Timing Problem

poidog · Post by **poidog** » Wed Feb 16, 2005 11:57 am

I have started having a really weird problem with my card/sequencer app. I am taking a simple audio clip and sending out of my sequencer to Pulsar, and then recording it back into the sequencer. Here is the weird part - when I record the new "resampled" signal it is actually ahead of the original signal. As in the timing is noticeably off, but the new rerecorded version is AHEAD of the original. It is like the reverse of latency. This happens no matter how simple I set things up. If I patch things directly from a ASIO source to an ASIO destination, I get the same result. There must some timing conflict between Scope and Sonar (my app) but I can't figure out what it is. I've had 3 months of stable operation, and this just started recently, and no I haven't added anything new to the system. Anyone have any ideas?

symbiote · Post by **symbiote** » Wed Feb 16, 2005 12:41 pm

Sounds like a "Play-Ahead" setting was activated within Sonar, maybe a negative delay on the track in question.

poidog · Post by **poidog** » Fri Feb 18, 2005 12:45 am

This doesn't sound like the issue. I contacted Cakewalk support and they insist that there is no such feature in Sonar. I completely reinstalled both Sonar and Pulsar, and still have the same issue - the re-recorded signal is ahead of the original. Any other suggestions? How does Pulsar synch to other apps?

Post by **garyb** » Fri Feb 18, 2005 3:30 am

no problem with sonar here.....some kinda...setting..thre's no syncronization involved(except for digital sync thru asio).

poidog · Post by **poidog** » Fri Feb 18, 2005 12:53 pm

I agree. There doesn't seem to be any settings in either Sonar or Scope to adjust the synchronization. And I have used Sonar with no problem for around a month, and I just started seeing this issue out of the blue.

But the problem most certainly exists. Signals routed through the Scope card are showing up ahead of the original signal. While I love the way my Pulsar sounds, and the synths and effects it has, I am getting close to pulling it out of my system. I've had it for three months and it has been one arcane technical issue after another. I spend a good portion of my music making time either completely down or troubleshooting the system.

symbiote · Post by **symbiote** » Fri Feb 18, 2005 1:34 pm

Unless you have some sort of time-travelling/time-warping unit close, or parasitic singularities floating around the house, the signals cannot get to one place before they get emitted. Thus, if they reach the SFP environement too early, they have to have been emitted too early also.

I'm pretty sure most sequencers have a per-track "delay" settings, for MIDI and/or Audio, which can be adjusted to positive or negative delay. Is this something that happens with other software than Sonar? Does it happen only when you record something, or it's detectable at playback too? Sonar might be putting the recorded info earlier in the track instead of at the spot it sampled it at.

poidog · Post by **poidog** » Fri Feb 18, 2005 2:56 pm

I really don't know if it happens with other software because Sonar is the only app I have for recording. Its impossible to tell with just a simple playback because there is no reference point.

I only notice it after a new track is recorded from another track in Sonar. In other words if I play back the original and new track they are way off (maybe 500 ms). But when I hear the playback during recording, it sounds fine. If I bounce tracks to a new track in Sonar, I do not see the problem. Bouncing should take all track settings and bounce it to a new track. If it was a track setting in Sonar that was causing the problem, then it would likely show up in the bounced track.

Here is a quote from Cakewalk support (not that I take their word as the gospel...)

"SONAR does not have the ability to adjust for audio timing, this is controlled by the soundcard. If you card does not have any buffer settings then it is not functioning properly. It may be due to its interaction with Windows."

Also, I'm using Windows 2000 in standard mode with no IRQ conflicts, using the intel processor and motherboard recommended by Creamware support. Other than using Win2k instead of XP I have built a new machine around this card according to the recommendations of Creamware.

Also one last piece of info. This has not always happened. Things worked fine for maybe a month. Doing exactly what I am describing with no problems. Then for no apparent reason, the problem started showing up in every project I have.

[ This Message was edited by: poidog on 2005-02-18 14:58 ]

[ This Message was edited by: poidog on 2005-02-18 15:01 ]

Post by **garyb** » Fri Feb 18, 2005 4:37 pm

this has nothing to do with the card. it likely has everything to do with the software environment and management of it. if all else fails, save your important files, format the hd and start over. strange behavior is almost always the result of bad installs, cracked software, corrupted software or poor is management. dunno about 2k, but xp has been nearly flawless here.

symbiote · Post by **symbiote** » Fri Feb 18, 2005 7:16 pm

On 2005-02-18 14:56, poidog wrote:
If it was a track setting in Sonar that was causing the problem, then it would likely show up in the bounced track.

Yeah, you're right. Can't really help then, as I've never seen anything like it, but I'd also suggest a clean install, and maybe an upgrade to XP. Win2K has been pretty dodgy for me audio-wise (works well on some machines tho,) while XP got a pretty unanimous positive responses from everyone I know doing audio on PC.

Worst case a clean install of 2k, but Windows in general has a notorious habit of working well a few months and then breaking down systematically (I have to try and use new words =P)

Make sure it's not an errant black hole wandering in the vincity, if you capture it in a jar you can sell it on eBay for more gear money

[ This Message was edited by: symbiote on 2005-02-18 19:18 ]