The first time that Sean and I gave Canon's Digital Photo Pro version 4.0.0 a workout, we declared that it wasn't ready for prime time
. DPP 4 is now at version 18.104.22.168 and, forced by the lack of Canon EOS Rebel T6s
support in DPP 22.214.171.124 (and presumably the same issue coming for the Canon EOS 5Ds
and 5Ds R
), I am once again working with the latest version of Canon's image processing software. While I'm not ready to make recommendations on use of the latest version, I want to communicate a change that you might notice in forthcoming image quality results
processed using DPP 4.
Any time a new RAW image processing engine is used, there is a possibility of the output being different. With each version of DPP we load, a comparison test is done to insure that our results are not impacted by the new engine. It has been a long time since the first digit of DPP's version number has been incremented and ... I was nervous. Primarily, I didn't want to leave the huge inventory of lens tests unavailable for comparison with the latest-tested gear.
Based on testing performed on the Canon EOS 5D Mark III
(supported on both versions of DPP), I'm reporting is that DPP 4 indeed changes the output. However, I think that you are going to view the change as positive and non-harming to comparisons with older tests going forward. Important is that I don't see any change in sharpness based on the same standardized test settings (Neutral Picture Style, Sharpness = "1"). However, there is a clear reduction in the artifacts seen in the resolution lines in the test chart. The lead image in this post was processed in the latest version of DPP 4 and the following image was processed in the latest version of DPP 3.
To best compare these images, control-click on each to open them in new browser tabs and toggle between the tabs. Or, download and open the images in your favorite image viewer. Show the images at 200% and then toggle between them. Very little difference other than the artifacts clearing will be seen.
From a color standpoint, I am not seeing any changes between the two RAW image processing engines. From a noise standpoint, DPP version 4 has a very slight edge at higher ISO settings. The edge is tiny enough to be meaningless to most and unless toggling instantly between images at 100%, probably few would detect the difference.
Testing against all of these small changes is very time consuming, but ... worth it for accurate evaluations.