PTravel
Senior Senior Member
I defer to the pro and stand corrected.
Lt. Bob
Spread the Daf!
there are a lot of common misconceptions about piano tuning. so no problem.I defer to the pro and stand corrected.
It's actually a pretty complex subject and there's a lot of essentially mythology about pianos since they're the only thing we use today that is essentially unchanged for the last 100 years. You can often take a part from a hundred year old piano and put it right in a brand new action with no modification!
T
toddyjoe
New member
Do not confuse the math and science of bit depth and sample rate with the implementation. Like miroslav or another poster alluded to earlier in this thread, you can have one piece of equipment operating at 24-bit that sounds audibly different (or better) than another piece of equipment running at 16-bit, but that does not mean you are hearing the difference between 24-bit and 16-bit. It means the design and construction of the two machines are not identical and one is likely hearing those differences (assuming those differences actually exist on a null test and are not subjective or perceived). Even within a single piece of equipment that operates at different bit depths or sample rates, you will have differences in how each bit-depth or sample-rate setting is accomplished and/or differences in how successfully that setting is accomplished that may cause audible differences. It would not be an unusual engineering compromise for an A/D or D/A converter (or a design component like its anti-aliasing filter) to be optimized for or perform better at one setting at the expense of another. Sometimes people even find these very compromises, shortfalls and colorations in a product to sound better to their ears (some Apogee converters, for example). The same design and component differences applicable in hardware apply to the design and programming differences in software and plugins. And if you are tracking separately at different bit depths and sample rates to make your comparison, you have added another variable that makes comparing bit depth or sample rate even less reliable.
Everyone has certain equipment, software, plugins, settings, etc. that sounds subjectively better to them compared to something else. That does not mean the objective math and science of bit depth and sample rates is wrong.
EDIT: By the way, this is not intended to be directed at you, PTravel. More of a general comment upon statements similar to yours that are frequently raised when this topic comes up.
Bobbsy
Boring Old Git
Well, I can think of one major one. Many engineers claim that the process of converting from 24 bit/96kHz to 16 bit/44.1 does more damage to the sound than starting that way in the first place. Engineers are an argumentative lot!
Quite seriously, while there may be reasons (to do with the noise floor) to work at 24 bit), the 44.1/48/96 argument is less clear cut. It has been demonstrated time and time again that in A/B/X testing even the golden ears can't hear the difference in the original recording when they don't know which is which. At the same time, I have to admit myself that once I start applying digital processing (reverbs and other effects) to the tracks I can sometimes hear a difference. To some extent this makes sense to me--once you start applying maths to your digital signal, a more accurate starting point MIGHT give more accurate results. However, I'm also happy to accept that I'm falling victim to my own aural placebo effect.
However, I need to always remember that Nyquist was far more clever than me...
B
Beck
Guest
I agree about being careful with what one is really comparing, but just for the record I take issue with any notion of "Objective math science" in this case. There's not much of that going around. There are no magic tests... "Null test" or anything else that is so definitive as many people believe. It's just another term that will be kicked around the web like "Nyquist" that doesn't solve anything... and doesn't mean anything because its mostly misunderstood. Ask most anyone to give any real insight to Nyquist and you'll never get anything other than the standard line that anyone can google. I've met few people other than myself that can tell you where Nyquist really came from and how regularly its misapplied. Science is only as good as a person's methodology and preconceptions going in... not to mention interpretation of what one thinks one sees.
One of the biggest mistakes made over and over throughout history is that the then current generation has the arrogance of believing there's nothing left to be discovered by future generations. And of course we all snicker when we read about how past civilizations have interpreted the world around them... 500 years ago, 100 years ago and even 25 years ago and less. The best scientist is also a student of history... a true renaissance man or woman. Then you try best you can to look back on the present from the future and thus you see the error of the past... which to us is the present at this moment in time.
Bobbsy
Boring Old Git
Well, Beck, partial agreement but partial disagreement.
First what I agree with you on: mathematics and engineering can't always measure everything. Sometimes you have to rely on ears and listening tests because you might not be measuring the right things.
Where I disagree though is your apparent disregard of structured listening tests. These are the ONLY way to be sure of what you're hearing. Just as I said the differences between effects at 44.1 and 96kHz may be real or may be an aural placebo effect, everyone suffers from the same thing. It is hugely easy to allow yourself to be led by preconceptions. On the other hand, I see no downside in A/B/X testing especially if it involves a sufficient number of people with different hearing and listening skills. I know I'm getting old and might miss something that younger ears would pick up. However this doesn't change the need for true blind testing to be sure of results. The brain is wired to alter perceptions in terms of what it expects--that's a normal human response that affects everyone. However, it does make listening tests highly suspect.
THIS video is excellent in it's explanation--frankly even just watching the first ten minutes (JJ Johnson and Poppy Crumb) gives a real feel for the issues involved.
First what I agree with you on: mathematics and engineering can't always measure everything. Sometimes you have to rely on ears and listening tests because you might not be measuring the right things.
Where I disagree though is your apparent disregard of structured listening tests. These are the ONLY way to be sure of what you're hearing. Just as I said the differences between effects at 44.1 and 96kHz may be real or may be an aural placebo effect, everyone suffers from the same thing. It is hugely easy to allow yourself to be led by preconceptions. On the other hand, I see no downside in A/B/X testing especially if it involves a sufficient number of people with different hearing and listening skills. I know I'm getting old and might miss something that younger ears would pick up. However this doesn't change the need for true blind testing to be sure of results. The brain is wired to alter perceptions in terms of what it expects--that's a normal human response that affects everyone. However, it does make listening tests highly suspect.
THIS video is excellent in it's explanation--frankly even just watching the first ten minutes (JJ Johnson and Poppy Crumb) gives a real feel for the issues involved.
Ethan Winer
Acoustics Expert
As far as I know everything physical can be measured, and certainly everything that can be heard can be measured. Now, there are limits to how far below the noise floor we can measure a signal, or how high up into the GHz range an oscilloscope can work using current technology. But audio is relatively low tech compared to that stuff. Think of what's involved to get a clear photo image of the surface of Mars back to earth! Further, a null test can provide absolute proof that two audio signals are identical. If someone believes that, for example, playing Wave files from different hard drives sounds different, this can be determined with 100 percent certainty using a null test. Whether said person will accept this as proof is another matter entirely! But it's still absolute proof.
--Ethan
PTravel
Senior Senior Member
I agree with you completely. However, I don't think the sole metric in choosing sample rate/bit depth is whether a wave file encoded at a lower rate/depth would "null out" one at a higher rater/depth. I simply don't see how manipulation of sound through application of effects and other processing prior to mastering is going to yield identical results regardless of sample rate and bit depth and, indeed, I can readily hear the difference with some kinds of processing. I'd welcome an explanation as to why I'm wrong, but I haven't heard one yet -- only that the human ear can't detect the difference between two wave files, one at 16-bit/44.1 KHz and the other at 24-bit/96 KHz. I'll accept that it's true, but it's not relevant to my concern.
Bobbsy
Boring Old Git
Badly phrased on my part. Of course anything physical CAN be measured but,
a) Sometimes we don't measure all the right things when doing the basics, and,
b) Measurements can't always determine what sounds "good" to individuals. For example, in theory a mic or a speaker would be best with a completely flat response but often some of the most popular have bumps and peaks that just happen to work well with certain sound sources. Subjective rather than scientific I know--but my expensive measurement mics sound very flat and boring with most real sources.
miroslav
Cosmic Cowboy
Yeah, this is a point that Ethan and I have danced around.
I don't doubt the majority of the math-n-measurements, but I don't feel that anyone should just use those things as a deciding factor for personal preferences.
As an example...I agree with Ethan that companies selling snake-oil remedies at absolutely ridiculous prices with ridiculous claims should be ashamed of themselves...but that said, if someone is going to feel all warm-n-fuzzy buying a $200 power cord...it's OK if they have the money, and it's OK if their personal perception is altered by that.
I don't see a need to always pound some new religion into them just to force them to abandon theirs.
I think we have to consider that audio...music...is mainly about individual perceptions and how it makes us feel, the mood it puts us in. There's no measurements or standards for that, that can be applied to each and every individual like the measurement of an audio spec. If you try too hard to take all the "magic" out of it...you may win your mathematical argument, and in the process just piss off a lot of folks and deflate their dreams.
I mean...it's like going around with a calculator proving to little kids that there really ain't no Santa Clause.
I think it IS important to leave a little "magic" and a little unsolved mystery in the mix...and not try to measure it all out to 0.0001%
I don't like taking a purely analytical perspective when working with audio and choosing and using audio gear.
I mean...yeah, sometimes I just like the damn thing with the fancy shaped knobs-n-buttons...and yeah, it makes me feel like my audio is improved going through it, and when I feel like that, I end up doing things a little better, playing a little better...etc....
Bobbsy
Boring Old Git
I guess I represent sort of a half way house between you and Ethan.
My mixes are done 90% by ear (and where I involve the numbers it's things like avoiding clipping or (important for me in a theatre situation) not letting the quiet passages get so quiet that they'll be lost in a sea of crackling sweet packets, moving lights and air conditioner noise. The rest is what sounds good to me.
However, surrounding that, there are a lot of peripheral things that CAN be determined by the numbers in proper double blind testing. Choosing among 44.1, 88.2 and 96 kHz sample rates is something where I figure I might as well work out if there's any real difference rather than persuading myself I can hear something even if there's nothing there. As noted previously, so far I've still got one minor niggle (whether or not certain effects render better at higher sample rates) but certainly in double blind testing I hear nothing...but thought I did when playing and knowing which sample was which.
On the other hand, I don't care if the numbers prove it's noisy and the low bit rate is icky, I like my little Alesis nanoverb better than some more expensive boxes for some things (live female vocals for one) despite the spec.
It's hard to put into words but a lot of things lend themselves to sorting out by the numbers and double blind testing (bit depth, samples rates, that $300 power cord) but other things can and should be a pure subjective choice (microphones, effects units, MAYBE monitors though that's risky) and so on. I can't put a hard and fast rule on it and will defend to the death my right to be arbitrary on which I want to measure and which I want to use ears on!
My mixes are done 90% by ear (and where I involve the numbers it's things like avoiding clipping or (important for me in a theatre situation) not letting the quiet passages get so quiet that they'll be lost in a sea of crackling sweet packets, moving lights and air conditioner noise. The rest is what sounds good to me.
However, surrounding that, there are a lot of peripheral things that CAN be determined by the numbers in proper double blind testing. Choosing among 44.1, 88.2 and 96 kHz sample rates is something where I figure I might as well work out if there's any real difference rather than persuading myself I can hear something even if there's nothing there. As noted previously, so far I've still got one minor niggle (whether or not certain effects render better at higher sample rates) but certainly in double blind testing I hear nothing...but thought I did when playing and knowing which sample was which.
On the other hand, I don't care if the numbers prove it's noisy and the low bit rate is icky, I like my little Alesis nanoverb better than some more expensive boxes for some things (live female vocals for one) despite the spec.
It's hard to put into words but a lot of things lend themselves to sorting out by the numbers and double blind testing (bit depth, samples rates, that $300 power cord) but other things can and should be a pure subjective choice (microphones, effects units, MAYBE monitors though that's risky) and so on. I can't put a hard and fast rule on it and will defend to the death my right to be arbitrary on which I want to measure and which I want to use ears on!
Lt. Bob
Spread the Daf!
that one's a shocker to me ..... I have one and I find it almost unusably nasty sounding. Just a metal can sounding reverb at its very best.
I use it in one situation where I have to set up two sets of gear
when I've got two solo gigs back to back in different locations and it's better than no reverb at all but only barely.
miroslav
Cosmic Cowboy
Well yeah, that's my point....it should be personal preference all the way.
But honestly, how many folks actually measure anything these days... especially the younger recording generations...?
I'm sure you do, and heck, I measure/test a lot of stuff. I mean, just setting up a tape deck requires a heck of a lot of measuring, and I don't just use my ears...it's all about the numbers.
However, I can still prefer the thing with the inferior numbers in some instances, and THAT is the point. So, proving which of two has better numbers doesn't and shouldn't settle everything in an absolute sense...which I think sometimes seems to be the intent when all the arguments for double-blind tests and measuring are presented.
Acknowledging that we can still prefer something with "less fidelity" often sounds like someone saying you're free to make a choice...but your choice is the worse of the two.....
.....so there's an underlying suggestion that better numbers should equal the better choice.That's that part I don't agree with.
Bobbsy
Boring Old Git
Well, not ALL of the nanoverb. In fact, just one preset--Plate 1 to be precise. It just works for me in the sort of theatre musical gig I do a lot of.
Bobbsy
Boring Old Git
Acknowledging that we can still prefer something with "less fidelity" often sounds like someone saying you're free to make a choice...but your choice is the worse of the two.....
That's that part I don't agree with.
You set me to thinking on this (well, I had to think about SOMETHING while doing laundry and grocery shopping) and may have come up with a better definition of where I decide measurements (or at least blind testing) is best and where personal preference is the appropriate deciding factor.
For me it comes down to the "reality" of the perceptions.
We'd all agree that different mics really do sound different. I don't like the C1000S; others swear by it. That's a valid choice. Similarly, I like Plate 1 on a Nanoverb but Lt. Bob hates the unit. Another personal preference.
However, hearing a difference between 16 and 24 bit recording? Highly controversial. On something like that, I'd rather depend on blind testing since it's so easy to convince myself that I hear differences that aren't there. We've all done that--I refer again to Ethan's video. Whatever you think of Ethan, the video makes some excellent points (and my psychologist wife confirms them for me). I don't want to delude myself with differences that aren't there and I don't want friends wasting money on $300 power cords that don't make a bit of difference in blind testing no matter what you thought your heard the first time you plugged it in because you desperately wanted to think you'd spent that money well.
Does that make any sense?
Lt. Bob
Spread the Daf!
I think everyone tends to think what they 'hear' is an absolute but without watching the video ( I NEVER watch vids on the 'puter since that takes so much time) there have been zillions of studies that show that we tend to hear what we expect to hear.
All of us.
As a long time piano tuner I can tell you flat out that your hearing can be dramatically different from day to day.
Doesn't seem like it should but don't forget ..... most of your hearing takes place in the brain which has control over the process and can thus affect it.
I go by what I think I hear without much regard for anyone else's opinion because I set these things up for what I want them to be.
But I'm not so foolish as to think that what I hear is an absolute ..... it's very clear that it's not.
All of us.
As a long time piano tuner I can tell you flat out that your hearing can be dramatically different from day to day.
Doesn't seem like it should but don't forget ..... most of your hearing takes place in the brain which has control over the process and can thus affect it.
I go by what I think I hear without much regard for anyone else's opinion because I set these things up for what I want them to be.
But I'm not so foolish as to think that what I hear is an absolute ..... it's very clear that it's not.
Farview
Well-known member
If any of it made a night and day difference, there would be no discussion.
Of course a higher sample rate will give you smoother pitch shifting and time stretching. That makes all the sense in the world when doing that sort of processing. A lot if the better plugins will upsample as part of their process for that exact reason. But if you generally make a habit of recording things that are supposed to be what they are, the higher sample rates don't buy you any 'quality'.
Of course a higher sample rate will give you smoother pitch shifting and time stretching. That makes all the sense in the world when doing that sort of processing. A lot if the better plugins will upsample as part of their process for that exact reason. But if you generally make a habit of recording things that are supposed to be what they are, the higher sample rates don't buy you any 'quality'.
bouldersoundguy
Well-known member
That doesn't bother me. What bothers me is when someone prefers one thing over another when there is no difference.
miroslav
Cosmic Cowboy
Yeah...I get where you're coming from, and I don't think the stuff in Ethan's video is "false"...but I then ask myself what is it that is being proven...what is the real intent?
To prove that box ABC is just as good as box XYZ at 1/4 the cost, and we don't need to waste our money, and that there's a scam being run on all of us by a lot of gear manufacturers....
...or is it to change how we listen, and that we should avoid/discard personal perceptions in favor of mathematical analysis for all of our ***subjective*** audio/music decisions....
...or both...?
While the former may occasionally help with some buying decisions (though you know, I still like that damn thing with the fancy shaped knobs-n-buttons, and I'll feel better buying it over the other)....
...I think the latter can put us in a creative box to a degree. That once you yield to the math and analytics, it makes you question every "perception".
"I think I'll use this mic because it sounds warmer/fatter to me....oh, but wait, does it really? Maybe I need to run some tests on my mics to see which one really has the bigger low-mids that we often associate with 'warmer/fatter'......???"
I mean....then it's kinda hard to go with the math for "these" decisions but not for "those"...or do we then live in some half reality/half fantasy world where we use the math when convenient, but avoid it when not.
I agree that audio and music is very math based....but then, so much of it done without ANY consideration of math, and it somehow all just comes out sounding good, that I don't think it's all that critical, unless you just feel the need to measure everything out, and when someone says "by ear" that immediately there needs to be double-blind null test done to prove or disprove whatever it is they are perceiving.
Yeah...some people find a comfort zone with the math, so they can put aside certain perceptions in order to do good audio and mnusic...and yet, some people completely ignore the math for the exact same reason, so they can do good audio and music.
So....coming around full circle.....I still say it's all about *personal preferences*.
miroslav
Cosmic Cowboy
Yeah...that's kinda what was being said in the link Mo Facta provided a few posts back....so then it again raises the question of which rates, and if the only negative argument against higher rates is that they eat up HD space and require more processing power, then that may not be enough, IMO, with all the new CPUs and mass storage devices out there that can handle it.
If there IS some effect on the audio...why NOT use it....even IF it's hard to notice 40dB down?
It's like saying..."there's a smudge on my window, but when I stand 20ft back, I can't see it", yet you still feel compelled to go clean it off.
Well...I get what you are saying...but there apparently IS a perceived difference to THEM, and as long as they are not making YOU go with their perception, I think one needs to allow them that bit of personal choice...that bit of "possibility"...since it may have a direct impact on how they feel/perceive the rest of their audio/music, especially if it's providing a positive effect even if it may be a placebo effect.
I think there's a lot of psychology/medicine/science that would support that approach for many things, since it's about the positive end-result.