Lies, none of you handbag swinging ladies have any shlongs.

Pics or it didn't happen.

Lies.

Unlike the sample rate article, I think the link quoted below misses the point of higher bit depths a completely different topic. Bit depths have to do with dynamic range and noise floor and a floating point formats eliminate or greatly reduce errors introduced by multiple mixing stages and dsp processing unlike integer formats that cause truncation and data loss. That's pretty proven despite what the article claims. Even 24 bit audio as an integer format is widely known to be more forgiving while recording (dynamic) music. 
BTW I didn't post that article because I agreed with everything in it. It however covered a lot of good info about sample rate theory. I particularly found the section interesting on how high sample rates can mask errors in converter design and therefore sound better. That's probably a large reason why *some* interfaces sound better at high sample rates. And also why a well designed interface with high quality converters can sound as good or better even at 44.1.
 

Goddard
Sigh... wotta buncha hooey. Facetious scientist indeed.
 
Surprise! ADCs actually sample at frequencies in the MHz even if they only output PCM streams at 44.1/48kHz. And DACs oversample 44.1kHz audio streams (in MHz) too! That's not ultrasonic, it's radio frequency (and relates to why the use of a CD player is prohibited at times on airliners). Not to worry though, decimation and lowpass (and often highpass) filtering fortunately keeps the out-of-band nasties from getting through (at least, its supposed to if things are designed and working properly).
 
Another surprise: the cheapo onboard 'high definition audio" codec chip inside the typical PC/Mac can handle 192kHz digital audio (such as one might find as a primary audio stream on a Bluray disc) just fine (by design).
 
Here ya go Noel, he's written all about bit-depth too:
 
http://www.sonicscoop.com/2013/08/29/why-almost-everything-you-thought-you-knew-about-bit-depth-is-probably-wrong/
 

I muddied the waters and contributed to the forum rubber-necking by asking that bit depth question, and re-triggered the 64-bit DPE thread! Now both topics are being analyzed ad infinitum.
 
I guess I at least have the comfort of the 44.1/48 sample rates, 24-bit files, and 32-bit processing being sufficient for... ever. Regardless of technology improvements, the point everyone fails to point out is that no one's hearing capacity improves... in fact, it gets worse. (I hope that doesn't trigger a hearing-aid technology discussion!)

Noel Borthwick [Cakewalk]
There is a white paper co authored with Intel on AVX optimization in SONAR.

Holy cow... this is a slam-dunk for the threads saying to use Intel over AMD with SONAR!

mettelus
I muddied the waters and contributed to the forum rubber-necking by asking that bit depth question, and re-triggered the 64-bit DPE thread! Now both topics are being analyzed ad infinitum.
 
I guess I at least have the comfort of the 44.1/48 sample rates, 24-bit files, and 32-bit processing being sufficient for... ever. Regardless of technology improvements, the point everyone fails to point out is that no one's hearing capacity improves... in fact, it gets worse. (I hope that doesn't trigger a hearing-aid technology discussion!)

I need to use a hearing aid in both ears, with one ear being much worse than the other, where can I get hearing aids that don't distort music so much?

I just turn the headphones really loud without hearing aids atm though, they work ok to do that for now.

Noel Borthwick [Cakewalk]
Unlike the sample rate article, I think the link quoted below misses the point of higher bit depths a completely different topic. Bit depths have to do with dynamic range and noise floor and a floating point formats eliminate or greatly reduce errors introduced by multiple mixing stages and dsp processing unlike integer formats that cause truncation and data loss. That's pretty proven despite what the article claims. Even 24 bit audio as an integer format is widely known to be more forgiving while recording (dynamic) music.

 
Hi Noel, welcome back to the party. Yes, there are many things wrong in that bit-depth article. And yes, that's a different topic, for another thread.
 

Noel Borthwick [Cakewalk]
BTW I didn't post that article because I agreed with everything in it. It however covered a lot of good info about sample rate theory. I particularly found the section interesting on how high sample rates can mask errors in converter design and therefore sound better. That's probably a large reason why *some* interfaces sound better at high sample rates. And also why a well designed interface with high quality converters can sound as good or better even at 44.1.

 
Well, it's reassuring to hear that you didn't agree with everything in it.
 
Can't agree about high sampling rates masking errors in converter design, which just smells too much of negative marketing aimed at competitors offering 192k-capable converters by a company trying to promote their own 96k-limited converters, which imo was unfortunately echoed by that facetious scientist blogger without any scrutiny. If anything, higher rate converters are more demanding to design (especially in VLSI), and designers don't design them simply to mask errors. Among reasons they can sound better is greater linearity, in that reconstruction filters in the DAC can work with more actual data during interpolation.
 
But yes, a well-designed high quality converter at 44.1 will almost always sound better than a poor one at whatever rate.
 
Btw, Lavry came out with another paper in which he specifically discusses oversampling:
 
http://lavryengineering.com/pdfs/lavry-sampling-oversampling-imaging-aliasing.pdf
 

Sooo...
 
Sampling rates higher than 44.1 through a "perfect" converter will not provide ANY benefit.
 
Many converters AREN'T perfect so increased sampling rates WILL help things sound better due to better accuracy.
 
96khz is where any practical benefits stop no matter how good/crappy the converter.
 
Over 96khz the higher inaudible frequencies can interact creating noise down in the audible range.
 
Ultra high sampling rates can lead to LESS accuracy because??? The current computers aren't fast enough??
 
60-70khz is considered optimal but those aren't standard settings so 88.2 is the closest compromise yet in SOME case 96khz MAY provide SOME minimal extra benefit.
 
44.1 will replicate the highest relevant accuracy possible which is proven by 44.1 being able to produce the same accuracy of a 96khz mix down even with poorly designed converters.
 
======================================================
 
So I may have bastardized some of that info but this has helped me better understand what the heck is going on with samplerates. I was going to start using 88.2/24 and may still do so but just to be ultra cautious or if I'm working on something really important I'll continue using 96/24.
 
One thing that I'm left wondering though... what exactly are these "converters" and what exactly constitutes a "poorly" designed one vs. a "well" designed one? This article almost makes it sound like more of a software algorithm as opposed to some kind of hardware thing which I assumed was the case. Or is it a a tag team type of thing where the algo and the chips or whatever makeup the resulting converter?
 
I could (should) probably just do my own homework but these types of subjects are just a liiiiitttle too sciency and dry for derpy little Beep.
 
Dammit, Jim! I'm a guitarist! Not a physicist!!! ;-p

I'm firmly convinced that the crap I record isn't worth the extra half percent in sound quality vs doubling my disk usage :)
I do feel my soundcard (1820M) performs well at 44.1KHz, and I'm sure I could "double" the sound quality of whatever I output by improving on my mixing skills, let alone my compositional and playing skills. Always consider the application, is all I mean.

Never hurts to dream and plan accordingly :)

Beepster
And I guess my answers may lie in this thread already. Probably should have read the whole thing before posting. Just wanted to type up my conclusions to help me remember.
 

Perhaps, but your answers skipped all the insults and pretentious knowledge posturing.

Given science is  usually not going to be followed. I record at 64000 Hz if I could, but Sonar will not. And I found out that at 96000Hz it to me sounds a bit better for some reason like for headroom a little more and that which I do not think I am imagining. I just do not like 88200Hz because it reminds me of being twice of 44100 Hz and I have recorded already at 48000Hz and Windows MP can play back at that speed. All 24bit, so although I have not tried 96000Hz yet, it was my next stop and going any higher just does not appeal to me.But the little I have recorded at 96000Hz did for some reason sound better to me.
??
I also am not in loudness wars of any kind. And that will be that also. Turn up the volume, because if you go anywhere and they play loud, it is tiring to me after awhile and I do not need to be pounded by sound to hear the beat or the music or anything else. I simply thing that some people have lost their hearing over the years, and nothing is going to really correct that.
Standing close (well for me) about 40 feet away at a concert and seeing the sky change to dark skies seem just to indicate that perhaps since it was really loud they might have had something to do with that. No where else was the sky starting to get dark except there in that area of the concert, the rest of the sky was brighter even if cloudy.
 
I decided then and there that I like my own volume and control of the output listening to music at anytime, and anything else was not going to influence me anymore.

 
Les Paul had echo and reverb and overdubs when he recorded, and even back then, it sounded louder than music had been before. I always laugh at the speaker in "Back to the Future" movie because that, that is funny when he blows it up.
 

John T

Beepster
Ultra high sampling rates can lead to LESS accuracy because??? The current computers aren't fast enough??
 

 
Yes, that's claim more or less, though the problem isn't computers per se, it's an issue for the analogue components in the converter too.

This is the one thing from the article that you could reasonably call contentious. It's Lavry's claim, and his reasoning is sound - faster sample rates potentially lead to less accuracy at each sample point, because the component voltages can't settle fast enough - but as far as I'm aware, nobody's actually blind tested this to see if it actually makes any difference, including Lavry.
 
The other side to this though, is that as in your summary of other points, there are no gains to ultra-high frequency, and that is tested and known.

Thanks for the confirmation on that. I guess at that point the only way to possibly resolve the theoretical issues is through quantum computing (which is like... wow man) and optic circuitry where light is used instead of electric current which although electricity technically moves at light speed is slowed down by the resistance of the conductive material.
 
At least that's my peabrained understanding of such things from various documentaries I've watched and articles I've read that I can barely grasp. Such topics fascinate me but also make me feel very dumb and irrelevant in the grand scheme of the cosmos.
 
For our humble sacks of mostly water trying to make pleasant sounding noise though really... it looks like we've gone WAY beyond what we actually need. In a very short time at that.
 
Not saying it was aliens but...
 
just kidding