capnkid
Optimus Prime
What is the difference? does it matter that much, if I am only recording guitar and vocals?
A
AGCurry
New member
As you may or may not know, you are referring to different sample rates, meaning the number of samples recorded per second.
The main advantage to higher sample rates is the ability to record and reproduce higher frequencies. For CDs, 44.1 kHz was chosen because it translates to 22,050 Hz - which is higher than most people (but not dogs) can hear. So, 96 kHz means a theoretical ability to record and reproduce up to 48,000 Hz.
Some people say they can hear the difference. I can't.
The main advantage to higher sample rates is the ability to record and reproduce higher frequencies. For CDs, 44.1 kHz was chosen because it translates to 22,050 Hz - which is higher than most people (but not dogs) can hear. So, 96 kHz means a theoretical ability to record and reproduce up to 48,000 Hz.
Some people say they can hear the difference. I can't.
bennychico11
...
i would consider the main advantage being a better representation of the wave form first....before i would think about the whole Nyquist thing. Mainly 'cause speakers and microphones hardly respond to much above 20kHz as it is.
Worry about recording a higher bit depth before thinking of getting a higher sample rate. You'll hear better results that way first.
Worry about recording a higher bit depth before thinking of getting a higher sample rate. You'll hear better results that way first.
Massive Master
www.massivemastering.com
There's SOOOO much debate on this...
Here's a thing or two to keep in mind - Conversion is a big part of the equation... I'll take a really nice set of converters at 44.1kHz over a "so-so" set at 96kHz.
If you're using decent converters and your careful to maintain the integrity of the sound, you almost can't go wrong by working in 24-bit at the target sample rate (44.1 is just fine for 99% of what most people do).
Here's a thing or two to keep in mind - Conversion is a big part of the equation... I'll take a really nice set of converters at 44.1kHz over a "so-so" set at 96kHz.
If you're using decent converters and your careful to maintain the integrity of the sound, you almost can't go wrong by working in 24-bit at the target sample rate (44.1 is just fine for 99% of what most people do).
RhythmRmixd
New member
Not to mention that in a DAW environment, 24 bit/96khz audio files eat hard disk space like Fat Albert at lunchtime. 
T
TimOBrien
New member
And will also tax your system... you'll be able to run a lot less tracks and plug-ins.
I'm sticking with 24/44.1 myself.
I'm sticking with 24/44.1 myself.
Teacher
New member
plug-ins tend to sound better at 96khz but a/bing a single track I don't think would yield much difference especially if your doing electronic music
JKestle
New member
my two sense...
I've also heard that mixing down from 96khz to 44.1khz is tough on some lower end converters because of the odd calculation.
88.2khz sample rate could be better because when mixing down to 44.1khz, the sample rate is simply divided by 2...instead of the odd 2.1768707 when going from 96khz to 44.1khz.
But in reality, I doubt you would be able to tell the difference between 96/88.2 & 44.1 with just a vocal/guitar mix.
I've also heard that mixing down from 96khz to 44.1khz is tough on some lower end converters because of the odd calculation.
88.2khz sample rate could be better because when mixing down to 44.1khz, the sample rate is simply divided by 2...instead of the odd 2.1768707 when going from 96khz to 44.1khz.
But in reality, I doubt you would be able to tell the difference between 96/88.2 & 44.1 with just a vocal/guitar mix.
F
fuzzrhythm
New member
I see a lot of people recording at 24 bits and 48.8 sample rate. is the advantage of 48.8 over 44.1 really that significant?
Massive Master
www.massivemastering.com
If the target frequency is 44.1, probably not. If the final SRC is happening in the box, it's most likely worse.
dgatwood
is out. Leave a message.
I've always heard that recording in 88.1kHz or 96kHz was preferable to lower rates despite the Nyquist theorem saying that there should be no difference, primarily because of poor quality filters.
In your sound card, to avoid aliasing (bogus low frequency mis-samplings of high frequency signals... think of car wheels spinning backwards on film and you get the idea), the frequency input to the ADC must (as much as possible) hit a brick wall at the Nyquist point. To achieve that, you do a low pass filter that has a certain dB cut per octave beyond a certain point. The sharper the filter, the better.
A lot of audio hardware doesn't do that well in this regard. Thus, you end up with excessive loss of high frequency content significantly below the Nyquist point.
That said, I've always wondered how one avoids aliasing when downsampling audio. The obvious mechanism of throwing out every other sample should sound positively awful.... Anyway....
In your sound card, to avoid aliasing (bogus low frequency mis-samplings of high frequency signals... think of car wheels spinning backwards on film and you get the idea), the frequency input to the ADC must (as much as possible) hit a brick wall at the Nyquist point. To achieve that, you do a low pass filter that has a certain dB cut per octave beyond a certain point. The sharper the filter, the better.
A lot of audio hardware doesn't do that well in this regard. Thus, you end up with excessive loss of high frequency content significantly below the Nyquist point.
That said, I've always wondered how one avoids aliasing when downsampling audio. The obvious mechanism of throwing out every other sample should sound positively awful.... Anyway....
MadAudio
Damned if I do
A practical difference is that is uses up twice as much disk space. If you record a lot, you may find yourself hurting for storage quickly!
IMO 96kHz for a home studio personal recording is overkill. If you're recording a CD that's going to be released, it may be worthwhile (especially if the chain going into those converters is good). I've gotten great results with 24 bit, 44.1 kHz. I only do 16 bit when required.
IMO 96kHz for a home studio personal recording is overkill. If you're recording a CD that's going to be released, it may be worthwhile (especially if the chain going into those converters is good). I've gotten great results with 24 bit, 44.1 kHz. I only do 16 bit when required.
dgatwood
is out. Leave a message.
Well, I have a rule: Disk space is free.
That's not saying that it doesn't cost money, but if you're ever in a situation where you have to decide whether to make a compromise or add more disk space, you should always add more disk space.
If you shop around, you can get hard drive space as cheap as $0.45 a gig. Put another way, at 96kHz, 24-bit, that comes out to $0.43 per hour of audio.... Disk space is free.
That's not saying that it doesn't cost money, but if you're ever in a situation where you have to decide whether to make a compromise or add more disk space, you should always add more disk space.
If you shop around, you can get hard drive space as cheap as $0.45 a gig. Put another way, at 96kHz, 24-bit, that comes out to $0.43 per hour of audio.... Disk space is free.
B
Bulls Hit
Well-known member
dgatwood said:
That's something I've wondered about too.
If you throw out every other sample of an 88.2 recording, wouldn't you just be left with what you would have had if you recorded at 44.1 in the first place?
P
PeteHalo
New member
dgatwood said:
I have a rule too: Take a look at the whole picture and not just a little part of it.
To add more disk space to your system is easy and cheap but you can't double the PCI bandwidth to move that double amount of data between memory and HD or soundcard. And to do something useful with that data you need to double the processor's horse powers which isn't that cheap either and if you have a dsp card you need two of those to handle same number of effects. The sonic improvement that most of your listeners will never even notice is IMHO not worth the extra cost.
regebro
Insane Genious!
No, because the maths done in filters and reverbs are more accurate.Bulls Hit said:
Is the difference audioble? Who knows. It's almost impossible to test, because it's almost impossible to exactly replicate two recordings.
MadAudio
Damned if I do
"Audioble." Wow, that one is going on the list, LOL!
I think you meant "audible."
I think you meant "audible."
X
xstatic
New member
Personally, I record at 24/44.1 . I am totally with MM on wanting quality 44.1 converters over cheap 96k converters. In my opinion the bulk of the difference lies in how whatever your converter is performs at the different sample rates. In fact, I would be more prone to using the 96k conversion on a smaller project rather than a larger one. The logistics involved in recording jsut a few tracks is much different than recording a band using 32 plus tracks. Also, something like an acoustic guitar and a vocal is more likely to reveal the smaller more intimate aspects of the recording than a full band which leaves some leeway for "covering" or "masking" certain imperfections whether they be concerning recording quality or musicianship. Th only time I run stuff at 96k is for those sparse recordings solely because I believe that those few small differences may be apparent, and because my system will handle 96k with a project of that size.
regebro
Insane Genious!
Wow, you figured that out? Good job.MadAudio said:
J
Jim Y
New member
48K gets used for several reasons...
Audio for DVD video should be 48k.
Some soundcards (Creative for example), only actually run at 48k. It is better to record and mix with this rate and use the DAW or editor program to do a final conversion to 44.1k for CD.
The projects audio may have arrived on 48k media (DAT, Minidisk etc).
If the card also supports 96K (higher Audigys do 48 and 96), a rate conversion to 44.1 will be more accurate from 96k.
Clues for Audio interfaces that only do 48 and 96, is that these will be the only rates provided by the ASIO driver and s/pdif output.
DirectSound and Wave(MME) are not always trustworthy, because Windows provides an SRC (sample rate convertor) for these. Creative cards also use their built in FX DSP to carry out the SRC. Neither of these will be as good as dedicated audio software at doing SRC.
As for working at 96k (or 88.2), the best reason I've heard is that it sidesteps issues with the soundcards converters. Most converters have a "brick wall" low-pass filter to stop them handling frequencies close to the sample rate (anti-alias). The filters work at half the sample rate. Unless very well made, these filters cause audible artifacts due to phase shifts. This is not obvious with standard equipment tests like frequency response or THD - you have to hear it.
The Anti-alias filter side-effects occur around the cut off frequency. With 48k this would be at 22k and so artifacts extend into the audible range below 20k. With 96K, the filter's at 48K and far fewer, if any, artifacts extend down below 20k.
If we go even higher and use 192K, the converters filter doesn't need to be a "brick wall". A gentler roll-off shelving type can be used which should be far more benign.
If you'd like to read a book on the more esoteric aspects of all this, have a look at "Mastering Audio" by Bob Katz. I'd warn you that although it's not a technical engineering book full of formulas, it's still not an easy read!
Audio for DVD video should be 48k.
Some soundcards (Creative for example), only actually run at 48k. It is better to record and mix with this rate and use the DAW or editor program to do a final conversion to 44.1k for CD.
The projects audio may have arrived on 48k media (DAT, Minidisk etc).
If the card also supports 96K (higher Audigys do 48 and 96), a rate conversion to 44.1 will be more accurate from 96k.
Clues for Audio interfaces that only do 48 and 96, is that these will be the only rates provided by the ASIO driver and s/pdif output.
DirectSound and Wave(MME) are not always trustworthy, because Windows provides an SRC (sample rate convertor) for these. Creative cards also use their built in FX DSP to carry out the SRC. Neither of these will be as good as dedicated audio software at doing SRC.
As for working at 96k (or 88.2), the best reason I've heard is that it sidesteps issues with the soundcards converters. Most converters have a "brick wall" low-pass filter to stop them handling frequencies close to the sample rate (anti-alias). The filters work at half the sample rate. Unless very well made, these filters cause audible artifacts due to phase shifts. This is not obvious with standard equipment tests like frequency response or THD - you have to hear it.
The Anti-alias filter side-effects occur around the cut off frequency. With 48k this would be at 22k and so artifacts extend into the audible range below 20k. With 96K, the filter's at 48K and far fewer, if any, artifacts extend down below 20k.
If we go even higher and use 192K, the converters filter doesn't need to be a "brick wall". A gentler roll-off shelving type can be used which should be far more benign.
If you'd like to read a book on the more esoteric aspects of all this, have a look at "Mastering Audio" by Bob Katz. I'd warn you that although it's not a technical engineering book full of formulas, it's still not an easy read!