Here's a strange phenomenon that I somehow uncovered in DPP4 (V4.17.10.0) while editing some .CRW raw files from an old Canon Digital Rebel.
After making some adjustments (which are somewhat limited compared to what I can do with .CR2 files) I did a Batch Process to a group of shots to convert to .JPGs and save. No hiccups there as best as I can recall. But a few days later when reviewing the .JPGs this one, single, odd image in the batch showed up. In some areas that appeared overexposed, the blown-out highlights displayed as solid black in the JPGs, the same as seen in this first screen-shot of my saved .CRW file
I can't imagine missing that while editing the raw images, but maybe I dozed off for a few seconds. Or I got distracted when the phone rang. Or more likely, I pasted a recipe from a previous file and for some reason it took longer than normal for the adjustments to show up onscreen.
So I opened the .CRW file with the saved changes in DPP4 again and with the Highlights/Shadows warning enabled, this is what I see. Which wasn't all that surprising to me except for the fact that I didn't see it this way while doing the original editing. Also, what are obviously blown Highlights are displayed in blue instead of red because the adjustments I made rendered them as black, as shown in this second screen-shot.
In this final screenshot I disabled the Highlights/Shadows warning and backed down the Highlight slider to -5.0 from +1.0 and what do you know? The previously black highlight areas disappeared and I started to pull up at least a tiny bit more detail in the highlights. Saved my raw edits and converted to a .JPG again and the results look just like the re-edited .CRW this time.
What I'm figuring is that all this strange behavior is probably due to operator error on my part. But just in case, I'm curious if there might be some other distinct differences between .CRW and .CR2 files that I'm not aware of. Maybe something in the areas of dynamic range or sensor designs or in-camera processor types. I know I wasn't doing myself any favors by starting off with an overexposed image. But does it mean that this Digital Rebel is showing its age by "misbehaving" in a more evil, permanent way? Everything else about it seems to still be working properly, though I have no intentions of using it as a substitute for my 60D.
1 ACCEPTED SOLUTION
I played with the brightness adjustment in DPP. When I moved it off "0" the black disappeared and the highlight warning was red. The RGB was 255,255,255. If I then started reducing the brightness the black appeared at -.10.
I deleted and redownloaded the file to start anew. In DPP I set the Brightness to "0" and used the Auto Gamma tool. It did a decent job of adjusting the image. In that tool the Highlights were set to -4.
Then I reset the Gamma tool and varied the brightness. It worked fine with no black appearing.
@BurnUnit wrote:I reopened the file in DPP4 and clicked the AUTO Gamma button like you did. You're right. It did a much better job than I expected! Usually I avoid using it because the results aren't pleasing. But I had made some manual gamma and brightness adjustments and think I may have been a little heavy-handed. In other words, I think I created at least part of my problems. Wouldn't be the first time. . . And I wish DPP4 had a "Midtone" slider in addition Highlights and Shadows.
My expectations going in were probably over-optimistic.I'm often amazed at the amount of improvement I can make with the newer CR2 files. But looking back I think that usually tends to be from opening up the shadows and futzing around with the midtones a bit. For something that's this over-exposed, the highlights just don't have enough image information to manipulate very much. I'm chalking the weirdness up to that and a bit of glitchy behavior from DPP4.
You can move the middle slider in the gamma histogram as a midpoint slider. I don't really understand that tool but from what I read a number of years ago in a forum posting ( maybe DPR) that tool is very popular in Japan.
