Re: Processing RAW using DxO - post your samples and tips - Part II
Larry Rexley wrote:
Still it's better than what I used to be able to get out of Canon's DPP 4 at ISO 3200.
Indeed. I don't think DPP4's noise reduction is any better than what the camera JPEG engine can do. I'm pretty happy with ART's noise reduction in general, it's just the extreme cases where DxO is significantly better.
Good point, processing is always a matter of taste of course. Even processing the same images six months apart I've found my preferences and workflows have changed.
That's one reason why I like the film simulations. Each RAW editor is going to have a default rendering, but film simulations allow you to scroll through different options to easily get different tone and color profiles and get away from that default look if desired. I originally started down this road last week playing with a trial of DxO Filmpack, but IMO it is overpriced for what it is. That's why I started looking into alternatives. I didn't expect to like ART as much as I have.
The biggest PP change was when I first switched to DxO PL. I had a well-established, efficient workflow in DPP 4 before switching, and for non-noisy images was very happy with the DPP results. But there was a set of noisy images I could never get right with DPP. After getting DxO PL I spent weeks going back through my images and re-processing many of them. With DxO it took far less time and I generally am happy with the results. Now I couldn't go back to DPP!
I've also used darktable and gotten very good results. There are some very nice controls in datktable and its astro denoise was better than DPP, but DxO Deep Prime is far better so I no longer use darktable.
I also used darktable for a couple years. I think the most recent version I tried was 3.8, before I got into DxO last year. The main problem I had with it was sharpening. Comparing with DxO and even DPP4 (with DLO), the darktable images were frequently not as sharp, sometimes very noticeably so. Cranking up the sharpening radius is prone to causing artifacts or just making it look artificial. That was one factor that originally sold me on DxO. darktable uses an unsharp mask, which probably explains the difference. RawTherapee/ART uses RL Deconvolution by default. I will admit that I am far from an expert in the differences between these algorithms, but the results seem to be just as good as DxO's. It would be interesting to do a more scientific comparison, but that's been my experience anyway.