Fader automation mis-aligned with audio

I’m sure this will turn out to be something exceptionally stupid on my side, but here goes…

I recorded a session about a week ago using Ardour 3.4 (on Fedora 19 + CCRMA), then I edited it over the course of the following few days, as I had time, using Ardour 3.5.14 (on Fedora 14). After several editing sessions, A3 seemed to be getting somewhat more crash-prone - a couple times just opening the session and hitting the Zoom-to-Session button was enough to make it seg-fault - but otherwise things were going fine. Tpward the end of editing, I decided to add some fader automation points so as to avoid an audible click that had occurred right before something I wanted to keep. The click waveform looked quite distinct, but just to be sure I had the right place, I positioned the play head just before it and hit play (indeed it was there), and then just after it (no more click). With that verified, I added automation points to set the fader at -inf some time before the click, and two more just after it, one at -inf, and the other back at 0. Very routine - I’ve done this sort of thing many times before.

When I played this, the automation worked - the fader moved - but to my surprise, the click stayed in the output. At first I looked for obvious errors, such as forgetting to set play mode on the fader “track”, but didn’t find anything. I experimented with sliding the automation points around in time, and found that there was an offset of about 200 ms between where I thought I was putting the automation, and where it actually seemed to be taking effect. Exiting Ardour and re-starting didn’t help - I tried it several times.

I’ve tried reproducing this in a new session, with no luck – it works fine, of course! :slight_smile: I guess if I never see it again it won’t lose any sleep, as this was one of the last few edits I needed to make on this session. But with my luck…

Any ideas how I could troubleshoot this? Is there anything in the .ardour file that would be relevant?

Thanks!
Don

I have a new session, recorded yesterday using 3.5.14, that seems to have the same problem. Perhaps coincidence, but It was created from the same template as the session mentioned in my original post. I don’t believe the session that was used to create the template had the issue - or if it did, I didn’t notice it somehow. I’ll try doing more experiments as I have time and post back with results…