24/192 Downloads ... and why they make no sense?

[crisnee]

"PS/ I don't have nothing against the CD media (I have a lot that I love), but if it can be improved. Why not?"

 

I'd much prefer "they" would improve the mastering/remastering and recording side of things, not to mention work on getting the best out of what we already have (redbook--if you've paid attention you'll have noticed that over the years the exact same cd has yielded better and better SQ as technology has allowed). And if not, at least give us a guide to finding the best masters/remasters/re-issues of the great music that's out there. What a morass, but let's add some more into the quagmire, what the heck, morass is always better than goodass, eh? What more can you ass for?

 

-Chris

 

 

[labjr]

"The players that play CDs have improved extraordinarily over that 30 year time frame, I might add."

 

Inexpensive CD players are better than ever too. And 24 bit files sound better than CD on inexpensive DVD players.

 

[Talk2Me]

""PS/ I don't have nothing against the CD media (I have a lot that I love), but if it can be improved. Why not?""

 

I agree 100%! But, increasing the resolution must be acknowledged as only a brick in the wall, if you will. The recording itself has to be improved, the playback equipment must be improved, and lastly our listening skills must be improved. Just increasing the bits and pieces only does not a game changer make. However, with certain recording hgher res does make a slight improvement. As an example, I down loaded Marvin Gay's "Let's Get in On" from HDTracks in 24/96 to my MacbookPro. And compared to MOG in 320 KBPS MP3 and also payed the CD from my Wadia S7i. I have to admit the 24/94 download sounded the best to me. The CD was a close 2nd and the MOG came in last. The difference? The high-res computer download had less "shimmer" than the commercial CD, which I am sure is not a "audiophile" recording.

 

[Daphne]

“There is not much 24/192 material available yet, and a lot of that is 'specialist' garbage recorded more for its own sake than because anyone actually likes it.”

 

 

A statement with attitude!

 

I, like many here will agree with your statement, we all know of the poor selection available, and what is not available. You have made this same statement on several threads. How often do you think we need to be reminded?

 

Perhaps you should consider placing this statement in your signature. But, if you feel that passionate about the subject please open a separate thread on “Dispose of Your 24/192 Garbage Here.” Then all the members of this forum will have a resource on exactly what to avoid. You could start off the thread with your All Time - Top 10 High Res Hate List. ;-)

 

 

“Vanity publishing by unknown, second rate 'artists' who usually go nowhere.”

 

 

Most every great singer or musician started out as a 2nd or even 3rd rate artist. They all get their start somewhere. Most will never make it into the ranks of mass global appeal, oh so worthy of high resolution download sales, but it’s strange that the big stars only seem to be available on Mp3 or standard Red Book. A most curious situation. Many, but not all, of the remainder will become first class artists that may never go anywhere, but should not be banned from recording and publishing at high resolution due to their Vanity.

 

In my observations, “vanity publishing” best applies to the 1 or 2 hit has-been from the 70s and 80s. After several decades of 12 step programs, they are finally clean and sober, have found the lost Jesus, and got their Mojo back baby! Being unemployable, over 60, and broke, it is the perfect time for a come-back tour and album. It does not matter if their voice is half gone, or they can barely play their instruments due to arthritis, we must show homage to greatness and purchase whatever recording is dished out. I would much rather listen to a group of excellent musicians playing good music and heading to the lost world of nowhere.

 

I can also attribute Vanity recordings to the record company. In their zeal to produce high resolution 24/192 recordings they have unwisely chosen to record uninspired musicians playing in odd venues. Oh, the sound quality is excellent, but the music is not something people will be attracted to. It is basically the vanity of the record label wanting to demonstrate what they can do. I often ask myself, with all the great artists and performances available, why can’t these companies find something truly decent to record?

 

Personally speaking, I have come across a number of recordings at 24/192 that I have found quite enjoyable and would not consider “garbage.” Just last evening I was listening to two Russian composers, recorded by Channel Classic Records. No second rate composers, or has-been musicians, but performed by a world renowned conductor and orchestra. The performance and sound quality was fantastic. There are some people who consider classical music as “garbage.” For then I can offer my sympathy, in the same manner that professional musicians I know offer me their sympathy for not being a fan of Bach. Oh well, on the other hand, I personally dislike Rap and Country music and consider it only appropriate for autopsies and alien abductions, but their fans would claim my appreciation of Asian music would fall into the same category.

 

 

“Already you are calling for 24/384. What's the point? And when, if that ever shows the faintest glimmer of being taken up by any more than one or two percent of available music, which even 24/192 has not yet reached, I suppose you will want 32/768.”

 

 

Mark, what could you be thinking, you are preaching and asking questions to the wrong group of revolutionary enthusiasts. I seem to recall back in 2008 when a few free samples of 24/192 appeared, the general consensus on this website was: we want them! We want them now! We know the recording studios are sitting on high res masters. Break them out and get them online immediately! Every few months for the past 3 years someone will feel compelled to issue the same demands, over and over again.

 

I agree with you completely on the point of 24/384. I don’t know of any 24/352.8 or higher downloads available, except for 2 free samples, and there are very few professional DACs available. All of this 24/352.8 resolution is dedicated to recording and mastering in DXD. Nevertheless, Mark, all it takes is a few sample downloads posted on a website. The call is out – higher res downloads are here, higher res downloads are here, rally round the flag boys! Over a short period of time, quite a number of members here will be instantly bewitched with that irresistible “must have” feeling.

 

 

The whole process will begin anew. Where are the DACs? Don’t the equipment manufacturers want to satisfy their customers? We want more downloads, the recording studios are sitting on the masters, give them to us now! I must have a USB interface because I only purchase PC notebooks, and despise Apple and iTunes with every living cell in my being. I am only willing to pay $600 for a notebook, and only interested in free music players and applications, but will seriously consider spending $500 on an “audio quality” USB cable. Chris, Chris, Chris, where are the reviews? I am going back to bit perfect 24/192 because all the music players suck, 24/384 makes no sense and therefore useless; my golden ears tell me so. Right up to the point when someone asks, where are the 32-bit downloads?

 

Oh well, life goes on, right along with the trials and tribulations of the audio enthusiast.

 

 

[elcorso]

Could be a good reference or not. I don't know the source of HDTracks: DSD or PCM.

 

But Lady GaGa’s ‘The Fame Monster’ sold more copies than Miles Davis 'Amandla'. Who would last forever Miles or Lady? Of course Miles, but the big money is now on Lady.

 

I agree that the new players and DACs help a lot. I have in my hands an 'very old' CD: Egil Kapstad "Cherokee", the original one (there is no reissues) from 1988, out of print, but you can get it on only now on MP3. A close friend who was on 2012 CES Las Vegas told me that was used for demo in several places. The sound is great, so was the recording!

 

Roch

 

[labjr]

"However, higher res does make a slight improvement"

"The high-res computer download had less "shimmer" than the commercial CD, which I am sure is not a "audiophile" recording."

 

 

So what are you saying, they would be equal if the CD was audiophile? I don't think the HDTracks version is a particularly special mastering. The Japanese CD versions were about the best CDs I've heard. However, The DVD-Audio version killed the best CD version I've heard. I'm suspect the HDTracks files were made from the the DVD-A files.

 

 

Audiophile or not, I don't think I've heard an album that I felt wasn't improved by higher resolution. And not just slight improvement.

 

Unless you have CDs that can't be replaced with higher resolution or found on vinyl,I see no reason to play CDs anymore.

 

[crisnee]

"Unless you have CDs that can't be replaced with higher resolution or found on vinyl,I see no reason to play CDs anymore."

 

Have you tried ripping the cds to wav/flac? Perhaps its your mode of play.

 

And when it comes to vinyl, you may like the sound better, but comparing them to high resolution? They're a whole different animal. It almost seems as if they should cancel each other out, so to speak.

 

As a matter of fact, I think you're implying, although unintentionally, that the problem with redbook cd sound is the playback mechanism, not the data. Think about it--I'm to lazy to explain right now.

 

-Chris

 

 

 

[Talk2Me]

I truly feel there are many members here who have not heard a state of the art CD play back system. I am not saying it is the best playback by any means, but really, some act here like hi-res is more important that the hi-fi systems, or the recording quality.

 

[crisnee]

"but really, some act here like hi-res is more important that the hi-fi systems, or the recording quality."

 

I was just reading an interview with one of the 47 Labs guys, and he was saying that the industry hasn't even been able to extract everything from Redbook so he didn't quite understand the obsession with hi-res (the interview was a few years old I think, but still).

 

One point he seemed to be making was that there's nothing wrong with hi-res in itself, but it does take the focus off of where it needs to be, perfecting the current system, and puts it on the new and "exciting" and a lot of marketing, to make more money by selling the consumer another "new and improved" format at even higher prices.

 

-Chris

 

 

[Paul R]

Redbook is already pretty hi res, compared to Mp3s that is. Why spend more time and effort on CDs, when more and more people are going to rip them anyway?

 

Just publish redbook quality files online and call it hires. (shrug)

 

Sooner or later they will figure out it costs more money to master those nice hires files they record in down to good sounding 44.1 than to just sell em as Hires.

 

 

Paul

 

 

 

[audiventory]

"I don't have nothing against the CD media (I have a lot that I love), but if it can be improved. Why not?"

 

I'm agree with you, elcorso: my collection now contains hi-res records preferably.

 

Interesting fact: new CD (since 2008 approximately) sounds better then old CD.

 

I guess what is more perfect recording technology (using 96 and 192 kHz for full cycle producing and downsampling/down-bitdepth in last stage).

 

In theory sample rate 44100 Hz allow full restoring of information 0 - 22050 Hz. But simplifying of software/hardware algorithms (by reason of high power of calculation) don't allow full restoring.

 

By this reason we are using 96 kHz and more.

 

For manufacturers (of hardware and audio records) it is profitable (sell new equipment, re-release of records). Sound become better also (profit for listeners-consumers).

 

May be in future will released perfect filter low frequencies for ideal restoring of 44 kHz sampling rate. This filter make high oversampling with interpolation. Profit - low using of volume of data storage. It will new marketing strategy. May by after 384 kHz for home audio.

 

[Talk2Me]

NAD is now marketing their new DAC, the M51 with 35-bit, 844 kHz. Go get 'em guys!

 

[PeterSt]

One point he seemed to be making was that there's nothing wrong with hi-res in itself, but it does take the focus off of where it needs to be, perfecting the current system,

 

Now *that's* a good statement !

 

[PeterSt]

Case 1) transient shorter than 22 µs - won't be encoded at all.

 

Correct.

 

 

Just that the "vertical" cannot be more vertical than smooth 22.7 µs long transition from value to another.

 

No. Digital depicts inifitely short time. Say 0.00000000µs.

Analogue has the 22.7µs time for it though. But it doesn't work like that, because there will be distortion first. But see further down.

 

 

to make sure that the transition speeds don't exceed these figures.

 

No. See further down.

 

That's the nature of PCM. If you want to use "minimal step" method, it's called DSD...

 

Almost correct. Why ? well, because I said exactly this, earlier in the thread.

But still not enough. Not enough, but better.

 

Of course no electronics can make anything completely vertical. But idea of the analog reconstruction filter after DAC is naturally to smooth out the steps in order to have nice smooth continuous wave instead of staircase

 

No. Because I am not talking about staircases. My picture shows that alright, but it came along with my comment that this was only to show that bigger vertical steps are in there than "can be done" (with the encoding means used) and that this picture was about a *frequency*.

 

So ...

 

Here we all keep talking right along eachother.

I'm not sure how many times I must repeat it (before I'm proven wrong), but we are not talking about the same ting.

 

A TRANSIENT is going from one point to the other, and it happens all over.

Your "proposed" transient Miska, is about one which goes down again. And this in within the perceived 0.5µs (etc.).

NO SUCH "TRANSIENTS" EXIST ... except in test signals.

 

Since such transients don't exist, what is the reason to talk about them, and keep talking about them. Point is :

 

The transients how I define them, sure do exist. Look in the files !!

So, they only go up (and most probably only go down with inverted absolute phase or what ever may cause e deep suck otherwise (I don't think that can exist, but alas - not related anyway)).

 

This is nothing like staircasing, although it may look the same to you when not interpreting matters in detail.

 

Now back again to the first quote in this post :

 

Case 1) transient shorter than 22 µs - won't be encoded at all.

 

Here you can only be referring to that "transient" which does not exist in music or any other sound. If you think it does, please point me to the example of it. Synths are not allowed (this is dangerous because synths are musical instruments too and they could do it; whether this is actually used - I don't think so).

 

So, right. That shorther than 22µs (etc.) "transient" will not be encoded. But the transient as per my definition sure will. It goes up and stays there, further riding (also during its further envelope) along the remainder of the wave. It will be encoded always. It is just a matter when the sampler comes by.

 

So ...

Paul and Julf maybe talking along eachother (one talking about "events", me not even being sure whether that will be my transient or yours), but we can all talk along eachother.

And quite forever as it seems.

 

A transient in the context of this thread goes up only.

 

A pulse of the short duration implied DOES NOT EXIST in music.

So let's stop about this one.

 

A transient in the context of this thread *will* be captured, but its timing can be right by coincidence only. This is similar to the 1:11025 chance I talked about earlier.

When it is caught in time (highly unlikely) the relative timing to the other transients/"music" it correct, as will its phase;

It must be considered that both are -thus- not correct always.

 

I'd even go as far that 176.4 depicts only 4 times more chance of it being correct, but it -thus- still never will be (correct). It's a moot think to think this will be "better".

However ...

 

When the "smear" is (explicitly !) applied as how I proposed it in an earlier post, suddenly your chances get way better. Still YMMV but it becomes a complete different thing.

 

Lastly and again - when this is all to be seen in the context of heavy normal filtering, I am out right away, plus that nobody should be talking about µs time differences. ms comes then first. A totally moot thing as a subject. Then.

 

... which all is different an uncomparable from/to perceived THD figures hence the frequency domain itself.

So let's not start about that, because it is a subject in parallel. Too bad the one solution implies the other. But for the timing issue ... please let's talk about it in its actual context. See above.

 

But where I'm wrong I will stand corrected. No probem with that.

 

Peter

 

[Paul R]

I was trying to remember where I had read about Peter St.'s "bumps" before, and found an ancient bookmark that led me to this old paper. I make no comment on it's validity.

 

http://www.cco.caltech.edu/~boyk/spectra/spectra.htm

 

And this one too...

 

http://www.csis.ul.ie/dafx01/proceedings/papers/duxbury.pdf

 

Paul

 

[Downrange]

I don't know how anyone can say 16 bit is hi-res in any sense of the word. Except maybe by comparison with MP3 - maybe that's the issue here. Even though I'm now getting the very best reproduction of ripped CD now I've ever heard, including that of a 2K player (thanks to the new V-Link 192 doing the bit arranging for a high end DAC), it is still easy to hear the problems of ALL 16 bit material.

Today, I've deliberately chosen two of Harry Pearson's "superdisc" CDs to be the "victims" - these appeared on a listing in 2010, alongside super SACD and LP material, as the "best of the best." I refer to the Minnesota/Oue Symphonic Dances, et al, and the Norah Jones "Never Too Late" CDs. A decent mix of material there, with both female vocals, studio-mixed, as well as a truly great recording of symphonic music, recorded in its original acoustic space. The mostly low level dynamic material of the Oue is not challenged too much by CD's 16 bit rendering, but, wait - there it is - massed strings and a brass crescendo! The Achilles heel of 16 bit. Interestingly, the SACD doesn't suffer from this anomaly...

The Jones is easier to parse. Because it's a hot studio mix, with emphasis on Norah's spectacular voice, the application of some eq and reverb by the masterer makes it easy to spot the hard-edged signature of 16 bit. Will Chesky please obtain the original analog or (hope, hope) 24 bit recording and resample and release this? It makes it so easy to demonstrate when one has both 16 and 24 bit versions.

Like with Diana Krall, for instance. I'll bet there are plenty of people who have her "Best of" CD. And I've got most of those cuts in 24/96 from HDTracks. It's just like shooting ducks in a barrel. Just play them, or better yet, just play key sections of each cut, a vocal peak, the piano intro, back-to-back. If your equipment is up to snuff, you'll never be happy with 16 bit again.

That's what this all comes down to. I'm willing to keep an open mind about 192, but I suspect it's not nearly as important as that all important bit depth. Even 24/44.1, as in the numerous BIS downloads of great classical material I've gotten, is an order of magnitude better-sounding than even the BEST 16 bit recordings, like the Oue and the Jones.

I'm now in search of an analytical tool to help make what is necessarily a mostly "subjective" process (and that subject to huge variations because of playback equipment limitations) into a more objective study. I don't think Audacity has enough capability, but I've not completed my evaluation of it yet. There are probably other very expensive analyzers out there that will show what's going on, but the budget disallows, alas. What is needed is something that shows graphically what happens when Redbook "runs out of bits," and runs screaming from the sound stage. High frequency square waves attacking the old eardrums, aye, that's what it feels like.

Someone said Barry D. had some tools on his site that clearly show 24 bit material - need to check that out. But, really, the old ears work fine, as Teresa has mentioned so often. Just listen... But, of course, we'll never get anywhere with that, will we? So it's back to the drawing board, trying to find a way to objectively pinpoint the problem with "not enough bits." We will press on.

 

 

 

[Fokus]

Over a century after the basics were laid.

 

More than 50 years since a complete and elegant proof was provided.

 

Thirty years after digital audio went mainstream.

 

And still such vast misconceptions...

 

Someone ought to make a study of this phenomenon.

 

 

A 10 microsecond transient has no components with a frequency lower than 50 kHz, so no, it would not get past a 21 kHz filter.

 

Wrong.

 

Take a signal, all zero for a long time, then maximum level during 10 us, then all zero again.

 

Take its Fourier transform. (Note, I wrote transform, not series. If you don't know the difference then you should read up on it, it is important for what follows.)

 

The Fourier transform gives the spectrum of the signal. Observe how it extends from the lows to the highs, with some peaks and nulls.

 

Take the limit case of above signal: a single needle of unit energy and zero duration. Its spectrum is a flat line. DC to light.

 

 

The easiest way to view a 10?s transient is as a 10?s square pulse. If you do a FFT on that pulse, you will see that it consists of half a wavelength of a 20?s period (50 kHz), plus higher-frequency harmonics. So the lowest-frequency component is 50 kHz.

 

See above. That is not what you will see.

 

 

 

Since such transients don't exist

 

Of course they exist. Perhaps not regularly in music,

but they can and do exist. And they are nothing to be afraid of. In fact they are not even remotely exceptional and actually they are quite handy for reasoning about signal processing systems.

 

 

 

Case 1) transient shorter than 22 µs - won't be encoded at all.

 

Wrong.

 

See above. Such a signal has spectral content below 22kHz. When the signal hits the ADC's anti-aliasing filter the part above 22kHz gets sliced off, and the part below 22kHz gets sampled.

 

 

 

Does a 16/44.1 file have sub 10us. timing precision? No, it does not. It cannot accurately place an event to within 10us.

 

Wrong.

 

It can and it does. Hands down.

 

The theoretical timing precision of 16 bit 44.1kHz is of the order of hundreds of picoseconds. In practice nanoseconds are easily achieved.

 

You people forget the action of the anti-aliasing filter and of the reconstruction filter.

 

The AA filter removes the input signal's spectral contents above the Nyquist rate, so that it can be sampled legally. In the time domain this amounts to a spreading out of the signal's shape.

 

Upon playback the reconstruction filter interpolates through the samples according to a well defined and prescribed function (Sinc, which can be implemented to arnbitrary accuracy). Post-reconstruction the smoothed and spread-out waveform has its center of gravity in exactly the same temporal location as the original unfiltered signal.

 

 

Picture this (I have no time for graphs now):

 

Two channels (i.e. stereo). One channel contains a needle impulse, and the other channel contains an identical impulse 1 microsecond delayed relative to the first one.

 

You pass this stereo signal through a 44.1kHz ADC/DAC chain; assume linear phase AA and AI filtering (which is the norm in industry).

 

At the (analogue) output you'll find two blobs. The centers of gravity of these two blobs are spaced exactly 1 microsecond.

 

If you don't believe me: take a decent audio editor, construct needle signals at, oh, 1MHz or whatever (this to emulate analogue), convert, and observe.

 

 

$&%$% ow shucks, I did it ...

 

Here are two 1 us-wide needles spaced 1 us in 1MHz sample space, or pseudo-analogue. Please look only at the dots, ignore the wiggles.

 

 

 

 

Here are the needles after sampling by a 44.1kHz ADC, using a none-too-exotic half-band linear phase AA filter. Again, look at the dots. Observe how that are not quite identical the two channels. That is the crux. (Oh, I also boosted the signal with 26dB, because most of the original energy was above 22kHz, lost in the AI filtering action.)

 

 

 

And here are the same samples after expansion to 1MHz with again a standard reconstruction filter, zoomed in on the central lobe of the signal blob. I realise now I should have taken a time delta of 5 us, because 1 us is hard to spot now. But look closely. The yellow cursor may help.

 

 

 

 

 

If you re-read Shannon's proof it is obvious. It demonstrates that with a band-limited input signal nothing is lost. This covers time information too. And timing accuracy and bandwidth are not related. Our ears can resolve down to microseconds, despite a limited bandwidth. And 44.1kHz digital can pass this, too.

 

 

 

May be in future will released perfect filter low frequencies for ideal restoring of 44 kHz sampling rate. This filter make high oversampling with interpolation. Profit - low using of volume of data storage. It will new marketing strategy. May by after 384 kHz for home audio.

 

1) the perfect filter has been known for decades.

 

2) what do you think DACs have been doing since the late 80s?

 

 

 

 

[Jud]

Interesting. Just been trying to read stuff on instruments with significant inharmonic components, e.g, percussion. How are these modeled / sampled / reconstructed?

 

[Fokus]

The same as all the rest.

 

Please folks, take this: there is no modelling involved with sampling. There are no restrictions to the input signal with as single exception that it be band-limited (i.e. low-pass filtered).

 

And while the required filters are theoretically impossible, they can be approximated with near-arbitrary accuracy in the real world. That is engineering.

 

 

 

 

[Jud]

How are inharmonics approximated (or are you saying they are not)? With harmonics, or in some other way? Please realize I'm asking from a position of great ignorance and would sincerely like to know.

 

[PeterSt]

http://www.cco.caltech.edu/~boyk/spectra/spectra.htm

 

Although I have difficulties with following the article in above link, I at least got this from it :

Jangling keys indeed will exhibit the (relative !) highest frequency content. BUT, this is how it's technically derived. These are all squares, and no way the "sine frequency" will be that high. Yes, to form the transients.

So what we must realize, is that this is about those transients, and for keys almost about transients only. Squares. This means that these transients should remain were we to perceive the "clearness" of the jangling keys. This is nothing about the "fineness" or refinement or "nice silk" sound, most people tend to think about, when talking HiRes. It is almost the other way around.

How ?

Because the transients will be flattened by the filtering; the less this happens, the more "crystal clear" (thus) HiRes should be.

(but I don't think it ever is ?)

 

And this one too...

 

http://www.csis.ul.ie/dafx01/proceedings/papers/duxbury.pdf

 

Paul

 

This one is better readable for me and I wonder whether this isn't in the standard toolbox of today's digital mastering engineers. I guess so.

 

 

 

[PeterSt]

At the (analogue) output you'll find two blobs. The centers of gravity of these two blobs are spaced exactly 1 microsecond.

 

Fokus, your very extensive outlay is really appreciated.

But what I don't see is what you try to prove with the spacing of your pulses sustaining (in two channels or in to subsequent time shifted runs).

Why would the spacing change ?

 

But now run the same on top of two different frequencies. 5Khz and 12 Khz will be fine. Through that, run the same pulses with same time shift.

I think it will be easy to find spots where the same shift is not there. The contrary, it will be tough to find the places where it is the same.

 

Or apply another nice test;

Run the test with the pulses so close that they run into eachother's sweep-up and -down. See what's left of the pulses and the level of the peaks.

What will it tell you ? well, that the allowed spacing in between the pulses depends on the filter length. Too bad.

 

May you still see results you can deal with, then please run it through analogue and just measure. Especially that last nice test.

 

There *is* no free lunch here.

 

Regards,

Peter

 

 

 

[Paul R]

Oh, I am sure it is. It just thought it would be interesting to other people. The math in the second paper sure helps to make things clearer.

 

The first paper was interesting because it is one of the first papers I know of that essentially says the high frequency harmonics present in most instruments may have something to do with how we perceive the sound of those instruments. The paper is more than 20 years old and we still argue about it today.

 

Back to sampling however, I find it slightly amusing that we talk about sampling, and everyone jumps immediately to filters and reconstruction of the signal. The actual samples themselves contain no information about what happens between the samples. A 10us transient signal completely between the samples will not be recorded. While looking for a nice graph to show this, I stumbled across this paper by Tim Westcott. Amazing, I don't seem to disagree with anything in this paper. It is, indeed, the same thing I have been saying.

 

http://www.wescottdesign.com/articles/Sampling/sampling.pdf

 

-Paul

 

 

[Julf]

"That is engineering."

 

Ah, yes. That black art practiced deep down in the depths of Mordor. By evil Engineers, who toil on dark devices designed to destroy the life and freedom of pure-hearted, fair-haired and blue-eyed Audiophiles.

 

[Julf]

"And this one too...

 

http://www.csis.ul.ie/dafx01/proceedings/papers/duxbury.pdf[/i]"

 

Indeed, interesting paper - I especially like the way he applies concepts from phase vocoding, something that was a very hot topic back when I studied music theory back in the early 90's.

 

The only thing is that I don't quite see how this applies to any of our discussion about resolutions and sample rates. Maybe you could help me by explaining the connection?