hi, On Wed, Sep 5, 2018 at 9:09 PM Heiko Bauke <heiko.ba...@mail.de> wrote: > > Hi Johannes, > > Am 05.09.2018 um 12:08 schrieb johannes hanika: > > > as to your question, integrate gmic or reimplement. does it make much > > of a difference? > > main differences: > > * using libgmic introduces an additional dependency. > > * libgimic is written in C++, GMIC module needs to be implemented in C++.
okay those two i don't care much. this sounds more like free updates to the algorithms. we've had a few c++ modules, so no real show stopper. > > can the code run on cropped regions of interest and > > in floating point? does it work for preview, i.e. does a downscaled > > image look similar when processed to first processing and then > > downscaling? > > For the sharpening filters: At the moment one has to scale the preview > to the scale of the intended export to get an reliable preview. Taking > into account the scaling of the preview and scaling the parameters > accordingly would be easy. However, I am unsure, if applying the > sharpening filters with scaled parameters really gives a better preview. > When it comes to sharpening, I think one always has to zoom in to > adjust the parameter appropriately. Therefore I neglected this issue so > far. right. we do very similar things for denoising/sharpen inside darktable. this is just something that has to be considered. > > any performance concerns in the interface (need to copy > > buffers etc)? > > GMIC stores image data quite differently than darktable. Thus > additional copying steps are needed, which is not that so nice. that sounds suboptimal. but how much does it cost? did you run darktable -d perf with something very minimal just to benchmark the cost of the buffer copy? if it's in the single digit millisecond range maybe we shouldn't bother porting the code but just use the library.. > > for R/L sharpening, maybe that would be a popular addition to the > > current sharpen module (so it'll be easy to discover)? > > Completely agree. cool! cheers, jo ___________________________________________________________________________ darktable developer mailing list to unsubscribe send a mail to darktable-dev+unsubscr...@lists.darktable.org
