Page 1 of 1

Direct - Button switch-necessity - SOMEHOW SOLVED ;-)

Posted: Fri Nov 14, 2008 12:53 pm
by zounds
I recently set up my new DAW, this time enabled the XTC-Mode, which I did not use before.
Had some troubles, but finally everything works fine now.

Now my question (maybe I just overlook something, help would be appreciated):

I for example record a loop in cubase with lets say Minimax as XTC-"VSTi". To record live with no latency I press the direct-button (to enable the directplay-monitoring). With the button pressed, I can hear what I play, but not what already has been played on that track (as in: I record the first cycle in the loop and want to hear that first cycle while playing in the second one to add some notes).
The problem: I don't hear the before played loop. I am able to off the direct-button while recording, and then I can hear the already recorded notes in the clip, but when I then want to play additional notes, they have the latency (as soon as I press the direct-button again, latency is gone, but also I cannot hear the "old" cycles anymore...

Does someone know a solution to this? (There are more annoying things with the direct-button, but this is by far the most annoying one).

Thanx in advance.

Re: Direct - Button switch-necessity - any "workaround"?

Posted: Sat Nov 15, 2008 12:57 pm
by zounds
Thanks for the hint.

At the moment my ULLI-Setting is at 7ms (44,1Khz). I tried to set a separate latency in this .ini-file for XTC, but got an error-message explicitely saying that I have to remove that entry (where I set the buffersize).

Cubase shows me a latency of 24ms in the devices-tab (Asioscope-driver with ASIO2 24bit).

I guess I'll try to set the overall ULLI-latency to 3ms, but I fear the latency in cubase will still be too big, since the latency I have now is really unplayable, even with slow notes...

I'll report later (will do that tomorrow).

PS: Maybe there's latency-problem because one of my Pulsars is a Pulsar I (but it's not the first in the row).

Re: Direct - Button switch-necessity - any "workaround"?

Posted: Sun Nov 16, 2008 3:45 pm
by zounds
Add the following entry into your /SFP/App/Bin/Cset.ini under item
[XTC-config]:

intBlkSize=x
that was exactly what i tried before. as i mentioned, the result was an error-message at the next start, something like iirc:

"there is an error in the cset.ini, please remove the entry intBlkSize=x"
(yes, the message was really that precise)
and the asioscope-driver couldn't be loaded any more.

I'll give this a second try within the next days and will report (some reallife-issues to solve before...)

Re: Direct - Button switch-necessity - any "workaround"?

Posted: Sun Nov 16, 2008 11:51 pm
by the19thbear
I got the same error message, but that was ONLY because i had written something wrong.. so my guess is that you should try writing it again :)

Re: Direct - Button switch-necessity - any "workaround"?

Posted: Mon Nov 17, 2008 1:43 am
by zounds
wow, that was really helpful now, the19thbear!

it was not a write-error and I got the same messages again:
Image

it seems scope had set the xtc-mode-latency automatically to the highest value (about 24ms), independent of my ULLI-setting 7ms:

Image

i tried other bulk-sizes and ended with scope accepting 13ms (512). f.e. 7ms (256) still got the error message. I've read here something about 13ms is the least value a pulsarI-card can handle in XTC-mode, but it can be faster if the pulsarI is not the first in the z-linked row (or on irq - dunno which irq the pci-ports have, have an acpi2-computer).
well, as you also can see on the second pic, the pulsarI is my last card in the row (4 dsps). maybe lower latency is only working by somehow turning the pulsarI completely off (and abandon the 4 dsps)?

however, i played around a bit with the 13ms now and the latency is now small enough to get viable results in live/loop-recording (it feels MUCH better than 24ms - i somehow think the latency was more than 24ms before, though the value was announced in the device-manager of cubase - where now 13ms is announced).
so i now also ended in not using the direct-button any more - problem solved!

(in case you know a trick to get even lower latency with my configuration i'd appreciate further comments :roll: )

Re: Direct - Button switch-necessity - SOMEHOW SOLVED ;-)

Posted: Mon Nov 17, 2008 11:57 am
by zounds
stardust, thank you so much, you did it!

i knew about the difference, but your post let me think it over.

i thought there is only 1 setting for the order of the cards, applying to both modes (sfp and xtc).
that was the reason i postet only the picture of the sfp-ulli-setting.

after thinking it over, i looked at the dsp-load in xtx-mode (i had the b-2003 xtc/vsti loaded and clicked on the "xtc"-logo for this). big surprise: the pulsar1 was listed first! i doublechecked with the sfp-mode-settings, which were still correct.

so i looked into the cset.ini for the BoardID-entries - there weren't any! (i bought the system around 2003 and let the installs do from the local music store. since i never used xtc-mode before, the error didn't appear before ;-) )

i then corrected the following entries:
[Host-Config]
Numboards=3
Board0=Host

(before it was Numboards=1)

[hw]
numboards=3
intBlkSize=64
[board0]
boardid=1
[board1]
boardid=2
[board2]
boardid=0

(added the boardid-entries and changed the bulksize to 64 as in 3 ms).

everything works fine now! cubase sx3 shows some 2.9xx ms in the device-tab and i cannot feel any latency any more. this is so great. you people here are really the greatest (and always were, for years!). is there any audio-related board with so knowing and helpful people elsewhere in the net? i doubt... :lol: :P

PS: i double checked the sfp-mode card-order again, it still was correct. whithout any entry i had 2 differing card-orders (wrong one in xtc and right one in sfp). with this one entry now (boardid) both orders are correct :-)

Re: Direct - Button switch-necessity - SOMEHOW SOLVED ;-)

Posted: Sun Feb 01, 2009 4:07 pm
by goaFromLivingRoom
I tried all entries described here, but i have still 24ms delay. I use Scope, PulsarII and Elektra. Scope is the host. What could be the problem?
Delay problems.doc
(69.5 KiB) Downloaded 190 times