Bit rate and depth

There's no debate on the bit depth... use 24-bit.

Sampling rate is another story. Some people say they can hear a different when they use higher rates, some people say they can't. Some will set the rate according to the target media... 44.1kHz if you plan to use the music on CD, 48kHz if you plan to put it on DVD, and higher for Blu-ray. And some say 88.2kHz is best because it down-converts the best to 44.1kHz. For some more info on it, check out this article...
http://www.soundonsound.c...icles/qanda-0712-1.htm

Scott

--
Scott R. Garrigus - http://garrigus.com - SONAR X2 Power! - http://garrigus.com/?SonarX2Power
* Author of the Cakewalk Sonar and Sony Sound Forge Power book series: http://garrigus.com/?PowerBooks
* Author of the Cakewalk Sonar ProAudioTutor video tutorial series: http://garrigus.com/?ProAudioTutor
* Publisher of the DigiFreq free music technology newsletter: http://digifreq.com/?DigiFreq
* Publisher of the NewTechReview free consumer technology newsletter: http://newtechreview.com/?NewTechReview

ardjunc

Thanks Scott.
 
I have been using 24/48 for my own recordings. I'm just not sure how high to go with the sampling rate when the target media is unknown.
 
My view is higher is probably better, but at what cost related to computer resources?
 
Ardjunc

Hi ardjunc,

I have an i5 760 PC with 4GB RAM and an M-Audio Fast Track Ultra.

So our system specs are really close.

A while back I did some extensive testing on my own system regarding Sample Rate. I posted the results and the server crashed and the post was lost unfortunately.

What I discovered was, higher sample rate didn't always provide better sound. Because of differences in DA converters, you are going to get different results on every system.

I discovered that the high end was less distorted and the low end was much more defined at 44.1kHz, and just the opposite at 99.6kHz.

I would suggest creating some test tones in something like Sound Forge at different sample rates. If I remember correctly I used a sweep from 20Hz ~ 20kHz @ 24bit at 44.1/88.2/96 (Since those are the only ones supported by my Fast Track Ultra). Then create projects in Sonar matching those Sample rates and import the track. Make sure you have it set to import at original bit depth.

Then see if you can tell a difference between them. I did see a difference when I did it, but I was surprised at the results. It was not necessarily better the higher the Sample rate went.

That said, I used to have an EMU-0404 PCI Audio Card that went to 192kHz. I loved the way it sounded, but in theory we shouldn't be able to hear the difference between sample rates. It was more than likely some color that the sound card was adding.

I use 96kHz all the time on my Fast Track Ultra. Reason is, if you track with Input Echo on so you can hear the effects real time, there is less latency the higher your Sample Rate. I found that setting my Fast Track Ultra to 88.1kHz causes problems with the driver, so I stopped using it. Scott is right, I have read a lot about people liking 88.1kHz because of how well it converts to CD format.

Good luck.

Bub.

According to converter guru Dan Lavry, the design of an ADC entails some compromises that result in the unit being optimized for a specific sample rate. Units intended for the pro market will therefore be optimized for 96KHz, while prosumer devices will be optimized for 44.1KHz. It's not at all unusual for home-studio users to report better results at the lower rate.

hey all. I have used 44.1 for a lot of years now, and i have never had any need to change. 

I am recording my current project in 88.2 to see if there is any real difference. 

I think perhaps too much stress is put on bitrates/sample rates and not enough on room acoustics,  performance etc etc

bitflipper


According to converter guru Dan Lavry, the design of an ADC entails some compromises that result in the unit being optimized for a specific sample rate. Units intended for the pro market will therefore be optimized for 96KHz, while prosumer devices will be optimized for 44.1KHz. It's not at all unusual for home-studio users to report better results at the lower rate.

Totally agree. I have a modded MOTU 24ioCore, that I've replaced all the OP-AMPs and and added an external BLA clock to for the best results. For the money, it's good with the mods. However, it is correct you said that 96Khz is what certain brands and designs are optimized for. Some design engineers who "qualify" specific A/D chips do so with one rate being priority. If they have a choice between a great-sounding 96Khz, and terrible-sounding 44.1khz vs. mediocre quality for all sampling rates equally, they're likely to go with what sells the unit, 96Khz being the best is can be. 

I'm ashamed to say that I can't hear the difference between 44.1 , 96 or anything in between :-/ 

Bibs

wizard71


I'm ashamed to say that I can't hear the difference between 44.1 , 96 or anything in between :-/ 

Bibs

That's because nobody can.

Guitarpima

wizard71


I'm ashamed to say that I can't hear the difference between 44.1 , 96 or anything in between :-/ 

Bibs

That's because nobody can.

People who still play disco music can. 

Bub



I use 96kHz all the time on my Fast Track Ultra. Reason is, if you track with Input Echo on so you can hear the effects real time, there is less latency the higher your Sample Rate. I found that setting my Fast Track Ultra to 88.1kHz causes problems with the driver, so I stopped using it. Scott is right, I have read a lot about people liking 88.1kHz because of how well it converts to CD format.

Good luck.

Bub.

Yes, I've noticed this, but don't understand the mathematics as to why latency is improved at higher sample rates.
I guess if your computer is up to the task, the higher sample rate seems best, regardless whether you will hear the difference or not.
 
 

I use 88.2 for what I would consider my more important projects and 44.1 for everything else.  I think the sound quality is slightly better at 88.2 but it's so slight that I can't say for sure. The performance is much more of a difference than the sample rate. Another thing to consider is the difference between using a 24 and a 32 bit rate. I hear no difference between the two but a huge difference between 24/32 and 16.

thanks slartabartfast, I had to read this a number of times to get it, and it's a good explanation you've given here,
I think . . . wait, let me read it again. Think I spend too much time in the artist part of my brain, let me switch to my math part.
The "bucket of samples" analogy is helpful, thanks again.
oops, now you've deleted part of it in your edit ?