Quadraphonic software decoders

[Andrej Falout]

Hello Quad enthusiasts,

I am wandering if anyone is aware of an Quad matrix decoder, implemented in software, that uses Variable Gain Matrix / steering logic processing (as used in Tate processors & similar) ?

I know of several basic software decoders, and even though they are better then most consumer hardware decoders of yesteryear, they still have quite poor channel separation. Plus, they all seem to process only 16 bit.

EG. what is the best software matrix decoder?

Probably reasonable to assume nobody wrote a CD-4 software decoder (?)

Thanks,

Andrej

[cepl...@aol.com]

Andrej,

  Am not sure what is out there, but keep in mind a lot of what you are wanting to listen to has allready been decoded and uploaded to the web in one form or another.  Make sure you aren't making extra work for yourself. 

- CP

--
--
You received this message because you are subscribed to the Google Groups "SurroundSound" group.
To post to this group, send email to Surrou...@googlegroups.com
To unsubscribe from this group, send email to SurroundSoun...@googlegroups.com
For more options, visit this group at http://groups.google.com/group/SurroundSound

---
You received this message because you are subscribed to the Google Groups "SurroundSound" group.
To unsubscribe from this group and stop receiving emails from it, send an email to surroundsoun...@googlegroups.com.
For more options, visit https://groups.google.com/d/optout.

[Andrej Falout]

<surrou...@googlegroups.com> wrote:

Andrej,

  Am not sure what is out there, but keep in mind a lot of what you are wanting to listen to has allready been decoded and uploaded to the web in one form or another.  Make sure you aren't making extra work for yourself. 

- CP

Very true. Problem is that "one form of another" part:

1. A lot of it was compressed using lossy methods (DTS)
2. Almost all of it was AD'd 16 bit (even if it was later converted into DVD-A)
3. ADs where mostly of amateur variety computer sound cards and the lik, even the 24 bit ones.
4. And most of all, huge majority is using sub-standard turntables, pick-ups and decoders, that where sub-standard the day they departed the factory, not to mention today.

After listening to Mike Oldfield's "Boxed" SQ set, in good condition, on good turntable with good preamp, going through 32bit AD through Dolby Pro Logic IIz (yes I know that part is not ideal, but even with limited compatibility it does a surprisingly good job) - I am absolutely sure that most available Quadraphonic digital conversion should sound a lot better then they do.

Obviously, the sound chain is only as good as it's weakest link, and mine is a Quad decoder. eBay made sure that every item on the planet finds a person with too much money in there pockets - anywhere on the planet. So hardware is not an option (for my hobby).

Therefore my question.

OTOH, I may invert this with a question:

"What is in your opinion, the best quality, publicly available (hint hint...) digital conversion of an Quad record"?

Thanks,

Andrej

[Lokkerman]

Before we re-invent the wheel this was Oxford Dickie’s domain; in fact Richard discovered many matrix sources including CD’s and I know also wrote a number of new fully tested algorithms for SQ/QS & matrix H decodes. Anyone know what Richard is up to now?

[Richard]

Hi Loks, long time no speak.

 

I thought i'd break my long silence here to keep those not aware of what i've been up to in my attempts to take decoding of DY/EV-4/SQ/QS/Matrix H & HJ/UHJ to the highest possible level.

 

Things have moved on considerably with most of my time having been spent on the the first four matrix systems, but if you don't mind i'l update you all on where i am with each of them and a little about each (just to update and correct a lot of the misinformation out there):

 

DY: The grand-daddy and the system that had it's first release in 1969/70 and last in 1980!!!! Like it's close bother EV-4 it has been impossible to decode in anything like quad until now. My process, as it stands at the moment, is DY/II and does a very respectable job but will soon be replaced by a new improved process.

 

EV-4: The second oldest of the systems with considerably more releases than it's brother. Decoding has just been upgraded to a process called 'Centaur' which brings certain areas that were difficult to control in EV/II now firmly under control which has improved separation and image stability, especially on complex mixes. It is planned to do the same for DY.

 

DY/EV4: There's a lot of rubbish talked about these two systems, so let me correct the one issue that keeps popping up. Neither systems are 'Phase' matrix systems like those that came after them and are NOT compatible with either QS or SQ. Right, off my soap box...

 

QS: Part or the 'RM' family, which so far consists of QM/QS/QX - none of which appear to be 100% compatible with each other. Decoding of this has gone through various versions and as of this very moment i'm working on a slight upgrade to QS/IIa.

 

SQ: The most popular of all the matrix systems, this is also the most complex of them all and has taken some time to get al the pieces right. The latest process , called 'Phoenix' takes SQ decoding to a level never thought possible as it tackles areas previously ignored in any other decoder.

 

Matrix H: Pass the sick someone. This Frankenstein matrix was the BBC's idea of how it should be done...  LOL. Thank god it died at the end of 1977. I finally cracked it a couple of years ago, process called H-77/II (because there were actually to Matrix H's)

 

UHJ: This is the one system that continues to this day. Cutting the story short i have the following Ambisonic process's: Ambi-UHJ/Ambi-45j/Ambi-SuperStereo.

 

That's very basically it, none of the process's are in the public domain and i'm sure many of you know where i hang out.

 

OD

---

This email has been checked for viruses by Avast antivirus software. www.avast.com

[cepl...@aol.com]

Richard,

  Great to hear from you.   Was worried that I hadn't heard anything in a while.  - CP

[Richard]

Hi mate

 

well not heard from you either, guess too busy. While we are chatting i'm having serious issues with the address (this one in fact) in that it's almost 9.9 spam so most gets auto deleted so it would be an idea to use the other email address until i create a new personal one.

 

I had real trouble doing the first reply

[Lokkerman]

OD

I am so pleased you are well. It has been a long time since we heard from you.

I think you need to speak to Andrej as he is another soul buddy. If you PM me then I will send you his email and you can converse off-line (that is on the highly technical stuff).

Very best wishes

Lokks

[Andrej Falout]

OD,

wow. I mean, Wow. Rally.

Can you give any details of the implementation (SW/HW, processing bit depth, I/O, etc) that you are using?

Also wondering if you ever considered CD-4?

Cheers,

Andrej

[Oxford-Dickie]

Hi

 

Well of course i could write absolutely anything down, it actually proves nothing. And that's one thing i detest, people stating things without proof. So i'm supplying the proof, the 'Phoenix' decoded copy of 'John Keating's - Space Experience'.

 

It's a pretty well know album and shows off the process very well. It's a DVD-A/V with MLP/DTS/DD streams, and the torrent is here:

 

https://onedrive.live.com/redir?resid=D19ED7DC334A1CA2!411&authkey=!APAY-q3TB1MwKFA&ithint=file%2ctorrent

 

By the way it was decoded from the quite rare SQ encoded CD.

 

OK, now to your questions. For reasons that may well become apparent early next year i won't go into the 'hows' but i'd like to pass on some of the facts of life concerning matrix material playback.

 

 

Most of the decoding comes from LP transfers, although these are many of the classical releases are done from CD which is preferable, and this is a very important area where a poor transfer will result in a poor decode, even with hardware decoders.

 

It is important to consider the Pick-up and the RIAA pre-amp as one unit that needs to be not only accurate in frequency response but channel balance and phase response as well. This last area is more important that frequency response and is normally ignored even by stereo purists. Unfortunately those single IC RIAA pre-amps that people use are far from accurate, you need to use something that preferably uses twin stage equalizing, whether it is active or passive doesn't seem to matter, as long as it's accurate and the output isn't inverted.

 

Any channel imbalance, which cartridges are the major culprit, will also affect the ability of the decoder to do it's job.

 

Of course it goes without saying that the cartridge must see the correct loading for optimum results, something that seems to have been completely forgotten nowadays. There can be quite dramatic phase shifts above 10k due to the response peak being incorrectly over or under damped.

 

If your planning on doing a transfer it needs to be fed into a high quality, low noise, soundcard with 24-bit A/D converters. It is important that the software used to record is set-up correctly. I've been sent many albums to decode where this has not happened, being used with basic 16 bit recording enabled. And if there's an  option for 32-bit processing, it must be enabled. Also make sure it is saved correctly once done, another area that people often over look.

 

My set-up is pretty modest really, you don't need to spend a fortune to get high quality results. My turntable is a Rega P2 fitted with an Audio Technica AT440Mla, which is the best cartridge i've ever owned. Correctly loaded it has the flattest phase response i've seen, and sounds superb. This is fed into a self-modified Azzur 551p pre-amp. The combination of these two items produced an almost perfect phase accurate copy of the LP needed for decoding.

 

From there the signal is fed into an M-Audio Delta 96/24 and recorded into Wavelab. The decoding process depends on the system being decoded, each has their own dedicated process's using various pieces of software, etc. None of the matrix systems are compatible so no one decoding process, be it hardware or software, can do the job accurately.

 

As to CD-4 i think it would be possible to do software decoding, bet there are some serious issues to be resolved. The major one is that of replicating ANRS noise reduction. It was originally developed for CD-4 and later adapted for use on cassette decks and is the real stumbling block, but who knows, one day...........

 

I seem to have rambled on again, sorry about that. Hope that helps a little

 

 

OD

[Lokkerman]

OD

I cannot agree more when it comes to RIAA pre-amps. Way back in the 80’s I used to build and design my own pre-amps and you mentioned the most important things; twin stage equalisation, (preferably with a pre-amp buffer so to try and have a stable signal not upset by the cartridge before you eq.), front end damping and phase response. As you know I found myself really sensitive to phase and perhaps this is the reason why I love R2R tape so much - as there is very little to upset in the replay chain, as it is normally integral to the product design. i.e a tape recorder not only records, it plays back and is designed to do so. It is also noteworthy that some vinyl records are known to have phase anomalies.

Andrej

 

This email has been checked for viruses by Avast antivirus software. www.avast.com

 

--

[Jan Bakker]

Thanks for the John Keating's - Space Experience album!!

Op woensdag 10 december 2014 12:04:21 UTC+1 schreef OxfordDickie:

[Jan Bakker]

Well done!! Great sound on this album, thanks again!

Op woensdag 10 december 2014 12:04:21 UTC+1 schreef OxfordDickie:

Hi

[Alex]

Wow.... I'm thrilled.  The sound quality of that CD is... buff.  Unexpected.

Thanks, from my cave.

---

De: Jan Bakker <jhr...@gmail.com>
Para: surrou...@googlegroups.com
CC: dreamingspir...@yahoo.co.uk
Enviado: Miércoles 10 de diciembre de 2014 15:26
Asunto: Re: [SurroundSound] Quadraphonic software decoders

Well done!! Great sound on this album, thanks again!

Op woensdag 10 december 2014 12:04:21 UTC+1 schreef OxfordDickie:

Hi

 

Well of course i could write absolutely anything down, it actually proves nothing. And that's one thing i detest, people stating things without proof. So i'm supplying the proof, the 'Phoenix' decoded copy of 'John Keating's - Space Experience'.

 

It's a pretty well know album and shows off the process very well. It's a DVD-A/V with MLP/DTS/DD streams, and the torrent is here:

 

https://onedrive.live.com/ redir?resid=D19ED7DC334A1CA2! 411&authkey=!APAY-q3TB1MwKFA& ithint=file%2ctorrent

 

By the way it was decoded from the quite rare SQ encoded CD.

 

OK, now to your questions. For reasons that may well become apparent early next year i won't go into the 'hows' but i'd like to pass on some of the facts of life concerning matrix material playback.

 

 

Most of the decoding comes from LP transfers, although these are many of the classical releases are done from CD which is preferable, and this is a very important area where a poor transfer will result in a poor decode, even with hardware decoders.

 

It is important to consider the Pick-up and the RIAA pre-amp as one unit that needs to be not only accurate in frequency response but channel balance and phase response as well. This last area is more important that frequency response and is normally ignored even by stereo purists. Unfortunately those single IC RIAA pre-amps that people use are far from accurate, you need to use something that preferably uses twin stage equalizing, whether it is active or passive doesn't seem to matter, as long as it's accurate and the output isn't inverted.

 

Any channel imbalance, which cartridges are the major culprit, will also affect the ability of the decoder to do it's job.

 

Of course it goes without saying that the cartridge must see the correct loading for optimum results, something that seems to have been completely forgotten nowadays. There can be quite dramatic phase shifts above 10k due to the response peak being incorrectly over or under damped.

 

If your planning on doing a transfer it needs to be fed into a high quality, low noise, soundcard with 24-bit A/D converters. It is important that the software used to record is set-up correctly. I've been sent many albums to decode where this has not happened, being used with basic 16 bit recording enabled. And if there's an  option for 32-bit processing, it must be enabled. Also make sure it is saved correctly once done, another area that people often over look.

 

My set-up is pretty modest really, you don't need to spend a fortune to get high quality results. My turntable is a Rega P2 fitted with an Audio Technica AT440Mla, which is the best cartridge i've ever owned. Correctly loaded it has the flattest phase response i've seen, and sounds superb. This is fed into a self-modified Azzur 551p pre-amp. The combination of these two items produced an almost perfect phase accurate copy of the LP needed for decoding.

 

From there the signal is fed into an M-Audio Delta 96/24 and recorded into Wavelab. The decoding process depends on the system being decoded, each has their own dedicated process's using various pieces of software, etc. None of the matrix systems are compatible so no one decoding process, be it hardware or software, can do the job accurately.

 

As to CD-4 i think it would be possible to do software decoding, bet there are some serious issues to be resolved. The major one is that of replicating ANRS noise reduction. It was originally developed for CD-4 and later adapted for use on cassette decks and is the real stumbling block, but who knows, one day...........

 

I seem to have rambled on again, sorry about that. Hope that helps a little

 

 

OD

 

 

 

OD,

wow. I mean, Wow. Rally.

Can you give any details of the implementation (SW/HW, processing bit depth, I/O, etc) that you are using?

Also wondering if you ever considered CD-4?

Cheers,

Andrej

---

This email has been checked for viruses by Avast antivirus software. www.avast.com

--
--
You received this message because you are subscribed to the Google Groups "SurroundSound" group.
To post to this group, send email to Surrou...@googlegroups.com
To unsubscribe from this group, send email to SurroundSoun...@ googlegroups.com

For more options, visit this group at http://groups.google.com/ group/SurroundSound

---
You received this message because you are subscribed to the Google Groups "SurroundSound" group.
To unsubscribe from this group and stop receiving emails from it, send an email to surroundsoun...@ googlegroups.com.
For more options, visit https://groups.google.com/d/ optout.

--
--
You received this message because you are subscribed to the Google Groups "SurroundSound" group.
To post to this group, send email to Surrou...@googlegroups.com
To unsubscribe from this group, send email to SurroundSoun...@ googlegroups.com
For more options, visit this group at http://groups.google.com/ group/SurroundSound

---
You received this message because you are subscribed to the Google Groups "SurroundSound" group.
To unsubscribe from this group and stop receiving emails from it, send an email to surroundsoun...@ googlegroups.com.

For more options, visit https://groups.google.com/d/ optout.

---

This email has been checked for viruses by Avast antivirus software. www.avast.com

[EoH]

Hi all,

Since changing providers I haven't been around here a lot. Only reading.

A lot of our upmix methods are based on QS although lately we use a complete different approach based on pre-separated source material and second order Ambisonics. Results are sometimes as good as Retail and sometimes even better !

Now I read something about how to decode quad in the best way and unfortunately that's not my thing, but...... we do have a software routine for UHJ encoded CD's.

With this routine I converted a lot of Nimbus records to multichannel .

This routine ( bidule) is not on my website, but if anyone is interested I can publish it........ 

Next year when my contract has finished I will change again to another provider and will be on the hub again too !

kind regards,

EoH

[EoH]

Hi,

What's the difference between  DS58f-2 and DS58p (both John Keating's Space Experience) ?

grtz,

EoH

[Lokkerman]

Well done OD quite an experience; truly amazing from CD.

Andrej

 

This email has been checked for viruses by Avast antivirus software. www.avast.com

 

[Oxford-Dickie]

It's just as good from LP...

[Lokkerman]

And so it should be J

[Andrej Falout]

Hi OD,

On Thu, Dec 11, 2014 at 12:07 AM, 'Oxford-Dickie' via SurroundSound <surrou...@googlegroups.com> wrote:

without proof. So i'm supplying the proof, the 'Phoenix' decoded copy of 'John Keating's - Space Experience'.

Sounds interesting, and I appreciate your sharing this with us. But I do have to confess that this aggressive synth is not agreeing with my ears - to the point that I cannot objectively listen to it. I accept that this is entirely my own fault.

I wonder how a bit, shell I say, "conservative" peace of music would sound using this processing - maybe DSOTM, Steely Dan, Sanatana, Mike Oldfield perhaps - something that was previously decoded too, so we can compare it with previous transfer ?

 

By the way it was decoded from the quite rare SQ encoded CD.

... therefore 16 bit. I wonder how would that compare with vinyl transfer.

 

 

OK, now to your questions. For reasons that may well become apparent early next year i won't go into the 'hows'

I am very interested to follow the events "early next year" - what is the best way/place to follow your future developments?

Most of the decoding comes from LP transfers, although these are many of the classical releases are done from CD which is preferable, and this is a very important area where a poor transfer will result in a poor decode, even with hardware decoders.

I do agree that poor vinyl transfers are a weak point, but would argue that a high quality 24/96 vinyl transfer is preferable to 16 bit source. To my ears - and I do confess to be an audiophile, not only surround-o-phile :)

My experience is that minimising tracking errors by using tangential arms and Shibata stylus can increase quality of matrix decoding from vinyl dramatically. Paired with a a good turntable, DA, and vinyl itself, of course.

Additionally, the fact remains that in large majority of cases, there is simply no choice, as the vinyl is the only available source.

 

 

It is important to consider the Pick-up and the RIAA pre-amp as one unit that needs to be not only accurate in frequency response but channel balance and phase response as well. This last area is more important that frequency response and is normally ignored even by stereo purists.

100% agree. This is the same reasoning that explains why tangential arm helps so much - much more then for simple stereo material.

 

Unfortunately those single IC RIAA pre-amps that people use are far from accurate, you need to use something that preferably uses twin stage equalizing, whether it is active or passive doesn't seem to matter, as long as it's accurate and the output isn't inverted.

 

Any channel imbalance, which cartridges are the major culprit, will also affect the ability of the decoder to do it's job.

 

Of course it goes without saying that the cartridge must see the correct loading for optimum results, something that seems to have been completely forgotten nowadays. There can be quite dramatic phase shifts above 10k due to the response peak being incorrectly over or under damped.

Loading is unfortunately something nobody likes to talk about this days. It's crazy. Just impedance and capacity matching is responsible for large part of overall analogue system sound. As any recording engineer that connected a microphone to a wrong preamp learned early on. Corresponding phase shifts will kill any chance of good matrix decoding. All 100% true. But it does not have to be that way.

 

 

If your planning on doing a transfer it needs to be fed into a high quality, low noise, soundcard with 24-bit A/D converters.

I Use Prism Sound Lyra and Dream AD-2/DA-2. I would discourage use of any PC soundcards for this purpose. As one can frequently find Apogee Rosetta AS96 on eBay for 300 USD or so, there is really no reason to.

 

It is important that the software used to record is set-up correctly. I've been sent many albums to decode where this has not happened, being used with basic 16 bit recording enabled. And if there's an  option for 32-bit processing, it must be enabled. Also make sure it is saved correctly once done, another area that people often over look.

Yup. Another one of my "favourites" is not adjusting analogue levels correctly, either hitting the digital 0db in AD (OMG!) or recording everything at -40db and then normalising it...

 

My set-up is pretty modest really, you don't need to spend a fortune to get high quality results. My turntable is a Rega P2 fitted with an Audio Technica AT440Mla, which is the best cartridge i've ever owned. Correctly loaded it has the flattest phase response i've seen, and sounds superb. This is fed into a self-modified Azzur 551p pre-amp. The combination of these two items produced an almost perfect phase accurate copy of the LP needed for decoding.

If you get your hands on a good tangential TT, such as Yamaha PX2/3. Pioneer PL-1000, Sony PSX800/555, Revox B291/790, several B&O models starting with 4002 (which BTW had a CD-4 decoder built in in some versions), H/K ST-6 (Rabco), .... try one. You will very likely appreciate the results.

Yamaha PX3 and Pioneer PL-1000 can regularly be seen on eBay for around 300 USD. Both are excellent.

 

 

From there the signal is fed into an M-Audio Delta 96/24 and recorded into Wavelab. The decoding process depends on the system being decoded, each has their own dedicated process's using various pieces of software, etc. None of the matrix systems are compatible so no one decoding process, be it hardware or software, can do the job accurately.

My OP was about finding available SW decoder(s) - can you talk about them, or is that part of the upcoming events?

 

 

As to CD-4 i think it would be possible to do software decoding, bet there are some serious issues to be resolved. The major one is that of replicating ANRS noise reduction. It was originally developed for CD-4 and later adapted for use on cassette decks and is the real stumbling block, but who knows, one day...........

 

Some useful info here: http://www.google.com/patents/US3991283, http://www.google.com/patents/US3936619 and few more...

In the meantime, what are the best options for decoding CD-4 - today?

 

I seem to have rambled on again, sorry about that. Hope that helps a little

It is very much appreciated. Thank you for sharing.

Andrej

[Andrej Falout]

Hi EoH,

is that "routine" a patch for a "Bidule" commercial application?

I have seen several free UHJ decoding apps, but did not try them so far. Any recommendations?

Thanks,

Andrej

--

[EoH]

Hi Andrej,

Yes, it's a layout for Plogue Bidule which is a commercial program.

But IIRC they have a fully functional trial version on their site, which takes a long time to expire......

The layout however doesn't contain any commercial VST, so is freeware.

My Swiss mate wrote it years ago for me , because I owned quite some UHJ encoded stereo CD's from Nimbus UK.

Results varied, but recordings from Nimbus vary a lot too ;-) (Cannot expect much from a 78 RPM transferred to CD complete with all the typical 78 records distortions.) The orchestral pieces came out quite good !

I can only recommend this one, because it's the only one I have ;-)

FWIW my mate also wrote several scripts to decode SQ and QS with Cooledit (Audition). Look on the quad forum for his name : kempfand !

And another thing he also made especially for me : SSSBV : Stereo to Super Sereo to B-format to Vienna encoder to multichannel WAV.

The routines for this were taken from a bought AES paper, so I'm not sure how things are with copyright after all these years.

My mate is out of running for 5 years now, so I cannot ask him......

My expertise in the development of layouts in Plogue Bidule was the critical note and corrector on what didn't sound good..... nagging till the endresult was perfect ;-) ;-)

And luckily kempfand had the technical skills to put my nagging into improved code.

As said in another message : my main hobby is upmixing those stereo CD's that have never been brought out in any form of surround.

kind regards,

Aart

Op donderdag 11 december 2014 09:01:36 UTC+1 schreef Andrej Falout:

[EoH]

Well, I did compare the f-2 and p version and........ there is hardly any difference !

I ripped from both DVDA's the first two tracks and compared them in Soundforge > It's very hard to discover any difference.

There was some in the RMS and Peak levels, but that difference was so small, not worth to mention....

I did however found it a bit strange that both are recorded on DVDA in 16/44.1 ..... I expected 24/96 ;-)

Anyway, when listening on my HTS I also couldn't hear a difference.......

grtz,

EoH

Op woensdag 10 december 2014 16:17:37 UTC+1 schreef EoH:

[Richard Ford]

Please do explain why you would expect to see 24/96 considering the source was a CD?

---

--

[Aart Nienhuis]

Because it was a DVDA.........

You received this message because you are subscribed to a topic in the Google Groups "SurroundSound" group.
To unsubscribe from this topic, visit https://groups.google.com/d/topic/surroundsound/J7xcnUiwpNc/unsubscribe.
To unsubscribe from this group and all its topics, send an email to surroundsoun...@googlegroups.com.

For more options, visit https://groups.google.com/d/optout.

--

 

-=For stereo to surround conversions visit http://www.dtsac3.com=.

[Richard Ford]

Just because it was a DVD-A doesn't mean the contents has to be 24/96. It can be whatever.

---

[Aart Nienhuis]

I know ! I only see 99% in 24/96.............. ;-)

[Richard Ford]

Well, just because it says it is doesn't mean it is. A couple of weeks ago i received proof that a major record company upsampled a title from 48k to 96k just to please those who won't buy it if it doesn't say 24/96.

So what did that upsampling do to the quality? It wasn't as pure as the 48k original, but as long as people think they are getting it then i guess it keeps them happy.

---

[Lokkerman]

Many  high sampling rate recordings appear to be up-sampled or at a higher sampling rate than originally recorded. My own experiments indicate that some up-samples (note: some) can sound better, but I don’t profess to know why.

 

With some recordings I don’t think we are getting a true master or a true up-sample; as I think they  could be using, for example, vintage digital multi-track, outputting mix-down in analogue; then converting back to digital.

 

I feel that some of the Hi-Res Madonna HDtracks are this way - as it is well documented:http://www.soundonsound.com/sos/sep07/articles/classictracks_0907.htm) that the “Like A Virgin” album was recorded on a Sony 3324 digital tape recorder with a Sony F1 for the two track ( I know you should get a similar result by a sample/rate conversion but where do the out-of- band signals come from without adding something?)

 

If you Google the Hoffman forums the threads appear to imply that the HDtracks master was from the analogue master -  so that’s how you get to a 24/192 master recording, without some form of sleight of hand,  - this was a DAD recording?

 

Anyway this thread reaches a consensus and that is that the 24/192’s sound better to quite a few. Let’s not talk psychology here as these guys are studio wiz’s and should know  better. And at that point I stop before I start another great debate.

 

Anyway a great free tool can be found here to analyse and see this visually: http://www.sonicvisualiser.org/

 

This tool Is developed by the University of London QMC. So it is for academic use but it doesn’t half quickly show you a number of things that you normally need a professional tool for.

 

From: surrou...@googlegroups.com [mailto:surrou...@googlegroups.com]
Sent: 11 December 2014 21:54
To: surrou...@googlegroups.com
Subject: Re: [SurroundSound] Quadraphonic software decoders

 

Well, just because it says it is doesn't mean it is. A couple of weeks ago i received proof that a major record company upsampled a title from 48k to 96k just to please those who won't buy it if it doesn't say 24/96.

[Andrej Falout]

Hey Lokks,

On Fri, Dec 12, 2014 at 12:13 PM, Lokkerman <phil.s...@gmail.com> wrote:

 

With some recordings I don’t think we are getting a true master or a true up-sample; as I think they  could be using, for example, vintage digital multi-track, outputting mix-down in analogue; then converting back to digital.

Ironic as it may be, this is actually in line with the idea of audio fidelity. It preserves, the best it can, the sound that artist/producer was hearing when they made creative decisions.

(I wonder how much fun would be to be able to go to the original studio, and record a master playing on the original house speakers - through the Neumann dummy head placed in the producer's char. LOL)

If they took the PCM out of the digital mixdown/master, then used today’s advanced upsampling methods, they would almost certainly get more juice out of it, but it would not be what was originally intended/heard.

 

 

I feel that some of the Hi-Res Madonna HDtracks are this way - as it is well documented:http://www.soundonsound.com/sos/sep07/articles/classictracks_0907.htm) that the “Like A Virgin” album was recorded on a Sony 3324 digital tape recorder with a Sony F1 for the two track ( I know you should get a similar result by a sample/rate conversion but where do the out-of- band signals come from without adding something?)

Mathematics :)

 

 

If you Google the Hoffman forums the threads appear to imply that the HDtracks master was from the analogue master -  so that’s how you get to a 24/192 master recording, without some form of sleight of hand,  - this was a DAD recording?

Some caution is needed here (and in other cases of early "digital" recordings). First, more often then not, a lot of tracking was done to analogue, often in various studios, and then transferred to digital multi-track in studio where record was mixed. Second, a lot of masters where done to both digital and analogue in parallel.

Third, a lot of mastering/cutting engineers of the time did not have digital anything, and a lot of them despised digital (with a good reason, at that point in time. Of course, they would never say so in public for obvious reason of having a mortgage and kids to feed).

That meant that a large number of digital masters where never used past mastering, for other purpose other then marketing (The famous "DDD" utopia). I think there was 1992 when I heard first fully digital mastering chain. It sounded awful. Manager of the facility told me not to worry, as they got it only to be able to tick the check-box of the marketing guys, and don’t plan to use it. Imagine a digital compressor in year 92 and be happy they didn’t.

All of the above also explains to a large degree why some of the early "digital" recordings actually sounded good. If they where digital to any significant degree. Such as Stevie Wonder JTTSLOP and Hotter then July, where Sievie's management on the covers is profusely thanking Sony for donating most of the hardware for recording studio he built in his new house. Not in those exact words, of course. (BTW, he was famous for graciously accepting prototypes of many vendors, for testing of course... bless him :) Yet, I have so far 3 separate, first hand witness reports, that both where tracked to analogue, and mixed in parallel to analogue and digital. From which point nobody ever recalls seeing the U-matic and/or Beta digital tapes. Ever...

So there. :)

Andrej

[Lokkerman]

Andrej

Lovely summary there and thanks for it  – you confirm what I always suspected ;-)

 

From: surrou...@googlegroups.com [mailto:surrou...@googlegroups.com] On Behalf Of Andrej Falout
Sent: 12 December 2014 00:18
To: surrou...@googlegroups.com
Subject: Re: [SurroundSound] Quadraphonic software decoders

 

Hey Lokks,

--

[Oxford-Dickie]

OK, have a few minutes before i continue the day job (converting quad material)..

 

I wonder how a bit, shell I say, "conservative" peace of music would sound using this processing - maybe DSOTM, Steely Dan, Sanatana, Mike Oldfield perhaps - something that was previously decoded too, so we can compare it with previous transfer ?

I have planned a demonstration disc set showing the the decoded versions of tracks from many albums, and the first will be SQ. It will also contain the same tracks so comparisons with other 'decoders' can be made. It has been on the back burner due to the work on the latest process's but i think i will start on it in the new year. I'll post here when it's done. (by the way, Steely Dan releases were in QS)

 

 

therefore 16 bit. I wonder how would that compare with vinyl transfer

 

It would be 'almost as good'. A good, untouched, digital copy is the best option when decoding phase matrix recording.

 

 

I am very interested to follow the events "early next year" - what is the best way/place to follow your future developments?

 

Hmmm... this is a little awkward. There are many who don't agree with the way i have been forced to go in relation to 'my work', plus i didn't come out to advertise 'my wares', so to speak so i'll leave that up to the moderator to decide whether this is the place to mention where i dwell.

 

 

My OP was about finding available SW decoder(s) - can you talk about them, or is that part of the upcoming events?

I've been working with Universal records on a project that has possibly been extended. I can't say anymore at the moment, it's all WIP. As to discussing the workings of the process's, i'm no longer inclined to release any information due attitudes from certain sectors who 'know better' but fail to prove it.

 

 

In the meantime, what are the best options for decoding CD-4 - today?

Well, surprisingly, at the heart of the JVC 4DD-5 (or one of it's clones) is a good demodulator. Out of the box it's ok, but it can be improved without to much work.

 

 

Hope that helps

 

OD

[Lokkerman]

OD

If you want to leave a link to find you – it’s up to you. I think what you do is very valid; my personal view is that I would let a few of the older quad releases creep into the wild a little so that people may understand your commitment more.

 

From: surrou...@googlegroups.com [mailto:surrou...@googlegroups.com]
Sent: 12 December 2014 09:37
To: surrou...@googlegroups.com
Subject: Re: [SurroundSound] Quadraphonic software decoders

 

OK, have a few minutes before i continue the day job (converting quad material)..

--

[Oxford-Dickie]

Hi

 

Thanks Lokks for clearing that up, and as to commitment, i think i need committing! With 451 releases, and counting - at least 3 a week, i never thought it would have gotten this out of control.

 

Your idea is certainly an interesting one, although it does pose a problem. I'd not be happy to have a lot of the earlier releases out there as an advertisement of what i'm doing now, the decodes are incredibly accurate in ways that no other form of decoding can emulate.

 

I'll try and explain. I've dug very into what SQ really is about and, as i mentioned before, it is far more complex than folk realise when  it comes to decoding.

 

The first area that all forms of software decoding (and a certain hardware one) fails to get anywhere near right is the extraction of the rear channels. I've tested every software version and they get nowhere. There is actually only one way to do it, and that's because of the way the encoding matrix was written. Very clever!

 

The other major area that no hardware or software decoder has ever attempted to deal with is the fronts. It is never discussed and i suspect Columbia wanted it that way. There are those who can't stand to listen to SQ decodes, even when done by a Tate. The reason is because the front channels of every decoder are actually just the raw stereo encoded tracks. It has been impossible to remove the complex phase information that are the rears.

 

I'm sure i don't need to tell you that having all of this information being played along with the decoded rears causes a 'bit' of a phase muddle, the effects of which colour every aspect of the aural presentation.

 

After tracking down that naughty phase issue on the Right Rear channel i decided to finally tackle the fronts. Long story short, i've managed to greatly reduce all of the rear channel information by a large margin. The obvious benefits are improved image location and stability, better rendition of low level information like reverb, etc (something that was always a complaint of SQ) and improved separation.

 

There are other areas that are improved as well, but you get the idea. There is a problem though. It now takes 3 - 4 times longer to do a decode. This has scuppered my idea of doing a mass re-issue of titles which does obviously affect your idea.

 

The only way around it is a far more powerful computer, but that costs so will have to wait until next year.

 

So you see why i'm not to happy about making the older titles openly available. What i will do is the afore mentioned demo-disc set and i'll put out a new title for all, again in the new year.

 

I'm busy doing Christmas releases, so must get on...

 

My home - IF you go there you MUST read the rules first. I don't reply to questions regarding them:

 

https://dreamingspiresquadarchive.wordpress.com

 

 

OD

[Malcolm Lear]

Hi OD,
Whilst you may have developed a method to decode SQ that produces good separation, I must disagree that the fronts have been ignored in previous hardware decoders. In fact its well known that SQ hardware decoders have always enhanced both front and rear channel separation to an equal degree. This includes the most basic dynamic front/back blend systems from CBS through to the Tate. The fact you have only now started to look at front/rear separation and dealing with them on an equal basis is quite amazing. You say it is never discussed and Columbia wanted it that way, what utter rubbish, please read the many papers and patents on various variable coefficient decoding systems which  includes the Tate and Shadow Vector.

If your decoding run time is so long, I guess you are still using scripts so I'd consider moving to a compiler and writing real code.