Processing Power

Post by **hubird** » Mon Sep 06, 2004 9:15 pm

if you think in therms of GHzs it's NEVER enough...
think musical, you always win

nprime · Post by **nprime** » Mon Sep 06, 2004 11:03 pm

Oh, I agree with everything you guys told him, but I thought I'd just put the question to rest, with that system you wouldn't have any excuses for your gear not being good enough, then you would have to have good ideas after that.

As you well know from some of my previous rants I am a devotee of simplicity.

More reverbs is always a bad thing, and if you can't hear them then what is the point exactly, to say you used them so subtlely that no one knows, I was a bit confused there for a minute.

Many is the time someone thought that a piece of gear was going to make great music; turns out that it takes humans working really hard!

R

cleanbluesky · Post by **cleanbluesky** » Tue Sep 07, 2004 4:54 am

On 2004-09-06 17:49, astroman wrote:
[definetely not - if it's faster then it's just a trick which leaves out every 2nd sample, possibly interpolating the remaining 2 neighbours.

That's not downsampling but just spoiling the sound - it's not about 'even' division but about placing a value at the proper position on a curve.

The 'distance' between the virtual points isn't constant, so every interpolated sample has a 50% chance to be at the right position

cheers, Tom

I would agree with this idea, that downsampling from 88.2 offers no advantage over 96, but I have seen people disagree and I do not believe there is a way to settle the argument decisively. Even ABX tests can be called into question.

On 2004-09-06 19:56, symbiote wrote

The thing with using 2 or 3 reverbs tho, is that you can send different amounts of each signal to each reverb. So by setting 2 or 3 reverb, you can accomplish a pretty wide palette of effects/sound/stuff. By using 5 or 7, you most likely end up setting up fairly identical reverbs on a few tracks, or smearing things up by creating too many different reverb tails that all kind of interact weirdly together.

One situation where I could see using 8-10 reverbs is setting up 10 of them, one for each octave, heavily band-limited. I'm sure that would be quite fun. But I can't try it now

Then again, using a reverb on an insert still has its place for custom jobs, like a 100% wet reverb output for weird atmospheric stuff. Of course you can do the same with aux sends by muting the channel playback, so I guess I'm just being rhetoric (and I also like to process reverb outputs.)

Symbiote had a good idea, using different levels on the sends will give a lot of variation. He also understands why I would want to use a lot of reverbs. There is always experimental need for more, as long as I wouldn't want to overdo it.

On 2004-09-06 22:02, nprime wrote:
To answer your question:
How about dual Xeon 3.06 GHz CPU's?

That should be enough power to do just about anything.

R

A dual or quadruple Zeon workstation would probably do the trick. Can't afford one yet though.

symbiote · Post by **symbiote** » Tue Sep 07, 2004 5:38 am

Hmm, I said I couldn't try 10 reverbs, but now that I think about it (takes a few days to wind up sometimes ^_^), I'm pretty sure I can load 10 Masterverbs in a mixer and start having fun (and yes, that's right, 5 "I" in one sentence! English is also really fun heh (I also like paranthesises (not sure this is a word) alot (Are you confused yet?))) Thanks to 18 DSPs, all the evilest experiments are now possible!

About Xeons, you sure could do anything with them. Someone should cook up (no pun intended) some sort of egg cooking/toaster-thing using all the heat coming from them! Your server farm can now o double up as a neat Parallel Cooking Array to serve your employees delicious meals! Big market for that!

valis · Post by **valis** » Tue Sep 07, 2004 8:35 am

I doubt dual/quad xeons (or even opterons) would give you the performance gains you're imagining...

BingoTheClowno · Post by **BingoTheClowno** » Tue Sep 07, 2004 10:33 am

I think the problem lies in the way the sound is processed: it is hard to use multiple threads to process one audio stream since the processing must be done sequentially.

nprime · Post by **nprime** » Tue Sep 07, 2004 11:29 am

I don't know how Steinberg has implemented the use of dual processors for Nuendo, but I can tell you that the studio I work at has just bought a dual 3.06 system and Nuendo is very happy using all that power!

However I don't see how it would help Pulsar, when it's all about the number of DSP's on the card.

R

BingoTheClowno · Post by **BingoTheClowno** » Tue Sep 07, 2004 11:37 am

Probably they are distributing all the graphics processing in such a way that minimizes the impact on the audio processing. Of course, they are probably also taking advange of the special SSE registers that could speed up math calculations.