Topic: 48 Khz or 44.1 Khz?

[cpclark]

i use to use 44.1 when i used the ad-20, then when i got the mp-2 and v2 i ran 48 and after transfering and converting to 44.1, i could tell a slight difference in quality on my playback systema after downsampling in soundforge 6, just my .02 and why not get that extra sample in your recording

[Sean Gallemore]

you guys are missing out on over 3,500 samples per second!

[F.O.Bean]

you guys are missing out on over 3,500 samples per second!

that-a-boy

[Cooker]

DVD-A can contain 48K PCM. I'm, glad i record at 48K.

Now if i had a JB3 i'd probably record at 44.1 - i wouldn't be into keeping two sets of master FLACs around for each show....

[phishn]

Ok..only thing that concerns me about 48 khz is that when you resample to 44.1 khz, you are throwing away 3500 samples per second with some kind of statistical averaging.  Wouldn't it be better to have the recording device write the bitstream straight out at 44.1?  In a nutshell, the downsampling will not be identical to the straight out 44.1 recording.

[chase]

bit for bit they won't be the same but it really depends on *how* it determines which bits get tossed.  i seriously doubt anyone could tell the difference between straight 44.1 and the same tape run at 48 then resampled.  i really think this isn't all that important.  if you want 48kHz, run at 48 and resample, if you want to save time, run at 44.1

btw greg, i finally got around to doing some more testing on the dmic that died.  it appears the noise was coming from one of the chips in the board.  i noticed that the underside of the board was partially fried underneath all the pins, and when i touched the chip it was extremely hot.  hopefully i can get someone in the EE lab to help me figure out what's wrong, sure glad i didn't trade it to someone for some blank discs.

[Karl]

There are plenty of people who can tell a recording that has been resampled.  If you are spending hundreds to thousands of dollars for high-end recording equipment, why would you hurt yourself by resampling? Personally, I do all of my recording at 44.1, because I listen to everything off of CD.  I would be missing quality if I recorded at 48 and then just resampled so I could listen with CD's.

[Lil Kim Jong-Il]

Go with the highest possible resolution capture.  

Later you can take resolution away, but you can't put it back.

[phishn]

how do you take resolution away from 48 and have the 44.1 file be identical to one that could be recorded directly to 44.1 k with the taping gear?  hope statistical averaging doesn't smooth out and remove any samples that are important to the recording.  This is my only arguement to NOT resample and record directly to 44.1 since thats what I and everyone else listens to most of the time.

[Lil Kim Jong-Il]

how do you take resolution away from 48 and have the 44.1 file be identical to one that could be recorded directly to 44.1 k with the taping gear?  hope statistical averaging doesn't smooth out and remove any samples that are important to the recording.  This is my only arguement to NOT resample and record directly to 44.1 since thats what I and everyone else listens to most of the time.

I doubt that you could ever resample later and get the exact same file that would be natively produced when recording at the lower rate.  I expect a 48k sample resampled to 44.1k to not sound the same as a native 44.1k recording but you might not hear that on low res systems or in the car.  So you might think that it is better to just record at 44.1k and save yourself the extra step and questionable results of a resample.

But don't assume that CDs/44.1k are here to stay and thats all you will want to listen to.  DVD is becoming the dominant format and DVD players are as cheap as CD players now.  At some point you'll want to step up to DVD.  When you do, you will want better source material - the material at 48K - for your DVDs.   At that point, upsampling from 44.1k may get you a better sound, but it won't have all the original information that you could have had if you recorded at 48k originally.

Look at it this way.  Alot of guys taped using high end RR recorders, even though they knew that would always use cassettes in the car and for trading.  So why bother with the big RR?  When CDs became ubiquitous, almost everyone could enjoy the higher resolution of those recordings that were moved to digital from the high resolution RR source.  So aren't we all better off that they didn't use cassettes to master those shows?

[dklein]

Does anyone know if converting on the fly through your soundcard is any better or worse, than doing a software conversion? I always assumed that hardware conversion wasn't as accurate...

I believe hardware conversion is generally not as good.  Software will employ more recent algorithms.  In the old days hardware had the advantage of speed.  Kinda like video today!

[Ed.]

There are plenty of people who can tell a recording that has been resampled.  If you are spending hundreds to thousands of dollars for high-end recording equipment, why would you hurt yourself by resampling? Personally, I do all of my recording at 44.1, because I listen to everything off of CD.  I would be missing quality if I recorded at 48 and then just resampled so I could listen with CD's.

fair enuf, but if you're dropping thousands and thousands of dollars on recording equipment and thousands more on playback equipment, why not add a computer into your playback equipment.  you can network it to your other computers or even store files on it...put a dvd drive and save your recordings as .flac files on that.  you can easily playback 24/48, 24/96, or even 24/192 flac files now on your playback system.

ed

[Willie T.]

All frequencies above one-half of your sampling rate become aliased and appear as lower frequncies in the sampled signal. This is an inaccuracy that can never be recovered when trying to reconstruct the origianl waveform.  So go for the 48 sampling rate, that way when you do down sample, you should should not have the aliasing that would occur from a direct 44.1 recording.  This link talks a little bit more about why the standard sampling rates were choosen. http://www.promastering.com/pages/techtalk_mac/tt-3_mac.html

[Willie T.]

here is another site that explains a little: http://eamusic.dartmouth.edu/~book/MATCpages/chap.2/2.3.samp_theory.html

[silentmark]

sorry to be off topic here but wtreynol, where is your avatar from, lol, funny stuff ... as far as my two cents goes, I do 44.1 and will continue to do so until the market says otherwise ...