The Difference Betweeen 96khz & 192khz - Page 3

Archived from groups: rec.audio.pro (More info?)

On Wed, 28 Sep 2005 03:22:22 GMT, "Lorin David Schultz"
<Lorin@DAMNSPAM!v5v.ca> wrote:

>I miss Fletcher.

We all do. Sign of the times, maybe,

"This Usenet Posting kills Fascists"

Times have changed, and rarely for the better,
But, they are a'changin'.

God bless us, every one.

Chris Hornbeck
"I had a perspective on the booing, too. You
can kill somebody with kindness." -Bob Dylan

Archived from groups: rec.audio.pro (More info?)

Anybody want to talk about vinyl versus CD ?

;-)

Archived from groups: rec.audio.pro (More info?)

David Morgan \(MAMS\) <mams@NOSPAm-a-m-s.com> wrote:
>
>Anybody want to talk about vinyl versus CD ?

Vinyl is obviously better because you can melt it into interesting shapes
with a hairdryer, if you want to make planters and silly hats. CDs have
too high a melting point to do this easily.
--scott

--
"C'est un Nagra. C'est suisse, et tres, tres precis."

Archived from groups: rec.audio.pro (More info?)

In article <dhe7od$2jo$1@panix2.panix.com>, kludge@panix.com (Scott Dorsey)
wrote:

[snip]

>
> Understanding the philosophy of science is not the same as understanding
> scientific principles or methods, any more than understanding the
> philosophy of food will help you mkae a good souffle.

The last undergraduate class I took was philosophy of science. It presented an
understanding of neither.

-Jay
--
x------- Jay Kadis ------- x---- Jay's Attic Studio ------x
x Lecturer, Audio Engineer x Dexter Records x
x CCRMA, Stanford University x http://www.offbeats.com/ x
x---------- http://ccrma.stanford.edu/~jay/ ------------x

Archived from groups: rec.audio.pro (More info?)

Chevdo wrote:

> I know more about science than you ever will, simpleton.

Ah, the truth revealed. Ignorant prick.

Lavry bases his objections to marketing sample rates on the basis of
_science_.

--
ha

Archived from groups: rec.audio.pro (More info?)

"hank alrich" <walkinay@thegrid.net> wrote in message
news:1h3ljzk.4l8deb15sdsirN%walkinay@thegrid.net...
> Chevdo wrote:
>
>> I know more about science than you ever will, simpleton.
>
> Ah, the truth revealed. Ignorant prick.
>
> Lavry bases his objections to marketing sample rates on the basis of
> _science_.


I had a bit of a run-in with Dan when he was on about that - damning people
for advertising 24 bit converters, while happily describing his own device
as '24 bits' and conceding that it wasn't really !

Maybe I misunderstood something somewhere...

geoff

Archived from groups: rec.audio.pro (More info?)

In article <1h3ljzk.4l8deb15sdsirN%walkinay@thegrid.net>, walkinay@thegrid.net
says...
>
>Chevdo wrote:
>
>> I know more about science than you ever will, simpleton.
>
>Ah, the truth revealed. Ignorant prick.
>
>Lavry bases his objections to marketing sample rates on the basis of
>_science_.
>

Well if he were basing his objections to sample rates on the basis of the
'science of marketing', he wouldn't be making objections. You can't even write
a single sentence that makes sense. Even Krueger isn't this bad with language,
and he's pretty bad!

Archived from groups: rec.audio.pro (More info?)

In article <E9L_e.252519$9A2.236486@edtnps89>, chev@dont.com says...
>

>I'd probably commit suicide before finding myself in a mental state whereby
I'd
>be seeking your approval by having you listen to something I've engineered.
>

Err, unless you paid me, of course. $40/hr for life coaching. No more
freebies.

Archived from groups: rec.audio.pro (More info?)

"Chevdo" <chevdo@dont.com> wrote in message...

> Therefore I should be glad when people sabotage their own
> intellect, because they have also sabotaged their ability to compete with me.
> But that doesn't mean I want to hang around and babysit them! I just check in
> once in a while to reassure myself that things are the same..


In each of your most recent posts, Chevy, you make serious implications
that the reality in your mind is that you are too good for mingling with others.

Select some heavily armored clothing when you finally enter the real world.
This place is a sandbox of puppydogs in comparison. Practice humility, it
will get you out of hot water one day.

Cheers,

DM

Archived from groups: rec.audio.pro (More info?)

In article <n8O_e.10696$WT3.9698@trnddc03>, mams@NOSPAm-a-m-s.com says...
>
>This place is a sandbox

That's the most astute observation I've seen you make! Oh wait, I already
said it.

Archived from groups: rec.audio.pro (More info?)

In article <ArO_e.862$4h2.18@newsread3.news.pas.earthlink.net>,
schopenhauer@thinkers.org says...
>
>Chevdo wrote:
>> And understanding scientific principles is not the same as implimenting
>> methods. There are plenty of people who impliment technology, design
circuits,
>> etc, who look up how to impliment it when they need to, and don't bother
>> learning any 'scientific principles' beyond ohms law. In fact, I'm accusing
>> everyone who has been trolling me of that, and I'm giving them the benefit
>> of the doubt that they even recall ohm's law.
>
>There are plenty of people who check their spelling before attempting to
>impress on the basis of academic depth.
>
>

But have you ever seen me put an apostrophe on the word "its" when using it as
a possessive noun for a non-human object?

Obviously I've got more important things on my mind than spelling. I'm trying
to keep track of my apostrophes right now! I did notice I was particularly off
with spelling tonight, however. I happened to have a headache earlier, but I
don't suppose that'll save me from being raked over the coals.. OH DEAR.

Archived from groups: rec.audio.pro (More info?)

In article <1128001220.927063.324610@f14g2000cwb.googlegroups.com>,
huwgareth@my-deja.com says...
>
>
>Chevdo wrote:
>> In article <1127912593.636004.243430@g44g2000cwa.googlegroups.com>,
>> huwgareth@my-deja.com says...
>> >
>> >
>> >Why do you think that being familiar with a practical joke played on
>> >postmodernism makes you know more about science than someone who
>> >doesn't keep up with Social Text (or Skeptical Inquirer)?
>> >
>>
>> I wonder if you googled it or actually knew what I was referring to. I'm
>> guessing you googled.
>
>You are the most arrogant idiot I have come across in usenet for some
>time. *Of course* I'm familiar with the Sokal controversy.

Oh yes, of course...

>It was
>reported in the New York Times, you dumbass!

So? Do you have any idea how many words are published in the New York Times
every day? Do you read all of them?


>And Skeptical Inquirer
>can't go an issue without reporting on it.

Hmm, I don't see them mentioning it very often.

> I also get Nature, which
>covered it in detail. You dumb, dumb adolescent moron. Just because
>your flatmates haven't heard of Sokal, don't assume that outside your
>group of acquaintances no-one else has.

Great, well as I pointed out, knowing about it doesn't guarantee that whatever
you think about it is accurate. Furthermore, I accused Shultz of not knowing
about it, not you. And I already answered your question as to why I thought
that was relevent in comparing our abilities to understand 'science'. That
accusation was issued in response to his claim that he is far more
scientifically literate than I am. What do YOU have to do with that? Nothing.



>
>
>> Many of the people who are familiar with that 'practical
>> joke' are scientifically illiterate, on both the 'skeptic' and 'post-modern'
>> sides of the debate.
>
>Actually, I studied Physics at Cambridge University. How's about you?
>

Autodidact. But who cares where you studied physics (no capital P, by the
way)? I never claimed to be more scientifically literate than you, or that
'no-one' knows about Sokal except me. I guess you just wanted to toot your
horn, because you are a putz?

I can hear the difference between 44.1 and 96k very easily, but that's with 32 bit floating point format. some times it's like "why does this recording sound so inferior". Oh, I recorded at 44.1 16 bit. I am suprised someone would feel there's no difference. Especially when you start itb processing.

In my humble opinion, 192Khz /24bit, is the way to go, now and for the future, and even with small common hi-fi systems at home... yes.

I just realized this a few months ago while digitizing some nice old LP albums :

1 - i first wanted to digitalise these LPs in 96Khz to have a better quality than CD : Because everytime I take out my old nice LP-player with a good diamond, i just find the sound terrific, plus i say myself everytime : "I've forgotten how GOOD the music can feel!" The habit of the 44Khz/16bit CD-poor-quality just make us deaf !

2 - So i put a nice LP on my LP-player, the computer, the sound card, and started to compare the music going out of the LP with the one going thru the in/out of my Little ECHO Indigo sound card at 96K, at 44,1K... And, even if it was a little less worst at 96K than at 44,1K, the comparison with the fine analogue LP sound, just made me realized how ugly the digitalization of this sound card was. (same EQ/ same amplifier output/ bypass tests)

3 - I then ordered a 800 Euros FIREFACE 400 (48khz/96Khz/192Khz) influenced by user opinions on forums. Well. Of course this great sound card is an other world than the little Portable ECHO, especially for mic inputs. But my LP sound going in and out from this card was still FAR, FAR, in term music sensation, from the original analogue LP sound. Plus i felt the sound a little less metallic at 48K than at 96K or 192K, don't ask me why..

So this is easy to shout that 96K or 192K (or even 48Khz) doesn't make any difference on the whole world, after listening to a card where 96K or 192K make the same sound... but wait for the following.

4 - I ordered the great 2000€ APOGEE ENSEMBLE, influenced by user opinions on forums. (knowing that i can try the cards for 15 days at my very friendly Playback.fr french reseller) First, 96Khz was not bad... and for the first time in my tests, there were no EQ changes introduced by the in/out digitalization of the sound card. Still the LP sound was really more natural and musical. Then 192Khz started to approach rather close to the original analogue LP sound fine definition... but still. I must admit that the analogue LP sound is a little bit warmer, a bit more musical, and finer in the high frequencies than the 192Khz recorded signal from the Ensemble. That said, i was not going to look for a 352Khz system, and i felt rather satisfied, but then knowing and surprised that, even in 2009, nothing accessible reach fully the quality of a good old analogue LP.

Now i can ear some of you guys reading this, saying it's all bullshits... And i think there may be a paradox here : "sound engineers ears" are usually taken by most people as the best kind of "ears" to judge CD quality vs 192Khz... But please note that tests like this are not so easy for sound engineers just because the sound engineers attitude is accustomed to distinguish noises, frequencies, changes in the sounds while working with them. And the problem is - except for some high frequencies for which you can say that they are more metallic on some cards than on the LPs - , the problem is that i mostly cannot ask to my brain to tell me exactly which noises this sound have that the other of poorer definition doesn't have, just because the two sounds are the same !!... they only have a difference of definition, but my brain is making all its best to SEPARATE the instruments from the orchestra and tries to minimize this difference between the two test sounds. If your saying yourself "try listening objectively": objectively your brain won't find any difference between the two sounds : every instrument is at the same place at the same time. But if you let yourself feel the music, then, in the case you like this music, you'll feel a big difference because you'll be able to feel more all the different instruments in the melody : i mean it's not because i cannot ISOLATE the difference that i can't FEEL the difference in the music pleasure. There is nothing exoteric here. Even if some Nyquist-theorem disciples say that "all above 48Khz is uninteresting" can be proved by a little formula, there's a too big step between, a little ideal formula on one side, and all the physics of the sound circuits AD and DA and the brain operation, on the other side, that science, in the future, still have to explain.

5 - I then started to digitalize 90 LPs at 192Khz /24bit. That's 2 Gb for one LP album! But don't be afraid, and think about it: You can still put 250 LP albums like this on the last 500Gb small 2.5″ portable HDs, it worth it.

Wow, during this recording session, i found in my stuff an old LP from the best argentinean jazz/tango violinist which is called Hernan Oliva, i didn't remember i had buy it, and never have listen it. But i knew by heart this wonderful music "Tristezas del Plata" from the CD version that i have since a long time. It was just incredible... to listen for the first time the LP version of this music. It was like a new discovery of the same recording. I could listen so finer emotions and density in his playing, what a pleasure music can be. Imagine you have the CD of the recording of a little concert of an acoustic band you love, and listen to it, often, in many places, for years... And then, AFTER knowing this recording by heart, imagine you could GO to see THE SAME little acoustic concert, having the band playing RIGHT IN FRONT OF YOU : it was exactly this sensation. ...and i was only listening thru my sister's sony regular hi-fi system with subwoofer and satellites, which is actually not that bad, but not even a luxurious system.

Regards,

Alexandre Flet
LoftTango.com

instrument notes aren't irregualr, they all just play a fundamental note, and overtones to infitnity.

Of course you can tell the difference, 192 being much better, but uses more memory.

Last things first- your comment on my "middle C" example: I did NOT suggest that an infinite sample was needed to reconstruct it within tolerable limits as you are considering, but responding to the earlier post claiming that a wave could be reproduced "perfectly" with "two samples". That's usually simply impossible; it can merely be well approximated if you have an unchanging waveform (i.e. a continuous tone, probably produced in a lab) and you're repeatedly sampling, so that in fact you really are using the information from far more than two samples...and remember that a continuous tone is the opposite of music.

According to information theory, that's really the nub of the issue: the shape of a single irregular wave cannot be measured by, or reproduced from, a mere two measurements. Quantitatively speaking, it's not possible to get more information out than are put in. And considering an audible sound made from a series of such irregular waves but each varying a bit from the previous one, as in a note of real music, then two measurements each cycle is still not enough.

Bringing up the "Fourier series" and casting doubts about my background is a bit of a red herring; he started with an infinite sum of sine and cosine functions - that's more than two sample measurements permit. You can't reasonably fit a Fourier with two or three measurements. (You can try fitting if you get a large number of measurements at, say, three samples per cycle, AND the sound doesn't change as you repeat the measurements over and over and over again; but Fourier doesn't work as well with transients.)

The reason 44.1 kHz is almost good enough, isn't because it can contain all the information we can hear; but rather because the sounds in music tend to continue quite similarly, not identically but pretty close, for significant fractions of a second.

To repeat, it is not true that "If you sample at 44.1kHz, then you can perfectly reproduce all information contained in the signal that is below 22.05kHz" unless you're listening to a steady dial tone. That's a misconception from misapplying math, from overlooking the assumption that the sound is steady. If the sound is unsteady - and it is in music - then you have to start all over doing a Fourier or whatever you want to do on the next cycle, and that means you need a high sampling rate. So if you can hear 10 kHz clearly, then 96 or 192 kHz should give you a good measure of each wave's shape.

With respect, I can't find anything fundamental in your post that adds to what has already been claimed earlier in this thread. However, you propose a neat empirical test. Thanks for proposing that "a 15 kHz sawtooth and a 15 kHz sine will sound identical". If you want to prove this point, perhaps you could post or link to these type of files, which could presumably be created with Black Cat Systems' Audio Toolbox (Mac) or Christian Zeitnitz' Soundcard Oscilloscope (PC). (Personally, I can barely hear anything at 8 kHz and nothing at all at 10 kHz, so files would have to be at 5 kHz or so for me to compare.)

The question is not whether we can hear a given high frequency in isolation, which is easy to measure, but whether, within the limits of our hearing, the shape of each of a series of individual waves affects their subjective quality of sound (regardless of whether you want to abstractly model the shape from the sum of harmonics or some arbitrary mathematical model). It's a matter of ear construction that may be only partially linked to high frequency audibility.

"A little knowledge is a dangerous thing. So is a lot." Albert Einstein

OK, now I know that you don't have a clue about digital signal processing or the mathematics of fourier series.

A waveform can be perfectly reproduced if you have at least two samples per cycle of the highest contained frequency. This is always true (with the slight caveat that if you have exactly two samples per cycle, you can have aliasing issues). All waveforms from any instrument are simply linear combinations of sine waves of varying frequencies - it's true that they aren't continuous tones, but they most certainly are combinations of continuous tones. This is what allows the Nyquist-Shannon sampling theorem to work on waves of any shape, not just sine waves.

As I said above, two measurements per cycle of the highest contained frequency. Since instruments have harmonics, this requires a much higher sample rate than simply twice the fundamental frequency. Our ears cannot hear harmonics above 20kHz though, so if you lose those, it does not affect the sound in any audible way. Because of this, a sample rate that is a bit more than twice 20kHz (such as 44.1kHz) can perfectly reproduce any sound to the limits of human perception.

You can reasonably fit a fourier with 2 or 3 measurements per cycle of the highest present frequency. You should understand this.

Bogus. It is more than good enough because it can contain all the information we can hear plus a bit more.

This is a pretty serious misunderstanding of digital signal processing. If you sample at 44.1kHz, you can perfectly reproduce all information in the signal that is below 22.05kHz, steady or not. You can catch transients, you can accurately show each wave's shape to the limit of human audibility (note that a square wave at 10kHz has harmonics well above 22.05kHz which will be discarded, but the difference is inaudible to the human ear).

I can't do it right now (the computer I am on doesn't have a sound card), but if I remember, I'll definitely try it tonight. I can also concoct other waveforms that should sound identical and try them.

(Also: your hearing ends at 10kHz? That sucks. Is it hearing damage from loud noises, or are you just supremely unlucky?)

The shape of each wave can be described as a series of sine waves. All of them which are above 20kHz are inaudible, and as such do not affect the sound. This has been tested before. This is also the point of my test above - since all harmonics of a sawtooth or square wave at 15kHz are above 20kHz, it should be indistinguishable audibly from the fundamental played alone (a 15kHz sine), a test which I will definitely try later tonight.

To me It makes a difference when you listen to the quality of the tunes I think 96hz is cd quality. Try turning the base up and tweeter settings to 100% and check out distortion levels and see if you can tell a difference betwen 96hz cd quality-128.wav and 192bit is mpeg maybe I don't remeber that well to tired to look it up goto www.wikipedia.org and read up on the specs!

I'm sorry to say I feel the same about your position, and further that you haven't given serious consideration to my earlier posts. With reluctance I will make one last attempt, against my better judgment, for the sake of others who may stumble upon this thread. First I should note that I have found in my classes it often leads to regrettable outcomes to teach advanced methods to students who do not fully understand the basic building blocks, as they are inclined to treat advanced methods (analogous to acoustic 'digital signal processing' and 'Fourier series') as magical black boxes and, consequently, in overlooking the assumptions in the methods they merrily use them inappropriately and come to unsubstantiated and frequently wrong results, and continue to do so even after they graduate. It is a strong human instinct to want to have a magical black box that purportedly produces "perfect" results; some adherents of each box get extremely zealous, and this can be taken advantage of by those promoting whatever it is; diet pills, cults, exercise videos, or hypothetically even digital signal processors. But enough philosophizing, let's look at a concrete example.

For a simple example, let's ignore aliasing and consider a single period of one 22-thousandths of a second during which we take two sample measurements (equivalent to one 44-thousandths of a second each, that is, at 44 kHz) using whatever measurement units of amplitude are convenient. Let's say the first measurement happens to come out at "20" and the second measurement is "80". What is the frequency? What "Fourier series" would you construct? Were we measuring a sine or a sawtooth? None of these questions can be answered. That's why 44.1 kHz sampling is not adequate to "perfectly" quantify 22 kHz sound, and the argument that the listening benefits observed from 192 kHz sampling can be dismissed on theoretical grounds is rubbish unless you're listening to a dial tone.

As to why my hearing stops at about 8 kHz, I don't know for certain, but it's quite satisfactory for pleasurable listening to music. I could hear up to 18 kHz when younger. I think it's likely the delayed result of occupational hearing damage. When a student I worked in a loud engine room for several summers, before hearing protectors were commonplace. Cheers.