Tom
Your PC is close to mine in specs. I have 1@1.3GHz Celeron, an 815 board and 512 MB RAM.
I rarely need more. Shure destructive editting of 2hours recordings can take some time, but apart from that my PC is definitely strong enough for my current needs.
PCI tester
Yes, there should be a way. It would be nice to have a precise mobo benchmark tool on scopeBut seriously, if such a device is trustworthy it might add well to Pulsar's toolset.
I really like the idea, worth elaborating a bit.
cheers, Tom
For those interested : the Pci-Timeline (1 pound) includes 40 delay lines.
When you load 2 you load 80 delays, with 3 120 delays etc ...... (and as much audio channels and vu meters).....
<font size=-1>[ This Message was edited by: spacef on 2004-02-06 05:22 ]</font>
-
samplaire
- Posts: 2464
- Joined: Tue Jun 05, 2001 4:00 pm
- Location: Warsaw to Szczecin, Poland
- Contact:
Can you imagine what kind of work you fund a computer with such device? Billions of operations per second. UhhhAhhh! That's a hard jobFor those interested : the Pci-Timeline (1 pound) includes 40 delay lines.
When you load 2 you load 80 delays, with 3 120 delays etc ...... (and as much audio channels and vu meters).....
This message has nothing to do with pci-bandwith..On 2004-02-05 08:38, valis wrote:
Supermicro P4DC6+
2x1.8Ghz Xeon
512Mb PC800 rdram
etc...
I can load 3 of them fine, but when I load the 4th I recieve the following message:
"A capacity limit of your DSP cards has been reached.
The specific cause is:
Not enough DSP memory: cannot connect
PC Master 32k Delay.Tap3 and Multi VU-Meter 16.in3...."
etc
That means, that there is not enough dsp memory!
yes that's what was explained a bit sooner.
I guess you have less than 15 dsp André ? that's why the delays fill up the memory before pci limit is reached. One day, may be i'll do a real benchmark tool with other internal modules (delays are not adapted, as they also use dsp memory itself, so the test may work but with 15 dsp minimum only) ... (this one just a vu-framer analog style, ie you can do the same with a mixtable vu-board and 40 delays
It would be also nice that someone with More than 15 dsp say if he can load 4 or more....
<font size=-1>[ This Message was edited by: spacef on 2004-02-06 06:46 ]</font>
I guess you have less than 15 dsp André ? that's why the delays fill up the memory before pci limit is reached. One day, may be i'll do a real benchmark tool with other internal modules (delays are not adapted, as they also use dsp memory itself, so the test may work but with 15 dsp minimum only) ... (this one just a vu-framer analog style, ie you can do the same with a mixtable vu-board and 40 delays
It would be also nice that someone with More than 15 dsp say if he can load 4 or more....
<font size=-1>[ This Message was edited by: spacef on 2004-02-06 06:46 ]</font>
You need more dsps with more ramOn 2004-02-05 08:16, kensuguro wrote:
? What is this?
I'm on P4P800 with PIV 3ghz 1gb ram
I get "DSP capacity reached" on the 3rd load. What does that mean?
@spacef:
in this setup : 1 p1 + 1 p2 (10dsps)
<font size=-1>[ This Message was edited by: Andre Dupke on 2004-02-06 06:57 ]</font>
mmm, i don't know how to interprete Wayne's results (unless it's a pci message and not dsp capacity).... (i guessed more dsp would have allowed to add at least one more Timeline, and avoid dsp capacity message due to added dsp memory but apparently it doesn't work that way .... i guess i have to check sfp dsp/pci architecture better (may be next breakfast).
<font size=-1>[ This Message was edited by: spacef on 2004-02-06 09:00 ]</font>
<font size=-1>[ This Message was edited by: spacef on 2004-02-06 09:00 ]</font>
Atomic : No i don't use that one, except may be on the last 2 vus.
What is used is a "normal" delay (the ones with 682 ms or so max dtime) which may use dsp memory at short delay time too (not sure about that, but it supposedly works the same as the Long delay).
"PC delays" have a minimum delay time of 190 ms (at 44.1), so anything under is normally dealt by the dsp.
I guess, the solution would be to use something else that does give hard time to the pci bus, but doesn't use dsp/dsp memory at all (ie no delays, unless i make different settings so there is no short delay dealt by the dsp, may be it's better for the dsp memory, but not sure as long as i haven't tried)....
<font size=-1>[ This Message was edited by: spacef on 2004-02-06 09:39 ]</font>
<font size=-1>[ This Message was edited by: spacef on 2004-02-06 09:41 ]</font>
What is used is a "normal" delay (the ones with 682 ms or so max dtime) which may use dsp memory at short delay time too (not sure about that, but it supposedly works the same as the Long delay).
"PC delays" have a minimum delay time of 190 ms (at 44.1), so anything under is normally dealt by the dsp.
I guess, the solution would be to use something else that does give hard time to the pci bus, but doesn't use dsp/dsp memory at all (ie no delays, unless i make different settings so there is no short delay dealt by the dsp, may be it's better for the dsp memory, but not sure as long as i haven't tried)....
<font size=-1>[ This Message was edited by: spacef on 2004-02-06 09:39 ]</font>
<font size=-1>[ This Message was edited by: spacef on 2004-02-06 09:41 ]</font>
Of course, since the delays you mention do not use PCI bandwidth or RAM, you have to use long delays: only these use PCI bandwidth. Some reverbs contain about 200 such PCI delay lines.On 2004-02-06 09:35, spacef wrote:
I guess, the solution would be to use something else that does give hard time to the pci bus, but doesn't use dsp/dsp memory at all
the delay i mention do use pci bandwith or ram (32K PC delay). it is also supposed to use dsp memory (just like the long delay, when delay time is under 190s at 44.1) (at least if i can read well) Any delay module "switches" to dsp memory, so what happens is normal seen the original purpose of the device.
Both atoms are the same (actually, the long delay is just a kind of good looking chain of the delay above, But i'll try to see if it is better whith longer delay time that may just avoid the dsp memory limit and concentrate on the pci (although it doesn't explain some of the test results?). 200 delays in a reverb, uuh, and it works::??? ! i look small with an equivalent of 40 or so delay lines in Echo2 (but they are managed dinamically, in order to avoid too much pressure on dsp, so it may be a bit different than for reverbs).
There could be atoms that i don't have too...(i doubt) (and what is sure is that i do not know everything neither )
<font size=-1>[ This Message was edited by: spacef on 2004-02-11 15:23 ]</font>
Both atoms are the same (actually, the long delay is just a kind of good looking chain of the delay above, But i'll try to see if it is better whith longer delay time that may just avoid the dsp memory limit and concentrate on the pci (although it doesn't explain some of the test results?). 200 delays in a reverb, uuh, and it works::??? ! i look small with an equivalent of 40 or so delay lines in Echo2
(but they are managed dinamically, in order to avoid too much pressure on dsp, so it may be a bit different than for reverbs). There could be atoms that i don't have too...(i doubt) (and what is sure is that i do not know everything neither
) <font size=-1>[ This Message was edited by: spacef on 2004-02-11 15:23 ]</font>
-
Counterparts
- Posts: 1963
- Joined: Tue Aug 19, 2003 4:00 pm
- Location: Bath, England
-
Counterparts
- Posts: 1963
- Joined: Tue Aug 19, 2003 4:00 pm
- Location: Bath, England