| View previous topic :: View next topic |
| Author |
Message |
opcode Jammer

Joined: 10 Aug 2005 Posts: 42
|
Posted: Fri Aug 12, 2005 8:24 pm Post subject: Using jamstix with DirectiXer in a high latency environment |
|
|
Hi!
I have come across a problem while using jamstix wrapped by DirectiXer 2.5 demo, using Sonar 4 as the host. When the latency is high (i.e. ~340ms) and MIDI-automation is used (I had an envelope for velocity to override the lock on the velocity slider in a manual audio jam with no audiom8 instances running), jamstix plays out of sync and hits random notes - so it seems.
I was successful at reproducing this problem starting with an empty project and inserting only a single audio track for jamstix.
This is likely to be a problem with DirectiXer, because it didn't occur with Cakewalk's own VST adapter, but I thought I'd post it anyway just to let you guys know. When the latency is considerably lower, the problem doesn't occur either. Nor when MIDI-automation is not used...
cheers!
opcode |
|
| Back to top |
|
 |
Ralph [RZ] Site Admin

Joined: 02 Jan 2008 Posts: 13332
|
Posted: Fri Aug 12, 2005 8:34 pm Post subject: |
|
|
Thanks for the heads-up, opcode! _________________ Ralph Zeuner
Rayzoon Technologies LLC
http://www.rayzoon.com |
|
| Back to top |
|
 |
wmountney Jammer

Joined: 06 Mar 2005 Posts: 36
|
Posted: Wed Aug 17, 2005 4:29 pm Post subject: Re: Using jamstix with DirectiXer in a high latency environm |
|
|
| opcode wrote: |
I have come across a problem while using jamstix wrapped by DirectiXer 2.5 demo, using Sonar 4 as the host. When the latency is high (i.e. ~340ms) and MIDI-automation is used (I had an envelope for velocity to override the lock on the velocity slider in a manual audio jam with no audiom8 instances running), jamstix plays out of sync and hits random notes - so it seems.
|
I've encountered this problem too at much lower latencies, say around 30ms, but it seems somewhat intermittent in that case. I kept hearing some doubled drum hits and I finally noticed that it was only happening in areas where the velocity automation was ramping up or down. However, I didn't find a way to reproduce it every time. I'll try increasing the latency more and see if I get the same results. |
|
| Back to top |
|
 |
opcode Jammer

Joined: 10 Aug 2005 Posts: 42
|
Posted: Wed Aug 17, 2005 9:06 pm Post subject: |
|
|
I have now found out that the problem is more serious than I thought it was. Even at lower latencies (~60ms), the problem is there - if not as noticeable. The beat is always recognizable, yet it's as if there is some kind of extreme humanization going on: Many Hits are too late and then there are several hits in a row to make up for it...
The reason why there was no problem with Cakewalk's VST adapter is that it only occurs when using MIDI output (which the Cakewalk VST adapter does not support) - and this time I was not using any automation. DirectiXer works fine with the internal sounds as well.
When I'm recording, which I do with a latency setting of 2.9ms, everything's fine. But later on, as I add more effects during mixing, I need to inevitably raise the latency, so I do hope there is something that can be done about this. Maybe I just have some weird settings, I don't know, because you'd think somebody would have noticed this before... (Or did I not search the forum thorougly enough? )
cheers!
opcode |
|
| Back to top |
|
 |
JohnR Jammer

Joined: 14 Jan 2005 Posts: 30 Location: Newcastle upon Tyne, England
|
Posted: Thu Aug 18, 2005 2:47 am Post subject: |
|
|
I had the same problems that you describe when using Cubase SX2 - this was before Jamstix was released. Ralph started sending debug versions to me to try to resolve the problem.
The problem was caused by high latencies, I didn't realise, but Cubase had somehow switched to the DirectX drivers - I was mixing at the time.
I don't have ASIO drivers for my card, but use ASIO4All (www.asio4all.com) which manages to keep my latency down to around 25ms (when I have lots of inserts) so the problem has gone away for me. Is this worth trying on your system? There's an old thread at KVR talking about Sonar ASIO drivers.
John |
|
| Back to top |
|
 |
opcode Jammer

Joined: 10 Aug 2005 Posts: 42
|
Posted: Thu Aug 18, 2005 6:38 am Post subject: |
|
|
I just tried using ASIO instead of WDM. At the maximum latency setting of 2048 samples (46ms), I do still notice the problem. It's not as obvious, but it's still noticeable.
And I'd prefer to stick with WDM because with ASIO, I can't go down to 2.9ms in Sonar.
cheers!
opcode |
|
| Back to top |
|
 |
Ralph [RZ] Site Admin

Joined: 02 Jan 2008 Posts: 13332
|
Posted: Thu Aug 18, 2005 7:36 am Post subject: |
|
|
The problems with high latencies that JohnR is (most likely) referring to where resolved with Jamstix 1.2.
The issue Opcode is pointing out may be caused by DirectiXer's handling of MIDI output, which may not be latency-corrected. We will look into the issue. _________________ Ralph Zeuner
Rayzoon Technologies LLC
http://www.rayzoon.com |
|
| Back to top |
|
 |
opcode Jammer

Joined: 10 Aug 2005 Posts: 42
|
Posted: Thu Aug 18, 2005 8:27 am Post subject: |
|
|
I have now tried subhosting my VSTi (VSampler3) in jamstix instead of using MIDI out, and that worked like a charm at any latency setting...
(I tried this with the Cakewalk VST Adapter only; DirectiXer remained silent when subhosting)
Unfortunately, that is not the solution for me, because I have a couple of kits where I need to set MIDI controller values (e.g. to mix dry/overhead/ambient amounts), and that can't be done when it's subhosted - or can it?
cheers!
opcode |
|
| Back to top |
|
 |
Ralph [RZ] Site Admin

Joined: 02 Jan 2008 Posts: 13332
|
Posted: Thu Aug 18, 2005 9:01 am Post subject: |
|
|
| opcode wrote: | Unfortunately, that is not the solution for me, because I have a couple of kits where I need to set MIDI controller values (e.g. to mix dry/overhead/ambient amounts), and that can't be done when it's subhosted - or can it?
|
We'll add this to the next beta. Here is a test Build with controller passthrough. Let us know how it works for you.
http://66.235.201.50/updates/jamstix1673.zip _________________ Ralph Zeuner
Rayzoon Technologies LLC
http://www.rayzoon.com |
|
| Back to top |
|
 |
wmountney Jammer

Joined: 06 Mar 2005 Posts: 36
|
Posted: Thu Aug 18, 2005 10:08 am Post subject: |
|
|
| opcode wrote: | | I have now tried subhosting my VSTi (VSampler3) in jamstix instead of using MIDI out, and that worked like a charm at any latency setting... |
This must be a different issue than what I was seeing. then, since I've always been using BFD subhosted. I'll try to come up with a reproducable recipe for what I was encountering. |
|
| Back to top |
|
 |
opcode Jammer

Joined: 10 Aug 2005 Posts: 42
|
Posted: Thu Aug 18, 2005 10:44 am Post subject: |
|
|
Thanks, Ralph! Works just fine. I've gotta say, your support is simply outstanding.
Looks like I won't be needing DirectiXer after all. I can't record MIDI, but I can always use the export function in the arranger, so that's cool...
cheers!
opcode |
|
| Back to top |
|
 |
Ralph [RZ] Site Admin

Joined: 02 Jan 2008 Posts: 13332
|
Posted: Thu Aug 18, 2005 10:50 am Post subject: |
|
|
We'll still look into the DirectiXer MIDI-out latency issue but if the problem lies within DirectiXer then we won't be able to fix it. _________________ Ralph Zeuner
Rayzoon Technologies LLC
http://www.rayzoon.com |
|
| Back to top |
|
 |
wmountney Jammer

Joined: 06 Mar 2005 Posts: 36
|
Posted: Sat Aug 20, 2005 2:48 am Post subject: |
|
|
| Zauni wrote: | | We'll still look into the DirectiXer MIDI-out latency issue but if the problem lies within DirectiXer then we won't be able to fix it. |
Ralph,
I've been playing around with these issues, and I think there are actually two different problems:
1. Using DirectiXer's MIDI out loopback to drive another VSTi or DXi frequently has timing problems, which seem to be worse at higher latencies, but always somewhat bad at any latency. I had never been using MIDI out, and so had never noticed this problem.
2. When NOT using MIDI out, just subhosting BFD in Jamstix (via DirectiXer), there seems to be an issue caused by MIDI automation at higher latencies. For instance, with latency around 50ms, if any type of automation change point falls exactly on a drum hit, it causes a double-hit, flamming effect on that hit. If you are using something like a sloping velocity envelope over a wider area it will affect all of the hits during the changing portion of the envelope. Depending on other factors, like the tempo, you may hear it as a double hit, almost synchronized with slight flange, or a more seperated flam effect. It is easiest to hear if you create a simple pattern with a single regular kick or snare hit, and turn down the ambience.
The problem seems to begin at 23.9ms latency, where at the next lowest notch of 20ms with my audio drivers it is OK.
Bill |
|
| Back to top |
|
 |
Ralph [RZ] Site Admin

Joined: 02 Jan 2008 Posts: 13332
|
Posted: Sat Aug 20, 2005 9:33 am Post subject: |
|
|
Thanks, Bill! We will research this... _________________ Ralph Zeuner
Rayzoon Technologies LLC
http://www.rayzoon.com |
|
| Back to top |
|
 |
wmountney Jammer

Joined: 06 Mar 2005 Posts: 36
|
Posted: Sat Aug 20, 2005 1:17 pm Post subject: |
|
|
| wmountney wrote: |
The problem seems to begin at 23.9ms latency, where at the next lowest notch of 20ms with my audio drivers it is OK.
|
On further testing, it appears it can happen even below 23.9ms latency, even at 18ms or 12ms, perhaps depending on other factors in the project. I guess the only thing that is safe to say is that the problem is gets worse as the latency goes up.
Bill |
|
| Back to top |
|
 |
|