Quoting Tobiasz Karon <[email protected]>:

> On 1 Mar 2014 20:28, "Raine M. Ekman" <[email protected]> wrote:
>> OTOH, now instruments produce sound also when muted. I exported your
> 50-tracker at a speed of 3 minutes/track from a generic x86 build of
> 0.4.15, which couldn't play it live with the drums. The self-built 0.9.91
> that can (almost*) play it with no dropouts is ATM taking close to 10
> minutes/track on the same box. Looks to be 8-9 hours in total... will get
> back to the sync thing tomorrow.
>>
>> *) Had to remove one resource-hungry Zyn track. Either that or compiling
> with some harder optimizations.
>>
> Interesting results you have here. I don't know what to think. Faster
> rendering means slower playback?

Not quite. More like "muting tracks doesn't speed anything up  
anymore". And track export is done through muting.

On to some forensics...
Before rendering, I added code to print out all tempo changes and it  
looks like the sync problem comes from the 780 BPM change at tick 0  
being delayed by up to 3 ticks in ~50% of the cases. All the other  
tempo changes seem to come at their correct times, or at least at the  
same tick on every render. (This in 0.9.91, the variation was a lot  
worse in 0.4.15)

The few milliseconds this translates to usually isn't much of a  
problem on playback or when you export the whole project, so it could  
be that nobody has noticed this jitter much before.

-- 
[email protected]
softrabbit on #lmms



------------------------------------------------------------------------------
Subversion Kills Productivity. Get off Subversion & Make the Move to Perforce.
With Perforce, you get hassle-free workflows. Merge that actually works. 
Faster operations. Version large binaries.  Built-in WAN optimization and the
freedom to use Git, Perforce or both. Make the move to Perforce.
http://pubads.g.doubleclick.net/gampad/clk?id=122218951&iu=/4140/ostg.clktrk
_______________________________________________
LMMS-devel mailing list
[email protected]
https://lists.sourceforge.net/lists/listinfo/lmms-devel

Reply via email to