Recording at 44.1 or 48

192kHz vs 44.1kHz!!!!!!!! Why 192kHz if their converters are so great - and it's the filter (not the sampling rate) that makes great sound? Is Apogee just playing everyone for a fool? Or do they know something that you don't?

You know, I think that they might well be playing people for fools - but they are also aware that their converters really do sound exceptionally good. And to be fair, it's not just the filters - the rest of the analog electronics is also significantly better too. But I have to say that if you talk to them about all of this - and I have - you get wry smiles...!

I'd like to see people here try the Rosetta, and then talk about what's best for recording.

I rented one, and I have to say that it performs very, very well. But there again, I also agree entirely and utterly with what Oretez says:

...and, again primarily for other forum readers, I tend to avoid any & all vendors who try to sell the position that any bit of gear (hard or soft) is panacea for recording,  any recording gear is far down the list from skill of practitioner (musician) & content.  If you accept 'room' as gear, that is more important then all the rest of the gear combined.  The mechanical transducer (i.e. microphone) is far more important then all other gear that follows it (in the signal path) . . . rest of stuff pretty much follows the same pattern.

Sample rates really aren't the issue here at all. What the Rosetta does is enable you to hear very clearly what it is you need to be adjusting, certainly, but it can't do this on its own; you need the rest of the chain to be there too. And in most circumstances, the rest of the chain is far more important. I've got converters in stuff here that to all intents and purposes, perform very similarly to the Apogee ones, but I know full well that it's the rest of the chain that really makes the difference - which is why purchasing Apogees can wait for a bit.

Fact: analogue tape is continuous sound

Actually - this is not quite true.  Analogue Tape is composed of particles (or domains) , each of which acts as a "little bar magnet" that has two possible magnetic states. N or S.  It is the sum of the magnetizations of all the particles within the magnetic gap of the read head that yields the final signal.  As these particles pass through the gap their contribution appears and disappears in a somewhat discrete manner.  Since the particles vary in size, their relative contributions vary as well so the process is randomized time and amplitude domain.  BUT, the resulting signal is not truly continuous at the microscopic level and is somewhat akin to an ADC/DAC conversion process with variable magnitude bits, variable sample rate, and with gaussian noise added.  HF bias serves to "dither" the signal to allow the aggregate collection of particles to come to a magnetic position most closely corresponding to the applied recording field, but they will still settle into a final magnetic position that differs from ideal by the remanance of the smallest few particles and the resulting playback signal is not truly continuous in either the time or amplitude domains.

Having done micro magnetic modeling for the data storage industry I am very familiar with the exact nature of these processes.

So, analog tape is NOT as analogue as you may think!!

Similarly, a 20 bit converter has 2^20 = 1,048,576 levels, which exceeds the number of particles within the gap of a tape head at any given time, so
this converter would be more "analogue" that analogue tape (on the amplitude axis).

HF bias serves to "dither" the signal to allow the aggregate collection of particles to come to a magnetic position most closely corresponding to the applied recording field, but they will still settle into a final magnetic position that differs from ideal by the remanance of the smallest few particles and the resulting playback signal is not truly continuous in either the time or amplitude domains.

I think that's called 'distortion', isn't it? But with the best will in the world, you cannot expect the cone of a loudspeaker (or even a headphone diaphragm) to follow these most microscopic of changes faithfully - a simple consideration of the relative mass of the moving parts rules this out completely. So to all intents and purposes the energy that is re-exciting the air does appear to be continuous, whatever form, and with with whatever weaknesses, the intermediary storage medium takes.

Yes, I realise that some extremely detailed analyses of what actually happens with magnetic domains took place when the computer industry attempted to cram more and more data onto HD's, etc - but it all came too late for analog audio recording using tape to benefit in the slightest, I'm afraid. It was pretty much dead in the water already as a general recording medium by the time the advanced research took place, AFAIK.

No would should contend that your recordings, via a V76, do not sound better, to your brain, then ones made without it.

I listen to what my musicians like. And to what they hear. I am a musician myself. Do musicians have different ears than engineers? This is not just "My Brain" in action here as you suggest. The musician is listening how the beat influences / drives a song. The engineer keeps dynamic energy alive. Do musicians / engineers hear the same things?

Musicians have a vision of how a song goes. Colors of sound. Energies. I've been there. Sculpting in 3D space. 4D when you include time. Do engineers see the music in a song beyond a physics theory? Lennon - against cries of EMI lab coat engineers - brought reel to reel tape recorders from sterile engineering rooms right into the heart of making music: the control room. ("All you Need is Ears" / George Martin) This brave act forever changed how music was made. Making life interact is important. Creative moments happen in absence of scientific law or reason.

"Leap of Faith" / Albert Einstein
"The mind can proceed only so far upon what it knows and can prove. There comes a point, call it intuition or what you will, where the mind takes a leap and comes out on a higher plane of knowledge, and can never prove how it got there. All great discoveries involve such a leap of faith."

I'm not even contending that hi-bit, hi-sample rate recording does not sound 'better' to your brain.  For other readers of this forum I'm just pointing out that there is no scientific basis for this interpretation.... And there are serious holes in the type of things you are trotting out for argument. Those arguments seem to raise a question as to how thorough or accurate your understanding of recording, including analog recording, is.

RE: "to your brain"
Consider the brains of several musicians I work with. People who went to Berkley Music College. People who have great ears. They acted as if they also heard a difference recording at 9632.

Regarding scientific basis / possibility for multiple interpretations of events in life:

Example 1. There's more than one scientific understanding of the existence of light.
"Light is both a particle and a wave. Whether light behaves like a wave or like a particle depends on how the light is observed."

Example 2. There's no scientific basis for Uri Geller bending spoons with his mind. But he does it. People have seen him do it. Yet since they can't do it does that make it not real?

Example 3. Bob Ludwig
".... when I first heard that the DVD Video was going to support recording at 96kHz/24 bit I got very excited. We were (to my knowledge) the first mastering studio in America to have 96kHz/24 bit recording ability. Some of our 96kHz gear has serial number 1!! We made the first CD ever done on a high sample rate Sonic Solutions Workstation. All the complaints about digital being inferior in sound to analog fall away when listening to 96kHz/24 bit recordings."

and, again primarily for other forum readers, I tend to avoid any & all vendors who try to sell the position that any bit of gear (hard or soft) is panacea for recording,  any recording gear is far down the list from skill of practitioner (musician) & content.

Though I love V76 harmonic distortion effect, when I work with people, I tell them to forget about perfection. My goal: get blood out of stone in a performance. That's what really counts. A great song. A good singer. Good players. These criteria are always forever upfront in anything I do in music / video.

Here's a great example of going nuts for a mic!!! Zemlin would live reading this. He used ribbon mics on his daughter. What a nice singer!!

I really like / and (FWIW) deeply value hearing everyone's arguments. The knowledge. Though I was a pre-med who went into film then music, I have deep respect for those who can explain the reason why things work.

Hey!!! someone "lend" me an Apogee!!!!

Brewer

Example 1. There's more than one scientific understanding of the existence of light.
"Light is both a particle and a wave. Whether light behaves like a wave or like a particle depends on how the light is observed."

What rot! It's a fundamental concept of quantum mechanics - not 'more than one understanding' at all.

Example 2. There's no scientific basis for Uri Geller bending spoons with his mind. But he does it. People have seen him do it. Yet since they can't do it does that make it not real?

I wouldn't use Geller as an example of anything, except how easy it is to fool people - read this, or any one of the countless other descriptions of his fakery - which involves fooling people into thinking they've seen something they haven't. A bit like the sales pitch for some high-sample-rate audio...

"All the complaints about digital being inferior in sound to analog fall away when listening to 96kHz/24 bit recordings."

Just as they tend to do when listening to the results you get from an Apogee... which, since this has happened to rather a lot of people now, rather proves the point, at least subjectively, that it's not being digital that is inherently the problem...

Hey!!! someone "lend" me an Apogee!!!!

You'll need a few other things as well if you want to appreciate fully how much better they really are. And, I have to say they are not the only good converters around. But unfortunately, the other thing that all these converters have in common, apart from exceptionally good sound, is that they are not cheap.

Creative moments happen in absence of scientific law or reason.

Since nothing happens in the absence of either of these - and their absence does not define creativity at all, I can't possibly agree with this statement. And what Einstein said was also somewhat misleading - virtually all of the 'great' discoveries were made by people who had a good grasp of what went before - and that is very important. You are not going to 'discover' something magical simply out of the blue - it really doesn't happen like that.

" All the complaints about digital being inferior in sound to analog fall away when listening to 96kHz/24 bit recordings."

You should also remember that it was many years ago that Bob said this and convertor technology has improved since then. We can throw more DSP at the digital filters so that the filtering is improved while the linearity has also improved.

There is still no conclusive scientific proof that we need a higher bandwidth than 44.1kHz offers. Indeed - mastering engineer Bob Katz attempted to set up some kind of experiment to see if this was the case in the mid 90's but abandoned the idea when he found that he couldn't hear the difference between an unfiltered signal recorded at 96kHz and the same signal filtered at 22kHz (or thereabouts). Bob is used to listening for minute differences in sound and is familiar with the sort of things to listen for so, if he can't tell the difference, then it is unlikely that many others would hear a difference either. There was also a famous paper from Oohashi, a Japanese researcher, that claimed that brainwave patterns were altered when subjects were exposed to ultrasonic signals but his results have not been found to be repeatable in subsequent experiments.

I don't doubt that you might hear a difference with your gear when changing from 44.1 to 96kHz but that points towards a problem with filtering or some other problem with the convertor rather than with the sample rate itself. It is certainly true that it is easier to design filters for 96kHz operation and that it probably the main reason for its adoption but don't dismiss the lower sample rates too easily.

Cheers

James.

No would should contend that your recordings, via a V76, do not sound better, to your brain, then ones made without it.

I listen to what my musicians like. And to what they hear. I am a musician myself. Do musicians have different ears than engineers? This is not just "My Brain" in action here as you suggest. The musician is listening how the beat influences / drives a song. The engineer keeps dynamic energy alive. Do musicians / engineers hear the same things?

Musicians have a vision of how a song goes. Colors of sound. Energies. I've been there. Sculpting in 3D space. 4D when you include time. Do engineers see the music in a song beyond a physics theory? Lennon - against cries of EMI lab coat engineers - brought reel to reel tape recorders from sterile engineering rooms right into the heart of making music: the control room. ("All you Need is Ears" / George Martin) This brave act forever changed how music was made. Making life interact is important. Creative moments happen in absence of scientific law or reason.

I'm going to fade out here with one more comment (and spending time to make it is something I'm going to regret the instant I post it).  You've been presented with pointers, on a silver platter (by Steve et al), towards areas, subjects in which you need to improve your theoretic understanding of audio, of magnetic storage, digital storage, etc.  before you continue to make statements, the foundations of which can be refuted by elementary physics . . . 

The ease with which musicians, artists in general, are seduced by fashion, over physics, is an internal struggle I deal with almost daily This means that all the comments you made are not merely familiar to me, but at times indistinguishable, to opinions I've expressed regarding slightly different parameters . . . nor am I done being foolish, making foolish statements has that is my typical pattern for learning . . . for me art has always been a process of hurling myself at brick walls (once in a while literally).  My use of 'to your brain' was never penned despairingly!  But, while it is, perhaps, permissible for an artist to submerge themselves in the subjective, if you as producer, engineer are offering a professional service you are going to need to apply objective filters to almost all aspects of a project.  Simple economics, long before aesthetics has to be addressed, demands this.  An average singer songwriter can expend four to five years and burn through six million dollars to capture 14 mediocre songs without coming up for breath!  Helping talent articulate their vision is seldom assisted by pandering to their preconceived prejudices (and that was deliberately redundant).  For every John Lennon (& Lennon was no Les Paul) there are probably a thousand Stanley Awkrights.  Failing to credit the experience, skill (not least of which was managerial) of George Martin does a tragic disservice to reality of what Beatles evolved to become.

I started this by suggesting that presenting a single objective reference to the things you want to believe could help us all reevaluate our positions.  You have insisted on airing a large selection of subjective, even derivatively subjective statements . . . Piling subjective upon subjective (particularly when theoretic underpinning is shaky, easily refuted, at best) in no way strengthens your case.  I also tried to, gently, indicate that talent (and more importantly their managers) with a modicum of intelligence (which/whom tend to fall into the camp that can and will pay for service) tend to eschew tech support that panders to their prejudice.  And it is this that is the lesson that should be taken home from both early & late Beattles, not that they created hit records using an obsolete, constant gain, broadcast mic pre . . . it is the worst kind of magical thinking to imply that clients are going to get into beatles territory because the studio has a V76 . . . & generally speaking clients that fall into that trap tend to avoid paying when they do not achieve fantasy results they expect and even one of those fights can damage long term health of any studio operation.

articulating artistic vision requires subtle, careful listening (to artist).  It does not require, nor is it assisted by adopting the artists' prejudices.  Reality is purely the intersection of those things that leave sign . . . and physics always kicks fashions ass in the end . . . usually sooner rather then later (a reason why marketing, marketers will always default to short term profit rather then long term growth)

Thank you. Really good stuff here. Any good books on audio theory for musicians / engineers?

Today: called a major recording studio here NYC (won't mention the name). It's featured in Mix alot. The key engineer (who's very famous) records 88.2 (as per his asst who I spoke with). The asst records at 44.1 so he can easily interchange between different studios. Mentioned our above discussion. He also emphasized converters. Apogee. Lavry

Made a test recording w/ AA 2.0 : clean guitar /  Fender Tele / EMG 85 pickups. Straight into Countryman FET / DI to Echoudio/Mona (XLR input as 1/4" inputs have pblms on the Mona).
Brief test recordings: 44.1kHz (21 sec) //  48Hz (32 sec)

BTW - The 48kHz test sounds brighter / more "air" in the mix. Could be that Mona has better/different converters per higher sample rates.
RE: 88.2 / 9632 - both record noise on the wav with AA 2.0 - 88.2 less than 9632. How can I record at these sample rates w/o noise in the wav file?

there is no reason that a recording at a higher resolution should sound different.  I own a Layla 3G and recording at any rate above 44.1khz sounds identical. 

There are also a couple holes in your test:

1) From reading your post, it seems like you have recorded two different performances, which is not reliable.
2) The 48khz perfomance was resampled, and both were compressed with a low-quality mp3 encoder (Audition's included encoder is low-quality)

If you want to try a more reliable test, you could get an external recorder, record your performance to that, then play it back, recording  to your DAW at 44.1khz and 48khz sample rates.  Then you would have to post an uncompressed (or losslessly compressed) file from both sample rates.  If there were any audible differences (which there should not be), then it could be a converter issue.


And some more thoughts on the issue, by J. Robert Stuart, the developer of MLP (the audio format on DVD-A discs):

http://www.meridian-audio.com/ara/coding2.pdf

In regards to sample rates above 44.1khz:

"The author has experienced listening tests which showed that the sound is degraded by the presence of normal (undithered) digital anti-alias and anti-image filters. He is also aware of careful listening tests indicating that any supersonic (i.e. >20kHz) content conveyed by 96kHz sampling is not detectable either in the context of the original signal or on its own...It therefore seems probable that we should concentrate even harder on the methods used to limit the bandwidth, rather than spending too much time considering the rapidly diminishing potential of program content above 20kHz."

In regards to 24bit:

"There is no convincing argument for using 24-bit data in a distribution format, [that even] 20-bit channel noise would be inaudible."

Could be that Mona has better/different converters per higher sample rates.

Well, the Mona converters are getting a little long in the tooth now... pretty good in comparison at the time, but things have certainly moved on. If you really want to find out what differences there are at different rates (and I bet there are), then you really need to do RightMark tests at different sample rates and compare them. Unfortunately the only ones available on the site are an incomplete test, and they don't indicate the response up to 20kHz. But even there you can see that the results generally aren't quite as good as the later generation Echo cards.

If you download Rightmark, you can do the tests yourself, but you have to bear in mind that this is a loopback test regime, so it indicates the cumulative response of the A-D and D-A converters. Chances are that you will find differences in the response above 15k at different sample rates, I'd guess. And that will be down to filtering differences at some point - you only ever get one set of converters per system, and they are run at different rates. And on earlier ones, that too could make a difference, because of changes in settling time, etc. And for all I know, the clock jitter could be different as well. All these things will make a perceived difference to the quality of the output.

The whole point is not that the differences you hear aren't real - they may very well be. All we are saying is that this isn't inherently caused by changes in the sample rate.

I own a Layla 3G and recording at any rate above 44.1khz sounds identical. 

Note: Mona converters are not Layla 3G converters. Layla 3G are not even Layla original converters.

When you are dealing with music on computers, the easily availability of several freeware ABX testing programs (e.g. WinABX, PCABX, a foobar200 pluging) makes arguing about audible differences pointless. Many a person who adamantly insisted he hears differences that most people don't has come to the uncomfortable experience of finding out what researchers have know for a long time now: you often hear more of what you expect to hear than what is really there. When you participate in correctly done ABX tests, many audible differences, such as those between 44.1kHz and 96kHz sampling rates, disappear. It's a learning experience, possibly comparable to learning to correctly hear many other aspects of audio.

For several years now, anytime I run across these discussions on various forums, I've been requesting a sample of music in which higher bit depths or sample rates do make an audible difference. I gotten arguments, but so far no one has offered any evidence.

The data is different. There are certain audible differences that can be demonstrated with test tones. This holds out the possibility that some real music somewhere also has audible differences, but nothing seems to have survived objective testing.

Suggest the following test:

1. Take any recording you know well, recorded at 44.1kHz

2. Upsample it to 96kHz (or whatever)

3. In a separate file at the same sampling rate, create 10 seconds of very loud pink noise

4. In FSE mode, select a chunk of this noise, above 23kHz, and copy it, then paste it at a random point into the nothingness above your music file

5. Play it back with your eyes shut, or not looking at your monitor screen

Can you even vaguely guess where the noise begins or ends? No - unless there's something wrong with your soundcard. If you've carried this experiment out as described you've just dropped in loads more energy 'up there' than would ever exist in a real life recording, yet you fail to detect it or notice any change, for good or bad, in the original. Your pet dog or bat might be howling, but you're none the wiser.

Yet, if you can't detect this walloping great big noise being even slightly detrimental (or otherwise) to your recording, how can anyone expect the tiny, tiny amounts of harmonic information which might be present in a regular recording above 20kHz to have any sonic impact whatsoever on the listener?

Can you even vaguely guess where the noise begins or ends? No - unless there's something wrong with your soundcard. If you've carried this experiment out as described you've just dropped in loads more energy 'up there' than would ever exist in a real life recording, yet you fail to detect it or notice any change, for good or bad, in the original. Your pet dog or bat might be howling, but you're none the wiser.

I think it would take more than just a dodgy soundcard - you'd need quite an exceptional pair of speakers to reproduce this anyway, as most have trouble even getting up to 20kHz in anything like a sensible manner, never mind beyond this. So to check this out properly, you probably need something like a cat* (which has better HF hearing than any dog), anyway, just to see if there is anything there at all.

* Izzy gets seriously hacked off if I run my ultrasonic leak detector anywhere near him. That runs at about 43kHz, and it's reckoned that dog hearing only extends to about 35kHz at best, despite claims that it can run up to about 40kHz. The most outrageous claim for cats that I've seen extends their response to 79kHz, but I've seen no proof of this, and can see no good reason why it should, either. But nevertheless, it's considerably further extended than dog hearing. Incidentally, cats also have a better LF response than dogs as well. Generally superior creatures all round, in fact...

I've no idea what my Rogers LS5/8s go up to, and unfortunately the cat is currently out and about - probably lolling in the sun wondering whether it's worth the energy bothering to chase lizards...