Log in
Log in
Forgot your password ?
or
Create an account
or
Log in using a Facebook account
EN
FR
NotificationsNo new notification
Sound Techniques
News
Articles
Media
Forums
Tutorials Tutorials & Tips

96,000kHz or 44,100kHz

  • 17 replies
  • 17 participants
  • 16,935 views
  • 4 followers
96,000kHz or 44,100kHz
#1
i have witnessed alot of arguements about wether recording at 96,000 or 44,100 gets a better sound. i personally see no point in recording at 96,000. it takes up too much ram, hard drive space and unless your ear is extremely acurate and i mean really really acurate you can't tell any difference. im interested in what you all think.
#2
I agree, I have recorded both ways, and I can tell no difference. I mainly just give my CDs to freinds and family, so they have never been reviewed by "experts".
#3
This is mathematics!! ;-)

I agree there is no big differences... but the reason of 96kHz recording is based on interferences behaviour:

do you remeber that:

cos a + cos b = 2*cos((a+b)/2)*sin((a-b)/2))

So if you are mixing two signals: the first with a [a1,b1] frequencies range and the second with [a2,b2] frequencies range, when you mix this two signals, the result is a new signal with [abs(a1-a2),b1+b2)] range.

It means that, if you are mixing two non-audible signals together(from 18Mhz to 25Mhz for ewemple), you can optain frequencies you can ear!!!

That is one reason it could be interesting having a 96kHz format.

The second reason is the quality of the filter and you convert from analog to digital, it's better having a tolerence between the frequencies of the filter and the limit of your ears. This is the reason proffessional format is 48kHz instead of 44kHz...

So I personaly choose 48kHz for the last reason, but this is true 96kHz is perhaps a bit too much for home studio recording ;-)

mmmm... am I clear? ;-)
#4
i was reading in a magazine that human ears cannot tell the difference between the two
#5
the human ear can only hear frequencies up to about 22Khz, so if something is sampled at anything faster that 44Khz (remember nyquist's theorem fs >= 2fc) then the human ear wont hear any difference. its only if you are adding effects that it would be beneficial to have eg 48Khz or 96Khz because you are basically performing mathematical algorithms on the signal and you are bound to get some unwanted frequencies appearing if you use a low sampling frequency.
so if you are not adding effects, use 44.1Khz, otherwise you are just waisting memory :?
#6
well said!
#7
And don't forget the speakers. It is very simple because if you are using hi-fi computer based speakers you don't hear any change, but if have the money to buy some Monitor speakers ( Good Quality Speakers i mean ) then you hear the changes even from 96 to 192!
#8
that is all thrue, but if you going to edit that recorded sound and add some effects you will hear the difference in final mix down
#9
Well in the analog world I can really hear the difference between a very very popular 20KHz response mixer and my 50KHz response Allen & Heath.

So I say buy more ram and a few hard drives. Put every project on its own hard drive. Cheaper than 2" tape.
#10
Human hear can not even hear 20KHz so 96 is useless but also I agree with mathematic so if you have space and speed go for 96, personally I use 44.1
bye
#11
Wouldn't 96 be better for recording the classical guitar with all it's harmonic overtones etc?
#12
96 khz and 192 khz are all very well, but are the converters in your kit capable of resolving at that speed without jitter ?
Look up 'jitter'
Only the top-end converters can do it, with super-accurate clocks e.g. Digidesign
Consumer gear doesn't have a hope.

John
#13
Two points; One, to record at 96 KHz you need to be good. This is going to point out every little mistake you make. Two, I have had more problems downsampling and upsampling than anything else in my engineering career. Just stick with the industry standard. If you really want some more "fidelity" which you won't even notice, go with 48KHz.
#14
i have only skimmed over the above, but if you are recording at 96 or 48, then remember that a cd can only contain 44.1 at 16bit!!
#15
I agree that 41,000 is the best from what i've eard here and ther places, however, i dont know what 96,000 sounds like so i cant really say.

Oh well, looks like ui guys know wat ur talkin about! lol

Whipped.
#16
96,khz all the way if you can afford it and especially if you want to match a commercial release. The reason for the higher sample rate is the quality of the truncated waveform when re-sampled down to 44,100khz. Its also important to have 24bits too (dithering).
#17

Quote: i have only skimmed over the above, but if you are recording at 96 or 48, then remember that a cd can only contain 44.1 at 16bit!!


Quote: 96,khz all the way if you can afford it and especially if you want to match a commercial release. The reason for the higher sample rate is the quality of the truncated waveform when re-sampled down to 44,100khz.


Don't know if it's true but I've always heard that since everything has to be converted back to 44.1 in the end then it's wiser to simplify the conversion to avoid potential problems, thus it's better to work in either 44.1 or 88.2 rather than 48 or 96... Now clearly you should hear more differences between 16 and 24 bits than you'd do between 44.1 & 48kHz.
#18
Ok, so just my two cents: :D:

Everyone is correct, the Human ear cannot hear above about 22kHz (this maximum range decreases over time).

As stated above, the math does workout to be that a higher sampling rate can lead to interesting results that audible. However, Lavry Engineering published a very interesting white paper about just this very question:

http://www.lavryengineering.com/documents/Sampling_Theory.pdf

In short, he points out that 96 is beneficial in terms of frequency response when recording. When recording at 96kHz, you get a close to flat response in recording over the audible range, versus sampling at less than 96kHz which can lead to a gradual decrease in gain across the higher frequencies. The lower the sample rate, the more aggressive the decline in sound levels of the higher frequencies.

It is a great read and he has also done many other papers.

http://www.lavryengineering.com/index_html.html

Go to Support, and scroll down and under White Papers are other articles. Very interesting (esp if you understanding the higher order math involved).

Ciao.