or Connect
Head-Fi.org › Forums › Equipment Forums › Sound Science › Why 24 bit audio and anything over 48k is not only worthless, but bad for music.
New Posts  All Forums:Forum Nav:

# Why 24 bit audio and anything over 48k is not only worthless, but bad for music. - Page 17

Quote:
Originally Posted by ab initio

It's sort of funny (at least, I get a huge kick out of the fact) that the least important frequencies in music are the ones that make music files so stinking big. It's the highest octaves that take up all the disk space, just for the sake of satisfying Nyquist theorem for those frequencies. Every extra octave makes the file double in size.

Because this thread is lacking in sample calculations (at least to my satisfaction), let's do one:

If we assume that we can hear from 20Hz to 20kHz, then we can hear a 10 octave range. The octaves are (in Hz):

Octave:      Start:        End:    Nyquist:           1          20          40          80           2          40          80         160           3          80         160         320           4         160         320         640           5         320         640        1280           6         640        1280        2560           7        1280        2560        5120           8        2560        5120       10240           9        5120       10240       20480          10       10240       20480       40960

Every time another octave needs to be captured in a recording, the resulting file size doubles. For the less technically inclined (do they every visit this forum?) we call a doubling in size with a unit increase "exponential growth", which is the the mathy way of saying something gets awfully expensive in a hurry. The theoretical minimum size of an uncompressed PCM file (per second) is the [Nyquist frequency] times [the bitdepth of the recording] times [the number of channels e.g., 2 for stereo].

If we ignored the top octave, and only recorded music up to 11025 Hz (i.e., sampled at 22.05 kHz), we would cut down the file size by 50%. If the music is band limited below 11kHz, then the recording  with sampling rate of 22050 Hz would still have exactly the same fidelity as a recording at 44100 Hz (or 88.2kHz, or 192 kHz, or 32 million THz, or whatever 30x super DSD is, etc.). This is throwing out 1 octave out of 10, i.e., the 10% of the range that bigshot mentioned above. Adding additional octaves beyond 20kHz extends the "musical range" by only a small fraction ("musical range" is in quotes, because nobody can even hear those frequencies) at the expense of doubling the file size for each extra inaudible octave.

Interestingly, in my samples above, you can somewhat see this effect - they're all 16 bit, 44,100Hz files, but the ones lacking the >12kHz content compress into a FLAC much better because of the missing high frequency information. The low pass filtered files are around 85-90% the size of the full files, since the high frequency information is difficult to compress. They would be even smaller if I resampled them at 25kHz or so after applying the low pass filter (which would still perfectly preserve the 0-12kHz information).

Quote:

And you know what else we dont need besides frequencies over 12khz?  Letters past W.  I mean who even uses them?  We only need the first 23.

That is actually a perfect example. The top octave is to the rest of the frequency spectrum as the letters X, Q and Z are to the alphabet. Approximately 10% of the alphabet, but responsible for a tiny percentage of words.

Most headphones and speakers don't perform well on the outer octaves of the spectrum of human hearing, neither in extension to the outer limits, nor in balance to the rest of the spectrum. Every day we listen to band limited response on TV, car stereos, portable stereos and radios and we don't complain. Is it nice to have balanced response all the way from 20Hz to 20kHz? Sure, why not? But it better not be a deal breaker for you or you won't find much to listen to until you get that USB port put in your skull! But the important thing is, if you can get balanced response within the range of 40Hz to 10kHz, you've gone the lion's share of the way to getting really good sound. I'd much rather have that than unbalanced with extension all the way out.

Many audiophiles spend all their time worrying about the specks of dust and ignore the elephant sitting in the corner. Relativity is an important thing to have, but it's in short supply in audiophile forums, that's for sure! It might be a psychological thing.

Quote:
Originally Posted by bigshot

That is actually a perfect example. The top octave is to the rest of the frequency spectrum as the letters X, Q and Z are to the alphabet. Approximately 10% of the alphabet, but responsible for a tiny percentage of words.

And yet, despite not being present all the time, they are clearly and demonstrably necessary for an accurate reproduction of a text (or a piece of music).

Quote:
Originally Posted by ab initio

Bigshot,

How much do you think that one's personal preferences in music affect what aspects of sound reproduction they value the most? If I understand correctly, you are particularly fond of classical music, which, I think, tends to have much less content in the highest octave of human hearing, than, say, modern rock or metal, which feature much more percussion (notably the cymbal crashes).

Depends on the piece.  I also listen exclusively to classical and while it's true that if I put my DSP viewer in JRiver to display everything between 9-16 kHz, for many pieces it stays eerily...black.  But not for all by a long shot.  For instance I'm now listening to Albinoni's concerto strings Opus 9 , number 4 in A major (admittedly a somewhat disingenuous choice)  and there's still quite some activity between 8-14 kHz (visual estimate).  If I have the time later today I'll take a couple of pieces from it, apply filters and post them so you can hear for yourself.

Quote:
Originally Posted by cjl

And yet, despite not being present all the time, they are clearly and demonstrably necessary for an accurate reproduction of a text (or a piece of music).

Absolute again. I'm speaking in practical terms, not armchair theory. We live in an imperfect world with imperfect transducers. You can get close as damn it in the core frequencies and it makes a big difference. And you can get close as damn it in the outside edges of human hearing and it doesn't make much difference at all. You can't have 100% of 100% all the time. You pick your battles. But it really helps if you have some perspective to choose which battles to fight and which aren't worth it. It's all relative.

I'm talking about stereo systems to listen to music on here by the way.

Quote:
Originally Posted by ab initio

How much do you think that one's personal preferences in music affect what aspects of sound reproduction they value the most? If I understand correctly, you are particularly fond of classical music, which, I think, tends to have much less content in the highest octave of human hearing, than, say, modern rock or metal, which feature much more percussion (notably the cymbal crashes).

I listen to a lot of different kinds of music, not just classical. Fidelity is fidelity, I suppose.

But I'm not talking fidelity in an abstract sense... I'm talking about relative importance... and that is the concept that isn't getting across. I thought the alphabet analogy was perfect, but it whooshed right on by.

One of the big problems I see in audiophila (and it's just as bad in the Sound Science forum as it is in the rest of Head Fi) is the tendency to look at numbers on a page and not relate them to sound. If I say "take a 1kHz tone at 45dB" how many people do you think would have an idea of what that would sound like? Would they know that it would drive them right out of the room, or would they nod their head in theoretical ignorance and say, "yes... yes..."? I see people going to the mat fighting about frequencies on a headphone chart that really don't mean much at all... and then totally overlook the remarkable performance of the same headphones through the entire range that matters. People buy headphones with V shaped response or "in your face" midrange and worry about the part of the chart all the way up at the very top. That makes no sense at all to me.

Because somewhere in a book it says "humans hear from 20Hz to 20kHz", that doesn't mean that every single frequency in that range is just as important to a guy sitting in a chair listening to music. But the OCD cuts in and people start worrying about molehills while the mountains go unattended. The perfect example of this is the old saw, "Audiophiles worry about the last few percent." Yes, that is undeniably true... so true in fact that they usually have great performance where it doesn't matter and mediocre performance where it does!

There are ways of prioritizing improvements in sound quality that don't involve factoring out to more and more decimal points. For instance, all things being equal, simply going from 2 channel stereo to 5:1 is a MASSIVE leap in sound quality. Add a good DSP to that and you are squaring that huge improvement. But people are more interested in speaking about noise floors or distortion at -90dB being theoretically audible. They worry about frequencies they won't even be able to hear in ten years! There are things about sound reproduction that most people don't even begin to address. Most of that involves room acoustics and sound fields.

Crazy town. I just want music to sound realistic. That isn't hard, and it doesn't depend on frequencies I can barely hear. In fact, I can get spine tinglingly realistic sound out of a 1914 Caruso record and my upright Victrola. I'll give you a hint how that is... My Victrola has a horn designed to project sound in a specific direction using the shape of the room to put a real dimensional acoustic envelope around the recorded singer. That can have more of an impact on realism than the drastically reduced frequency response of an acoustic recording.

I know these are things that aren't published as specs or discussed in hifi forums. But it's all relative to the degree of realism you achieve with your system.

Edited by bigshot - 5/6/14 at 5:59pm
Quote:
Originally Posted by bigshot

That is actually a perfect example. The top octave is to the rest of the frequency spectrum as the letters X, Q and Z are to the alphabet. Approximately 10% of the alphabet, but responsible for a tiny percentage of words.

O Fortuna,
velut Luna
statu variabilis,
semper crescis
aut decrescis;
vita detestabilis
nunc obdurat
et tunc curat
ludo mentis aciem;
egestatem,
potestatem,
dissolvit ut glaciem.

Sors immanis
et inanis,
rota tu volubilis,
status malus,
vana salus
semper dissolubilis;
obumbrata
et velata
mihi quoque niteris;
nunc per ludum
dorsum nudum
fero tui sceleris.

Sors salutis
et virtutis
mihi nunc contraria;
est affectus
et defectus
semper in angaria.
hac in hora
sine mora
cordae pulsum tangite!
quod per sortem
sternit fortem,
mecum omnes plangite!

Cheers

Quote:
Originally Posted by bigshot

Absolute again. I'm speaking in practical terms, not armchair theory. We live in an imperfect world with imperfect transducers. You can get close as damn it in the core frequencies and it makes a big difference. And you can get close as damn it in the outside edges of human hearing and it doesn't make much difference at all. You can't have 100% of 100% all the time. You pick your battles. But it really helps if you have some perspective to choose which battles to fight and which aren't worth it. It's all relative.

I'm talking about stereo systems to listen to music on here by the way.

I'm talking in practical terms too. I'll be the first to agree with you if you say that something like 256kbps lossy compression sounds very nearly the same as lossless, or that fancy cables are pointless, or that all well-designed solid state amps sound pretty much exactly the same (or, more accurately, do sound exactly the same within the limits of the loads and power levels they are designed to output). However, here you are claiming that in practical terms, the upper octave doesn't matter, and that is simply wrong. I've already provided examples of sound where the upper octave clearly matters, and cutting off the upper 8kHz of human hearing range strongly and noticeably affects the output. This isn't a subtle thing.

(Any change that makes it sound like I've thrown a heavy blanket over my speakers, even when I'm listening casually with a ton of background noise (the A/C is running right now) is not something I'd consider "impractically small")

Edited by cjl - 5/6/14 at 6:20pm

Never was big on the pomp of the Carmina Burana but it reflects the mood in the discussion here I guess.

Quote:
Originally Posted by ab initio

Cheers

Edited by GrindingThud - 5/6/14 at 6:20pm

I'm guessing you guys didn't see what I did there....

This only goes to prove that we don't need Q, X, or Z.

Cheers

Edited by ab initio - 5/6/14 at 6:24pm

I like this version better...

Yes I did (but the relative prevalence of that letter in Latin is quite high, cherry picking).

I saw and it made me laugh..... I still am.
Edit.....never saw cjl's version....now I'm laughing harder.
Quote:
Originally Posted by ab initio

I'm guessing you guys didn't see what I did there....

This only goes to prove that we don't need Q, X, or Z.

Cheers

Edited by GrindingThud - 5/6/14 at 6:28pm

It seems to me, if you can do this without X, Q or Z then you don't need them at all!

`Twas brillig, and the slithy toves
Did gyre and gimble in the wabe:
All mimsy were the borogoves,
And the mome raths outgrabe.

"Beware the Jabberwock, my son!
The jaws that bite, the claws that catch!
Beware the Jubjub bird, and shun
The frumious Bandersnatch!"

He took his vorpal sword in hand:
Long time the manxome foe he sought --
So rested he by the Tumtum tree,
And stood awhile in thought.

And, as in uffish thought he stood,
The Jabberwock, with eyes of flame,
Came whiffling through the tulgey wood,
And burbled as it came!

One, two! One, two! And through and through
He went galumphing back.

"And, has thou slain the Jabberwock?
Come to my arms, my beamish boy!
O frabjous day! Callooh! Callay!'
He chortled in his joy.

New Posts  All Forums:Forum Nav:
Return Home
Back to Forum: Sound Science
• Why 24 bit audio and anything over 48k is not only worthless, but bad for music.