Hi,

I just upgraded to UFRaw 0.15 and noticed that performance when
rendering previews has deteriorated severely. Operations that used to
complete almost instantly sometimes take close to a full minute to
complete, making the software close to unusable.

It turns out that the parallelization of the rendering is at fault,
disabling OpenMP in the application significantly improves performance.

To be more specific, it is the parallelization of the 'develope'
procedure that is causing the performance issues when rendering
previews. If I understand things correctly, the count parameter to that
procedure is related to the number of pixels to process. Looking at the
procedures called, I highly doubt that the current threshold of 32
pixels(?) for parallelization would offset the cost of thread creation
and the synchronization at the end of the parallel region. I disabled
the parallelization of that particular region and the performance
returned to what I'm used to from older versions.

Since it was the slow previews that bothered me the most, I played
around a bit with parallelizing that part. Making render_preview_image
render several subareas in parallel instead seemed to yield a better
result.

Sorry for complaining so much... I really like UFRaw and hope it
continues to an invaluable image processing tool!

//Andreas


------------------------------------------------------------------------------
Come build with us! The BlackBerry(R) Developer Conference in SF, CA
is the only developer event you need to attend this year. Jumpstart your
developing skills, take BlackBerry mobile applications to market and stay 
ahead of the curve. Join us from November 9 - 12, 2009. Register now!
http://p.sf.net/sfu/devconference
_______________________________________________
ufraw-devel mailing list
[email protected]
https://lists.sourceforge.net/lists/listinfo/ufraw-devel

Reply via email to