Subject:Do I sample up or not?

Posted by: luces

Date:6/23/2003 8:00:52 PM

Howdy again. I was wondering what would yield the best results....

I definitely prefer working at 24/96 for obvious reasons, so would it be better for processor performance and general audio quality to take any samples/loops that are at 44/16 and change the sample rate and bit depth, or should I just let Acid w/ the processor do all the work?


Let me be more clear. Whenever I do a project I try to get the playback to be at 24/96 and then do all of the dithering and whatever in Sound Forge, because I definitely hear a positive difference in quality. Also, I have heard arguments that 48/24 is more than sufficient and sometimes prefered to 24/96.
Anyone care to chime in and help me weigh the pros and cons??? Thanks, luces

Subject:RE: Do I sample up or not?

Reply by: Erik_Nygaard

Date:6/24/2003 5:46:50 AM

As you know, more bits and higher sampling rates come at a CPU/disk-price.
I prefer to work at 24/44.1 for regular audio/CD-work. Keeps the audio-quality high while preserving CPU and avoiding resampling. Also resampling/dithering from 96 to 44.1 may add artifacts, it will probably be better to resample 96 to 48 in this case (whole numbers).

Subject:RE: Do I sample up or not?

Reply by: Ted_H

Date:6/24/2003 9:17:00 AM

Resampling audio that is 16 bit/44.1KHz up to 24 bit/48KHz or 96KHz is not going to improve the sound quality at all. It will sound exactly the same.

Ted

Subject:RE: Do I sample up or not?

Reply by: Iacobus

Date:6/24/2003 7:43:45 PM

The only way you'll benefit from upsampling to 24-bit/96 kHz is if you plan to add effects afterward. Otherwise, the file(s) will sound exactly the same.

Using 24-bit is definitely recommended. Using a sample rate as high as 96 kHz is optional in some circles. It's that whole Nyquist frequency thing again.

Theoretically, only half of the sample rate is accurately reproduced. If the sample rate is 44.1 kHz, then about 22 kHz is reproduced accurately—the alleged upper ceiling of human hearing. (There are those who claim to hear above 20 kHz. This will obviously be a neverending issue.)

I'd say upsample anyway. Based upon my experimentation, ACID has an easier time with samples that match the project's overall resolution and sample rate.

HTH,
Iacobus

Subject:RE: Do I sample up or not?

Reply by: spongebob

Date:6/25/2003 8:53:25 AM

Sorry if this sounds like a stupid question, but I'm a bit confused because there are multiple places to adjust the sample rate and bit depth.

Should this be done under file > properties? Should it be done in my Mia soundcard console? Both? Are there other places I should be looking? Does it make a difference?

...and speaking of multiple places to make adjustments, does it make a difference where the buffer size is adjusted? Under options > preferences? Soundcard console? Both?

Thanks,
sponge

Subject:RE: Do I sample up or not?

Reply by: Iacobus

Date:6/25/2003 12:19:08 PM

There are no stupid questions here.

You can't change the actual resolution and sample rate of the sample itself from within ACID. You can, however, bounce down the sample and it will take on the properties of the overall project. (If the project is set at 24-bit/96 kHz, then your bounced-down sample will be 24-bit/96 kHz.)

I'm not familiar with Mia's software console. There should be definitely a setting for sample rate, and usually that setting can either be locked or left alone (it will usually just adjust itself to the sample rate of the project that's playing back).

If you were using the ASIO driver model, usually latency and buffer size is adjusted via the soundcard's own software console. (Any changes made there are reflected in ACID once you apply changes.) If using WDM, you can try adjusting both the playback buffering slider and the advanced properties (both under the Audio tab under Options>Preferences) in addition to your soundcard's software console/mixer.

If I also remember correctly, didn't Echo have a proprietary driver model called PureWave? If possible, I'm pretty sure you should use that model if you have to go with anything other than ASIO.

HTH,
Iacobus

Subject:RE: Do I sample up or not?

Reply by: dkistner

Date:6/27/2003 8:46:56 AM

For what it's worth, I keep my Echo console locked to 44,100 because I read somewhere that Windows is bad about using up everything it can get if you don't lock it, and that can lead to some iffy performance. I think there's something about this in the Echo manual. I use a 32-bit float bit-rate (I record my voicelines first out of Chainer before importing the audio tracks into Acid), and I use the Pure Wave drivers in Echo's console.