing to fix the original problem.
So, why bother to do it all?
-RW-
On Aug 17, 9:55 am, "VF" <vf.surro...@gmail.com> wrote:
> What grill linked to, although useful info, was irrelevant to the main, important discussion (which I thought we resolved), which was, *should you lower the (digital) recording gain of an already-clipped signal*….to which the answer is, NO.
>
> The answer to your question, George, is yes. There IS such a thing as digital input gain control and I was wrong before.
>
george...@comcast.net
Your argument is based on the assumption that the SACD is clipped. My argument is based on the assumption that it is not.
Lowering the signal by 12dB is a ridiculously huge amount. We need an attenuation of 2 or 3 dB on a 24 bit file. That is less than 1 bit out of 24.
The reason for doing so is to prevent the clipping that exists on the finished product. Clipping not only produces a distorted sound but can damage loudspeakers.
george...@comcast.net
Your assumption that the signal is already clipped is incorrect.
Lokkerman
Surely if this is done; i,e, that we know what is coming off the disc is matching what is being output to the bitstore (HDD) then we know that some other artefact is at play.
george...@comcast.net
A known non-clipped SACD. That's a great idea. Calibration is needed. How could one be determined?
grill
DVDA. Of course, DSD->LPCM conversion is another story.
On aug. 17, 21:47, georgeshan...@comcast.net wrote:
> A known non-clipped SACD. That's a great idea. Calibration is needed. How could one be determined?
>
>
>
> ----- Original Message -----
> From: "Lokkerman" <phil.steep...@gmail.com>
> To: surrou...@googlegroups.com
> Sent: Monday, August 17, 2009 12:38:15 PM GMT -07:00 US/Canada Mountain
> Subject: [SurroundSound] Re: Digital SACD Ripping Guide
>
> Can I just raise a point; are we sure that the levels are not set in the DAC, when it processes the SACD, and does the conversion to the PCM output? It would appear to me that a calibrated output from a calibrated source is what we require, afterall this is what would be done with an analogue set-up.
> Surely if this is done; i,e, that we know what is coming off the disc is matching what is being output to the bitstore (HDD) then we know that some other artefact is at play.
>
jolson
another?
On 17 Aug, 21:47, georgeshan...@comcast.net wrote:
> A known non-clipped SACD. That's a great idea. Calibration is needed. How could one be determined?
>
> ----- Original Message -----
> From: "Lokkerman" <phil.steep...@gmail.com>
> To: surrou...@googlegroups.com
> Sent: Monday, August 17, 2009 12:38:15 PM GMT -07:00 US/Canada Mountain
> Subject: [SurroundSound] Re: Digital SACD Ripping Guide
>
> Can I just raise a point; are we sure that the levels are not set in the DAC, when it processes the SACD, and does the conversion to the PCM output? It would appear to me that a calibrated output from a calibrated source is what we require, afterall this is what would be done with an analogue set-up.
> Surely if this is done; i,e, that we know what is coming off the disc is matching what is being output to the bitstore (HDD) then we know that some other artefact is at play.
>
Lokkerman
george...@comcast.net
Copying files from one location to another is a totally different subject than what we're discussing in this thread. In this thread we're discussing VF's process of recording files into his PC from a standalone player. His player is playing an SACD through SPDIF into a recorder software with digital input gain stage control. The gain control is currently set too high resulting in clipping on the finished files. I recommend he monitor the recording level and adjust the gain until the the files have the maximum signal level without clipping. The Herbie Hancock recording would benefit from a 2dB reduction in gain. The DSOTM recording would benefit from a 3 dB reduction in gain.
Hope this helps,
george...@comcast.net
I'll bet the major recording studios have such a disc or file. Trouble is VF is pioneering a new home based technology and he hasn't the benefit of those tools, yet. He may be the one to invent such tools for others of us to use someday.
----- Original Message -----
From: "Lokkerman" <phil.s...@gmail.com>
To: surrou...@googlegroups.com
Sent: Monday, August 17, 2009 3:33:26 PM GMT -07:00 US/Canada Mountain
Subject: [SurroundSound] Re: Digital SACD Ripping Guide
there must be a test sacd - let's all dig around (frantically ) to find one, but there must be one that can have all the test parameters like a test tape, which I have to calibrate R2R recorders....
george...@comcast.net
I was thinking about this also. Do you think an album like Sea Change or Tommy might be helpful. If the DVDA was not clipped, could we assume the SACD was not clipped?
VF
Ok. It looks like I'm getting bad luck from all sides at the moment.
In wanting to get to the bottom of this issue, I have looked deep into the complicated and obscure world of DSD to PCM conversion, and I have made a shocking discovery. THE OPPO (yes, the "wonderful" Oppo) IS CLIPPING ALL SACDS BY 3dB!
Read this!: http://www.diyaudio.com/forums/showthread.php?postid=1441829#post1441829. Read that thread from the bottom of that page to the end of the thread. I am disgusted.
I have made tests by recording the analog output signal from my AV receiver connected to the pure DSD signal output from the Oppo, and it revealed the original shape of the DSotM SACD :(.
Just to make sure it wasn't the S/PDIF output board that was clipping but in fact the Oppo player itself (grr), I recorded the same analog output signal from my AV receiver but this time with the Oppo feeding it its 24/88.2 PCM output (this is through its HDMI output, not the mod board), and it came out clipped like the straight digital rip through the mod board. I level matched the recordings and calculated that is approximately 3 dB worth of clipping.
Note that it's only the SACDs that are loud enough in the first place that end up clipped! *Some* SACDs, like the wonderful Steve Reich, are still well within the range and have loads of space to breathe in. So for nice cool discs, the Oppo might ramp it up by 3dB, but it's still nowhere near the 0dB mark, and it's not suffocated in any way. But whatever's a hot SACD disc and close to the limit, the Oppo will clip!
As for accuracy of original gain levels for the disc, I wonder how analog rippers try to work out what the original levels of the disc was. I gues they can't. In analog, it's all just relative and they probably record classical titles at a much higher level than the original disc - with the Oppo digital method, you can at least figure out what the level of the disc was mastered at. Of course, it doesn't matter how close it is to 0db, as long as it doesn't touch it, but I'm jut sayin...
The noticable track in DSotM where the Oppo's clipping occurs, is "money" - the barbaric saxophone solo near the end. I can thankfully say though, that not much of the 40 minutes of DSotM ends up being clipped. But I, more than anyone (well whatever), am NOT satisfied with an imperfection like this so I am pretty cut about this discovery, to say the least.
But really, does this make my whole investment of this ripping rig and my mammoth ripping efforts worthless? Is this whole gig a deception that's been feeding a placebo of superiority to analog rips? Of course not! The audible difference I have heard between my Oppo PCM rips and other analog rips has ranged from "definite improvement" to "huge difference" - sometimes, as much as the difference between say an MFSL master and a bad master. And others have reported the same results too. Use your own ears whether you think my 3dB clipped digital recording of DSotM sounds better than the russian analog recording or not.
BUT it is a shame. I'd also like to ask everyone (because I really don't know): how bad is 3dB worth of clipping? How much of an *audible* degradation does it bring? I don't have time to take listening tests to try and detect a difference, but I'd like to know how bad this clipping is, in the real scheme of things. Right now I honestly don't know how significant this problem is.
But no matter how *IN*significant it is, I can't bear the thought of making clipped recordings, and I will *just not have it* in the long run. And guess what, fellas. I have an announcement to make. There's hope! I didn't want to tell the group this until much later, but in light of this setback I'd like to make an announcement. I am developing a method to rip the *pure DSD signal* from SACD, forgetting about PCM conversion forever. I am working on a number of solutions for pure DSD playback of these DSD rips, including a DSD codec for the PC/HTPC and SACD-R (that's right, an SACD ISO file that you burn to DVD....it does exist, and you would play it back in your SACD player).
On top of that, for those who will still need to experience these rips in the PCM realm, I will release software which will allow us to convert my pure DSD files to 24-bit 192kHz. The result, when you compare it with the Oppo's DSD to PCM conversion that I currently have to deal with (grr), will be much superior, and at a higher resolution as well. And this is 24-bit 192kHz *in multichannel* too - good timing for the growth of blu-ray and its superior capacity to DVD-A, perhaps :P. So how does this all sound? I hope it serves as a consolation. There are good days ahead for us, folks! :D
BUT.....I cannot give a guarantee of WHEN this pure DSD method will get off the ground. I am hoping it will be within the next 4 months. Lots of things are being worked out right now, but I can tell you it is all positive. The DSD world is really going to be opened up, I KNOW that. It's just a matter of when, not if. Please be patient. And I would also like to request....please do not ask me for any more details about this pure DSD stuff, it is rather touchy stuff and I just hope Sony don't read this. Thanks for understanding.
So to sum up...arn: you're on holidays right now, but congratulations (and sorry for dismissing your theory): you WERE right about DSD containing sounds that are too loud for PCM to handle - the decimation process needs to be done in a way to avoid this problem when converting DSD to PCM.
And the final question to ask, is: which SACDs really ARE clipped (non matter how purely you rip it)? Some may be "hot" and "close to the limit" ones like DSotM, but some ones are clipped whether you like it or not. I guess the definitive way to check, is to take an analog recording like I did with "Money". Mind you, it must be a DSD to analog path that you're recording to PCM, otherwise it might very well be clipped just like the Oppo's PCM output. In fact, I wonder for a second, if many analog rips out there are taken from the outs of SACD players whose DACs are PCM only. I shudder at the thought. DSD to PCM to analog to PCM....not something I'd want to experience...
And at this point I just wonder what the state of PCM output is for many SACD players out there, in general. The DSD to PCM conversion world turns out to be a messy one, and not a clean one as I suspected. Friends, if you want to listen to your physical SACDs with good fidelity, *use a pure DSD solution*! Investigate the DACs in your receiver/player, make sure it's not converted to PCM at ANY stage of the process! And to those not able to get into SACD, or still wanting quality PCM rips that they can play in places SACD and DSD cannot go, you can now look forward to something better than what I currently provide. The current rips I'm doing will not be the definitive SACD rips in the long run.
Mind you, I am NO less discouraged to continue to rip discs under the current digital PCM method. It still is opening up a lot of titles and new music for everyone in a quality that is in most cases shoulders above what we've had access to before. I have about 150 titles coming up in the next 4-5 months that I will be proudly releasing, and I hope you enjoy them!
Also.......George....how did you figure out that DSotM was 3dB over? Hats off to you!....is there a way to fiddle with a clipped recording and figure out by how many dB it is clipped? If so, that's another way to probably work out if an SACD was originally all the way to 0dB or not. One thing is for sure: some of the rock titles I've recorded recently, ARE clipped in their original form. e.g. an Incubus disc I recently did. I just took an analog recording to make sure, and yes the original shape is clipped, although not as much as my pcm rip, ofc :/.
THE END.
Thanks for reading my big, FAT message. :)
george...@comcast.net
Damn. VF don't be discouraged. You are at 95%. We were just hoping to push that to 99%. Your conversions are great. I listen to them. Can you install the conversion card into a Denon? Would that make a difference?
Soundforge is the tool I use for examining the wave forms. It has the capability to lower the signal until the clipped peaks are restored. Then one can measure the amount of attenuation required.
I know Ziggy Stardust is a cool disc. One has to really crank it to hear it in its full splendor. It even says so on the back cover.
VF
No mate, we’re only at about 80%, IMO :). The difference between DSD and PCM is quite a noticeable one, to my ears and with my gear. Elshagon has noted the difference too – as do many other audio lovers.
And with pure DSD, we will also avoid ALL the problems which arise with conversion to PCM as we have discovered.
Can you install the conversion card into a Denon? Would that make a difference?
I do not have soldering skills myself, I have others who do that for me. I do not have any Denon players either, and I do not trust the PCM output of ANY player now. Sure, I could try and see, but it’s not worth buying anything just to test out. From now on it’s pure DSD all the way, baby.
VF
“Your assumption that the signal is already clipped is incorrect.”
George….I must say that I am VERY annoyed that you still believe this. It I simply false information.
Don’t you realise that this is actually embarrassing for you, saying this? You are coming across as a major noob if you (still!) think the signal coming out from the Oppo is NOT clipped, but that it is fine and that the problem lies in the recording (m-audio S/PDIF input) side.
Everyone else on this board understands that the signal being fed into my m-audios is already clipped, and I have even now discovered and PROOVED (and admitted) that the Oppo gives a CLIPPED signal, and that in half the cases, it’s NOT the disc.
Now you’re suggesting that I try another player. That suggests that you understand that the player IS the problem, not the recording gain. So if you;’re saying that, will you accept that the signal IS “already clipped”, now?
And to make sure we’re talking about the same thing, you do realise we’re talking about the SIGNAL, NOT the bits on the original SACD disc, right?
Lokkerman
Also worryingly I noted in your earlier remarks that you mentioned 192/24 from DSD, DSD klogically decimates to 88/24,(that was the design) to reach 192 you would need to upconvert. I think these matters we are investigating now are designed in this medium and refer back to Stan Lipshtiz paper I gave the lnk for a couple of days ago...
VF
I’m not sure about you (I wouldn’t want to wrongfully assume), but I’m no expert in DSD to PCM conversion, I can only declare that I’ve discovered how crazily complex it is.
How about 176.4kHz? That’s a clean decimation, right?
And if there are professional DSD DAWs like Sonoma or Sadie or pyramix or digital audio denmark, that can convert DSD to 192kHz PCM (afaik…not just 176.4), there must be algorithms which do it well. But if in the end I find 176.4 is the best result from DSD, then I’d suggest that option. The DXD format (which cleanly converts to both DSD and PCM) is 352.8kHz, and maybe for a good reason.
Another useful link for anyone interested in this complex world:
http://www.gearslutz.com/board/high-end/400961-dsd-pcm-dither-now-later.html
And what the heck….let’s share the information and in doing so we can all probably learn more about this world.
http://www.korgcanada.com/eng/support/downloads/pdf/AudioGate_Users_Guide.pdf
In this document it states “1-bit audio files may contain louder signals than PCM. When converting 1-bit audio files to PCM files, you might need to set GAIN settings down by approximately -3.0dB to avoid clipping (see page 13)”.
Anyway, in that gearslutz thread (and btw gearslutz is where all the dsd experts hang out, you can find lots of great dsd info there) - in this thread they talk about audiogate and how it isn’t likely the best software algorithm available for DSD to PCM conversion. But I guess, the lowering gain in the “converter” doesn’t reduce bits, as it happened BEFORE the PCM it put out. It’s a matter of how BEST to convert DSD to PCM. This is where it’s very complex.
Anyway, I won’t comment on which software I intend to release to the public….I’ll just say that you’ll likely get something better than audiogate :).
As for firmware….it hasn’t been fixed afaik. I’m trying that beta 980 firmare tonight, just in case. And I have to transfer bowie’s heathen too :).
grill
energy and time to keep going on.
On aug. 18, 07:11, "VF" <vf.surro...@gmail.com> wrote:
> Ok. It looks like I'm getting bad luck from all sides at the moment.
>
> In wanting to get to the bottom of this issue, I have looked deep into the complicated and obscure world of DSD to PCM conversion, and I have made a shocking discovery. THE OPPO (yes, the "wonderful" Oppo) IS CLIPPING ALL SACDS BY 3dB!
>
> Read this!:http://www.diyaudio.com/forums/showthread.php?postid=1441829#post1441829. Read that thread from the bottom of that page to the end of the thread. I am disgusted.
>
> I have made tests by recording the analog output signal from my AV receiver connected to the pure DSD signal output from the Oppo, and it revealed the original shape of the DSotM SACD :(.
>
> Just to make sure it wasn't the S/PDIF output board that was clipping but in fact the Oppo player itself (grr), I recorded the same analog output signal from my AV receiver but this time with the Oppo feeding it its 24/88.2 PCM output (this is through its HDMI output, not the mod board), and it came out clipped like the straight digital rip through the mod board. I level matched the recordings and calculated that is approximately 3 dB worth of clipping.
>
> Note that it's only the SACDs that are loud enough in the first place that end up clipped! *Some* SACDs, like the wonderful Steve Reich, are still well within the range and have loads of space to breathe in. So for nice cool discs, the Oppo might ramp it up by 3dB, but it's still nowhere near the 0dB mark, and it's not suffocated in any way. But whatever's a hot SACD disc and close to the limit, the Oppo will clip!
>
> As for accuracy of original gain levels for the disc, I wonder how analog rippers try to work out what the original levels of the disc was. I gues they can't. In analog, it's all just relative and they probably record classical titles at a much higher level than the original disc - with the Oppo digital method, you can at least figure out what the level of the disc was mastered at. Of course, it doesn't matter how close it is to 0db, as long as it doesn't touch it, but I'm jut sayin...
>
> The noticable track in DSotM where the Oppo's clipping occurs, is "money" - the barbaric saxophone solo near the end. I can thankfully say though, that not much of the 40 minutes of DSotM ends up being clipped. But I, more than anyone (well whatever), am NOT satisfied with an imperfection like this so I am pretty cut about this discovery, to say the least.
>
> But really, does this make my whole investment of this ripping rig and my mammoth ripping efforts worthless? Is this whole gig a deception that's been feeding a placebo of superiority to analog rips? Of course not! The audible difference I have heard between my Oppo PCM rips and other analog rips has ranged from "definite improvement" to "huge difference" - sometimes, as much as the difference between say an MFSL master and a bad master. And others have reported the same results too. Use your own ears whether you think my 3dB clipped digital recording of DSotM sounds better than the russian analog recording or not.
>
> BUT it is a shame. I'd also like to ask everyone (because I really don't know): how bad is 3dB worth of clipping? How much of an *audible* degradation does it bring? I don't have time to take listening tests to try and detect a difference, but I'd like to know how bad this clipping is, in the real scheme of things. Right now I honestly don't know how significant this problem is.
>
> But no matter how *IN*significant it is, I can't bear the thought of making clipped recordings, and I will *just not have it* in the long run. And guess what, fellas. I have an announcement to make. There's hope! I didn't want to tell the group this until much later, but in light of this setback I'd like to make an announcement. I am developing a method to rip the *pure DSD signal* from SACD, forgetting about PCM conversion forever. I am working on a number of solutions for pure DSD playback of these DSD rips, including a DSD codec for the PC/HTPC and SACD-R (that's right, an SACD ISO file that you burn to DVD....it does exist, and you would play it back in your SACD player).
>
> On top of that, for those who will still need to experience these rips in the PCM realm, I will release software which will allow us to convert my pure DSD files to 24-bit 192kHz. The result, when you compare it with the Oppo's DSD to PCM conversion that I currently have to deal with (grr), will be much superior, and at a higher resolution as well. And this is 24-bit 192kHz *in multichannel* too - good timing for the growth of blu-ray and its superior capacity to DVD-A, perhaps :P. So how does this all sound? I hope it serves as a consolation. There are good days ahead for us, folks! :D
>
> BUT.....I cannot give a guarantee of WHEN this pure DSD method will get off the ground. I am hoping it will be within the next 4 months. Lots of things are being worked out right now, but I can tell you it is all positive. The DSD world is really going to be opened up, I KNOW that. It's just a matter of when, not if. Please be patient. And I would also like to request....please do not ask me for any more details about this pure DSD stuff, it is rather touchy stuff and I just hope Sony don't read this. Thanks for understanding.
>
> So to sum up...arn: you're on holidays right now, but congratulations (and sorry for dismissing your theory): you WERE right about DSD containing sounds that are too loud for PCM to handle - the decimation process needs to be done in a way to avoid this problem when converting DSD to PCM.
>
> And the final question to ask, is: which SACDs really ARE clipped (non matter how purely you rip it)? Some may be "hot" and "close to the limit" ones like DSotM, but some ones are clipped whether you like it or not. I guess the definitive way to check, is to take an analog recording like I did with "Money". Mind you, it must be a DSD to analog path that you're recording to PCM, otherwise it might very well be clipped just like the Oppo's PCM output. In fact, I wonder for a second, if many analog rips out there are taken from the outs of SACD players whose DACs are PCM only. I shudder at the thought. DSD to PCM to analog to PCM....not something I'd want to experience...
>
> And at this point I just wonder what the state of PCM output is for many SACD players out there, in general. The DSD to PCM conversion world turns out to be a messy one, and not a clean one as I suspected. Friends, if you want to listen to your physical SACDs with good fidelity, *use a pure DSD solution*! Investigate the DACs in your receiver/player, make sure it's not converted to PCM at ANY stage of the process! And to those not able to get into SACD, or still wanting quality PCM rips that they can play in places SACD and DSD cannot go, you can now look forward to something better than what I currently provide. The current rips I'm doing will not be the definitive SACD rips in the long run.
>
> Mind you, I am NO less discouraged to continue to rip discs under the current digital PCM method. It still is opening up a lot of titles and new music for everyone in a quality that is in most cases shoulders above what we've had access to before. I have about 150 titles coming up in the next 4-5 months that I will be proudly releasing, and I hope you enjoy them!
>
> Also.......George....how did you figure out that DSotM was 3dB over? Hats off to you!....is there a way to fiddle with a clipped recording and figure out by how many dB it is clipped? If so, that's another way to probably work out if an SACD was originally all the way to 0dB or not. One thing is for sure: some of the rock titles I've recorded recently, ARE clipped in their original form. e.g. an Incubus disc I recently did. I just took an analog recording to make sure, and yes the original shape is clipped, although not as much as my pcm rip, ofc :/.
>
> THE END.
>
> Thanks for reading my big, FAT message. :)
>
> Sent: 18 August 2009 14:08
> To: surrou...@googlegroups.com
> Subject: [SurroundSound] Re: Digital SACD Ripping Guide
>
> I was thinking about this also. Do you think an album like Sea Change or Tommy might be helpful. If the DVDA was not clipped, could we assume the SACD was not clipped?
>
>
>
> ----- Original Message -----
> From: "grill" <gr...@index.hu>
> To: "SurroundSound" <surrou...@googlegroups.com>
> Sent: Monday, August 17, 2009 2:11:47 PM GMT -07:00 US/Canada Mountain
> Subject: [SurroundSound] Re: Digital SACD Ripping Guide
>
> I think VF's ripping method can be partly checked with a non-clipped
> DVDA. Of course, DSD->LPCM conversion is another story.
>
> On aug. 17, 21:47, georgeshan...@comcast.net wrote:
> > A known non-clipped SACD. That's a great idea. Calibration is needed. How could one be determined?
>
> > ----- Original Message -----
> > From: "Lokkerman" <phil.steep...@gmail.com>
> > To: surrou...@googlegroups.com
> > Sent: Monday, August 17, 2009 12:38:15 PM GMT -07:00 US/Canada Mountain
> > Subject: [SurroundSound] Re: Digital SACD Ripping Guide
>
> > Can I just raise a point; are we sure that the levels are not set in the DAC, when it processes the SACD, and does the conversion to the PCM output? It would appear to me that a calibrated output from a calibrated source is what we require, afterall this is what would be done with an analogue set-up.
> > Surely if this is done; i,e, that we know what is coming off the disc is matching what is being output to the bitstore (HDD) then we know that some other artefact is at play.
>
> > On Mon, Aug 17, 2009 at 7:31 PM, < georgeshan...@comcast.net > wrote:
>
> > Your argument is based on the assumption that the SACD is clipped. My argument is based on the assumption that it is not.
>
> > Lowering the signal by 12dB is a ridiculously huge amount. We need an
>
>
> tovább »- Idézett szöveg elrejtése -
Neil Wilkes
Some SACD are heavily brickwalled to -6dBFS, and are still distorted -
see Layla for a classic example. Add in the ultrasonic crapola, and
it's a dodgy format whatever way you look at it.
For example, have you looked at the Depeche Mode SACD/DTS-DVD discs? I
have them all, and without exception the dynamic range is actually
higher off the DTS streams than off the DSD ones by almost 100% - DSD
dynamic range is around 70dB at best whilst the DTS stream is at
100-110dB.
DSD was never developed as a playable system, but an archival one -
never meant to be used as a consumer format.
Remember it is a 1-bit system using heroic noise shaping to make it
listenable, and that ultrasonic noise bleeds into the audible range.
Email me for comparisons - I have some very revealing data here that I
found for myself - not just copying what others have said.
Having said that, I do agree there are some duiscs that are very good
indeed:
Avalon, Boys & Girls for example. Also 461 Ocean boulevard. I have
many more that are a sonic nightmare though.
On Aug 5, 2:06 pm, RW <rlwainwri...@gmail.com> wrote:
> VF is absolutely correct about the "gain" issue. Some discs are
> simply engineered very "hot" and there isn't jack shit you can do
> about it. I believe his methodology of taking the signal as-is and
> simply getting it onto a disc we can all enjoy is the best way to
> approach it. If it was recorded too hot, that's the way it is,
> hombre, and you gotta live with that. Even getting the original
> master tape wouldn't help you any, the signal is fucked right there
> for all eternity. The fault lies in the original production, live with
> it...
>
> -RW-
VF
that I myself have a lot to learn and am no audio engineer and epert, but
rather a gradually more knowledgeable enthusiast, and indeed that we are
here to educate each other, as I admit you have done towards me), I'd like
to state the FACT that a clipped signal does not necessarily mean it's
reaching all the way to 0dB. If you simmply attenuate a clipped signal it
will remain clipped, despite it being softer.
Clipping, essentially, is all about the SHAPE of the waveform. NOT how may
decibels it is! You really don't seem to get that basic FACT.
I do not understand your process that you've described with regards to
sourceforge. It sounds like some simple process and you think you're fixing
the waveform when in fact it's far from it. Why don't you google around and
actually learn the facts about de-clipping. I just did so and found this
useful link: http://goldwave.ca/forums/viewtopic.php?f=1&t=2903
Note these words of truth: "Compression/limiting is difficult (or
impossible) to un-do". It's no easy process. I've known this for years, and
it's all I've ever needed to know about the issue - since I have no desire
to fix clipped recordings...I'd rather go find nice vinyl rips/MFSL releases
instead.
I remember you saying ages ago how you said you "fixed" one of my clipped
SACD recordings and "restored" its natural dyamic range. Now, I seriously
doubt you did anything of the sort. I would only believe you if you gave me
a screenshot of a before and after, for me to see what you do to "fix" it
(i.e., restore it to its original shape). I could then make an analog
recording of the pure DSD output from my Oppo, showing the full original
"shape" to compare your "restoration" to.
It seems like you claim that simple attenuation just "fixes" the clipping.
that is utter rubbish. You're speaking nonsense. It's ok to admit that
you're wrong George. I certainly have done that with MANY points in this
discussion. You're simply wrong. the next step is for you is to realise this
and then admit it. There's no ego here....not with me, anyway.
-----Original Message-----
From: surrou...@googlegroups.com [mailto:surrou...@googlegroups.com]
On Behalf Of grill
Sent: 18 August 2009 18:33
To: SurroundSound
Subject: [SurroundSound] Re: Digital SACD Ripping Guide
gfv
clipping the DSOTM sacd, made me curious about the conversion in the
new Oppo BDP-83. I got my hands on a BDP-83 recently, but I must still
set it up. One of it's features seem to be a selection between DSD to
DAC or DSD converted to PCM and then to DAC, for the analog outputs.
Can someone tell me at what offset into the "Money" track, the
clipping is most noticeble? I want to evaluate the DSD to PCM
conversion of the BDP-83 or maybe someone else with a BDP-83 can do it?
VF
I know you're very knowledgeable about DSD's issues and problems. I would
like to ask you what *audible* difference you say this noise that bleeds
into the audible range, actually has.
I think I'm starting to see the picture. My approach with determining what
effects these things have on what you hear is to apply the science of audio
with the subjective "evidence" from, yep, what I hear.
People say (and I agree) that DSD is "smoother" and that PCM has more
brilliance and sparkle. I'm thinking this is because PCM (say 96kHz) has
active audio information in the supersonic frequencies, but DSD only has
noise.
The claim that whatever is in the supersonic frequencies affects what is in
the audible range seems to be a very debatable one, but due to hearing
differences between DSD and PCM first hand, I'm going to believe that they
do.
So, we lose some sparkle in DSD, but there are other things we gain.
Whatever it is in DSD's science, (might be the narrow analog-like pulse
response, flat frequency response, I'm not too sure at this point), DSD has
more *warmth* and depth - and detail in the mid range. In *some* respects,
it's closer to pure analog sound, and PCM, with its much wider pulse
response, or whatever it is, doesn't measure up as much there.
And also - if DSD has this distortion caused by the noise, might it be a
distortion that people actually like?....and one observation I have made
about DSD, is that it is very "musical". Maybe this is because the heart of
music happens in the mid frequency range and that is where DSD is at its
best.
But as soon as I first experienced DSD (late last year), my major initial
reaction was that it was very *beautiful* and quite musical. But this is all
subjective, isn't it...well I acknowledge that and don't mind it, because
this is all about *how it sounds*, isn't it.
The whole point that DSD was meant as an archival format and not really a
playback format, the noise issue, and the maximum dynamic range issue, is
all negated by the higher resolution variant of DSD, DSD128. With the
sampling rate being double, at 5.6448 MHz, the noise can be pushed up to
frequencies twice as high - putting it far away from the audible band. And
because of this and the way the noise shaping works, the dynamic range is
doubled as well.
And it gets better than that. There have been recent advances in digital
audio, and we now have the DXD format, the best of both worlds. It stops the
endless "DSD, or PCM?" issue and trumps them both. For people's info, read
about DXD here:
http://www.digitalaudio.dk/technical_papers/axion/dxd%20Resolution%20v3.5.pd
f
http://www.digitalaudio.dk/ax24_present.htm
But, sure....SACD is stuck with DSD64 and that's what we're talking about.
And it does have its limitations and issues. But it DOES have some things
that PCM *doesn't* have. I like this extra something that DSD offers, I like
it too much, to ignore it. And it's certainly better than 16-bit audio,
that's for sure.
But here comes a question I'd like to ask (another shameless noobish
question): does this dynamic range issue even matter? Does music even use
that range, and if I compare a "110dB range track" to a "60dB range track"
will it be noticeable?
Either way - sonically, SACD has something unique to offer (that PCM can't),
and I like that part of the audio "spectrum" that it trumps PCM on. I like
it very much. And that's what matters here.
VF
If you can test it (by recording the FL+FR analog outs with the player to
"PCM to DAC") that'd be fantastic. You need to compare it to the pure DSD to
analog output and zoom out (vertically) far enough to see if the shape of
the loud section (which is a broad gentle arc) is flattened or not.
Start recording from about 4:15 for about 60 seconds, that'll cover this
loud section.
If it turns out the BDP-83 doesn't clip the PCM upon conversion, I
definitely can't afford to get one of these beasts though, let alone pay to
get it modded. :/
-----Original Message-----
From: surrou...@googlegroups.com [mailto:surrou...@googlegroups.com]
On Behalf Of gfv
Sent: 18 August 2009 19:21
To: SurroundSound
Subject: [SurroundSound] Re: Digital SACD Ripping Guide
jolson
Horse, doomed not to have everyone to understand how digital gain when
recording over SPDIF works. Especially when we don't know everything
ourselves...
But I know Neil Wilkes know (just about) everything there is to know
about recording/mixing/mastering, so maybe you can say what are facts
and what are not. Which of these statements are correct?
* When recording the output from three spdif from a modified Oppo to
three M-Audio cards with spdif, with no further mixing/processing, the
recording gain should always be 0dB.
* Having digital recording gain set to 0dB cannot in itself create
clipping.
* When recording multiple tracks, one after another, from (for
instance) a MIDI instrument for later mixing to stereo, recording gain
needs to be lowered to (for example) -12dB. This compensates for the
added volume when mixing (for example) 8 tracks together for a CD.
* If you send a clipped signal out of the Oppo, changing gain on the
recording side can never compensate for that.
jolson
recording would be if either the original SACD was clipped, or if the
DSD->PCM conversion created the clipping. Then I wrote the second
explanation off as impossible, since many people would have heard that
clipping when playing SACD's on their Oppos. It would affect everyone
who uses Oppo's DSD to PCM conversion (which is a majority I believe,
not many AVR's take DSD and the older Oppo's also made the conversion
before sending the signal through the DACs to multi-channel analogue
out) if the SACD is within -3dB.
But it seems I was wrong, Oppo may be at fault. I will tell a friend
who is at the Oppo forum, if it's fixable by FW then at least the Blu-
ray player's FW is still heavily in development.
On 18 Aug, 07:11, "VF" <vf.surro...@gmail.com> wrote:
> Ok. It looks like I'm getting bad luck from all sides at the moment.
>
> In wanting to get to the bottom of this issue, I have looked deep into the complicated and obscure world of DSD to PCM conversion, and I have made a shocking discovery. THE OPPO (yes, the "wonderful" Oppo) IS CLIPPING ALL SACDS BY 3dB!
>
> Read this!:http://www.diyaudio.com/forums/showthread.php?postid=1441829#post1441829. Read that thread from the bottom of that page to the end of the thread. I am disgusted.
>
> I have made tests by recording the analog output signal from my AV receiver connected to the pure DSD signal output from the Oppo, and it revealed the original shape of the DSotM SACD :(.
>
> Just to make sure it wasn't the S/PDIF output board that was clipping but in fact the Oppo player itself (grr), I recorded the same analog output signal from my AV receiver but this time with the Oppo feeding it its 24/88.2 PCM output (this is through its HDMI output, not the mod board), and it came out clipped like the straight digital rip through the mod board. I level matched the recordings and calculated that is approximately 3 dB worth of clipping.
>
> Note that it's only the SACDs that are loud enough in the first place that end up clipped! *Some* SACDs, like the wonderful Steve Reich, are still well within the range and have loads of space to breathe in. So for nice cool discs, the Oppo might ramp it up by 3dB, but it's still nowhere near the 0dB mark, and it's not suffocated in any way. But whatever's a hot SACD disc and close to the limit, the Oppo will clip!
>
> As for accuracy of original gain levels for the disc, I wonder how analog rippers try to work out what the original levels of the disc was. I gues they can't. In analog, it's all just relative and they probably record classical titles at a much higher level than the original disc - with the Oppo digital method, you can at least figure out what the level of the disc was mastered at. Of course, it doesn't matter how close it is to 0db, as long as it doesn't touch it, but I'm jut sayin...
>
> The noticable track in DSotM where the Oppo's clipping occurs, is "money" - the barbaric saxophone solo near the end. I can thankfully say though, that not much of the 40 minutes of DSotM ends up being clipped. But I, more than anyone (well whatever), am NOT satisfied with an imperfection like this so I am pretty cut about this discovery, to say the least.
>
> But really, does this make my whole investment of this ripping rig and my mammoth ripping efforts worthless? Is this whole gig a deception that's been feeding a placebo of superiority to analog rips? Of course not! The audible difference I have heard between my Oppo PCM rips and other analog rips has ranged from "definite improvement" to "huge difference" - sometimes, as much as the difference between say an MFSL master and a bad master. And others have reported the same results too. Use your own ears whether you think my 3dB clipped digital recording of DSotM sounds better than the russian analog recording or not.
>
> BUT it is a shame. I'd also like to ask everyone (because I really don't know): how bad is 3dB worth of clipping? How much of an *audible* degradation does it bring? I don't have time to take listening tests to try and detect a difference, but I'd like to know how bad this clipping is, in the real scheme of things. Right now I honestly don't know how significant this problem is.
>
> But no matter how *IN*significant it is, I can't bear the thought of making clipped recordings, and I will *just not have it* in the long run. And guess what, fellas. I have an announcement to make. There's hope! I didn't want to tell the group this until much later, but in light of this setback I'd like to make an announcement. I am developing a method to rip the *pure DSD signal* from SACD, forgetting about PCM conversion forever. I am working on a number of solutions for pure DSD playback of these DSD rips, including a DSD codec for the PC/HTPC and SACD-R (that's right, an SACD ISO file that you burn to DVD....it does exist, and you would play it back in your SACD player).
>
> On top of that, for those who will still need to experience these rips in the PCM realm, I will release software which will allow us to convert my pure DSD files to 24-bit 192kHz. The result, when you compare it with the Oppo's DSD to PCM conversion that I currently have to deal with (grr), will be much superior, and at a higher resolution as well. And this is 24-bit 192kHz *in multichannel* too - good timing for the growth of blu-ray and its superior capacity to DVD-A, perhaps :P. So how does this all sound? I hope it serves as a consolation. There are good days ahead for us, folks! :D
>
> BUT.....I cannot give a guarantee of WHEN this pure DSD method will get off the ground. I am hoping it will be within the next 4 months. Lots of things are being worked out right now, but I can tell you it is all positive. The DSD world is really going to be opened up, I KNOW that. It's just a matter of when, not if. Please be patient. And I would also like to request....please do not ask me for any more details about this pure DSD stuff, it is rather touchy stuff and I just hope Sony don't read this. Thanks for understanding.
>
> So to sum up...arn: you're on holidays right now, but congratulations (and sorry for dismissing your theory): you WERE right about DSD containing sounds that are too loud for PCM to handle - the decimation process needs to be done in a way to avoid this problem when converting DSD to PCM.
>
> And the final question to ask, is: which SACDs really ARE clipped (non matter how purely you rip it)? Some may be "hot" and "close to the limit" ones like DSotM, but some ones are clipped whether you like it or not. I guess the definitive way to check, is to take an analog recording like I did with "Money". Mind you, it must be a DSD to analog path that you're recording to PCM, otherwise it might very well be clipped just like the Oppo's PCM output. In fact, I wonder for a second, if many analog rips out there are taken from the outs of SACD players whose DACs are PCM only. I shudder at the thought. DSD to PCM to analog to PCM....not something I'd want to experience...
>
> And at this point I just wonder what the state of PCM output is for many SACD players out there, in general. The DSD to PCM conversion world turns out to be a messy one, and not a clean one as I suspected. Friends, if you want to listen to your physical SACDs with good fidelity, *use a pure DSD solution*! Investigate the DACs in your receiver/player, make sure it's not converted to PCM at ANY stage of the process! And to those not able to get into SACD, or still wanting quality PCM rips that they can play in places SACD and DSD cannot go, you can now look forward to something better than what I currently provide. The current rips I'm doing will not be the definitive SACD rips in the long run.
>
> Mind you, I am NO less discouraged to continue to rip discs under the current digital PCM method. It still is opening up a lot of titles and new music for everyone in a quality that is in most cases shoulders above what we've had access to before. I have about 150 titles coming up in the next 4-5 months that I will be proudly releasing, and I hope you enjoy them!
>
> Also.......George....how did you figure out that DSotM was 3dB over? Hats off to you!....is there a way to fiddle with a clipped recording and figure out by how many dB it is clipped? If so, that's another way to probably work out if an SACD was originally all the way to 0dB or not. One thing is for sure: some of the rock titles I've recorded recently, ARE clipped in their original form. e.g. an Incubus disc I recently did. I just took an analog recording to make sure, and yes the original shape is clipped, although not as much as my pcm rip, ofc :/.
>
> THE END.
>
> Thanks for reading my big, FAT message. :)
>
> Sent: 18 August 2009 14:08
> To: surrou...@googlegroups.com
> Subject: [SurroundSound] Re: Digital SACD Ripping Guide
>
george...@comcast.net
Yes. As you can see, that was written before you announced that the player was clipping the signal. My statement was about the SACD; not the signal. Sorry for the confusion. Should have said "Your assumption that the SACD is already clipped is incorrect."
george...@comcast.net
I'm well aware clipping does not necessarily mean the signal is past 0dB. If I gave you that impression, I'm sorry. I don't fix the clipped discs, just measure them. Soundforge has the tool to do this. Been using it for years.
My statements about lowering the gain to prevent clipping are staightforward. If I ever stated that simple attenuation of a clipped signal would restore the peaks to the clips, that would be rubbish and I agree with you. I don't believe I ever stated that though.
----- Original Message -----
From: "VF" <vf.su...@gmail.com>
To: surrou...@googlegroups.com
Sent: Tuesday, August 18, 2009 3:08:06 AM GMT -07:00 US/Canada Mountain
Subject: [SurroundSound] Re: Digital SACD Ripping Guide
From: surrou...@googlegroups.com [mailto:surrou...@googlegroups.com]
On Behalf Of grill
Sent: 18 August 2009 18:33
To: SurroundSound
Subject: [SurroundSound] Re: Digital SACD Ripping Guide
VF, you've made me very, very excited. I wish you good health, lots of
energy and time to keep going on.
> From: surrou...@googlegroups.com
[mailto:surrou...@googlegroups.com] On Behalf Of
georgeshan...@comcast.net
> Sent: 18 August 2009 14:08
> To: surrou...@googlegroups.com
> Subject: [SurroundSound] Re: Digital SACD Ripping Guide
>
> I was thinking about this also. Do you think an album like Sea Change or
Tommy might be helpful. If the DVDA was not clipped, could we assume the
SACD was not clipped?
>
>
>
> ----- Original Message -----
> From: "grill" <gr...@index.hu>
> To: "SurroundSound" <surrou...@googlegroups.com>
> Sent: Monday, August 17, 2009 2:11:47 PM GMT -07:00 US/Canada Mountain
> Subject: [SurroundSound] Re: Digital SACD Ripping Guide
>
> I think VF's ripping method can be partly checked with a non-clipped
> DVDA. Of course, DSD->LPCM conversion is another story.
>
> On aug. 17, 21:47, georgeshan...@comcast.net wrote:
> > A known non-clipped SACD. That's a great idea. Calibration is needed.
How could one be determined?
>
> > ----- Original Message -----
> > From: "Lokkerman" <phil.steep...@gmail.com>
> > To: surrou...@googlegroups.com
> > Sent: Monday, August 17, 2009 12:38:15 PM GMT -07:00 US/Canada Mountain
> > Subject: [SurroundSound] Re: Digital SACD Ripping Guide
>
> > Can I just raise a point; are we sure that the levels are not set in the
DAC, when it processes the SACD, and does the conversion to the PCM output?
It would appear to me that a calibrated output from a calibrated source is
what we require, afterall this is what would be done with an analogue
set-up.
> > Surely if this is done; i,e, that we know what is coming off the disc is
matching what is being output to the bitstore (HDD) then we know that some
other artefact is at play.
>
> > On Mon, Aug 17, 2009 at 7:31 PM, < georgeshan...@comcast.net > wrote:
>
> > Your argument is based on the assumption that the SACD is clipped. My
argument is based on the assumption that it is not.
>
> > Lowering the signal by 12dB is a ridiculously huge amount. We need an
>
> ...
>
> tovább »- Idézett szöveg elrejtése -
VF
Amazing. Mission accomplished.
I can retire now.
:P
Lokkerman
Is the brickwall filtering set by the recording engineer when mastering DSD? If so could the placement of the brickwall on the SACD have an effect on the clipping and how this is processed by the DSP? - especially when output via DSD or PCM.
It is rather eveident to me that some of the sound properties are due to noise shaping and listening to the two versions of DSOTM (Parsons/Guthrie) it is apparent that some form of noise shaping is taking place on the latter, that clearly is not there on the former.
As for clipping the issue may be that because a DSP (see below) is used so high up in the food chain, straight from the SACD/DVD-A we may not get to reduce/remove it without modifying the action of the DSP.
Can some one please point out eactly where the clipping appears on money as I may have stumbled on another design quirk that appears to be on my Denon.
Back to the Oppo - this uses a Mediatek MT1389 DVD front end chip - looking at the data sheet - before the audio signal is output - be it DSD or PCM, it is subjected to and Audio DSP and SACD processor, hence my comments about this previously, that the setting of the levels is done prior to the output. This to me means that even the DSD output has some change and thus proving that outboard DACs can only be effective when the signal is digitally unmodified.
Now there are many groups that are based around the construction of the firmware and I suggest that a modification in this area to remove this anomaly would in fact the next course of action, the site I found that documents the firmware of the ARM processor were most concerned with video and little work had been done on Audio .
I also suggest that a good example of a disc v. disc comparison could be Brothers In Arms as this was issued as both DVD-A and SACD. We may then get to see whether there are some serious artifacts introduced by the processing with SACD.
VF
contacted Oppo a year ago who they responded saying it was in Mediatek's
hands and that mediatek would get back to Oppo. Nothing has happened since
then. We ought to contact Oppo and bring up the issue once more. TBH I'm
tired and discouraged and can't even be bothered wording up an email to
them. If someone else can email, (multiple people would be great...people
power :P) that'd be great.
I'm already moving on to pure DSD and would never listen to my SACDs in PCM
anyway.
So... maybe it can be fixed in firmware but Oppo haven't come up with a
solution after more than 18 months of first being notified. I know they've
got good technicians, great support and if any company would fix this, it
would be Oppo. Maybe we should drive up more attention on avsforum etc?
Imagine the uproar....maybe.
-----Original Message-----
From: surrou...@googlegroups.com [mailto:surrou...@googlegroups.com]
On Behalf Of jolson
Sent: 18 August 2009 22:26
To: SurroundSound
Subject: [SurroundSound] Re: Digital SACD Ripping Guide
Lokkerman
Lokkerman
What I have noticed is that my Denon appears to be soft clipping with it's proprietary "AL24 processing +", so that I cannot truly hand on heart state that on the Denon I can hear it, so next step it's the PC.
What I am getting to is a conclusion that the only way to playback DSD correctly, without ultra high end gear, which is caused by the chipset used, is to copy the SACD as DSD file and playback the DSD file on a PC. THis I'm afraid is way off yet.... or is it??
grill
a good tool for declipping.
george...@comcast.net
Mine too. We got to the bottom of the issue. I think this was a really good discussion. It was complicated discussing these concepts over email, but the subject matter was top notch.
grill
The CD layer of DSOTM SACD is originally clipped. I checked "Money"
and found a few clips at 0 dB, of which one spreads through even 14
samples of clip lenght (right ch: 05:00.614).
grill
revealed with Sound Forge :-)
>
> In wanting to get to the bottom of this issue, I have looked deep into the complicated and obscure world of DSD to PCM conversion, and I have made a shocking discovery. THE OPPO (yes, the "wonderful" Oppo) IS CLIPPING ALL SACDS BY 3dB!
>
> Read this!:http://www.diyaudio.com/forums/showthread.php?postid=1441829#post1441829. Read that thread from the bottom of that page to the end of the thread. I am disgusted.
>
> I have made tests by recording the analog output signal from my AV receiver connected to the pure DSD signal output from the Oppo, and it revealed the original shape of the DSotM SACD :(.
>
> Just to make sure it wasn't the S/PDIF output board that was clipping but in fact the Oppo player itself (grr), I recorded the same analog output signal from my AV receiver but this time with the Oppo feeding it its 24/88.2 PCM output (this is through its HDMI output, not the mod board), and it came out clipped like the straight digital rip through the mod board. I level matched the recordings and calculated that is approximately 3 dB worth of clipping.
>
> Note that it's only the SACDs that are loud enough in the first place that end up clipped! *Some* SACDs, like the wonderful Steve Reich, are still well within the range and have loads of space to breathe in. So for nice cool discs, the Oppo might ramp it up by 3dB, but it's still nowhere near the 0dB mark, and it's not suffocated in any way. But whatever's a hot SACD disc and close to the limit, the Oppo will clip!
>
> As for accuracy of original gain levels for the disc, I wonder how analog rippers try to work out what the original levels of the disc was. I gues they can't. In analog, it's all just relative and they probably record classical titles at a much higher level than the original disc - with the Oppo digital method, you can at least figure out what the level of the disc was mastered at. Of course, it doesn't matter how close it is to 0db, as long as it doesn't touch it, but I'm jut sayin...
>
> The noticable track in DSotM where the Oppo's clipping occurs, is "money" - the barbaric saxophone solo near the end. I can thankfully say though, that not much of the 40 minutes of DSotM ends up being clipped. But I, more than anyone (well whatever), am NOT satisfied with an imperfection like this so I am pretty cut about this discovery, to say the least.
>
> But really, does this make my whole investment of this ripping rig and my mammoth ripping efforts worthless? Is this whole gig a deception that's been feeding a placebo of superiority to analog rips? Of course not! The audible difference I have heard between my Oppo PCM rips and other analog rips has ranged from "definite improvement" to "huge difference" - sometimes, as much as the difference between say an MFSL master and a bad master. And others have reported the same results too. Use your own ears whether you think my 3dB clipped digital recording of DSotM sounds better than the russian analog recording or not.
>
> BUT it is a shame. I'd also like to ask everyone (because I really don't know): how bad is 3dB worth of clipping? How much of an *audible* degradation does it bring? I don't have time to take listening tests to try and detect a difference, but I'd like to know how bad this clipping is, in the real scheme of things. Right now I honestly don't know how significant this problem is.
>
> But no matter how *IN*significant it is, I can't bear the thought of making clipped recordings, and I will *just not have it* in the long run. And guess what, fellas. I have an announcement to make. There's hope! I didn't want to tell the group this until much later, but in light of this setback I'd like to make an announcement. I am developing a method to rip the *pure DSD signal* from SACD, forgetting about PCM conversion forever. I am working on a number of solutions for pure DSD playback of these DSD rips, including a DSD codec for the PC/HTPC and SACD-R (that's right, an SACD ISO file that you burn to DVD....it does exist, and you would play it back in your SACD player).
>
> On top of that, for those who will still need to experience these rips in the PCM realm, I will release software which will allow us to convert my pure DSD files to 24-bit 192kHz. The result, when you compare it with the Oppo's DSD to PCM conversion that I currently have to deal with (grr), will be much superior, and at a higher resolution as well. And this is 24-bit 192kHz *in multichannel* too - good timing for the growth of blu-ray and its superior capacity to DVD-A, perhaps :P. So how does this all sound? I hope it serves as a consolation. There are good days ahead for us, folks! :D
>
> BUT.....I cannot give a guarantee of WHEN this pure DSD method will get off the ground. I am hoping it will be within the next 4 months. Lots of things are being worked out right now, but I can tell you it is all positive. The DSD world is really going to be opened up, I KNOW that. It's just a matter of when, not if. Please be patient. And I would also like to request....please do not ask me for any more details about this pure DSD stuff, it is rather touchy stuff and I just hope Sony don't read this. Thanks for understanding.
>
> So to sum up...arn: you're on holidays right now, but congratulations (and sorry for dismissing your theory): you WERE right about DSD containing sounds that are too loud for PCM to handle - the decimation process needs to be done in a way to avoid this problem when converting DSD to PCM.
>
> And the final question to ask, is: which SACDs really ARE clipped (non matter how purely you rip it)? Some may be "hot" and "close to the limit" ones like DSotM, but some ones are clipped whether you like it or not. I guess the definitive way to check, is to take an analog recording like I did with "Money". Mind you, it must be a DSD to analog path that you're recording to PCM, otherwise it might very well be clipped just like the Oppo's PCM output. In fact, I wonder for a second, if many analog rips out there are taken from the outs of SACD players whose DACs are PCM only. I shudder at the thought. DSD to PCM to analog to PCM....not something I'd want to experience...
>
> And at this point I just wonder what the state of PCM output is for many SACD players out there, in general. The DSD to PCM conversion world turns out to be a messy one, and not a clean one as I suspected. Friends, if you want to listen to your physical SACDs with good fidelity, *use a pure DSD solution*! Investigate the DACs in your receiver/player, make sure it's not converted to PCM at ANY stage of the process! And to those not able to get into SACD, or still wanting quality PCM rips that they can play in places SACD and DSD cannot go, you can now look forward to something better than what I currently provide. The current rips I'm doing will not be the definitive SACD rips in the long run.
>
> Mind you, I am NO less discouraged to continue to rip discs under the current digital PCM method. It still is opening up a lot of titles and new music for everyone in a quality that is in most cases shoulders above what we've had access to before. I have about 150 titles coming up in the next 4-5 months that I will be proudly releasing, and I hope you enjoy them!
>
> Also.......George....how did you figure out that DSotM was 3dB over? Hats off to you!....is there a way to fiddle with a clipped recording and figure out by how many dB it is clipped? If so, that's another way to probably work out if an SACD was originally all the way to 0dB or not. One thing is for sure: some of the rock titles I've recorded recently, ARE clipped in their original form. e.g. an Incubus disc I recently did. I just took an analog recording to make sure, and yes the original shape is clipped, although not as much as my pcm rip, ofc :/.
>
> THE END.
>
> Thanks for reading my big, FAT message. :)
>
> Sent: 18 August 2009 14:08
> To: surrou...@googlegroups.com
> Subject: [SurroundSound] Re: Digital SACD Ripping Guide
>
> I was thinking about this also. Do you think an album like Sea Change or Tommy might be helpful. If the DVDA was not clipped, could we assume the SACD was not clipped?
>
>
>
> ----- Original Message -----
> From: "grill" <gr...@index.hu>
> To: "SurroundSound" <surrou...@googlegroups.com>
> Sent: Monday, August 17, 2009 2:11:47 PM GMT -07:00 US/Canada Mountain
> Subject: [SurroundSound] Re: Digital SACD Ripping Guide
>
> I think VF's ripping method can be partly checked with a non-clipped
> DVDA. Of course, DSD->LPCM conversion is another story.
>
> On aug. 17, 21:47, georgeshan...@comcast.net wrote:
> > A known non-clipped SACD. That's a great idea. Calibration is needed. How could one be determined?
>
> > ----- Original Message -----
> > From: "Lokkerman" <phil.steep...@gmail.com>
> > To: surrou...@googlegroups.com
> > Sent: Monday, August 17, 2009 12:38:15 PM GMT -07:00 US/Canada Mountain
> > Subject: [SurroundSound] Re: Digital SACD Ripping Guide
>
> > Can I just raise a point; are we sure that the levels are not set in the DAC, when it processes the SACD, and does the conversion to the PCM output? It would appear to me that a calibrated output from a calibrated source is what we require, afterall this is what would be done with an analogue set-up.
> > Surely if this is done; i,e, that we know what is coming off the disc is matching what is being output to the bitstore (HDD) then we know that some other artefact is at play.
>
> > On Mon, Aug 17, 2009 at 7:31 PM, < georgeshan...@comcast.net > wrote:
>
> > Your argument is based on the assumption that the SACD is clipped. My argument is based on the assumption that it is not.
>
> > Lowering the signal by 12dB is a ridiculously huge amount. We need an
>
>
> tovább »- Idézett szöveg elrejtése -
ArnoldLayne
forward, me thinks. Some, like Steve Reich and Toto, come out perfect.
Others maybe not, but it still gives us access to a fairly big out-of-
print surround catalog. I have an Oppo DV980H myself and I agree with
that SACD doesn't sound perfect on it. But good enough for making me
prefer it to CD anytime. I can live with small amounts of distorsion
on certain albums for now, even though some day I'll finish upgrading
my system to once again handle pure DSD. And I have the same view on
the DSD to PCM transfers. It's something to enjoy for now, until we
can access DSD streams. One day...
>
> leer más »- Ocultar texto de la cita -
>
> - Mostrar texto de la cita -
grill
Head Hunters digital SACD rip has only very few clips with few samples
in the following tracks and channels:
5.1:
2Lf/Rf, 3Lf/Rf (-0.1 dB)
4Lf (-0.6 dB)
4Rf (-0.4 dB)
2.0:
2R, 3L/R (0 dB)
Another (analogue) SACD rip of HHHH from hdtracker (ru) contains
clipping, too.
Lokkerman
Let's go back to what Arnoldlayne said - it's a lot better than CD, - my comments - we're just fussy nowadays perhaps :)
Note that DXD is PCM based coding -:)
me - I've always been analogue but then I couldn't store and play all my collection in seconds - swings and roundabouts I suppose and today's digital is close to reality enough - so i am not whingeing just being grateful that it beats the hell out of my old tape- recorder that my mates christened, the Bolan wig making machine - this recorder created more chewed hairdo looking tape piles than good copies of music....
VF
SACD is most definitely an originally clipped one. The Oppo's 3dB of
clipping didn't actually do any damage with this one, because there's still
2dB of headway above the waveform (meaning 5dB of space on the original
SACD). BUT the whole waveform shape is still clipped!
So basically, it's an originally clipped SACD, but it's not even "hot".
What's the point in clipping unless you want to make it louder? The stupid
mastering engineer/producer not only didn't achieve the advantage of
clipping (if you can call loudness an "advantage"), but they also damaged
the recording by taking out audio information with the dynamic compression
at the same time.
Oh well. At least the other Al Di Meola disc I did the other day, Flesh on
Flesh, came out ok.....IIRC :P. And most discs I've been transferring in the
last few days have been non clipped ones, so it's not all bad.....take
heart! (maybe I'm saying this more to myself than anyone else).
BTW, no one mentioned how bad 3dB of clipping actually is, *audibly*. If no
one really knows anyway (or can tell....I don’t have time to make my own ear
tests), maybe I shouldn't be so down about my current clipped releases. ;)
:) Basically, I won’t believe a claim by someone unless I can hear what
they're claiming with my ears. And then I use common sense and weigh up
whatever other factors there are in the picture to make a decision. So in
THIS case, I won't worry until someone tells me they sensed compression on
one of my releases when comparing it to the original SACD :). (and I'd cross
check to make sure it is indeed clipped, otherwise I can tell them it's a
"PCM vs the original DSD" issue :) ).
So basically....why get hung up on technical issues (a least in this case,
where I cannot fix it) when you (possibly) cannot hear the difference
anyway. What I CAN hear the difference between, is a digital PCM transfer
and an analog one, and I'm sure the difference is helped by all the
collective intricate steps I take to optimise the digital process. So ofc,
I'll technically strive for perfection, in the name of Audiophila.
VF
Lokk: that “broadband hash” you speak of is the infamous DSD “noise”. Read up about it. It’s part and parcel of the DSD64 format, for all that it offers.
As for the clipping with MJ thriller, I wouldn’t be surprised. I haven’t received that disc to transfer yet so I cannot see for myself if the original disc is clipped.
And grill….the herbie disc may be clipped for all I know. But if you find clipping in others of Macromaggot’s releases (he’s a fine analog ripper), let me know. If so, he is recording a DSD to clipped PCM to analog signal. To PCM. Not the greatest.
@lokk again….The biggest problem with the SACD format is not DSD itself (DSD is its calling card), but the fact that the majority of discs are recorded, mixed and mastered in PCM. All these discs should be DVD-Audio’s, but due to everything that happened, half the industry went to SACD and so we have many PCM SACDs :/. Even so, the principle is that every time you convert it from one format to another, you lose something in the process, so my goal to share the original “pristine” DSD (even if it not so pristine, really), is still a valid one – it’s nothing more than the pursuit of accessing the closest thing to the original source as possible.
So lokk, what you’re discovering day by day, is that DSD to PCM conversion is a mess. Pure DSD on its own, is quite a nice thing and I love it very much. But I was just doing my own analog recording tests today (at 192kHz) from the oppo’s analog out, my amp’s out etc….and both these instances of PCM conversion seemed to have advantages and disadvantages :S (this is from looking at their spectrograms, at least).
Yes you’re right…DXD is based on PCM. Technically, it is closer to PCM than DSD. But although it starts out as a PCM signal, it then undertakes major changes in its encoding settings and hardware used to encode/decode it, namely taking on the impulse and frequency response of DSD, such that the end result is the sum of both PCM and DSD’s (sonic) advantages.
Don’t be slow to complain about digital audio – we’re Audiophiles! DXD, by the looks of it, is something that finally advances to challenge analog such that even Michael Fremer (world’s no. 1 vinyl freak) might accept it as flawless :). The only question is, when will we have DXD equipment and recordings?….might be 10+ years away.
But I know what you’re saying…we’re here primarily to enjoy the music, that’s the important thing!
VF
“What I am getting to is a conclusion that the only way to playback DSD correctly, without ultra high end gear, which is caused by the chipset used, is to copy the SACD as DSD file and playback the DSD file on a PC. THis I'm afraid is way off yet.... or is it??”
No, for now, you can achieve this with an SACD player using a *pure DSD solution*. Like the Oppo’s pure DSD output. Which is marvellous and I enjoy it with my HDMI 1.2 pure DSD DAC receiver. We’re just dealing with DSD to PCM issues here. There should be NO processing for SACDs EVER done, it means it will be converted to PCM or something else horrible will be applied to the audio.
I found this info on sa-cd.net and it sounds like it’s a list of good pure DSD DAC AV receiver models (arn, you know any info on this?):
“# HDMI from version 1.2 up supports DSD but although several receivers are compatible with this HDMI version, so far few of them decode DSD. Models that do include Sony’s STR-DA5300ES (called TA-DA5300ES in Japan), Marantz SR60001, SR7001, SR8001, SR7002 and SR8002, Yamaha’s RX-V661, RX-V861, RX-V1700, RX-V2700, RX-V1800, RX-V3800 and RX-Z11 (DSP-AX661, DSP-AX761, DSP-AX861, DSP-AX1700, DSP-AX2700, DSP-AX1800, DSP-AX3800 and DSP-Z11 in some regions), Denon’s AVR-3808 and AVR-4308 and Onkyo’s TX-SR805, TX-SR875 and TX-NR905 receivers.”
Maybe it’s a list of players that simply officially accept DSD over HDMI 1.2, but they still might convert it to PCM before analog conversion :/
From: surrou...@googlegroups.com
[mailto:surrou...@googlegroups.com] On Behalf Of Lokkerman
Sent: 19 August 2009 00:14
To: surrou...@googlegroups.com
Subject: [SurroundSound] Re: Digital SACD Ripping Guide
Sorry got the mails out of sync
on the last emails but suffice to say the issue could affect the DSD output
because of where the DSP/SACD processor is placed, and this may be intrinsic to
all DVD-A/SACD players, it could be that on the Oppo BDP-83 that some elements
(of these devices) may be switched out - this needs to be verified.
grill
Here's one of the few clips found in macroMaggot's Herbie rip (track 2
- Lf):
http://yfrog.com/0chhhhtrack2lfclipj
Concerning DSOTM SACD I'm not absolutely sure that the stereo track
"Money" is MUCH heavier clpped in your digital rip as compared at all
dBs to that of the original CD layer. Someone having the tools needed
for clip statistics might support or reject this hypothesis.
>
> tovább »
>
> image001.gif
> < 1kBMegjelenítésLetöltés- Idézett szöveg elrejtése -
Lokkerman
Me, and I think you do now, have gripes about the need to convert from PCM to DSD and vice-versa and what annoys is the fact that there is always a commercial war over each medium; - we are picking up the pieces and you are leading the way - just WELL DONE on this:)
I also agree on your sentiments about +3dB - just a bit of VU bashing with maybe mild overtones of THD but NOT the harsh clipping that we abore, these are very short transients (most of these clips are so brief) so most people will hear them as edginess, and I think this was done by the sound engineers perhaps with this in mind.
Soz about Thriller - it's stereo so missed off the list, besides now on the next list....
Finally last point and most serious, - although the SACD players state that they output pure DSD the fact that there is within the disc controller (DVD/BD) chip an SACD and DSP controller function before the DSD stream ( and for the matter PCM), is outputted. Therefore the analogy is like listening to a record deck with the RIAA in circuit in situ before the pre-amp. Acceptable so long as the circuit is 100% accurate and that there are matched gains and no imparts. Hence my point about high end as they use custom logic to ensure that the output is untainted.....
GREG CARRIGAN
I am by no means a DSD expert, but in the reading I did on the subject back in the day, I understood that the DSD process produced high frequency clipping above the 100 kHz range (so high that no human could hear it - but perhaps dogs could). Wouldn't the spectographs you're all talking about ( of the DSD output) be showing this? You're right that it is a hot signal, but isn't the frequncy of the signal so high that you cannot actually hear it?
I ask this since even my Denon 5910 player has a SACD/DSD frquency filter setting of 60 KHz (still way higher than the human ear's ability to hear it) that you can switch on/off. The manual; tells me that it's important to have the filter set 'on' since the high frequency (clipping I assume) could damage my tweaters.
Is this not the case?
G
From: Lokkerman <phil.s...@gmail.com>
To: surrou...@googlegroups.com
Sent: Friday, August 21, 2009 7:11:41 PM
robert chewter
Because my view on music sharing is that it should be public, not private, I have decided to share, in full detail, how I rip SACDs digitally - with the intention that anyone can follow in my footsteps and learn how to make a Digital SACD Rip without having to go through the pain and learning/perfecting process that I had to.
I've finally gotten round to making the guide, so here it is. Jolson has expressed serious intentions to join me in the new effort, so here you are, too, my friend! Can't wait for you to join me!
If anyone has ideas to improve or would like to critique the process - I'd love to hear it. After all, our audiophile pursuit is about perfection in sound - for enjoyment in music.
I've attached a html file of the guide - in case you can't see the formatting that I've done for the guide. (makes it way easier to read). I very well might make some little edits to this guide over time, so I'll host it at some webpage eventually.
- - -
Digital SACD Ripping Guide
Introduction by HiResOrNothing (what a great guy, eh? :P):
"In the past we have only been able to rip SACDs by recording the multichannel analog outs of an SACD player and into our analog sound cards. Many people have used horrible Creative Audigy cards, and quality has been disappointing at best. Other rips that we have seen have been nicer, using the best that analog sound cards can offer from the likes of M-audio, Lynx, etc. However, quality has still suffered due to the multitude of problems that arise with analog capturing in general. Unless the SACD player used for the transfer is a very expensive one, the analog transfer quality will leave a lot to be desired.
"But we now can celebrate in a new "standard" for SACD ripping. We can now capture a pure 24-bit PCM digital signal which is converted directly from the pure DSD stream from the disc - and this is achieved by using a special "modifed" SACD player. These modified players been around ever since the SACD format has, but they have always been extremely expensive (like, try a $2000 Denon player which you have to order from Switzerland). But things have progressed and a few companies are offering a modified version of the famous Oppo DV-980H player, which is well known for its affordability yet uncompromised quality - especially with regards to digital SACD output!
"I have purchased such a player to make this SACD album available to you in the highest quality currently possible. The Oppo offers pure DSD output over HDMI 1.2 but it also gives the option for PCM output. The Oppo internally converts the DSD to PCM at 24-bit/88.2kHz and the "mod" captures this PCM signal and outputs it through three stereo S/PDIF (coaxial) jacks. To capture this glorious digital signal, I am using three M-Audio Delta 1010LT sound cards simultaneously. They each have a single stereo coaxial input, so in order to record all six channels simultanesously, three cards must be used. I'm using Adobe Audition 1.5 in multitrack mode to simultaneously record the 6 channels in real time, then I map the channels with Audition's "multichannel encoder", and export the individual mono wav channels. After that I split the album into the tracks and encode it to MLP with surcode, and author it to DVD-A with Discwelder Chrome. No editing of the audio signal is made in Adobe Audition, this is a straight digital capture of the signal and as close as you can get to copying an SACD disc-to-disc at the present time."
- - -
The following is a detailed guide from HiResOrNothing, of how to digitally rip SACD with the above method and equipment. Other equipment, software, or variations in the editing/cutting/encoding process may be used, and get the same results quality-wise. But this guide is for exactly how I do it.
Setting up the Modified Oppo 980H
- It all starts with buying a modified Oppo player. You can get them from www.switch-box.com.
- First, it is important to understand that while we are ripping a digital signal, this is still real-time recording and the signal we are capturing is indeed a processed signal. We are still not copying the information bit for bit, but instead we are ripping a signal which has been fed through the Oppo player, processed by its internal components, including the chip which converts the DSD signal to a PCM one.
- Before you start recording, get to know the Oppo player very well, read the manual and learn how to operate it and play back SACD discs before you continue.
- These are the settings you need to set in the Oppo player, to achieve the best processing and output quality of the PCM signal. Assuming the rest of the settings are from Factory Default, do the following:
General Setup Page
- Resolution - set to "1080p"
Speaker Setup Page
- Downmix - set to "5.1 CH"
- Front, Center, Surround Speaker - set to "Large", and set Subwoofer to "On"
- Channel Delay - set to "0ft"
- Make sure Channel Trim and Audio Delay are set to "0dB" and "00".
Audio Setup Page
- LPCM Rate - set to "192K"
- Pro Logic 11 - turn off
- HDMI Audio - set to "Off"
- SACD over HDMI - set to "PCM"
- Make sure EQ Type, Sound Field, Audio tone are set to "None", "Off" and "00".
Digital Volume Control
- On your Oppo remote you will see a "+" and a "-" button to control the volume of the Oppo. Make sure you keep this digital volume at 100% (level 20). Lowering the output volume will actually lower the output signal quality. The way digital volume control works, is that it reduces the volume by taking out bits from the audio signal - it actually removes audio information, and thus lowers the quality of the output signal! Volume control should ALWAYS be done in the analog domain (your amplifier), not the digital domain. So leave it at full volume, where no quality loss will occur.
- If you find a disc has dynamic clipping (with the amplitude touching the 0dB mark), it is not because the Oppo output signal is too "hot" and should be turned down, it is because the (stupid) mastering engineer/producer applied dynamic compression to the original disc. Lowering the volume of the Oppo will only output a softer, but still clipped signal, and reduce the quality at the same time. All digital formats (and even vinyl!) have the ability to suffer from the loudness war :/.
You will also need three quality coaxial S/PDIF cords to connect the Oppo to the Delta Cards.
Installing/configuring the Delta Cards
- You will need a motherboard with three PCI slots for the three Delta cards. You will also need relatively high performance components (fast CPU, RAM, SATA drive preferred) to ensure that no system lag occurs and gets in the way of accurate and error-free capturing of the high bandwidth 6 channel digital signal - and using three PCI cards at once with such intensity.
- Have plenty of HDD space, have a fresh system install, and only install programs that you will absolutely need for the system.
- The OS that you must use when using the Delta cards is Windows XP SP2. There is actually a bug whereby multiple M-audio cards sometimes have trouble syncing together and pops/clicks in the audio can even occur. So do not use Vista, do not use Windows 7, do not use XP SP3, use XP SP2. And you must use the 5.10.00.5057v3 version of the driver, as well! Also, READ THE USER MANUAL before you even touch the cards with your bare hands - learn a bit about what the cards can do, and how they generally work, and what you should and shouldn't do with the cards - e.g. if you attach the breakout cables to the cards while the computer is powered, you could damage the cards and void your warranty.
- Once you're ready, install the three cards on the system. Once installed, it's time to configure them in the Delta Control Panel. You will see the three cards in the H/w installed area on the right, which you simply select to individually configure each one.
In the Control Panel go to the "Hardware Settings" tab.
- Set the Master Clock to S/PDIF In.
- Set the S/PDIF Sample Rate to 88200.
- Set the MultiTrack Driver Devices to Single and In-Sync.
- Make sure these settings are configured for all three Delta Cards.
As for setting up playback with the M-Audios, consult the user manual and work it out yerself ;).
Setting up Adobe Audition
- I use Adobe Audition 1.5 to do the recording. Later versions are obviously fine to use (as are many other programs), but the settings might be different to what I outline in this guide.
- Go to Options -> Device Order. Click on Recording devices tab. Remove any devices previously selected for use in Audition, and click on "Use >>" to allow the three S/PDIF input devices. They will appear as:
M-Audio Delta 1010LT S/PDIF
M-Audio Delta 1010LT S/PDIF (2)
M-Audio Delta 1010LT S/PDIF (3)
- Click OK.
- Now we need to set the default recording settings to the right resolution. Go to Edit View, and click on the record button. Select 88200 as the sample rate, stereo as the channels, and 32-bit (float) as the Resolution. Click OK to start recording, then stop the recording and restart Adobe Audition.
- Now go into Multitrack mode. Here we will do the recording, using three stereo "tracks" for the three S/PDIF cards - totalling to 6 channel. Link FL+FR to M-Audio Delta 1010LT S/PDIF, C+LFE to M-Audio Delta 1010LT S/PDIF (2), and Rs+Ls to M-Audio Delta 1010LT S/PDIF (3). It might take some trial and error to match them up, but it's better if you can make it logical. In most multichannel mixes, the fronts are the loudest, C and LFE are pretty easy to isolate and then the remaining two, usually softer than the fronts, are the backs.
- So, right click in the Track 1 area, and choose Properties. Make sure M-Audio Delta 1010LT S/PDIF is selected as the input device, and change 16-bit to 32-bit. It should be stereo by default.
- Then do the same for Track 2 and 3, as per the suggested signal mapping above.
- Lastly, arm the tracks for recording by clicking on the red "R" button for each track. Now Adobe Audition is set for recording!
Recording
- Connect the three stereo S/PDIF outputs from the modded Oppo (FL+FR, C+LFE, Rs+Ls) to the three Delta S/PDIF inputs.
- When recording, disconnect any other audio connections from the back of the Oppo player apart from the S/PDIF, to remove any possible extra unnecessary processing/interference in the Oppo. This might include analog or Toslink/Coaxial connections. HDMI can still be connected though, as we have already turned HDMI Audio off in the Audio Setup Page anyway.
- On the recording computer, do not run any other application while you record with Audition, and make sure the system (RAM especially) is plenty free for the recording process!
- Just before you record, set the Oppo player to Audio-Only Mode! This turns off the video processing and output, reducing the inteference between the video and audio signals - helping to optimise the signal output of the Oppo. The screen will turn off after 5 seconds and you will see A.ONLY illuminated in the Oppo front panel. When you stop a disc and put a new one in, it goes back to normal mode so make sure you set it to A.ONLY before every recording! And give it a few good few moments before you start recording too - let the player focus all its processing power for the audio.
- With the three Tracks in Audition Multitrack mode configured and armed, press record, then start playing back the SACD in the Oppo! Now the magic happens!
Always listen to your capture all the way through to make sure there were no disc skips/glitches! Repeat if anything went wrong. Be sure to clean the disc before recording to prevent anything happening.
Other Tips and Hints
- It's a good idea to get the length of silence at the beginning of the first track exactly correct (when you edit your album). To make a good release, you should get it right, and it might also help when cutting the album's track points later on. To get the length of silence down to the milisecond, start playback of the SACD by pausing the disc in track 1, and pressing the "Previous" button on the Oppo player itself (instead of the remote). This will create a little click in the audio signal which Adobe Audition will pick up in the recording. You can then later use this click to remember where the album starts, exactly. But note that the click doesn't occur in exactly the same place in all the channels (even in stereo). If you zoom right in you will see this. So make sure to cut the album just after the last click that occurs, wherever it is amongst the channels.
- Annoyingly, Audition seems to not remember the device settings very well. Sometimes you have to keep going back to the Device Order settings and adding the S/PDIF input devices in again. Also, remember to check EVERY setting in Audition before every time you record - make sure all three tracks are set to 32-bit, (not 16-bit!) etc. You can also see down the bottom the current setting for recording - it should say 88200 * 32-bit Mixing.
- For the process of setting everything up, or if you want to easily monitor/play back your recordings in Audition (either during or after recording), it's a good idea to connect the analog outs of the Deltas to your multichannel receiver - to listen to the channels in your surround system. Simply use the WavOut 1/2 device for each Delta card as the three stereo pairs, matching them accordingly into your receiver/amp's multichannel analog input. Then in Audition, in the same way that you set the S/PDIF inputs to be recording devices and then set them in the track properties, do the same thing for the analog playback devices and set them for each track's output.
Wrapping Up
- Once you have your glorious digital capture on your computer, you can export the tracks to mono wav channels. Do this with Audition's Multichannel Encoder at the bottom of the View menu. Map the three tracks to their pertaining channels, and select the correct export format (6 mono wav channels, 24-bit). It will then export the channels for you. For the C+LFE track, move the Sub Channel Level to "100", it is 0 by default.
- One you have the mono channel files, you can load them back in Audition (I load them as six individual tracks in MultiTrack mode), and go on and edit the tracks for your release. When you have them loaded as six individual tracks, you can then pan them and monitor the recordings with any stereo setup that you edit the tracks with. If it's just a stereo mix, you can simply save Track1 to a stereo file in Wave Edit mode and then edit it as a stereo file as is.
- If your intention is to author it to DVD-A with DiscWelder, the easiest way to cut the tracks is to set track points in Audition (right click at the tip of the cursor bar where you want to cut a track and select "Insert CD Track Master"), then write down the track points (right click at any track point and "Go to Cue List", then copy and paste those cue values somewhere and change them into Discwelder's time format), then in DiscWelder import the full-length album and then insert the track points in there. This is also the only way to do gapless albums properly in DiscWelder.
- So once you edit the beginning and end of the album and determine the track points, you're ready to encode the mono channels to MLP, FLAC or whatever you desire. There is NO need to edit the tracks in any other way, the aim of this method is to rip the disc as close as possible to the original. If there's a problem, re-transfer the disc, DO NOT edit anything in the audio.
- So there you are. If you have the courage to take on this new method of ripping SACDs, now you know how to do it!
Yours,
HiResOrNothing
Lokkerman
88.2 times by 32 = 2.8224 or put it this way 44k1 times by 64, so any convert from DSD, should for be in these multiples - I' crap at explaining so WIKI is at hand
http://en.wikipedia.org/wiki/Direct_Stream_Digital
VF
How can DSD be a great system for archiving if it can't be then translated to something good for playback? I don't get that logic...
What's "VU bashing"?
Don't worry about Thriller. Someone else was supposed to send it to me but then they got too busy and couldn't send it in time :( - so at the time I knew you had it but didn't think I needed it from you. I was really hoping to have that for this batch but I'll have to wait for next time.
Yes, I'm interested in learning more about how the DSD is *processed* in the SACD/PSP chip for all these players. THAT is the next step. I can only hope the Mediatek in the Oppo does it well. (I dont really want to publically discuss details of what I will be doing, but I will likely be modding both my Oppo 980 and my Pioneer 696 and seeing what result is given by both of them). But with my EARS I can testify to a beautiful natural DSD sound, or at least, something of the like when compared to its PCM output…which is a testament to the notion that even if it's from a PCM source, it still is better than converting it to PCM yet again...
But anyway, I'm not sure what can be tampered with the DSD signal though? Being a 1-bit system, standard DSP (EQ etc) operations cannot be performed on it, unless it's converted to PCM (as I assume you know). Afaik it can only be cut/joined and have some VERY basic things done to it. In the sadie/sonoma DSD workstations, if they have to make any "must" edits, the DSD is temporarily converted into some 8-bit hybrid DSD format, before being converted back to DSD again. The mediatek chip obviosvly wouldn't do this, so I can't see how any major tamperings can occur.
In fact, I wonder if all clipped SACDs are from clipped PCM masters. Maybe it's impossible to apply dynamic compression to a DSD signal (as it classifies as an "edit")?... Maybe there's some DSD-compatible limiting function that they can do in the SACD authoring programs...not sure yet...but I strongly doubt it. Changing the audio in such a way is soethign that would not be possible with the single bit DSD signal. If this is true, then DSD truly does have this one "advantage" in that if you make a pure DSD disc, right from recording to mastering to pressing, it's impossible to apply clipping. Problem is, they don't care, do they.....and they have to do all their mixing, anyway.....well, DXD ftw.
But anyway...In the Oppo, all the speaker settings and audio settings like EQ, channel level and channel delay, only pertain to the PCM output. Half of them are greyed out when DSD is selected, and the rest don't make a difference anyway. E.g. if you turn the subwoofer off, restart the player, play through DSD, the subwoofer is still employed. I also tested channel levels and there was no difference. And likewise, with HDMI 1.2 DSD-capable receivers, if you want bass management or any other EQ setting with your receiver (like speaker calibration) to be used with SACDs, you need to compromise and use some sort of pcm mode. Those nice tweaks don't apply to the "pure" mode. So this gives more credence to the fact that a DSD signal can't have any DSP applied to it.
But, there's still lots of mysteries with this, and I won't rule anything out of course....so, do you have a link to info talking about a DSP in the Mediatek? And I will research further into the chips and what can be done to a DSD signal.
Lokkerman
VU bashing is a reel tape term; very rarely did technicians record at 0 VU, to try and maximise the S/N ratio, also prompted by the use of low noise high output tapes, enegineers recorded in the red. It could be a long paragraph so I'll stop here....
Not sure what can be done with a DSD signal but the mechanism is in place; see the datasheet 2 form the chip used in the Oppo.
We don't know whether the DSD is in some way corrupted causing the clipping unless you can capture the pure DSD as a file prior to any playback device processing and then analyse the file, we could do with a DVD Explorer for SACDs. And then we would be able to paybacK DSD anyway so it's quite circular.
Thriller :)
Lokkerman
bobchewie
except your using using spdif to make sure its digital...does it make
it cleaner then..ok so why does oppo only do 88.2 and not 96/24
then..are mojave upsampling ( a no no in my book)..anyone know what is
coming out of the analogue ports of a samsung hd950? is it a clean
signal ie no noise..maybe thats downside to analogue transfer then..in
the past in audtion 1.5 after recording into line in..upon playback in
audtion id hear an odd sound..sounded like windrushing very odd..fact
is it werent no music thats for sure..anyone else had
this..also ;correcting for drift' is that a ggod thng or bad thng as
getting channlels in sync (yes i use 'remove silence' to line all 3
stereos up)...
>
> read more »
VF
Well yes, you’re right about the *possibility* of the DSD itself being clipped by the Oppo…until it is disproved, I won’t rule it out. But a lot of evidence has pointed to the theory that the PCM conversion is where the clipping occurs.
Yes, parsing the disc itself IS the ultimate way to compare the original DSD file to one processed by the Oppo, and ultimately, extracting the original 2CH/6CH files (SACD’s equivalent of AOB files) is the best quality version (exact digital copy) you could get anyway. But until we have SACD Explorer 2010 ( ;) ) we have to use other means to find out if we’re getting a non-clipped DSD signal.
And anyway…we won’t need that to find out. We will be able to compare an extracted DSD signal and convert it to PCM with PC software, and compare THAT to an extracted PCM from the Oppo version. Like “Money” from DSOTM. If the converted PCM file from the pure DSD extraction is not clipped, then we know that the Oppo does the clipping at the PCM conversion stage, not DSD.
Furthermore, we can ask Oppo technicians, gather info (like we’re doing now), ask higher experts and make a conclusion right now.
Ok so I’ve looked at the (excellent) DataSheet, but where does it indicate that the DSD signal is DSP’ed? And likewise, the second PDF you provided…..not sure how it relates to DSD (doesn’t even mention the word DSD or PCM….it’s likely only talking about the PCM realm).
I have asked DSD experts on gearslutz whether you can add dynamic compression or edit DSD in any way, and this was their response: http://www.gearslutz.com/board/mastering-forum/416441-1-bit-dsd-dsp-operations.html
I am sure the Oppo does little more than processing the PSP, reading the signal, unpacking the DST to DSD data, and passing the resulting un-editable 1-bit signal on to the rest of the player.
VF
Yes lokk, I cannot see how the DSD signal can be level changed in ANY way.
See this thread: http://forum.merging.com/viewtopic.php?f=4&t=3389&start=0
I’ve now personally concluded that the pure DSD output of the Oppo is indeed pure –and unmolested by ANY DSP process. Due to the very nature of DSD and what you can’t do to it. I think I’ve scientifically come to the right conclusion. Let’s see what you think.
Lokkerman
I can't disagree with that and it does seem a logical conclusion. BTW the reference in the data sheet I was referring to was the block diagram of the internals which shows the DSP and the SACD procesor sytem on the system bus.
For the sake of completeness I will still try and find out what the actions of these two buddies are - it is not beyoond science to decimate and then reconvert although methinks it is not happening here:)
VF
I appreciate your research.
Yes I think that would be a bit riotous to convert to PCM within the DSD processing chip itself:). The PCM output is not so outrageous when you think the main drawing card of the Oppo (in terms of SACD playback) is its pure DSD output anyway. It’s the only reason (apart from the mod and this whole project) I upgraded from my Pioneer (with crappy SACD output) to the Oppo after all.
I’m certain it doesn’t happen anyway – if I (and others like elshagon who have carefully listened) hear a certain difference between the original SACD (pure DSD) and the PCM, the DSD must be unprocessed and pure.
But yes let me know what you find in your research. Feel free to privately email me it if you want. I will find it useful anyway :).