Building a studio computer

[aubreyz]

Quote:That's the theory. In practice you have to down-sample to 44.1 if you're going to CD (or MP3, or whatever), and the difference between a 96kHz recording and one at a standard sample rate is pretty subtle.I work at 24/44.1.I got into this thread very late, so responding to this point may be a little OT by now. I agree totally with the difference between 96k and 44.1 being subtle, but the quality of the AD convertors makes the difference even more apparent. When I was using a MOTU the difference was hardly noticeable. Now with the digidesign 192 convertors, the difference is more noticeable, and that's not even the best convertor on the block. For virtual tracks... I stay at 48 or 44.1 depending on what the final product will be, but tracking vocals or live instruments on anything major... 96k is worth it. Nuts and bolts work i still use 48 or 44.1 for everything, but if it is something critical 96k does make a difference.

[nickbatzdorf]

I use a Waves L2 for A/D, which is not the absolute best either, but it's still pretty darn good. The difference between 96 and 44.1 on it when you go down to 44.1 is microscopic. If I were recording lots of live tracks, then I'm sure the difference would add up, and the Digidesign 192 probably sounds better still at 192. But for what I'm doing (MIDI + overdubs) I decided it wasn't even worth thinking about.The audio interfaces on all my machines are 44.1/48 (MOTU boxes and Frontier Designs on my Windows machines); at one point I had a 96k interface here, but I didn't bother to buy it.Remember, I'm talking about 96k down-sampled to 44.1. And yeah, I probably should have tried 88.2, but alack.

[nickbatzdorf]

To clarify: Waves L2 -> S/PDIF -> MOTO 2408.

[nomiyah]

I admit I'm still a little confused. I understand that 44.1 is an acceptable standard and a safe bet. So I'm OK with that. But if I'm recording vocals or live instruments, would it sound better to record at a higher rate?Nomi

[aubreyz]

Quote:I use a Waves L2 for A/D, which is not the absolute best either, but it's still pretty darn good. The difference between 96 and 44.1 on it when you go down to 44.1 is microscopic. If I were recording lots of live tracks, then I'm sure the difference would add up, and the Digidesign 192 probably sounds better still at 192. But for what I'm doing (MIDI + overdubs) I decided it wasn't even worth thinking about.The audio interfaces on all my machines are 44.1/48 (MOTU boxes and Frontier Designs on my Windows machines); at one point I had a 96k interface here, but I didn't bother to buy it.Remember, I'm talking about 96k down-sampled to 44.1. And yeah, I probably should have tried 88.2, but alack.I've really wanted to test if 88.2 down to 44.1 sounds any different than 96k--- the whole divisible number math thing, but I really doubt there could be much of a difference. Real world, everyday stuff 48k for me (I do a lot of post work and 48k is what my Avid editors call for).Here's something that I need to AB test to see if it's all in my head, but I use Logic kind of like Giga and stream back into Pro Tools. I could swear that when I stream out 48k to the 192 and it upconverts to 96k that the result sounds better than 48k to 48k. That makes no sense, and I'm willing to admit it may be a perception thing, but maybe there is some kind of clocking thing that makes a difference. Anyway.. OT, but I do plan to AB that and see if it's my ears or my head

[aubreyz]

Quote:I admit I'm still a little confused. I understand that 44.1 is an acceptable standard and a safe bet. So I'm OK with that. But if I'm recording vocals or live instruments, would it sound better to record at a higher rate?NomiNomi,4 out of 5 dentists surveyed can't really tell a difference, so I wouldn't worry about it.

[nomiyah]

Quote:Nomi,4 out of 5 dentists surveyed can't really tell a difference, so I wouldn't worry about it. You just posted a couple different times saying you can hear the difference.

[aubreyz]

Quote:Quote:Nomi,4 out of 5 dentists surveyed can't really tell a difference, so I wouldn't worry about it. You just posted a couple different times saying you can hear the difference.Yes, you are correct, but I couldn't hear much of a difference at all until upgrading to a better AD converter. I'm very lucky to be able to record at 96k without a big difference in resource availability. You've probably picked up on the difference between a TDM and native Pro Tools system by poking around the digidesign forums. I run 4 FW800 drives, HD Accel-3 on a dual 2.0 Mac G5 with 7gigs of ram( and that's not even that powerful a system anymore). I can run a full 96k session (64+ stereo tracks and tons of plug-ins) with few problems. On a typical native system, it would not be worth the resource hogging to try and keep everything at 96k. The differences I'm talking about are very subtle, and without the computer and hardware power to pull it off, I wouldn't feel cheated at all by recording at 44.1.I go to a lot of trouble to get that little bit of extra air in vocals, or a small difference in the overall depth of a mix. But I must concede that Nick has produced some of the best sample libraries available, and I respect his opinion even more than mine (which says a lot, as I'm pretty stuck on myself )It may have sounded flippant, but the point was, don't get bogged down in a sample rate debate. Knowing what to do is more important than the format you do it on.

[nickbatzdorf]

nomiyah, if you have an interface that works at 96k, there's no reason not to use it... unless you find your computer straining, in which case you can stop using it and go back to 44.1/48.AubreyZ, you're thinking of Nick Phoenix. I'm Batzdorf.

[aubreyz]

Quote:nomiyah, if you have an interface that works at 96k, there's no reason not to use it... unless you find your computer straining, in which case you can stop using it and go back to 44.1/48.AubreyZ, you're thinking of Nick Phoenix. I'm Batzdorf. My bad. Got my Nicks crossed. I think now I'll go congratulate Nick Phoenix on how great EQ magazine is... just to even the score