Is 24/96 really better than 24/48?
Rich
higher sampling rate of 96khz when most mics don't go much beyond 20 khz
which is under the effective frequency response of 24/48. I know the
argument about upper harmonics etc. but if a mic can't pick them up what's
the difference? Ok localization. But that's not that big of a deal in a
multitrack situation where tracks are generally mono anyway and the mixer
determines the stereo placement. Tel me how uninformed I am so I can
justify the added expense of 24/96.
Rich
Bill Roberts
Rich wrote:
>
> I'm sure I'm missing something here but how do you take advantage of the
> higher sampling rate of 96khz when most mics don't go much beyond 20 khz
> which is under the effective frequency response of 24/48. I know the
> argument about upper harmonics etc. but if a mic can't pick them up what's
> the difference? Ok localization. But that's not that big of a deal in a
> multitrack situation where tracks are generally mono anyway and the mixer
> determines the stereo placement.
The fact is you need a higher sampling rate than the highest
frequency to represent it digitally. You need at least twice,
but if you try only twice or just over, this causes problems
in practice. 44.1 or 48 kHz is sufficiently over twice so
as to work pretty well.
However the one comparison I've made (96 kHz vs 48 kHz)
on a Pioneer DAT was clearly audible. No huge deal but
it was better for sure. Many others have heard the same
thing on many other types of equipment.
Who cares what the technical reason is. Probably has
to do with being able to use lowpass filters that is
further removed from the audio band, and/or with
a gentler slope, but really, how does the reason
matter to the user?
> Tel me how uninformed I am so I can
> justify the added expense of 24/96.
You probably can't justify the expense unless everything
else in your recording chain is already primo. Otherwise
the money would be better spent on better mics, better room,
better preamps, better EQ, better compressors, better mixer,
etc.
-- Bill
nos...@sirius.com
mag) that most converters do a pretty poor job of the "brick-wall"
filtering necessary to limit response to stay under the Nyquist limit,
meaning that even though you may have theoretical response to 22 kHz
with a 44.1 kHz sampling rate, the resulting fidelity up there in the
high frequencies isn't so hot.
One way to solve this, according to the article, is to improve the
engineering and components on current D/A and A/D converters. Another
way to "hide it under the rug," however, is to increase the sampling
rate, so that these inaccuracies are shoved way up there in the
frequency spectrum, meaning that theoretically, you should have
cleaner sound even in the audible range, not to mention the
(perceptible?) upper frequencies...
BTW, just because a mic is speced at a certain upper frequency
response limit doesn't mean there isn't still some information above
that.
"Rich" <ri...@attainment-inc.com> wrote:
>I'm sure I'm missing something here but how do you take advantage of the
>higher sampling rate of 96khz when most mics don't go much beyond 20 khz
>which is under the effective frequency response of 24/48. I know the
>argument about upper harmonics etc. but if a mic can't pick them up what's
>the difference? Ok localization. But that's not that big of a deal in a
>multitrack situation where tracks are generally mono anyway and the mixer
>determines the stereo placement. Tel me how uninformed I am so I can
>justify the added expense of 24/96.
>Rich
>
>
Rich
>The fact is you need a higher sampling rate than the highest
>frequency to represent it digitally. You need at least twice,
>but if you try only twice or just over, this causes problems
>in practice. 44.1 or 48 kHz is sufficiently over twice so
>as to work pretty well.
Yeah the Nyquist theorem or whatever.
>However the one comparison I've made (96 kHz vs 48 kHz)
>on a Pioneer DAT was clearly audible. No huge deal but
>it was better for sure. Many others have heard the same
>thing on many other types of equipment.
>Who cares what the technical reason is. Probably has
>to do with being able to use lowpass filters that is
>further removed from the audio band, and/or with
>a gentler slope, but really, how does the reason
>matter to the user?
It matters because we may just be talking about converters that are better
and might be perceived as better even if run at 24/48. if you're talking
about accurate A/B blindfold testing that's one thing. Otherwise, if it's a
subtle difference it might be chalked up to gear slut syndrome where it "has
to be better". It matters because working in 96 tends to decrease track
count and increases storage requirements and, yes, cost. I thought that the
current digital filters available pretty much removed the problems
associated with earlier analog filters used in convertors.
In any case, sounds like you might agree with me that upping the frequency
response of the recording is moot in terms of capturing those mixtures of
upper harmonics. If teh mic can't hear them how high the convertors can go
is irrelevant. Maybe relevant in terms of more benevolent filtering but not
in terms of actually capturing any relevant audio.
Rich
Bill Roberts
Rich wrote:
> >However the one comparison I've made (96 kHz vs 48 kHz)
> >on a Pioneer DAT was clearly audible. No huge deal but
> >it was better for sure. Many others have heard the same
> >thing on many other types of equipment.
> >Who cares what the technical reason is. Probably has
> >to do with being able to use lowpass filters that is
> >further removed from the audio band, and/or with
> >a gentler slope, but really, how does the reason
> >matter to the user?
>
> It matters because we may just be talking about converters that are better
> and might be perceived as better even if run at 24/48.
What I mean is, if you get a sound that you find better when
you run the same converters at 96 kHz than 48 kHz, or a sound
that is better from Brand A which runs at 96 vs Brand B that
runs at 48, well, the 96 sounds better regardless of the
reason behind it. If that reason is because Rupert Neve
said a prayer for it, and nothing more, still if it sounds
better that is what matters.
> if you're talking
> about accurate A/B blindfold testing that's one thing.
Yeah, it is having another person switch the machine. It is
something that is too close to be able to do yourself and
be sure.
> Otherwise, if it's a
> subtle difference it might be chalked up to gear slut syndrome where it "has
> to be better". It matters because working in 96 tends to decrease track
> count and increases storage requirements and, yes, cost.
Well, the reason still doesn't matter from the practical point of
view: if the 98 kHz piece of gear sounds better it doesn't matter
to the client what the technical reasons might be. It does matter
to designers of other equipment.
> I thought that the
> current digital filters available pretty much removed the problems
> associated with earlier analog filters used in convertors.
Maybe so, some have told me so.
> In any case, sounds like you might agree with me that upping the frequency
> response of the recording is moot in terms of capturing those mixtures of
> upper harmonics.
Not necessarily, that doesn't seem a proven point.
> If teh mic can't hear them how high the convertors can go
> is irrelevant.
Irrelevant if there is no beneficial change thereby allowed
in the analog circuits before the converter.
> Maybe relevant in terms of more benevolent filtering but not
> in terms of actually capturing any relevant audio.
Well, if the filtering is more benevolent, then the relevant
audio could be more accurately captured.
Again, I don't know the reason.
Personally 96 kHz is not a big priority with me, nor are
converters claiming more than 20 bits. I do believe that
mixers etc. should use at least 24 bits for sure to handle
the data.
-- Bill
Rich
nos...@sirius.com wrote in message <3804b27...@news.sirius.com>...
>My understanding is (largely based on a recent editorial in Recording
>mag) that most converters do a pretty poor job of the "brick-wall"
>filtering necessary to limit response to stay under the Nyquist limit,
>meaning that even though you may have theoretical response to 22 kHz
>with a 44.1 kHz sampling rate, the resulting fidelity up there in the
>high frequencies isn't so hot.
I thought designs had advanced in this area.
Another way to "hide it under the rug," however, is to increase the
sampling
>rate, so that these inaccuracies are shoved way up there in the
>frequency spectrum, meaning that theoretically, you should have
>cleaner sound even in the audible range, not to mention the
>(perceptible?) upper frequencies...
if this is the case, what's the thinking behind (it's coming)192khz
sampling? Putting filtering in an even more inaudible area?
>
>BTW, just because a mic is speced at a certain upper frequency
>response limit doesn't mean there isn't still some information above
>that.
The operative word here is "some". How much upper frequenciy/harmonics
gathering goes on above 20khz with available mics?
Rich
Scott Dorsey
>My understanding is (largely based on a recent editorial in Recording
>mag) that most converters do a pretty poor job of the "brick-wall"
>filtering necessary to limit response to stay under the Nyquist limit,
>meaning that even though you may have theoretical response to 22 kHz
>with a 44.1 kHz sampling rate, the resulting fidelity up there in the
>high frequencies isn't so hot.
This was the case a decade ago. This is no longer the case.
>One way to solve this, according to the article, is to improve the
>engineering and components on current D/A and A/D converters. Another
>way to "hide it under the rug," however, is to increase the sampling
>rate, so that these inaccuracies are shoved way up there in the
>frequency spectrum, meaning that theoretically, you should have
>cleaner sound even in the audible range, not to mention the
>(perceptible?) upper frequencies...
This is what oversampling does. It runs the converter at a very high
rate, then downsamples, so the filtering is mostly done in the digital
domain and only a shallow-slope analogue filter is needed. A typical
modern A/D runs a lot faster than 96 Ksamp/sec but it samples down before
producing output.
--scott
--
"C'est un Nagra. C'est suisse, et tres, tres precis."
Terry Demol
LPF's etc the newer 24/96 AD and DA chips basically
have better performance.
The newest Crystal 24/96 AD chips have 120dB DR as do
their latest DA chips. Same goes for Burr Brown multibit
DAC chips the 24/96 has slightly better DR and THD specs.
Same goes for Analog Devices 24/96/192 DA chips.
The performance bar gets higher constantly.
Having said this it is sad to see great chips with great
specs
in cheap little boxes with poor power supplies and analog
stages. Thats where a big difference in performance exists.
Hence the fact that some older format converters sound very
good.
Regards,
Terry Demol
Stephen St.Croix
<ri...@attainment-inc.com> wrote:
> I'm sure I'm missing something here but how do you take advantage of the
> higher sampling rate of 96khz when most mics don't go much beyond 20 khz
> which is under the effective frequency response of 24/48. I know the
> argument about upper harmonics etc. but if a mic can't pick them up what's
> the difference? Ok localization. But that's not that big of a deal in a
> multitrack situation where tracks are generally mono anyway and the mixer
> determines the stereo placement. Tel me how uninformed I am so I can
> justify the added expense of 24/96.
> Rich
24 bits good. Very good.
96 kHz totally stupid. Today's oversampled convertors no longer have those
old brickwalls, and 96 cannot be properly divided to get 44.1, what the
real world uses- 96 is useless.
________________
Stephen St.Croix
Dave Martin
Stephen St.Croix wrote in message ...
>24 bits good. Very good.
>
>96 kHz totally stupid. Today's oversampled convertors no longer have those
>old brickwalls, and 96 cannot be properly divided to get 44.1, what the
>real world uses- 96 is useless.
>
>________________
>Stephen St.Croix
Well, except that most digital recording is not destined for CD, but for
broadcast or film (with broadcast having by far the lion's share). I've
heard and read several opinions stating that 48kHz is the 'professional'
standard - 44.1 is 'just for music. CD's work at 44.1kHz - the entire
broadcast and post production world is at 48kHz. 96 divides quite well to
get to 48. And unless I'm mistaken, most of the 96kHz converters actually
give you a choice or 96 or 88.2 kHz.
Dave Martin
Digital Media Associates, Inc.
Nashville, Tennessee
dave....@nashville.com
party jammer
experiences with dogs will
maximize our living experience on this tainted orb 24/7. If your ears
say yes ,then surrender to the rhythm....and the forbidden dance......on
this forbidden planet.
ROCK ON, PARTYJAMMER!!!!
James Quinn
not, discuss elsewhere), then there is also the problem of "oversampling". All
the filters (including all the digital processing, that is) are doing a load of
maths on the differences between successive samples. If you squeeze even more
samples into the same period, then the differences between the samples gets
smaller.
So what? Well, at a given bit resolution the performance of the filtering is
reduced. Also you start needing filters with larger sample windows. So, to
avoid a reduction in filter quality, you tend to need greater bit resolution
AND even more processing power.
I think a term "diminishing returns" may apply...Surely the real way forward is
to use non-linear quantisation (cf. telephony) or is that an evil term??
BTW the converter's job is to convert - a filter does the filtering. We can
make quite sharp filters if we want to, although it normally means paying for
phase response, charging energy or time delay or whatnot.
Jim
nos...@sirius.com wrote:
> My understanding is (largely based on a recent editorial in Recording
> mag) that most converters do a pretty poor job of the "brick-wall"
> filtering necessary to limit response to stay under the Nyquist limit,
> meaning that even though you may have theoretical response to 22 kHz
> with a 44.1 kHz sampling rate, the resulting fidelity up there in the
> high frequencies isn't so hot.
>
> One way to solve this, according to the article, is to improve the
> engineering and components on current D/A and A/D converters. Another
> way to "hide it under the rug," however, is to increase the sampling
> rate, so that these inaccuracies are shoved way up there in the
> frequency spectrum, meaning that theoretically, you should have
> cleaner sound even in the audible range, not to mention the
> (perceptible?) upper frequencies...
>
> BTW, just because a mic is speced at a certain upper frequency
> response limit doesn't mean there isn't still some information above
> that.
>
Marc Lindahl/Sonorus
----------
>From: "Dave Martin" <dave....@nashville.com>
> I've heard and read several opinions stating that 48kHz is the 'professional'
>standard - 44.1 is 'just for music.
Actually, it's an AES standard, for the 'preferred' sample rate, which is
stated as
48kHz (over 44.1k, 96k, etc.).
> 96 divides quite well to
>get to 48.
Last year's AES, someone from Sonic Solutions did a paper on this -- it's
possible to do as good sample rate conversion to odd ratios as it is to even
ones. So it's a matter of the quality of implementation, so it could be
just as crappy with a 2:1 decimation if it's badly done.
Marc Lindahl <ma...@sonorus.com>
president, Sonorus Inc.
http://www.sonorus.com
Dave Martin
Brian Tankersley wrote in message <380814DD...@home.com>...
>
>And doing the best job starts with a bit of philosophy. Mine is
>this....listen like the end user will so that you and he/she share a
>similar musical experience. I don't use Genelecs, Dynaudio, etc. I use
>older, relatively lo-fi and unexciting Fostex coax. I find I work harder
>to get happy on that type of speaker. No hype.
Unfortunately, that road gets pretty dark. Outside of work, my main
listening is in the truck. FM radio through the stock Delco system in a GMC
Sonoma. I aslo have the drivers side window cracked, so you'd have to add
that noise... My home stereo has a pair of Minimus 7s.
So your speakers are better than either of my non-work listening
environments. Should you lower your mixing standards to my listening
standards? I would hope not. Or we could talk about my friend john's
listening environment. He put his $15K surround system together, and it
sounds great. Should you mix for his environment?
There is no 'typical' end user.
>
>As a matter of fact, much of what I currently hear sounds midrangy and
>"woofy" to me. I think it's a combo of too much compression and speakers
>too hyped on the top and bottom. The current crop of self powered "super
>hi-fi nearfields" is the culprit, IMO. I'm sure many would disagree.
I agree that there's definitely something wrong, and compression (or more
accurately), L1 and Finalizer type limiting is a great deal of the problem.
I haven't thought about the need for a Genelec EQ curve (I know some
mastering guys who have a stock NS10 curve), but maybe it would help.
>
>I even debate about monitoring 24 bit when mixing, though I certainly
>print the mix at 24 bits for the obvious reasons. Should I monitor
>dithered to 16 bit so I'm actually hearing the final result? A valid
>question, IMO, since nobody else will hear it at 24 bits any time soon.
>
You probably check it at some point, don't you? But I would think that even
checking the mix through a boom box would work, since I doubt that a boom
box is capable of 24 bit resolution (or 20 or 21, whatever the Paris
actually has).
>Again, there are many different ways to approach this subject. But
>"doing the best we can" does not necessarily mean listening in a fashion
>that no one who buys our music (thereby paying our salary) can hear. It
>may mean just the opposite.
Perhaps not, but, does it mean igoring that small percentage of the
population who do listen through audiophile systems? In a recent Audio Media
review, Doug Mitchell talked about pulling up what he thought was a finished
mix on the M&K Surround system, and being shocked to hear some major
problems (that were not audible on his usual mix speakers). Some percentage
of the record buying public will have a monitoring system that good.
Philosophically, do you ignore that audience as too small to worry about?
As a musician doing live gigs, I never played for the 'average' customer - I
played for me (and I'm a lot bigger critic of my playing than even most
players would be). As an engineer, I tend to take the same approach, 'Within
the time, money and client constraints, what can I do to make this project
sound as good as it can sound? And within those parameters, I work.
>
>Paris sounds killer to me at 24/44.1. Better than any other digital I've
>heard, IMO. 96K is pretty low on my list. Different strokes.
>
Cool. I'd still like to hear it the next time you do a PT/Paris/3348
comparison.
Marc Lindahl/Sonorus
----------
In article <s0fb28...@corp.supernews.com>, "Dave Martin"
<dave....@nashville.com> wrote:
>
>I think I need to buy some more books. Does the AES have a standards book,
>or is it somewhere on the website if they're still putzing around with it?
Yes, you can buy all that stuff from the AES (http://www.aes.org).
Unfortunately -- I would gladly pay another $5/year if they'd just post all
the stuff as pdf's on their website for free!
>chairs one; if and when the AES agrees on a standard in some area of audio
>engineering, do they ever actually print it?
Yes, they print reams of them. And some are actually used and followed :)
Marc Lindahl/Sonorus
----------
In article <3808F279...@home.com>, Brian Tankersley <gbt...@home.com>
wrote:
>
>So how many people 'round here have actually heard Paris in a controlled
<snip>
>years. They're retired in favor of the stock 24 bit cards you can buy
>for Paris.
What, are you a Paris salesman? Or just an unbiased observer?
Marc Lindahl/Sonorus
----------
In article <3807C296...@home.com>, Brian Tankersley <gbt...@home.com>
wrote:
>Is that why I can hear (and object to) the SRC built into Sonic
>Solutions and consequently don't use it?
That's probably why they spend so much time researching a new one. I don't
think the one in the paper is implemented yet, but maybe it is on their new
DSP board -- is that the system you are referring to?
>Maybe the guy who wrote the paper should take another crack at his own
>algo. Right now, I'd say Sonic is not a good advertisement for
>transparent sounding SRC.
Actually the guy is a female... I guess you didn't take a look at the paper
:)
Mikko Helin
a CD player system which with it's built-in CD-ROM drive,
decoder software and 24bit/96kHz multichannel converters
is able to play back all kind of digital audio stored
on CD's, ISO9660/Joliet and FAT16/32 file systems,
.WAV files, .MP3's, and supports loading of OS/flashing ROM
for updates. I mean, Terratec is already quite close to
this concept with their M3PO player, see
www.terratec.net
http://www.terratec.net/products/m3po/m3po_overview.htm
http://www.terratec.net/products/m3po/m3po_details.htm
They just need better hardware (support for 24bit/96k
digital audio) and they are there.
-Mikko
Brian Tankersley <gbt...@home.com> wrote:
> A question. Anybody want to hazard a guess at how many years it will be
> before more than 5% of consumers own a system doing 96K audio?
> My first guess might be "never", but that's probably too pessimistic, so
> I'll safely guess 5 years.
> Regards,
> Brian T
Marc Lindahl/Sonorus
----------
In article <7uet3o$6bp$1...@baker.cc.tut.fi>, Mikko Helin <NOSPA...@uta.fi>
wrote:
>I guess it's only a question of time when someone starts making
>a CD player system which with it's built-in CD-ROM drive,
>decoder software and 24bit/96kHz multichannel converters
>is able to play back all kind of digital audio stored
>on CD's, ISO9660/Joliet and FAT16/32 file systems,
>.WAV files, .MP3's, and supports loading of OS/flashing ROM
>for updates.
You're basically describing Alesis' new Masterlink! Check out their
website!
Dave Martin
Marc Lindahl/Sonorus wrote in message <7ueg67$v2$1...@maceo.dti.net>...
>What, are you a Paris salesman? Or just an unbiased observer?
No, Brian's a happy Paris user.
Dave Martin
Marc Lindahl/Sonorus wrote in message <7uegcr$118$1...@maceo.dti.net>...
>>I think I need to buy some more books. Does the AES have a standards book,
>>or is it somewhere on the website if they're still putzing around with it?
>
>
>Yes, you can buy all that stuff from the AES (http://www.aes.org).
>Unfortunately -- I would gladly pay another $5/year if they'd just post all
>the stuff as pdf's on their website for free!
So would I; Most of the standards information will be useless to me - and
most of the rest I would need to know once or twice for an article or
review. And since AES keeps refining the standards, I gather whatever set of
standards papers I buy will be out of date quickly. A web page would
definitely be more valuable to me.
>
>>chairs one; if and when the AES agrees on a standard in some area of audio
>>engineering, do they ever actually print it?
>
>Yes, they print reams of them. And some are actually used and followed :)
Oh well; It looks like I'll be searching the AES site again with a credit
card handy...
Keith Sklower
Brian Tankersley <gbt...@home.com> wrote:
}Is that why I can hear (and object to) the SRC built into Sonic
}Solutions and consequently don't use it?
}Maybe the guy who wrote the paper should take another crack at his own
}algo. Right now, I'd say Sonic is not a good advertisement for
}transparent sounding SRC.
Brian,
*Which* SRC are you using? Sonic's comes in several flavors.
If you have the no frills - can't adjust - do it on the fly as
you read in from DAT at 16 bits, then I could see why you might have
an objection (although it's pretty clean for the level of stuff
I deal with).
If you opted for the more elaborate (read expensive) SRC for
the regular Sonic System (now called Sonic Studio), you
can set lots of different parameters, including the number of
taps on the filter, stuff dealing with the decimation filter,
and if you were willing to not get it in real time, you could
set the number of taps even higher, to be as about as good as
you could get with 56 bit fixed point arithmetic.
And, the SRC that was referred to in the paper was written using
double precision floating point arithmetic, and is currently
only avaible on the new "High Density" system. (It is unfortunately
not real time either. Gotta do a separate pass for now. Damn . . .)
The guy who wrote the paper is Andy Moorer (j...@sonic.com).
Has a Ph.D. from Stanford in computer music.
Caveat:
I'm not a sonic dealer. I don't own stock. I *do* own and use
one of the systems and don't want to see the company go belly up,
since I've got a maintenance contract. Also, the company has
behaved more than scruplously fairly to me and some of my friends.
Robert Orban
>
>Mark Plancke wrote:
>> Of course you overlook the fact that harmonics are created at mixdown
>> that have nothing to do with the fact the a microphone can't "hear"
>> these harmonics.
>
>This is true. I've never worked with one of these high-sampling-rate
>systems, nor am I an audio.pro; but I have spent some time doing
>computer synthesis (csound) and I have heard some pretty ugly
>aliasing at times... often in cases when there's no microphones
>involved _at all_.
>It is very easy to digitally generate extremely high frequencies
>when doing dsp. You can tell your dsp system to produce a sine wave
>of 500 kHz and it will happily try to comply. Anything above nyquist
>frequency shows up aliased to a lower frequency in your output.
>Working at 96 or 88.1 or whatever would help: doubling the sampling
>rate doubles the nyquist freq. so there's no aliasing until you go
>past freq. of 44 - 48 kHz, and it wouldn't even alias to anything
>under 20 kHz until you got way higher than that.
>That adds a lot of breathing room for digital processing like
>harmonizers, distortion, whatever. I would guess that this matters
>to people who write dsp code for plugins or for effects boxes...
>make their lives a little easier I'd think.
Yes, aliasing is a significant problem with non-linear processing in DSP, and
has to be considered when one is designing the algorithms. For example, we run
the clippers in our Optimods for AM and FM broadcast at 128kHz SR to minimize
these effects.
Luis Cypher
Alec?
>.have you or have you not heard Paris at 24 bits in
>controlled setting? I believe that was the question that prompted your
>smart-alec post.
>
>Marc Lindahl/Sonorus wrote:
>>
>> ----------
>> In article <3808F279...@home.com>, Brian Tankersley <gbt...@home.com>
>> wrote:
>> >
>> >So how many people 'round here have actually heard Paris in a controlled
>> <snip>
>> >years. They're retired in favor of the stock 24 bit cards you can buy
>> >for Paris.
>>
>> What, are you a Paris salesman? Or just an unbiased observer?
>>
Brian Tankersley
Contemporary Christian record and the #1 Country dance record last week
using Paris, for instance. I also talk nice about other gear I really
like, as in Focusrite, T.C., Tube Tech, etc, but I don't sell any of
that either.
What are you, a Sonorus salesman or something? Oh yeah, I guess you are
at that. I don't suppose that was you pimping your cards on several NGs
I've read was it?
I make zero if somebody buys equipment, so I guess that makes me a
little more objective than you, eh? As for reading
comprehension......have you or have you not heard Paris at 24 bits in
controlled setting? I believe that was the question that prompted your
smart-alec post.
Oh yeah, you sell 24/96 stuff, don't you? I guess you really wouldn't
want to take a chance on it sounding better.
Regards,
Brian T
Brian Tankersley
current Sonic SRC, since that was a while back.
I'll check it out a little more thoroughly next time I master. Thanks
for the info.
Regards,
Brian T
Keith Sklower wrote:
>
> In article <3807C296...@home.com>,
Brian Tankersley
Tough day at the office here. Rereading my earlier post on
rec.pro.audio, it looks a little harsh. Sorry about that. Seriously.
More water and less coffee for me.
Regards,
Brian T
Marc Lindahl/Sonorus wrote:
>
> ----------
> In article <3808F279...@home.com>, Brian Tankersley <gbt...@home.com>
Mikko Helin
"Alesis' revolutionary CD24T (patent pending) AIFF-compatible
technology." So it is not what I'd like to see, no more
patent stuff and such, just a regular PC in small case
with a CD-ROM, LCD display and some buttons. Just take
some Intel low-power Pentium CPU, their PCI chipset, some
RAM and certain PCI I/O chipset for multichannel
24/96 digital audio (yes, the IC Ensemble's Envy24),
and some serious converters. Main point is the price,
it shouldn't be only for pro's, but for all.
-Mikko
Marc Lindahl/Sonorus <ma...@sonorus.com> wrote:
> ----------
> In article <7uet3o$6bp$1...@baker.cc.tut.fi>, Mikko Helin <NOSPA...@uta.fi>
> wrote:
>>I guess it's only a question of time when someone starts making
>>a CD player system which with it's built-in CD-ROM drive,
>>decoder software and 24bit/96kHz multichannel converters
>>is able to play back all kind of digital audio stored
>>on CD's, ISO9660/Joliet and FAT16/32 file systems,
>>.WAV files, .MP3's, and supports loading of OS/flashing ROM
>>for updates.
> You're basically describing Alesis' new Masterlink! Check out their
> website!
Stephen St.Croix
<Flet...@mercenary.com> wrote:
> Stephen St.Croix wrote:
> >
> > In article <7u2asu$9fr$1...@grandprime.binc.net>, "Rich"
> > <ri...@attainment-inc.com> wrote:
> >
> > > I'm sure I'm missing something here but how do you take advantage of the
> > > higher sampling rate of 96khz when most mics don't go much beyond 20 khz
> > > which is under the effective frequency response of 24/48. I know the
> > > argument about upper harmonics etc. but if a mic can't pick them up what's
> > > the difference? Ok localization. But that's not that big of a deal in a
> > > multitrack situation where tracks are generally mono anyway and the mixer
> > > determines the stereo placement. Tel me how uninformed I am so I can
> > > justify the added expense of 24/96.
> > > Rich
> >
> > 24 bits good. Very good.
> >
> > 96 kHz totally stupid. Today's oversampled convertors no longer have those
> > old brickwalls, and 96 cannot be properly divided to get 44.1, what the
> > real world uses- 96 is useless.
> >
> > ________________
> > Stephen St.Croix
>
>
> What about 88.2?
Sorry, have not been here in a while. 88.2 makes complete sense to me, no
SRC artifacts, and all the advantages
________________
Stephen St.Croix
Stephen St.Croix
<ma...@sonorus.com> wrote:
> ----------
> In article <3808F279...@home.com>, Brian Tankersley <gbt...@home.com>
> wrote:
> >
> >So how many people 'round here have actually heard Paris in a controlled
> <snip>
> >years. They're retired in favor of the stock 24 bit cards you can buy
> >for Paris.
>
>
> What, are you a Paris salesman? Or just an unbiased observer?
>
>
> Marc Lindahl <ma...@sonorus.com>
> president, Sonorus Inc.
> http://www.sonorus.com
I am, in fact, one of the creators of PARIS, which is why I feel that I
can not jump in and say what Brian T just said. I agree with him totally,
but I hate it when manufacturers come in and argue standards questions
solely in an attempt to increase sales, so I post only what I have learned
and heard in a couple of years of engineering and producing (well, maybe
more than a couple).
I have voiced my feelings about the damage that SRC does to music in my
column and elsewhere. I would do ANYTHING to avoid gearboxing audio.
________________
Stephen St.Croix
Stephen St.Croix
mic...@ERASEmytekdigital.com wrote:
> >
> > 24 bits good. Very good.
> >
> > 96 kHz totally stupid. Today's oversampled convertors no longer have those
> > old brickwalls, and 96 cannot be properly divided to get 44.1, what the
> > real world uses- 96 is useless.
> >
> > ________________
> > Stephen St.Croix
>
> Steve, sorry, but you must have outdated information.
I kind of doubt that.
> There are SRC's that are very good in converting 96>44.1 such as Prism AD2 or
> new Sonic algorithm.
For 44.1 CDs 96 is useless. NO SRC algo or box, including the new Sonic
algorithm, sounds as good as not using SRC. The raw concept itself reveals
that. One cannot resample something and actually expect it to be as
accurate as the original.
> Some top mastering enginners reported getting a better
> final 1644 master off a 2496 material processed in 2496.
After SRC? There are always *some* people who will report anything you want.
>
>
> Michal @ Mytek
>
>
> *********************************************
> http://www.mytekdigital.com >>>>>>
>
> 24 bit 48/96 k Digital Audio Converters
>
> DAW 9624 (tm) , D-Master 9624 (tm)
>
> *********************************************
________________
Stephen St.Croix
Peter Bruce Wilder
I have to be on the side of "NO to SRC" on this one. I had a great digital audio
engineer / programmer explain this to me in great detail a few weeks ago. What it
comes down to, and this is glossing over all the heavy formulas, is the "rounding
off" necessary to bring 96kHz to 44.1. The math is significant, and it just isn't
as elegant as halving it from 88.2.
My two cents - and the dig engineer works very high up in R&D for a major media -
theme park corporation plus has more than a few fingerprints on several milestone
products that have defined the course of the music industry. Unfortunately, as I
don't have his permission, he will have to remain nameless.
warm regards to all. May we all peacefully coexist, seeking the same thing:
accuracy.
- Peter
Peter Bruce Wilder
dba Ergo Communications
http://www.together.net/~pbwilder/homepage.htm
James Perrett
compressor and CD writer in a box. It had analogue and digital inputs
and basic editing functions. Seemed to do a similar job to the Alesis
box. I can't remember the price or the name of the company at the
moment.
James.
Mikko Helin wrote:
>
> It's better than M3Po, but unfortunately utilize only
> "Alesis' revolutionary CD24T (patent pending) AIFF-compatible
> technology." So it is not what I'd like to see, no more
> patent stuff and such, just a regular PC in small case
> with a CD-ROM, LCD display and some buttons. Just take
> some Intel low-power Pentium CPU, their PCI chipset, some
> RAM and certain PCI I/O chipset for multichannel
> 24/96 digital audio (yes, the IC Ensemble's Envy24),
> and some serious converters. Main point is the price,
> it shouldn't be only for pro's, but for all.
>
> -Mikko
>
> Marc Lindahl/Sonorus <ma...@sonorus.com> wrote:
>
> > ----------
> > In article <7uet3o$6bp$1...@baker.cc.tut.fi>, Mikko Helin <NOSPA...@uta.fi>
> > wrote:
>
> >>I guess it's only a question of time when someone starts making
> >>a CD player system which with it's built-in CD-ROM drive,
> >>decoder software and 24bit/96kHz multichannel converters
> >>is able to play back all kind of digital audio stored
> >>on CD's, ISO9660/Joliet and FAT16/32 file systems,
> >>.WAV files, .MP3's, and supports loading of OS/flashing ROM
> >>for updates.
>
> > You're basically describing Alesis' new Masterlink! Check out their
> > website!
>
Marc Lindahl/Sonorus
----------
In article <pr-191099...@ip201.laurel4.md.pub-ip.psi.net>,
p...@nospam.intdevices.com (Stephen St.Croix) wrote:
>
>For 44.1 CDs 96 is useless. NO SRC algo or box, including the new Sonic
>algorithm, sounds as good as not using SRC. The raw concept itself reveals
>that. One cannot resample something and actually expect it to be as
>accurate as the original.
This is not true. There is no theoretical difference between 88.2-)44.1 and
96->44.1 sample rate conversion, it's just an implementation problem dealing
with a different fraction. And, all modern (read, sigma-delta) converters
do sample-rate conversion, from the nominal 6.144MHz capture rate.
Fortunately, some real math brains are finally applying themselves to the
audio domain, and we're starting to get the theoretical groundwork for what
may be called 'lossless signal processing'.
I think the situation is similar to a few years ago -- who would have
thought you could run a small-to-medium sized DAW entirely on the CPU of a
desktop computer?? Or record hours of audio on a single hard drive? We're
used to the limitations of previous generations of signal processing.
Anything digital can be lossless. This is because when it comes down to it,
you have a finite representations, and once something is captured in that
representation, any operation you do on it will still be finite. This of
course leaves open the question of the quality of the capturing and
subsequent reproduction stages...
The basic example is adding 9 + 9 = 18. You need an extra digit to
represent the sum, but you're not losing anything -- it's exact. This
concept extends to the realm of signal processing. If the size of the
storage given to the numbers is big enough, then the numbers can be exact.
So now, it becomes a question of processing speed and memory size. Two
things which have been exponentially increasing for the last 20 years or so.
Therefore, it's a matter of time (and effort) to achieve these goals.
Then we will see an SRC with ideal properties -- which I would imagine would
be defined as 'limiting the signal bandwidth, with no other side effects',
though who knows, you may want a bit of 'tube warmth' or 'head bump'
simulated in there for good measure :)
Bob Smith
>
> Anything digital can be lossless. This is because when it comes down to it,
> you have a finite representations, and once something is captured in that
> representation, any operation you do on it will still be finite. This of
> course leaves open the question of the quality of the capturing and
> subsequent reproduction stages...
>
> The basic example is adding 9 + 9 = 18. You need an extra digit to
> represent the sum, but you're not losing anything -- it's exact. This
> concept extends to the realm of signal processing. If the size of the
> storage given to the numbers is big enough, then the numbers can be exact.
> So now, it becomes a question of processing speed and memory size. Two
> things which have been exponentially increasing for the last 20 years or so.
> Therefore, it's a matter of time (and effort) to achieve these goals.
>
Just a slight nitpick here, but could you give the exact representation
of x/3 when x is some silly value like 10? In binary 1010/11? Then
multiply that result back by 3 (11 binary) and see if the result matches
the input. Things have improved greatly, but there still are limitations
and compromises. Picking the best tradeoffs is the trick to giving good
results.
bobs
Amateur Recordist
--
Bob Smith - BS Studios
rsm...@bsstudios.com
http://www.bsstudios.com
Stephen St.Croix
<ma...@sonorus.com> wrote:
> ----------
> In article <pr-191099...@ip201.laurel4.md.pub-ip.psi.net>,
> p...@nospam.intdevices.com (Stephen St.Croix) wrote:
> >
> >For 44.1 CDs 96 is useless. NO SRC algo or box, including the new Sonic
> >algorithm, sounds as good as not using SRC. The raw concept itself reveals
> >that. One cannot resample something and actually expect it to be as
> >accurate as the original.
>
> This is not true. There is no theoretical difference between 88.2-)44.1 and
> 96->44.1 sample rate conversion, it's just an implementation problem dealing
> with a different fraction.
> snip
> Marc Lindahl <ma...@sonorus.com>
> president, Sonorus Inc.
> http://www.sonorus.com
I dont know you from Adam, and you may be a great guy, but this is
bullshit. There can be no larger difference, theoretical and actual than
the difference in a perfect division by two and ANY other divisor. You
know that.
I am having trouble with your using this thread to defend your own
products, so this will be my last response. I trust that you know who I
am, my history, and my research. I am very vocal in my column about 96 and
what an unfortunate mistake this all is
________________
Stephen St.Croix
Hot Jamms
96 is great for EQ otherwise I find that
24/48 is very good.
Hot Jamms
Shaun
> The reason no bigtime DAW is at 96K yet is simple. Your 64 track system
> just became a 32 track system. It's an HD subsystem issue more than an
> electronics issue, I'm sure. There's no way Digi can get 64 tracks of
> reliable throughput from any affordable HD setup at 96K, and won't be
> for a while.
> <snip>
Now take that back! I played back 64 tracks of 24/48 on ProTools 4.3.1 from ONE
Cheetah 9 gig LVD drive and it's (activity light) duty cycle was about 50%, using
a #2 DAE buffer on a 9600/450 G3.
--
Shaun Wexler,
Hellsgate Studios
Robert Orban
p...@nospam.intdevices.com says...
ANY sample rate conversion (including 2x), whether up or down, creates
non-linear artifacts. If the SRC is upward, then you create images and
then have to use the digital equivalent of a reconstruction filter after
the conversion to remove the images. If the SRC is downward, you have to
use the digital equivalent of an anti-aliasing filter _before_ the
conversion. 2x up and 2x down are no different in this regard. However,
they offer a significant advantage in that you can use "polyphase
half-band filters" to do the anti-aliasing or reconstruction. These
filters are approximately twice as computationally efficient as filters
needed for non 2x conversions, so, for a given amount of DSP power, you
can achieve higher quality conversion. Also, if you play your cards right,
you are likely to retain some of the original samples in the converted
output (although this is not really as advantageous as it appears).
Given enough DSP processing power, you can do arbitary SRC to any accuracy
desired. This really means that you can get the passband as flat as you
want and remove the images or aliases to whatever degree you desire. For a
given amount of DSP power, you will get better results with 2x up or 2x
down. But, other than cost, 2x has no _intrinsic_ advantages.
hank alrich
Neat. Now try it at 24/96 and tell us what you get. <g> "Cause that's
what Brian said. Nevermind the activity light.
--
hank - secret mountain
Note: the rec.audio.pro FAQ is at http://recordist.com/rap-faq/current
Read it and reap!
r_sach
--
| Robert Sach NYU
| "C:/DOS, C:/DOS/RUN, RUN/DOS/RUN"
-----------------------------------------------------------------:-)
Brian Tankersley
was uninformed. Thanks for the info......I take it back.
Regards,
Brian T
Shaun wrote:
>
> Brian Tankersley wrote:
>
> > The reason no bigtime DAW is at 96K yet is simple. Your 64 track system
> > just became a 32 track system. It's an HD subsystem issue more than an
> > electronics issue, I'm sure. There's no way Digi can get 64 tracks of
> > reliable throughput from any affordable HD setup at 96K, and won't be
> > for a while.
>
> > <snip>
>
> Now take that back! I played back 64 tracks of 24/48 on ProTools 4.3.1 from ONE
> Cheetah 9 gig LVD drive and it's (activity light) duty cycle was about 50%, using
> a #2 DAE buffer on a 9600/450 G3.
Geoff Wood
scientifically proven.
--
Geoff Wood
ge...@paf.co.nz
r_sach <r_s...@nyu.edu> wrote in message
news:r_sach-2610...@osh1-21.yesic.com...
Will Gallant
Particularly to those who, years ago before digital was ubiquitous, gave
serious thought as to why their analog mixes sounded different than a
digital rendering of the same.
Willg...
Geoff Wood wrote in message ...
R Krizman
they think--or would that blow all the idle speculation <g> ?
Rick Krizman
KrizManic Music,
Venice, CA
r_sach
rkri...@aol.com (R Krizman) wrote:
>Is it too much to ask if anyone has actually heard 24/96, and if so, what did
>they think--or would that blow all the idle speculation <g> ?
We have a Sonic HDSP system. The difference between 24/48 and 24/96 can
even be heard by "layman" ears.
hank alrich
> Is it too much to ask if anyone has actually heard 24/96, and if so, what did
> they think--or would that blow all the idle speculation <g> ?
Well, I was gonna try an' lissen to it, but the charming salesagent at
GC told me that I'd need speakers twice as big as the ones I now have
just to fit all the extra sound in there and I couldn't convince Lanis
that such a purchase was sensible, even in pursuit of higher technology.
But hey, what does _she_ know?
Stephen St.Croix
rkri...@aol.com (R Krizman) wrote:
> Is it too much to ask if anyone has actually heard 24/96, and if so, what did
> they think--or would that blow all the idle speculation <g> ?
>
> Rick Krizman
> KrizManic Music,
> Venice, CA
My opinions on the subject are based only on what I have heard myself on
far too many occasions.
________________
Stephen St.Croix
Dale Goudey
since some will hear an improvement and some will not. It is also of
interest to take a survey and see what the others think, since this
will indicate the market demand for the new formats.
My vote: I experience a VAST increase in listening enjoyment with
24/96.
Criticizing 24/96 as being worthless is not appropriate if you haven't
taken the trouble to give it a listen.
>R Krizman wrote in message
R Krizman
>Well, I was gonna try an' lissen to it, but the charming salesagent at
>GC told me that I'd need speakers twice as big as the ones I now have
Not to mention the requisite ear surgery .
R Krizman
>My opinions on the subject are based only on what I have heard myself on
>far too many occasions.
Specifically what have you heard and what did you think.?
I heard the new DVD-A player at AES playing 24/96 5.1 stuff through the new M&K
powered surround system and was favorably impressed. It seemed like the medium
itself was out of the loop and the music sounded as good or bad as it actually
was. Whether there was a verifiable difference big enough to warrant a whole
change in my technology, I couldn't say.
Stephen St.Croix
rkri...@aol.com (R Krizman) wrote:
> p...@nospam.intdevices.com (Stephen St.Croix) wrote:
>
> >My opinions on the subject are based only on what I have heard myself on
> >far too many occasions.
>
> Specifically what have you heard and what did you think.?
snip
>
> Rick Krizman
> KrizManic Music,
> Venice, CA
I listen to everything made and everything I can get to that is being
developed, over and over. I have to, for my column, for PARIS, and for my
studio
24 bit is very impressive, but the difference in 96k converted to 44.1 and
a good high rate delta-sigma printing 44.1 is questionable, at best
________________
Stephen St.Croix
R Krizman
>I listen to everything made and everything I can get to that is being
>developed, over and over. I have to, for my column, for PARIS, and for my
>studio
>24 bit is very impressive, but the difference in 96k converted to 44.1 >and a
good high rate delta-sigma printing 44.1 is questionable, at best
No doubt, but with the advent of DVD-A why does 96k need to be converted down?
I think the question is whether 96k sounds "better-enough" to warrant the
technology upgrade, from studio to consumer format. What do you think of the
sound of 96k per se?
Arny Krüger
"Dale Goudey" <gou...@earthlink.net> wrote in message
news:7v71ce$938$1...@holly.prod.itd.earthlink.net...
> The question itself should really be answered by each individual,
> since some will hear an improvement and some will not.
I agree! Therefore I am working to make some sample data files for
www.pcabx.com that are the identical same program material recorded
at 24/96 with my CardD Deluxe and some measurement mics, and then
sampled down to 16/44.
Meanwhile, folks can amuse themselves by trying to hear the effects
of cutoffs at 5, 9, 12, 15 and 18 KHz, in anticipation of what it
would be like to raise the high end limit from 22 KHz to 48 KHz.
They can likewise try out 8-15 bits versus 16, in anticipation of
24.
Stephen St.Croix
rkri...@aol.com (R Krizman) wrote:
It is the SRC that I object to, and considering that DVD-A is not much of
an advent, 44.1 is still the final target for 99 percent of all projects
done today.
Going from 16 to 24 bit offers an incredible increase in audio quality
(the math reveals why) with a small incease in DSP overhead and disk
space.
Going from 44.1 to 96 kHz offers a truly minor quality increase as long as
it stays at 96 forever. But it more than doubles the space used on disk,
and taxes the data throughput capabilities of any computer. Any given
system can do less than half as much DSP. And there is NO real world
increase in dynamic range or amplitude resolution.
And thats IF you stay 96.
96 is a gift from manufacturers so they can sell you all new gear... again.
________________
Stephen St.Croix
Brian Tankersley
files for digital audio. That's not a temporary trend....it's
inevitable.
So, how many consumers will want to duplicate and convert their entire
music collection from their audio only DVD down to at least 44.1/16 so
that they can hear it in their car, Walkman, Rio, etc? That's
right....not many. They *might* put a DVD-A in their car, but I doubt
it.
And if they are desiring to listen to it on a variety of portable
formats, including all the above, how many of those formats would have a
snowballs chance in hell of revealing any audio differences we're
talking about? None, or nearly none. You guys really think the
electronics/headphones on a Walkman or boombox care about 44.1 vs 96?
Come on, get real. There is **vastly** more difference in sound from
model to model in the store than their is from 44.1 to 96 on the *same*
model. Explain to the buyer why 96K is vitally important to him. Good
luck.
Lastly, are we so out of touch that we don't know who buys this stuff we
make? Do you think old farts with the disposable income to max out a 5.1
home theatre and collect audio only DVDs to listen in that one spot will
ever buy enough CDs to support the music biz? Not. Young buyers support
the music biz, always have and always will.
Get in an online video gaming chat room, like Kali, and start talking
with the 15 year olds about 96 vs 44.1 and DVD-A like it matters.
Prepare for some rude language back. Then start talking about the
*music* itself.....current bands, and get ready for plenty of interest
and passion about the *music*. That sums the whole discussion up to me.
96K is an answer in desperate search of a valid question. Close to
NOBODY currently buying music as a consumer, is complaining saying, "You
know, I just can't listen to this music because the phase smearing from
that 22K brick wall filter is ruining the whole experience for me". Yup,
I hear that from teenagers daily, don't you? That's your customer,
ultimately, whether you ever meet them or not.
Regards,
Brian T
Mark Deffebach
Crystal and AKM are continuously moving forward. 96k is hear to stay.
48k converters are old products with little support. However, most 96k
converters can be run a 48k as well 96k and still retain the better SNR
and dynamic range specs.
If I can design a system with 120dB SNR and lower THD by using a 96k
converter versus a 100dB SNR and higher THD using a 48k converter,
which do you want?
What if I use the 96k converters and offer 44.1k, 48k and 96k, all with
better specs than older 44.1k/48k gear.
P.S. My designs end up in fixed installation sound systems, but the
trend is to meet the higher specs expected within the recording
industry.
Sent via Deja.com http://www.deja.com/
Before you buy.
gearmaster
SHIT. For the same reasons, I would rather use 88.2 than 96. It has
been said in this thread that there is no difference in SRC from either
88.2 or 96 to 44.1 I find this hard to believe. here's my idea for a
quick conversion algorithim from 88.2 to 44.1:(written in java)
int count=0;
new AudioFileBuffer new441 = new AudioFile(44100);
public AudioFile get441(AudioFile old882)
{
while(old882.getSample(count) != null)
{
if (count%2 == 0)
{
new441.append(old882.getSample(count));
count++
}
}
return new441.toAudioFile();
}
Which basically grabs every other sample out of the 88.2 file and puts
it in the 44.1 file. If you don't like that approach, how about:
int count=0;
new AudioFileBuffer new441 = new AudioFile(44100);
public AudioFile get441(AudioFile old882)
{
while(old882.getSample(count) != null)
{
if (count%2 == 0)
{
new441.append((old882.getSample(count) + old882.getSample
(count+1))/2);
count++
}
}
return new441.toAudioFile();
}
which returns the average of 2 adjacent samples in the original 88.2
file as a single sample in the output 44.1 file.
Using either of these techniques would be better than having to create
a guess as to what the audio signal "might" have been doing at those
several hundreds of miliseconds per second where there was no overlap
in the sample timing between 98 and 44.1
I have no background in DSP algorithims, only in math and music and
computers. Please don't bash me too hard if both of these algorithims
have been previously shown to suck badly.
Rick Krizman
Brian Tankersley wrote:
>
>
> >So, how many consumers will want to duplicate and convert their entire
> >music collection from their audio only DVD down to at least 44.1/16 so
> >that they can hear it in their car, Walkman, Rio, etc? That's
> >right....not many. They *might* put a DVD-A in their car, but I doubt
> >it.
>
I was under the impression that a DVD-A disk would play back on a conventional
CD player--that the disk would contain a 16/44.1 file that the player could
read. Perhaps I'm mistaken. I agree, nobody's going to do that whole "convert
my collection" thing that they did when CD's came out.
>
> >96K is an answer in desperate search of a valid question. Close to
> >NOBODY currently buying music as a consumer, is complaining saying, "You
> >know, I just can't listen to this music because the phase smearing from
> >that 22K brick wall filter is ruining the whole experience for me". Yup,
> >I hear that from teenagers daily, don't you? That's your customer,
> >ultimately, whether you ever meet them or not.
>
This type of argument constantly comes up whenever we discuss quality issues
about which consumers have no knowledge. But you know, in the course of your
day you do all sorts of things to the sound of a recording--change reverb patch
on snare, notch out 400 hz on a guitar, maximize your levels-to-disk, etc--and
no teenager cares about that either, but you do it anyway, if for no other
reason than that you know that all these things you do to audio will add up to a
more satisfying listening iexperience for you, your client, and hopefully the
unwashed masses, whether they know it or not. It's not really the consumer's
job to think about all this, so the fact that they don't comment on the "phase
smear from the 22K brickwall filter" has absolutely no bearing on this
discussion.
I believe that DVD's, if only because of their greater efficiency, will
eventually replace CD's. CD's will continue to play on DVD players, but the new
format will not necessitate the same restrictions on new product. The newer
media will accomodate surround, video, higher sample rates in any number of
variable proportions. Whether or not 96K sounds better to the people who are
making records will be a factor in deciding how much of the increased bandwidth
of DVD's will be devoted to it.
>
Jonny Rash
All this technology talk can be crippling to the creative
process, getting me all flustered about whether I have the
right gear, right settings, right computer, right cables, right
whatever. Who Cares! I want to make great music and write
good songs!
Brian Tankersley <gbt...@home.com> wrote in message
news:38186C11...@home.com...
> Absolutely right on. Consumers are heading for smaller, not larger, data
> files for digital audio. That's not a temporary trend....it's
> inevitable.
>
Fletcher
>
>
> It is the SRC that I object to, and considering that DVD-A is not much of
> an advent, 44.1 is still the final target for 99 percent of all projects
> done today.
Having less than no clue what an SRC is...I can't comment, but yeah,
44.1 is still the fianl target for 99 percent of all projects done
today...except...when 96k is all the rage and everyone will have to go
restock their entire back catalog...again. So as long as the now 40
somethings, going on 50 somethings have disposable income for DVD
players...there will be product made for those players...so the labels
can re-release more and more Bachman Turner Overdrive records, and the
oldies stations can still be called "Classic Hits".
>
> Going from 16 to 24 bit offers an incredible increase in audio quality
> (the math reveals why) with a small incease in DSP overhead and disk
> space.
>
> Going from 44.1 to 96 kHz offers a truly minor quality increase as long as
> it stays at 96 forever. But it more than doubles the space used on disk,
> and taxes the data throughput capabilities of any computer. Any given
> system can do less than half as much DSP. And there is NO real world
> increase in dynamic range or amplitude resolution.
Here's where you got me...I'm a moron with this stuff, so go slowly...I
didn't think 96k was supposed to offer more dynamic range. Hell, I
own/use enough compression that anything greater than 6db of dynamic
range is a waste of my time...however, I have this incredible desire to
actually hear a Hi-Hat that sounds like a Hi-Hat, a bass that actually
sounds like a bass...won't more dots in the picture help me with this?
As for the disc space thing...I thought I read that some guy in Japan
just invented a blue laser...now blue, being a 1/4 the wavelength of
red, will give you 4 times more storage capability on the same
disc...won't it?
>
> And thats IF you stay 96.
>
> 96 is a gift from manufacturers so they can sell you all new gear... again.
>
> ________________
> Stephen St.Croix
And don't forget back catalog...it's really the key to sucess...think
about it, when was the last time Motown had something really big in the
charts...but they sell a whole lot of back catalog now don't they.
--
Fletcher
Mercenary Audio
TEL: 508-543-0069
FAX: 508-543-9670
http://www.mercenary.com
Rick Krizman
Jonny Rash wrote:
> Thanks Brian for putting it all in the proper perspective.
> All this technology talk can be crippling to the creative
> process, getting me all flustered about whether I have the
> right gear, right settings, right computer, right cables, right
> whatever. Who Cares! I want to make great music and write
> good songs!
>
By all means! You don't need 24/96 to make and record great music. You
don't even need digital recording at all. You don't need expensive Neve
boards and Studer machines either, if you are of the analog persuasion. But
nonetheless, there are people in this world, in this group, who are committed
to sonic excellence as their vocation. There are people who quite
legitimately opt to spend the money to record on a Studer/Neve rather than
1/2" eight track. There are people who buy Lexicon 480's instead of 300's,
and so on. Like it or not, all this "technology talk" comprises a great
portion of what this group is all about. Therefore, addressing the question
of higher sample rates is certainly worth discussing here. Nothing regarding
this right now is a foregone conclusion, including the question of whether
96k will even be the eventual upper limit for sample rates.
Rick Krizman
KrizManic Music
Steve Jackson
Fletcher's post (thanks Fletcher).
Forgive me if I am wrong, but from what I remember of digital signal processing
(Fourier transforms and all that rot) from my school days, the motivation for
going to a higher sampling rate is that you can use a lowpass filter with a more
gentle slope (less phase distortion) to recreate the sampled waveform. This
assumes that you are sampling above the Nyquist frequency, to avoid aliasing.
For the 20kHz bandwidth audio spectrum, this would put the Nyquist frequency at
40 kHz. So what am I missing here? Why all the buzz about (what in Hell is the
motivation for) going to 96kHz from 44.1 or 48 kHz? What am I missing??? Did I
say that already?
Steve
p.s. I don't know what SRC is either. Educate me please.
p.p.s. G.D.I.H.A.
Will Gallant
Willg...
Steve Jackson wrote in message
Stephen St.Croix
<s...@hpessj.fc.hp.com> wrote:
> I couldn't find the original post to reply to, so I'll just jump in here on
> Fletcher's post (thanks Fletcher).
>
> Forgive me if I am wrong, but from what I remember of digital signal
processing
> (Fourier transforms and all that rot) from my school days, the motivation for
> going to a higher sampling rate is that you can use a lowpass filter
with a more
> gentle slope (less phase distortion) to recreate the sampled waveform. This
> assumes that you are sampling above the Nyquist frequency, to avoid aliasing.
> For the 20kHz bandwidth audio spectrum, this would put the Nyquist
frequency at
> 40 kHz. So what am I missing here? Why all the buzz about (what in
Hell is the
> motivation for) going to 96kHz from 44.1 or 48 kHz? What am I
missing??? Did I
> say that already?
>
> Steve
>
> p.s. I don't know what SRC is either. Educate me please.
>
> p.p.s. G.D.I.H.A.
SRC= sample rate conversion
>
> Fletcher wrote:
>
> > Stephen St.Croix wrote:
> > >
> > >
> > > It is the SRC that I object to, and considering that DVD-A is not much of
> > > an advent, 44.1 is still the final target for 99 percent of all projects
> > > done today.
> >
> > Having less than no clue what an SRC is...I can't comment, but yeah,
> > 44.1 is still the fianl target for 99 percent of all projects done
> > today...except...when 96k is all the rage and everyone will have to go
> > restock their entire back catalog...again. So as long as the now 40
> > somethings, going on 50 somethings have disposable income for DVD
> > players...there will be product made for those players...so the labels
> > can re-release more and more Bachman Turner Overdrive records, and the
> > oldies stations can still be called "Classic Hits".
> >
> > >
> > > Going from 16 to 24 bit offers an incredible increase in audio quality
> > > (the math reveals why) with a small incease in DSP overhead and disk
> > > space.
> > >
> > > Going from 44.1 to 96 kHz offers a truly minor quality increase as long as
> > > it stays at 96 forever. But it more than doubles the space used on disk,
> > > and taxes the data throughput capabilities of any computer. Any given
> > > system can do less than half as much DSP. And there is NO real world
> > > increase in dynamic range or amplitude resolution.
> >
> > Here's where you got me...I'm a moron with this stuff,
Yeah, right
> > so go slowly...I
> > didn't think 96k was supposed to offer more dynamic range. Hell, I
> > own/use enough compression that anything greater than 6db of dynamic
> > range is a waste of my time...however, I have this incredible desire to
> > actually hear a Hi-Hat that sounds like a Hi-Hat, a bass that actually
> > sounds like a bass...won't more dots in the picture help me with this?
> >
96k is not supposed to offer more TDR, but it is yet another myth that is
just starting to appear based on a lame theory that more dots resolves
twice as much noise or the same noise at half the level. Not true at all.
As for more dots helping in general, yes they certainly will, but... there
is NO amplitude resolution increase at all, and when you consider that
just one bit increase doubles the amplitude resolution, and when you
consider that amplitude resolution limitation is what we hear most in
modern systems, and...
96k more than doubles the data storage demands yet yields only a minor
improvement (much less than even one bit more amplitude res), and...
that improvement is lost once you SCR TO 44.1, whats the point?
Use 88.2 instead, if you simply must use twice the disk space. Then at
least the SRC makes more sense
> > As for the disc space thing...I thought I read that some guy in Japan
> > just invented a blue laser...now blue, being a 1/4 the wavelength of
> > red, will give you 4 times more storage capability on the same
> > disc...won't it?
> >
Actually, the Japanese have been having a lot of trouble getting a
non-liquid blue to be reliably manufacturable. A LOT of trouble. But four
years ago a 24 year old kid in the USA (!) did get one working, and it was
manufacturable as well. But it never appeared, and three companies that I
consult for that were going to make blue light recorders have become
mysteriously quiet about it and no products have appeared. But this is not
my disc point- I was talking about the original DAWs hard disk and the bus
bandwidth.
> > >
> > > And thats IF you stay 96.
> > >
> > > 96 is a gift from manufacturers so they can sell you all new gear...
again.
> > >
> > > ________________
> > > Stephen St.Croix
> >
> > And don't forget back catalog...it's really the key to sucess...think
> > about it, when was the last time Motown had something really big in the
> > charts...but they sell a whole lot of back catalog now don't they.
> > --
> > Fletcher
> > Mercenary Audio
> > TEL: 508-543-0069
> > FAX: 508-543-9670
> > http://www.mercenary.com
Fletcher- have you ever snapped a throttle cable? I did last week- no fun.
________________
Stephen St.Croix
Les Cargill
Steve Jackson wrote:
>
> I couldn't find the original post to reply to, so I'll just jump in here on
> Fletcher's post (thanks Fletcher).
>
> Forgive me if I am wrong, but from what I remember of digital signal processing
> (Fourier transforms and all that rot) from my school days, the motivation for
> going to a higher sampling rate is that you can use a lowpass filter with a more
> gentle slope (less phase distortion) to recreate the sampled waveform. This
> assumes that you are sampling above the Nyquist frequency, to avoid aliasing.
> For the 20kHz bandwidth audio spectrum, this would put the Nyquist frequency at
> 40 kHz. So what am I missing here? Why all the buzz about (what in Hell is the
> motivation for) going to 96kHz from 44.1 or 48 kHz? What am I missing??? Did I
> say that already?
It's a psychoacoustics question. Is there useful information in the octave
between
20Khz and 40 Khz? Some say there is, some say there ain't. Some say the
additional
bit depth of 24 bits is all you need, some say the increase in sampling rate
helps. And
unfortunately, all we can listen to is implementations, which may or may not
bear
out the theory.
>
> Steve
>
> p.s. I don't know what SRC is either. Educate me please.
>
> p.p.s. G.D.I.H.A.
>
> Fletcher wrote:
>
> > Stephen St.Croix wrote:
> > >
> > >
> > > It is the SRC that I object to, and considering that DVD-A is not much of
> > > an advent, 44.1 is still the final target for 99 percent of all projects
> > > done today.
> >
> > Having less than no clue what an SRC is...I can't comment, but yeah,
> > 44.1 is still the fianl target for 99 percent of all projects done
> > today...except...when 96k is all the rage and everyone will have to go
> > restock their entire back catalog...again. So as long as the now 40
> > somethings, going on 50 somethings have disposable income for DVD
> > players...there will be product made for those players...so the labels
> > can re-release more and more Bachman Turner Overdrive records, and the
> > oldies stations can still be called "Classic Hits".
> >
> > >
> > > Going from 16 to 24 bit offers an incredible increase in audio quality
> > > (the math reveals why) with a small incease in DSP overhead and disk
> > > space.
> > >
> > > Going from 44.1 to 96 kHz offers a truly minor quality increase as long as
> > > it stays at 96 forever. But it more than doubles the space used on disk,
> > > and taxes the data throughput capabilities of any computer. Any given
> > > system can do less than half as much DSP. And there is NO real world
> > > increase in dynamic range or amplitude resolution.
> >
> > Here's where you got me...I'm a moron with this stuff, so go slowly...I
> > didn't think 96k was supposed to offer more dynamic range. Hell, I
> > own/use enough compression that anything greater than 6db of dynamic
> > range is a waste of my time...however, I have this incredible desire to
> > actually hear a Hi-Hat that sounds like a Hi-Hat, a bass that actually
> > sounds like a bass...won't more dots in the picture help me with this?
> >
> > As for the disc space thing...I thought I read that some guy in Japan
> > just invented a blue laser...now blue, being a 1/4 the wavelength of
> > red, will give you 4 times more storage capability on the same
> > disc...won't it?
> >
> > >
> > > And thats IF you stay 96.
> > >
> > > 96 is a gift from manufacturers so they can sell you all new gear... again.
> > >
> > > ________________
> > > Stephen St.Croix
> >
> > And don't forget back catalog...it's really the key to sucess...think
> > about it, when was the last time Motown had something really big in the
> > charts...but they sell a whole lot of back catalog now don't they.
> > --
> > Fletcher
> > Mercenary Audio
> > TEL: 508-543-0069
> > FAX: 508-543-9670
> > http://www.mercenary.com
--
Les Cargill
http://home.att.net/~lcargill/
Les Cargill
Fletcher wrote:
<snip>
> And don't forget back catalog...it's really the key to sucess...think
> about it, when was the last time Motown had something really big in the
> charts...but they sell a whole lot of back catalog now don't they.
Not as much as they used to. The back catalog market is slowing
down, last I heard. Only took 17 years... so maybe it's time
to do it again. Worked last time, right?
The whole A&R paradigm now is to have indie back catalog before somebody
even makes it to the majors.
It's tough for new bands to compete with Smokey Robinson for shelf space.
I wonder what the correlation with CD coming out in 1983 and the gradual
disappearance of new, good bands is? Will DVD make that worse? I'm no
Luddite, but there *seems* to have been an effect.
Except for the 1991 Seattle thing, the whole industry seems to have been
declining
and consolidating steadily since about 1980, when the Baby Boomers got into
real estate. IMO, anyway.
Stephen St.Croix
<tcd...@chilli.net.au> wrote:
> Stephen St.Croix wrote in message ...
> >In article <19991027204014...@ng-fi1.aol.com>,
>
>
> >
> >It is the SRC that I object to, and considering that DVD-A
> is not much of
> >an advent, 44.1 is still the final target for 99 percent of
> all projects
> >done today.
> >
> >Going from 16 to 24 bit offers an incredible increase in
> audio quality
> >(the math reveals why) with a small incease in DSP overhead
> and disk
> >space.
> >
> >Going from 44.1 to 96 kHz offers a truly minor quality
> increase as long as
> >it stays at 96 forever. But it more than doubles the space
> used on disk,
> >and taxes the data throughput capabilities of any computer.
> Any given
> >system can do less than half as much DSP. And there is NO
> real world
> >increase in dynamic range or amplitude resolution.
>
>
> I disagree. 96k represents a good step forward in the type
> of HF response acheivable with digital whilst maintaining
> transparent
> LPF's on the best DAC chips.
> Maybe with typical pro quality gear and cheap bitstream DACs
> (90% use them) you are right, but with more expensive
> multibit
> DACs and mastering / hi end quality build the difference is
> very real and 192 would be even better.
>
> No 44.1 sample dig system gets the top end *really* right,
> and is testament to the millions of CD's that sound like
> trash.
>
> Go to a live unamplified symphony or choral performance
> and listen to the HF presentation, even when it gets really
> loud its natural and not edgy or forced. Go back and listen
> to any digital system... the 96k will come much closer.
>
> Regards,
>
> Terry Demol
I disagree with your disagreement. I am pretty new to this NG, but isnt
that how its done here?
Those CDs sound like trash because 90 percent of them are around 12 to 14 bits.
And 44.1 is what CDs are. I would rather digitize directly to 44.1 with a
delt-sigma than do it at 96 and SRC the final mix to 44.1. Transparent
LPFs are not the issue and have not been for years, with oversampled
delta-sigma covertors using very, very mild LPFs with excellent phase
response and in-band linearity (much milder than the filter you would have
to use to clear 96!)
So...
With all other variables equal- recording at 96 and listening at 96 would
sound better than recording at 44.1 and listening at 44.1, BUT the world
currently listens at 44.1, like it or not, and any 96 gains are more than
lost during SRC to 44.1. And 44.1 17 bit will blow the doors off 96 16
bit.
My point is that there is much better use of hard disk and CD space than
96k, like increasing the word length by another bit or eight.
Regards back,
________________
Stephen St.Croix
Jerry Gerber
--
Sincerely,
>
> By all means! You don't need 24/96 to make and record great music. You
>don't even need digital recording at all. You don't need expensive Neve
>boards and Studer machines either, if you are of the analog persuasion.
But
>nonetheless, there are people in this world, in this group, who are
committed
>legitimately opt to spend the money to record on
>to sonic excellence as their vocation.
>Rick Krizman
>KrizManic Music
>
>
I suppose that the higher sample rate will allow for a potential smoothing
and naturalizing of the high frequency response if the filter is designed
well. This will give the music producer new options in regard to how to
orchestrate if there is a greater smoothness and roundness of the high
notes. We cannot hear beyond 18-20khz, but I imagine we can hear the effect
of the sum total of the harmonics at superaural levels on the overall
"aesthetic sense" of the high-end. So what is this thing about SRC from
96khz to 44.1? Is this a problem with some software and hardware but not
others? Is it an issue?
Jerry Gerber
http://www.slip.net/~jgerber/
jge...@slip.net
Ken Kantor
Stephen,
If you have back issues of AUDIO available, Bob Katz and I debated
96/24 in July of 98. I agree totally with your position, and you
state it eloquently.
--
Ken Kantor
Vergence Technology, Inc.
www.nhtpro.com
www.anxioushippy.com
Stephen St.Croix <p...@nospam.intdevices.com> wrote in message
news:pr-281099...@ip132.charleston2.sc.pub-ip.psi.net...
hank alrich
> I disagree with your disagreement. I am pretty new to this NG, but isnt
> that how its done here?
Oh, hell, no, man, you've got it all wrong! (Jus' kiddin'...)
Terry Demol
davi...@my-deja.com
> I never used 48k, because SRC from 48 to 44.1 has always sounded like
> SHIT.
Synchronous or asynchronous conversion? Was the interpolator a true
sinc approximation, or was it some gawdawful linear interpolation? How
long was each polyphase filter? How many digits of precision were
carried? Was the algorithm properly dithered? There are many, many
ways to make a SRC algorithm that sounds like shit.
> For the same reasons, I would rather use 88.2 than 96.
I would rather use 88.2 too, because given a fixed number of
CPU cycles, I'd get an interpolation filter twice as long.
(See the half-band filter discussion earlier in the thread.)
> It has been said in this thread that there is no difference
> in SRC from either 88.2 or 96 to 44.1 I find this hard to believe.
When Mark Lindahl said this, he was speaking theoretically.
The 96 --> 44.1 case has some extra implementation difficulties,
but these should not stand in the way of a competent DSP designer.
They might raise the cost of a hardware box slightly, all else
being equal. The trouble is, there's plenty of DSP code today
that's being written by nitwits. To wit:
> here's my idea for a
> quick conversion algorithim from 88.2 to 44.1:(written in java)
<snipping mercifully>
> Which basically grabs every other sample out of the 88.2 file and puts
> it in the 44.1 file.
Which basically aliases everything above 22 kHz down into the passband.
If you want to write a SRC algorithm that sound like shit, this
is certainly the cannonical way.
>If you don't like that approach, how about:
<snipping bad code>
> which returns the average of 2 adjacent samples in the original 88.2
> file as a single sample in the output 44.1 file.
... leading to a sin(f)/f frequency response having its first
null at 44.1 kHz. I suppose you could call it an anti-alias
filter of a sort, but it's only 3 dB down at 22 kHz. Moreover
it's still 1 dB down at 13 kHz, which probably isn't exactly
what you had in mind. Dude, a two-tap anti-alias filter ain't
gonna cut it. Try 64 taps for a start.
> Using either of these techniques would be better than having to create
> a guess as to what the audio signal "might" have been doing at those
> several hundreds of miliseconds per second where there was no overlap
> in the sample timing between 98 and 44.1
Maybe YOU have to guess. But, thanks to Nyquist, I can compute it
to any arbitrary precision, limited only by the original noise floor.
> I have no background in DSP algorithims, only in math and music and
> computers. Please don't bash me too hard if both of these algorithims
> have been previously shown to suck badly.
Well, then consider yourself corrected, not bashed.
David L. Rick (doing his best Dick Pierce imitation)
Seventh String Recording
ignore the spam-bait in the header and reply to:
dr...@hach.com
Stephen St.Croix
<jge...@slip.net> wrote:
>
SNIP---
> So what is this thing about SRC from
> 96khz to 44.1? Is this a problem with some software and hardware but not
> others? Is it an issue?
>
> Jerry Gerber
> http://www.slip.net/~jgerber/
> jge...@slip.net
Currently it is a problem with every single software and hardware
approach, and it is a real issue.
________________
Stephen St.Croix
Fletcher
>
>
> Fletcher- have you ever snapped a throttle cable? I did last week- no fun.
>
> ________________
> Stephen St.Croix
Ahhh, roadside McGuyver shit...gotta live for it!!
Throttle cable is easy...run the longest part you can find to the carb,
wrap it around the frame [if you can cut down the cable housing, use
that so the cable doesn't bind on the frame], then put on a glove and
wrap the cable around a couple fingers [I'm runnin' a "shorty E"]...as
long as you don't have a serious need for the front brakes, you can get
home.
Try snapping a clutch cable 7 miles and 11 traffic lights from home...in
traffic, with a 90 degree turn and a 30 degree hill at the end of the
road. It brings new fun and life to 'slam shifting', and 'light
timing', 'breakdown lane cruising' and 'cop avoiding'.
If you get stalled...you're fucked...you can try to start the ride in
neutral, but you better have it rev'ed pretty high when you kick it into
first, and you best be prepared to watch that front tire go airborn as
first grabs...first time I ever desired D.O.T. helmet!!
Another reason to love my little Ironhead...I lost the 'starter throw
out gear' last summer...150 miles from home...nice when you can kick
'um...lost the rear connection on my shifter linkage this summer too
[forward controls]...tied it together with a shoelace...did I ever tell
you the one about losing *half* the voltage regulator...it would run,
just wouldn't start.
Gee, this is *way* more fun than blue lasers and storage space...
Fletcher
> Why all the buzz about (what in Hell is the
> motivation for) going to 96kHz from 44.1 or 48 kHz? What am I missing??? Did I
> say that already?
>
>
> p.s. I don't know what SRC is either. Educate me please.
>
Steve, from reading all the other posts...I think I figured it
out...it's a code thing...SRC=Sample Rate Conversion...did I get it
right? 'cmon...I wanna be in the club too!!!
Mike Rivers
> Forgive me if I am wrong, but from what I remember of digital signal processing
> (Fourier transforms and all that rot) from my school days, the motivation for
> going to a higher sampling rate is that you can use a lowpass filter with a
> more
> gentle slope (less phase distortion) to recreate the sampled waveform.
That used to be the reason, back when the only way to make a low-pass
filter was to string a bunch of resistors and capacitors together so
that it would become a higher ratio voltage divider as the frequency
went up. With the discovery and development of oversampling and digital
filtering techniques, the behavior of the filter below its corner
frequency is no longer a problem.
> This
> assumes that you are sampling above the Nyquist frequency, to avoid aliasing.
> For the 20kHz bandwidth audio spectrum, this would put the Nyquist frequency at
> 40 kHz. So what am I missing here?
No assumptions allowed - it's the law - actually two laws. One says
that you can't accurately reproduce anything with a frequency content
equal to or greater than half the sampling frequency. The other says
that you cannot have any frequencies equal to or greater than half the
sampling frequency either going in or coming out.
So, what you're missing here is everything above (to be safe) 20 kHz.
While the argument has never been settled as to what, if anything, we
normal humans actually hear above 20 kHz, there is evidence (though not
necessarily overwhelming or gathered in a terribly scientific manner)
that when listening through a system with extended bandwidth, things
just sound better sometimes.
So, by moving to 96 kHz, you're opening up one of the known
restrictions. It doesn't mean that everything will automatically sound
better, but it means that there's potential there that's completely
ruled out when sampling at 48 kHz.
> p.s. I don't know what SRC is either. Educate me please.
[S] tephen
[R] anting
[C] roix
or
[S] ample
[R] ate
[C] onversion
This is what the discussion is really about. The fact that you might
have marginal, if any, improvement by sampling at 96 kHz is overshadowed
by the inaccuracy introduced by converting the sample rate back to
44.1 kHz so that today's music lovers can play it in their present day
CD players.
--
Mike Rivers (I'm really mri...@d-and-d.com)
Kevin Roll
In article <3819A5...@mercenary.com>, Flet...@mercenary.com says...
> Throttle cable is easy...run the longest part you can find to the carb,
> wrap it around the frame [if you can cut down the cable housing, use
> that so the cable doesn't bind on the frame], then put on a glove and
> wrap the cable around a couple fingers [I'm runnin' a "shorty E"]...as
> long as you don't have a serious need for the front brakes, you can get
> home.
I had a bike in college... lost the throttle cable about 5 miles from
campus. My friend did exactly what you describe to get it back home.
> Try snapping a clutch cable 7 miles and 11 traffic lights from home...in
> traffic, with a 90 degree turn and a 30 degree hill at the end of the
> road. It brings new fun and life to 'slam shifting', and 'light
> timing', 'breakdown lane cruising' and 'cop avoiding'.
Yep, did that one too... but at least I was only across campus.
--
Chew's Eye Shop Pages: http://www.mp3.com/chewseyeshop
http://www.bge.net/chews
Personal Homepage: http://oh.verio.com/~kroll
Brian Tankersley
convenience/functionality issue. Replacing your "Dark Side Of The Moon"
LP or with a CD happens because you don't have a turntable any more
and/or you desire the convenience of instant access and greater
durability of a CD, plus the fact that you can hear it in your car,
Walkman, etc, as well as the den.
My point is this; unless a new format presents a clear *functionality*
advantage over CD, I predict it's going nowhere. People just don't care
enough to spend the dough for something unless they can immediately
go...."Ahhhhhh, I get it."
So *maybe* 5.1 DVD-A will gain a few fans, *not* because it's 96K but
because it's an immediately recognizable difference in
functionality....namely 5.1. Plain stereo DVD-A would go nowhere
instantly without a functional difference from current CD. Nobody cares,
save a few audiophiles, and they can't support 1% of the music industry.
Does mp3 sound worse than a CD? Yup. Do people want it? Yup. Why? It's
more convenient. I'm not an mp3 proponent at all, but I'm simply
illustrating my premise. If we want 96K to prevail (not sure I care that
much personally), we better wrap it up inside a format that consumers
immediately want because of an improvement *functionally speaking* or
it's not gonna happen anytime soon.
Remember, we all thought CDs sounded worse than LPs when they showed up,
but there was no arguing the functionality improvement. Look who won. If
we're not careful as an industry, mp3 will become the CD sequel,
sounding way worse in the process. Meanwhile, we're screwin' around
navel gazing over 96/24. Wake up......THEY DON'T EVEN CARE A LITTLE BIT.
We would be far better off coming up with a way to deliver at least what
fidelity we have now in a form the consumer prefers....and soon.
Otherwise, we will be moaning about how good things *used* to sound when
we had CDs. Think I'm kidding? File this post and lets talk about it
again in a couple of years.
Regards,
Brian T
Fletcher wrote:
>
> And don't forget back catalog...it's really the key to sucess...think
> about it, when was the last time Motown had something really big in the
> charts...but they sell a whole lot of back catalog now don't they.
Rick Krizman
"Stephen St.Croix" wrote:
>
> With all other variables equal- recording at 96 and listening at 96 would
> sound better than recording at 44.1 and listening at 44.1, BUT the world
> currently listens at 44.1
I think we need to look beyond the next 15 minutes.
Rick Krizman
KrizManic Music
Akashah
but there was no arguing the functionality improvement. Look who won. If
we're not careful as an industry, mp3 will become the CD sequel,
sounding way worse in the process. Meanwhile, we're screwin' around
navel gazing over 96/24. Wake up......THEY DON'T EVEN CARE A LITTLE BIT.
We would be far better off coming up with a way to deliver at least what
fidelity we have now in a form the consumer prefers....and soon.
Otherwise, we will be moaning about how good things *used* to sound when
we had CDs. Think I'm kidding? File this post and lets talk about it
again in a couple of years.
Regards,
Brian T >>
Geez, Brian
Your post makes perfect sense and absolutely scares the shit out of me.
HELP!!! MOMMY!!!!!
-ak
Jay - Atlanta Digital
<ke...@secant.com> wrote:
> > Try snapping a clutch cable 7 miles and 11 traffic lights from home...in
> > traffic, with a 90 degree turn and a 30 degree hill at the end of the
> > road. It brings new fun and life to 'slam shifting', and 'light
> > timing', 'breakdown lane cruising' and 'cop avoiding'.
>
> Yep, did that one too... but at least I was only across campus.
You should have seen the little mini vise grip and grey tape thing my
riding buddy and I devised to get me home after a clutch cable broke.
Amusing memory, but I hope never have to do it again!
-Jay
Atlanta Digital
www.promastering.com
Jay - Atlanta Digital
<rkri...@mediaone.net> wrote:
Agreed. Most agree that 16/44.1 is not enough, so why limit ourselves to
it in the future because it is the most common format now? I occasionally
make a cassette, but I'm glad that I'm not limited to this lowest common
denominator. It seems like an American trend lately to pick the lowest
quality and bring things down to that level instead of trying to bring
things up to the highest level. I guess I'm a closet altruist and
optimist, but I think it's OK to have the 24/96 system at home, but still
listen to 16/44.1 in the car. I can't hear all the detail over the
traffic noise anyways, but I don't think I should have to listen to that
at home - any more than I'd want to simulate taffic noise to even the
playing field!
-Jay
Atlanta Digital
www.promastering.com
Jay - Atlanta Digital
<Flet...@mercenary.com> wrote:
> I
> didn't think 96k was supposed to offer more dynamic range.
Increase the bits and you increase the dynamic range - increasing the
sample rate gives you other imptovements, dynamic range not being one of
them.
-Jay
Atlanta Digital
www.promastering.com
Jay - Atlanta Digital
rkri...@aol.com (R Krizman) wrote:
> Is it too much to ask if anyone has actually heard 24/96, and if so, what did
> they think--or would that blow all the idle speculation <g> ?
>
> Rick Krizman
> KrizManic Music,
> Venice, CA
I have heard it several times, and continue to hear it when the situation
warrants utilizing the capability at the mastering studio. 24/96 is
clearly better than 16/44.1, and easily discernible in a decent monitoring
environment. When compared to 24/48 (or properly dithered 20/48 - but we
won't get into marketing bits again) it is less significant, but still
worth the effort in my sometimes nit-picky opinion.
The question is: is the improvement due to the increased bit rate, or the
increased sample rate? Or are both equally important? Or is the high
sample rate only important because of the easier filtering? I believe
that 96K may be slightly overkill, but then again, who would have approved
64K as the new standard? I think that's the reasonable real world point
of diminishing returns (see AES papers and Bob Stuart's wrintings), but
not necessarily the end-all be-all. As DSP power increases, and storage
space and speed becomes less of an issue, I think 24/96 will not seem a
daunting format. These things change and improve so quickly these days
that I would not discount 24/96 based on data requirements. With this in
mind, I don't mind slight overkill just to be on the safe side.
As for 88.2 making more sense if you need to make a 44.1 copy for the car,
this is not a problem. DVD-Audio supports both rates, as do virtually all
of the high rate A/D and D/A converters out there. It is simply up to us
as the pros to decide if we want to use 88.2 over 96 on our projects.
There is absolutely no detriment to the quality in using 88.2 over 96.
-Jay
Atlanta Digital
www.promastering.com
David Satz
news:znr941199418k@TRAD...
> With the discovery and development of oversampling and digital
> filtering techniques, the behavior of the filter below its corner
> frequency is no longer a problem.
I wish that were generally the case, but it all comes down to
the specific ways in which those filters are implemented.
Their quality varies considerably--not all of them are good.
There are linear filters whose audible quality is markedly
superior to some oversampling digital filters.
To throw more light into this general discussion I'd like to
suggest a look at this very informative AES paper:
http://www.nanophon.com/audio/antialia.pdf
--entitled, "The benefits of 96kHz sampling rate formats
for those who cannot hear above 20kHz."
> While the argument has never been settled as to what, if anything, we
> normal humans actually hear above 20 kHz, ...
The "argument" has never been "settled" only because the losers
just don't get it. The experimental literature is quite unanimous:
essentially no adult hears anything of significance above 20 kHz.
My apologies to those whose belief systems may be bruised by this
assertion, but I'd urge them to read some responsible scientific
literature on human hearing (i.e. not "Stereophile" magazine)
before taking it out on me.
(Mike, let me hasten to say that I'm not directing anything at all
against you here--I don't even know what your views may be,
and am merely attaching my message at this point in the thread
because you wrote as clearly as you did, which I appreciate.)
> ... there is evidence (though not necessarily overwhelming or gathered
> in a terribly scientific manner) that when listening through a system
> with extended bandwidth, things just sound better sometimes.
Well said, in my opinion. We cannot cut off all signals above 20 kHz
just any old way we please. Some low-pass filters sound lousy, while
others are so close to transparent that it takes special signals and
special conditions to hear whether they're in the circuit or not--and
even then only a few people hear them, and even then only with
practice, and not all the time.
Thus it's understandable that a mythology has grown up in audio
that "the world above 20 kHz" must be carefully preserved. Well,
doing so is surely one way to avoid lousy low-pass filters! But it's
a situation in which the perfectly valid evidence of our ears has
led many people's (apparently much slower) brains to an
unsupportable conclusion about ultrasonics.
> So, by moving to 96 kHz, you're opening up one of the known
> restrictions. It doesn't mean that everything will automatically sound
> better, but it means that there's potential there that's completely
> ruled out when sampling at 48 kHz.
See the above-referenced article, please. The quality of sound
in the audible range is the main issue. A move to 96 kHz could
benefit audible sound, but only when good 96 kHz designs are
compared with less-than-optimal implementations of 44.1 kHz
PCM. That's a comparison that is likely to lead to false
conclusions in a bg hurry (as I believe it already has).
The fact that so much 44.1 kHz audio sounds bad is no real
indictment of 44.1 kHz audio--you have to judge systems by
the _best_ sound they can each produce. If one is clearly better
than the other, then (and only then) you've demonstrated a
real difference in the capability of two approaches. Until
then, you're only comparing specific implementations.
--regards
Rick Krizman
Brian Tankersley wrote:
>
>
> My point is this; unless a new format presents a clear *functionality*
> advantage over CD, I predict it's going nowhere. People just don't care
> enough to spend the dough for something unless they can immediately
> go...."Ahhhhhh, I get it."
>
> So *maybe* 5.1 DVD-A will gain a few fans, *not* because it's 96K but
> because it's an immediately recognizable difference in
> functionality....namely 5.1. Plain stereo DVD-A would go nowhere
> instantly without a functional difference from current CD. Nobody cares,
> save a few audiophiles, and they can't support 1% of the music industry.
I agree that consumers will not rush to buy a DVD-A player and pay higher
prices for DVD's just because they're 24/96. DVD-A will either bomb or take
over. One scenario is that DVD players will eventually replace both CD
players and VCR's. A very efficient functional advantage. Music DVD's will
be priced comparatively to today's CD's. No physical reason why they can't
be. Maybe charge more for surround or video enhancement--probably not more
for higher sample rates because, as you say, who cares. But since there's
plenty of storage space available on the disk, why not use higher sample
rates? (If nothing else it will reduce the pressure on artists to make
albums that are hours long.)
>
>
> Does mp3 sound worse than a CD? Yup. Do people want it? Yup. Why? It's
> more convenient. I'm not an mp3 proponent at all, but I'm simply
> illustrating my premise.
To me MP3's are sort of like baseball trading cards. They're really not all
that convenient, they sound like crap, but they have a sort of Pokemon
allure to them.
> If we want 96K to prevail (not sure I care that
> much personally), we better wrap it up inside a format that consumers
> immediately want because of an improvement *functionally speaking* or
> it's not gonna happen anytime soon.
DVD - - one disk, one player, for all your entertainment needs.
> We would be far better off coming up with a way to deliver at least what
> fidelity we have now in a form the consumer prefers....and soon.
Well what fidelity do we have now? If you have a Paris system you have
24/48, right? Well currently CDs and MP3's don't support that, but maybe
it's not that big a thing. OTOH, sombody who's recording Studer/Neve and
mixing to 30 ips half inch may well think they have something that's
considerably higher fidelity than what CD's can offer. And believe it or
not, there are still people doing that. The DVD revolution may well occur
quietly without any consumer fanfare--the next time you buy a video player
or the next time you buy a CD player (yours will eventually break), it'll be
a DVD player. Heck, it'll probably even play a CD you burn with your MP3
collection. At that point, if 96k ,or even 192k, sounds better than 48 or
44.1, there would be no reason not to go there. Except, of course, for the
hassle for computer-based project studio owners. (And as far as the issue
of doubling your storage costs--people record at 30 ips 2" rather than 15ips
for the sake of sonic differences that the common consumer certainly don't
care about)
>
> Otherwise, we will be moaning about how good things *used* to sound when
> we had CDs. Think I'm kidding? File this post and lets talk about it
> again in a couple of years.
I can't imagine moaning about that, but who knows. I'm definitely putting
this discussion into a time capsule. The manufacturers and standard-setters
still have plenty of opportunity to screw it up.
Rick Krizman
KrizManic Music
spiff1242
and which is 96?
--
Check out my playing at http://members.home.net/spiff1242/
> > The question itself should really be answered by each individual,
> > since some will hear an improvement and some will not.
>
> I agree! Therefore I am working to make some sample data files for
> www.pcabx.com that are the identical same program material recorded
> at 24/96 with my CardD Deluxe and some measurement mics, and then
> sampled down to 16/44.
>
> Meanwhile, folks can amuse themselves by trying to hear the effects
> of cutoffs at 5, 9, 12, 15 and 18 KHz, in anticipation of what it
> would be like to raise the high end limit from 22 KHz to 48 KHz.
> They can likewise try out 8-15 bits versus 16, in anticipation of
> 24.
>
>
Tonebarge
Fletcher wrote:
>
> Having less than no clue what an SRC is...I can't comment,
Doh, yes you +do+ know. That's the problem with short-speak and acronyms. It
means "Sample Rate Conversion", specifically that nasty, nasty one down to 44.1
from 96. Ouch!
> ...won't more dots in the picture help me with this?
>
I don't believe I ever chimed in on this topic but this be a good time. I agree
with this completely. Give me a +huge+ sampling rate, please. To the ones who
claim that the sampling rate doesn't matter and the new low pass filters are
great, etc., etc., please answer me this. Why don't I get back a clean 1kHz
square wave when I put one in? Those of you who haven't tried this and can get
access to a square wave generator should give it a go. Put a 1kHz square wave
into your favorite digital doodad. Use a dual trace scope and feed an original
of the square wave to one channel of the scope and the output of your doodad
(with what's left of the square wave) to the other channel input of the scope.
Now, ask yourself. Do they even look remotely alike? Now send both signals to
an audio monitor chain and listen. Tell me, why do they sound different when we
aren't supposed to hear the high? The only answer I know works is, "Gimme mo'
dots."
Regards,
TB
P.S. I've been looking for a sig file line. This be it.
"Gimme mo' dots"
David Satz
news:381A48A3...@iscweb.com...
> Why don't I get back a clean 1kHz square wave when I put one in?
Because a truly "square" wave requires infinite bandwidth, and
that never exists. Square waves recorded on the best 30 ips analog
recorder can't be square either, for the same reason as they aren't
on a DAT or a CD--nor are square waves recorded onto vinyl.
Fortunately it doesn't matter. A 1 kHz square wave low-pass-filtered
at 20 kHz sounds just like a square wave limited to any greater
bandwidth, all other things being reasonably equal. If you hear a
difference, as you claim, then [a] I'm glad that you're letting your
ears rather than your eyes be the judge, but [b] you evidently have
something set up wrong in your comparison--levels not correctly
matched, parasitic oscillation, clipping or slew rate limiting, etc.
The comparison between, say, an 8 kHz sine and square waves
is a standard first-semester psychoacoustics experiment. There
_is_ no audible difference, since a properly generated square
wave has only odd harmonics, and the first odd harmonic of
8 kHz is beyond the human hearing range.
Blind Joni
won't if it's not worth the extra everything that's involved.
John A. Chiara
SOS Recording Studio
Albany, NY
"Survivor of the Slums"
Mark Deffebach
As a newbe here, an electrical engineer, a musician, and an ear with a
LP corner at least 19kHz (could be better, I need better equipment to
test it), jus'ow many dot chew need fo dem square waves to sound the
same to your ears?
Can you hear the difference blind folded? ( for these sample rates
44.1kHz, 44kHz, 96kHz, 192kHz, 384kHz, 768kHz, 1536kHz, 3072kHz,
6144kHz, 12.288Mhz
Can you say marketing bits/rates.
There is a point where not only your ear's frequency response, but your
ears sensitivity to things like Hauss effect stops. Perhaps some of
the difference you hear in the squarewave and its digital recording are
effects of the analog signal conditioning that all digital equipment
uses to match audio input levels to the ADC levels and to bring DAC
output levels back up to something pro audio respects (+24, +30 dBu?)
Oh yeah, don't forget that your noise floor is the sum of noise at all
frequencies in the pass band. The higher your sample rate, the wider
the pass band, the louder your noise, all other things being the same.
I also find it strange to complain about SRC and not dithering, which
happens with any volume adjustment performed in the digital realm. Not
the mention all the audio processing done the the digital domain (more
dithering and SRC).
Given some set of dots, your ADC and reconstuction filter fits a curve
to the dots. Your SRC does the same thing then makes new dots at a
different rate.
:)
Bill Roberts
You really should IMO read the link that was posted earlier
on this thread: it was to a paper by Julian Dunn.
http://www.nanophon.com/audio/index.htm
(the first paper)
-- Bill
Stephen St.Croix
<Flet...@mercenary.com> wrote:
> Stephen St.Croix wrote:
> >
> >
> > Fletcher- have you ever snapped a throttle cable? I did last week- no fun.
> >
> > ________________
> > Stephen St.Croix
>
> Fletcher wrote:
>Ahhh, roadside McGuyver shit...gotta live for it!!
>
>Throttle cable is easy...run the longest part you can find to the carb,
>wrap it around the frame [if you can cut down the cable housing, use
>that so the cable doesn't bind on the frame], then put on a glove and
>wrap the cable around a couple fingers [I'm runnin' a "shorty E"]...as
>long as you don't have a serious need for the front brakes, you can get
>home.
>
Snapped that throttle cable 57 miles from home in the deep woods- on a
Mikuni, 3/8 inch from the end.
>Try snapping a clutch cable 7 miles and 11 traffic lights from home...in
>traffic, with a 90 degree turn and a 30 degree hill at the end of the
>road. It brings new fun and life to 'slam shifting', and 'light
>timing', 'breakdown lane cruising' and 'cop avoiding'.
Not too good.
My first time for a clutch cable was in Arizona, sitting right behind a VW
Beetle (of coursr), revved up to make noise. Bike walked right up the back
of the bug and went over the front. This is no bull, it happened-
unfortunately
We gotta talk someday
Oooohhh. NOW I read the other posts, and it looks like we are all trying
to break a cable farther and farther from home
________________
Stephen St.Croix
Stephen St.Croix
clip
> [S] tephen
> [R] anting
> [C] roix
>
clip
> Mike Rivers (I'm really mri...@d-and-d.com)
Oh, thats nice. I am not ranting. I am didactically pontificating.
________________
Stephen St.Croix
Harvey Gerst
>In article <znr941199418k@TRAD>, mri...@d-and-d.com wrote:
>> [S] tephen
>> [R] anting
>> [C] roix
>> Mike Rivers (I'm really mri...@d-and-d.com)
>
>Oh, thats nice. I am not ranting. I am didactically pontificating.
Stephen,
There's only room for one pontiff in this newsgroup. Be in front of the
saloon at sundown. Bring your "didactic"!! If you don't have one, "rant"
it.
Harvey Gerst
Indian Trail Recording Studio
http://ITRstudio.com/
Mike Rivers
> To me MP3's are sort of like baseball trading cards. They're really not all
> that convenient, they sound like crap, but they have a sort of Pokemon
> allure to them.
Yeah, you have to be a member of the "I really, really want this bad"
club in order to get them. They're enough of a pain to acquire that
there's a certain perceived value even though they don't cost
anything. But as a preview of something that you can (and should, if
you like and want it) purchase, I think it's a cool idea. And it'll
be even cooler when we all have megabyte/sec Internet connections in
our homes.
> The DVD revolution may well occur
> quietly without any consumer fanfare--the next time you buy a video player
> or the next time you buy a CD player (yours will eventually break), it'll be
> a DVD player. Heck, it'll probably even play a CD you burn with your MP3
> collection. At that point, if 96k ,or even 192k, sounds better than 48 or
> 44.1, there would be no reason not to go there.
We're seeing the same thing now in audio I/O interfaces for computers.
Who still makes a 16-bit sound card, except for those that come
bundled with computers that will probably never see a mic input unless
it's for an Internet telephone hookup? You can't buy a 16-bit
converter chip any more, so why not get a 24-bit converter. Some day
the analog circuitry will be good enough so you can take advantage of
the additonal word lenght, and for now it's no more expensive than the
old 16-bit solution. Fortunately, you can turn it off and save the
disk space.
> I can't imagine moaning about that, but who knows. I'm definitely putting
> this discussion into a time capsule. The manufacturers and standard-setters
> still have plenty of opportunity to screw it up.
Don't throw away your cassette deck. <g>
--
I'm really Mike Rivers (mri...@d-and-d.com)
Mike Rivers
In article <u51vVSlI$GA.255@cpmsnbbsa03> DS...@msn.com writes:
> Mike Rivers <mri...@d-and-d.com> wrote in message
> news:znr941199418k@TRAD...
>
> > With the discovery and development of oversampling and digital
> > filtering techniques, the behavior of the filter below its corner
> > frequency is no longer a problem.
>
> I wish that were generally the case, but it all comes down to
> the specific ways in which those filters are implemented.
> Their quality varies considerably--not all of them are good.
Of course anyone can screw up good technology. The fact is that when
they do it right, it's better than the old way. And of course we're
only willing to spend our money on things that are done right, aren't
we? <g>
> The "argument" has never been "settled" only because the losers
> just don't get it. The experimental literature is quite unanimous:
> essentially no adult hears anything of significance above 20 kHz.
That doesn't seem to matter, but I really don't want to start this
debate again. The fact is that people hear things differently when
there's material present above 20 kHz. Whether they're actually
hearing above 20 kHz is irrelevant. The fact that there's something
up there affects what we hear in whatever range we are physically able
to hear. And there's plenty of measurements that show that musical
instruments produce frequencies above 20 kHz, so something's getting
lost if you can't record and reproduce those frequencies.
> (Mike, let me hasten to say that I'm not directing anything at all
> against you here--I don't even know what your views may be,
No offense taken.
> The fact that so much 44.1 kHz audio sounds bad is no real
> indictment of 44.1 kHz audio--you have to judge systems by
> the _best_ sound they can each produce.
Hey, that's for sure. It's the poor quality of 4-track cassette
Portastudios that have led today's recording hobbyists to think that
ADATs sound so great, and also think that analog recording can't
possibly sound good. It saddens me to think that the small turnaround
today is largely fad-followers and not really golden ears.
Roger W. Norman
(or virtulessness) of 96 khz sampling, to which based on his obviously
strong feelings on the matter, I must say changed my mind. 96 has always
been stupid for audio in the first place, but I guess because of the 48 kHz
standard in video 96 became the de facto new standard. But 24 bits works
better at either 44.1 or 48, and since you don't have to "gearbox" (yes, I
read your columns) it just makes far more sense to go 24/44.1 and leave it
at that until you're ready for the final product, which, as you've stated,
is still 99% 16/44.1. I don't find your posts a rant, nor do I even
perceive you as being on a soapbox. Just plain simple intelligence applied
correctly.
--
Roger W. Norman
SirMusic Studio
301-585-4681
"Never try to teach a pig to sing. It only wastes your time and annoys the
pig."
Robert Heinlein
Stephen St.Croix <p...@nospam.intdevices.com> wrote in message
news:pr-301099...@ip217.laurel.md.pub-ip.psi.net...
> In article <znr941199418k@TRAD>, mri...@d-and-d.com wrote:
>
>
> clip
>
>
> > [S] tephen
> > [R] anting
> > [C] roix
> >
>
>
> clip
>
>
> > Mike Rivers (I'm really mri...@d-and-d.com)
>
> Oh, thats nice. I am not ranting. I am didactically pontificating.
>
> ________________
> Stephen St.Croix
>
Brian Tankersley
>
> I agree that consumers will not rush to buy a DVD-A player and pay higher
> prices for DVD's just because they're 24/96. DVD-A will either bomb or take
> over. One scenario is that DVD players will eventually replace both CD
> players and VCR's. A very efficient functional advantage. Music DVD's will
> be priced comparatively to today's CD's. No physical reason why they can't
> be. Maybe charge more for surround or video enhancement--probably not more
> for higher sample rates because, as you say, who cares. But since there's
> plenty of storage space available on the disk, why not use higher sample
> rates? (If nothing else it will reduce the pressure on artists to make
> albums that are hours long.)
>
Good point about available space vs album length. It's always puzzled me
how most records stayed at 10-12 songs for a decade after CDs showed
with 74 minutes of time available. Record company royalty issues and I
guess plain old habits. Of course, an all in one unit that caught DVD,
DVD-A, CD and like it or not, mp3 would be a plus to the consumer
assuming it didn't cost a big premium. But, again, there's the
functionality advantage propelling sales, not fidelity.
> To me MP3's are sort of like baseball trading cards. They're really not all
> that convenient, they sound like crap, but they have a sort of Pokemon
> allure to them.
>
I'm 43 and grew up playing guitar and bass along with every Hendrix, Led
Zep, Deep Purple, Black Sabbath, etc record I could get my hands on. I
was in *LOVE* with rock and roll. What did I listen on? It didn't matter
much...mostly a pretty bad turntable that, looking back, would hurt me
now. It was about the band, the vibe, the music.
It's easy to be a pro audio engineer and belittle the way kids decide to
listen to music. It's also dangerous. Old Fart Syndrome is lurking just
around the corner from that headspace, whatever your age. That warm
feeling of "I know better" that dismisses the changing reality of time.
Until solid state memory gets so cheap and large it replaces moving
parts, expect demand for smaller audio file sizes, whatever the format,
to grow. That's the way it is. Likening that trend to Pokemons reminds
me of much of my parents reaction to my love affair with music. No
offense intended, seriously.
>
> Well what fidelity do we have now? If you have a Paris system you have
> 24/48, right? Well currently CDs and MP3's don't support that, but maybe
> it's not that big a thing. OTOH, sombody who's recording Studer/Neve and
> mixing to 30 ips half inch may well think they have something that's
> considerably higher fidelity than what CD's can offer. And believe it or
> not, there are still people doing that.
Oh, I believe it all right.
I've spent my whole career recording "Studer/Neve". I built the first
input/output transformerless Studer 24 track in the world (an old 80
Narrow) in about '82, AFAIK. I've owned Studer, MCI, Otari (peew),
Trident, etc and spent literally years of my life cutting and mixing in
studios like Oceanway, Woodland, yadda yadda yadda. Hotrodded ATR-102s
and mixed plenty of 1/2" 30IPS records along the way.
My reference points have been hi-fi analog all along the way. I often
track to 16 or 24 track analog and 48 Sony simultaneously locked up so
drums and whatever can go to analog while the rest go to dig.
I would never have considered a DAW replacing that gear based upon what
I heard. Small, narrow and usually pretty lifeless, IMO. Then I heard
Paris in a controlled setting. A sonic paradigm shift for DAWs. So I'm
now producing sonic output that myself, my clients and my mastering
engineer prefer to 20 years of analog efforts.....at 24/44.1. Done
right, it don't sound bad at all.
Maybe I just suck as an engineer (with #1 records in a few different
genres), but my ears say 24/44.1 can sound great **when it's done
right**. Maybe that explains the difference in our opinions. I might
agree we *need* 96K if I hadn't heard this. BTW, I don't sell Paris and
don't *even* care if readers of this post buy it or not. I just really
like the way it sounds.
>
> I can't imagine moaning about that, but who knows. I'm definitely putting
> this discussion into a time capsule. The manufacturers and standard-setters
> still have plenty of opportunity to screw it up.
>
> Rick Krizman
> KrizManic Music
Yup, let's have a look again, say Christmas 2000?
Regards,
Brian T