Hi,
Thanks a lot for testing, I have just uploaded new sync patch without extra
timer for this sync (on host side), so
sync latency should be now lover overall, and less crackling, as things are
now synced whenever are changed.
https://sourceforge.net/tracker/index.php?func=detail&aid=3601658&group_id=105168&atid=640436#
Now all data from host are passed into shared memory, but there is still
room for improvements, because in this setup, there is lots of useless
messages still being sent from host to plugin e.g. tempo etc.
Also there is and extra option in VST_sync_shm.h header file though to
subtract latency from host sync positions, well not sure if this is
computed right, and I think some plugins may compute such latency for
themselves, so it is disabled by default now.
BTW as you noted
3) I noticed this too, but so far was not able to do anything with it.
2) maybe this is sample buffer size dependent?
Best regards
Mike C.
On Wed, Jan 30, 2013 at 9:53 PM, Chrissy McManus <chrissy.m...@hotmail.co.uk
> wrote:
> hi Mike, I have patched the very latest VST_Sync patch and this time I
> have installed to my /usr/bin/lmms directory (which makes testing easier,
> no more changing files for me),
> anyway i see good improvements with this patch, like ;)
>
> (* note LMMS default install on linux is /usr/local/bin/lmms *)
>
> (1) the clicks/pops are almost gone when using dblue_glitch (i think the
> actual sounds used have an effect here)
> (2) i can seem to load far more vst and vst(i) into a project with less
> cpu overhead
> (3) the export seems flawless (what you hear in the song is what you get
> exported)
>
> some things that need looked at in general and i believe they are NOT
> relevant to your patch, but i will list them here;_
>
> (1) every vst/(i) gives this warning RemotePlugin::DebugMessage: failed
> getting shared memory
> (2) exporting a song with even the smallest buffer size i.e 64 frames
> always gives a silence at the start of 0.002ms or roughly 76samples
> (testing in audacity) : the frames size has a direct effect on this
> , : this could be a soundcard issue tho?
> (3) when opening a saved song with vst/(i) already in it, sometimes the
> terminal will show 2 instances being created although only 1
> actually shows up, perhaps this is harmless behaviour but worth
> looking at?
> (4) Automation is messed up at the moment, eg open a new song , goto
> bb_Editor add an automation track and drag the volume control
> of the triple-osc into that track, there is no automation showing
> what should be 100%, it is the same for all test i have done regarding any
> automation
>
> (* note i can report these as separate bugs if needed* )
>
> keep up the good work and p.s i hope the vst_sync makes the next release
> as i think it is almost there ;)
>
>
> thanks Mikobuntu ;)
>
------------------------------------------------------------------------------
Everyone hates slow websites. So do we.
Make your web apps faster with AppDynamics
Download AppDynamics Lite for free today:
http://p.sf.net/sfu/appdyn_d2d_jan
_______________________________________________
LMMS-devel mailing list
LMMS-devel@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/lmms-devel
