|
Who actually writes in 192kHz?
|
View this Thread in Original format
| Magnus |
I saw the Deadmau5 thread in the music discussion forum where he discusses degrading audio quality. Apparently he writes his tracks in 32bit at 192kHz. He says this:
in fact, here in front of me i have "deadmau5 - FML.wav" in all it's glory. WAV, 32bit float, 192kHz
deadmau5 - FML
8m:09.889s
Wave IEEE float signed 32 bit,
192000Hz
12288Kbps, Stereo.
717 MB (752,472,064 bytes)
I know there have been numerous threads on the subject of what rate we all write in and most agree on 24bit at 44kHz but this got me thinking, do any of you actually write in 192kHz? What kind of system would it take to write a track at this rate? My soundcard can be set to do this but there is no way my CPU could handle this. I have a pretty beefy system and even at 96kHz, my project quickly kills my CPU with only a few VSTis added so I cannot imagine 192kHz. Working with a file of that size alone would be a pain. Just curious what you all thought of this. |
|
|
| ponsshin |
it's bull^^ nobody writes music at such high sampling frequency. You would overload your computer with just 10 samples lol.
He just uses the settings everybody uses and made the output raw file using crazy settings just to show off and demonstrate his point.
Deadmau5 fail.
edit: that also means that he didn't use any outside samples as they're 16 bit and 44.1 kHz! LOL |
|
|
| MrJiveBoJingles |
| I don't think most digital synths even run at 192 kHz, do they? I don't see the point in writing a track in such a high sampling rate. Higher bit rate has an obvious rationale: lower noise floor. |
|
|
| mysticalninja |
| deadmau5 has the money for a system that can do it, so why shouldnt he |
|
|
| Kismet7 |
Many upper end sound interfaces, some DAW sequencers, and high grade PCs can output this. It does not mean all the samples he used within the project have to be 32-bit @ 192khz, not because computers could not handle this, but because most sample libraries dont make samples of that size and fidelity, for space reasons and practical use reasons. It is kinda pointless past 24-bit 96khz for samples. Yet there are producers and engineers who wont work with anything less than 24-bit for samples, for good reason.
But for bounces I think it makes perfect sense to use as much resolution as your software/hardware allows. So if you have the equipment it is possible. For mix test/listening bounces might want to stick to 24-bit or even 16-bit to conserve space, and then the final mix at 32-bit 192khz or as much as your setup allows.
Im thinking the new i7 generation based cpus,ram,and motherboards will able to do this with ease. |
|
|
| MrJiveBoJingles |
| Humans can't hear higher than 22 kHz, which is the upper bound represented by the 44.1 kHz sampling standard. Most humans in their twenties or older can't even hear past 18 - 20 kHz, and people who have spent a lot of time in clubs are probably on the lower end of that range unless they wore ear plugs every time. So I'm curious: what's the point of representing frequencies that human beings can't even hear? |
|
|
| ponsshin |
| quote: | Originally posted by MrJiveBoJingles
Humans can't hear higher than 22 kHz, which is the upper bound represented by the 44.1 kHz sampling standard. Most humans in their twenties or older can't even hear past 18 - 20 kHz, and people who have spent a lot of time in clubs are probably on the lower end of that range unless they wore ear plugs every time. So I'm curious: what's the point of representing frequencies that human beings can't even hear? |
i've read the same argument on the vinyl vs. digital debate. The article said something about vinyl being able to represent much higher frequencies than digital files because of its analogue nature. Some tracks with instruments such as trumpets which have a very rich timbre at frequencies as high as 30 kHz sounded much better on vinyl.
The mathematical law shows that to be properly represented a certain frequency has to be sampled at a frequency that's at least twice as high. Most people think exactly twice as high is ok which is true BUT your mid high and high frequencies will be poorly defined unless you use very high sampling frequencies.
So yes, it's useless to represent frequencies that can't be heard but that's not the point of using 192000Hz. It's just to get more definition on high frequencies;) |
|
|
| MrJiveBoJingles |
| quote: | Originally posted by ponsshin
i've read the same argument on the vinyl vs. digital debate. The article said something about vinyl being able to represent much higher frequencies than digital files because of its analogue nature. Some tracks with instruments such as trumpets which have a very rich timbre at frequencies as high as 30 kHz sounded much better on vinyl. |
I don't buy that one bit. High frequencies get rolled off of vinyl as you play it more anyway. Don't get me wrong, I like the sound of vinyl, but I put that down to the subtle imperfections and the fact that engineers can't compress things to death quite as much due to physical limitations, not to "being able to represent much higher frequencies."
Further, even if the "represent higher frequencies things" were true, it would be totally irrelevant. Why? Because the vast majority of sound systems don't even play anything above 22 kHz. So anyone who listens to audio at super-high sampling rate on a normal or even most very good sound systems and claims to hear something "special" is simply deluded, as those super-high frequencies aren't even coming out of their speakers, LOL.
| quote: | | Most people think exactly twice as high is ok which is true BUT your mid high and high frequencies will be poorly defined unless you use very high sampling frequencies. |
Evidence? Link it if you've got it. |
|
|
| ponsshin |
Read again what I've wrote: I said that the purpose of using high sampling frequencies was only to better define the higher range of the sound spectrum. I never implied that a speaker could send 44 kHz!
Here is the evidence: let's say the blue sine wave is the signal we want to sample. For the example we'll say that this frequency equals to 20 kHz.
The drawing on top represents the 20 kHz signal being sampled at twice its frequency. 40 kHz seems fairly high enough, close to the frequency used in all mp3s but when you look at it, the 20 kHz is very poorly represented (only a couple values per period).
The drawing at the bottom shows the same signal except it's being sampled at a frequency that's ten times as high. It seems crazy to sample something at 200 kHz but look at the way it's well defined now: 10 values instead of 2.
Hope that helped;) |
|
|
| MrJiveBoJingles |
| I already know the theory, thanks, I'm just wondering what practical difference you think this makes to the sound, considering that everything above 17 kHz is pretty much just noise and hiss. Most people can't even distinguish between different tones at that high level. Audiophiles can spout off vague stuff about "poorly defined frequencies" all day, but can they tell the sampling rates apart in a blind test? |
|
|
| ponsshin |
| quote: | Originally posted by MrJiveBoJingles
I already know the theory, thanks, I'm just wondering what practical difference you think this makes to the sound, considering that everything above 17 kHz is pretty much just noise and hiss. Most people can't even distinguish between different tones at that high level. Audiophiles can spout off vague stuff about "poorly defined frequencies" all day, but can they tell the sampling rates apart in a blind test? |
Again this is not about frequencies above 17 kHz. Try comparing it to HDTV: you're not gonna spend thousands of $$ on a HDTV to see all the little details in the Bluray disc of Iron Man are you? No, what you're looking for is a crystal clear picture and incredible image flow.
Well it's kinda of the same thing in audio. I agree though that using sampling frequencies above 44.1 kHz is useless for the average listener on an average sound system but like the HDTV and the Bluray player you gotta spend money to enjoy the quality. |
|
|
| MrJiveBoJingles |
| quote: | Originally posted by ponsshin
Again this is not about frequencies above 17 kHz. |
Okay, then why bring up the fact that the 44.1 kHz standard might be questionable for representing 20 kHz frequencies? I assume what we're talking about here is distortion, right? While I concede that there might be some distortion at the upper end of the range, I have to wonder whether that will be noticeable on any sound system simply because of the facts about human hearing. Could you tell apart a slightly distorted bunch of 17 - 20 kHz signals from a "normal" one when both of them will sound pretty much like hiss and noise? I'd be willing to put some money on "no."
And if you want to argue that this isn't actually about high frequencies, that frequencies below the high range will be distorted or poorly represented by the 44.1 kHz, then you need to actually give a reason for thinking that rather than merely asserting it. So, why would the 44.1 kHz standard distort frequencies that weren't near the upper bound of representation? |
|
|
|
|
