As mentioned in previous post, I went to Artproof to make some test prints of my upcoming exhibition. Writing down some random takeaway thoughts.
A word of warning. Our discussion with Artproof guys both started and ended with the conclusion that the material is good enough that if I want, we can make the final prints right here, no need for any additional adjustments. But there is room for improvement on some pictures and of course I'm going to use it. I'm not into spending big bucks on prints when I know I haven't given my absolute best. So the remainder of the post is discussing these minor places for improvement that 98% of the people won't even notice.
The whole process looked like this - first we went through the photos together with the printing expert and made some corrections. These were actually quite small, like lifting the very dark areas a bit on some pictures and making them deeper in some others, setting the white point where there's more room on the highlights side, etc.
Then we cropped narrow strips of most critical parts of the pictures and printed them. We set the size so that when the pictures would have been printed whole, the size would be 100 x 125 cm. One of my pictures has been printed 3 x 6 m for outdoor advertising banner, but I haven't had a chance to look at it closely. This time we looked at 100 x 125 cm prints with a magnifying glass under gallery spotlights.
That's revealing I would say, revealing being a modest word here.
- I've always struggled with the pattern noise of my Canon's sensor. At this size, the patterns are at their best. Once you notice, you can't forget it and concentrate on the picture any more. Part of it is psychological effect, but still. Nasty, nasty, nasty. Although the printing expert said that very few people will notice it, I'm still unhappy. I need to start over with post-processing of some pictures. The below example is overprocessed to the point my wife could see the pattern, my tolerance for it being far lower. And it only gets worse when blowing it up to 200+% needed for big prints. The point is, it's there in the data.
- It's unfair and totally unscientific comparison, but the one picture in my series made with Sony RX100 Mk3 actually looked better than the rest from Canon 5D Mk3. Grain is much less noticeable, tonal transitions are linear even in the places that have been heavily processed (and by heavily I mean +1 stops of exposure compensation and some clarity). If the term non-linear transition makes you wonder what the heck I'm talking about, then here you are, an example of state of the art Canon sensor:
- We also looked at some RAW files and the printing guy showed me why it makes sense to set sharpening to zero for the workflow and sharpen only at the end for the right output size. Much less noise and weird artifacts (which amplify with every additional processing step). Grain can and should be added later, knowing final printing size. Having some unobtrusive grain actually makes the tonal transitions smoother. Example: 200% crop of sky on ISO 100 image, left - unmodified image, right - sharpening +40, clarity +50 (not unheard of amounts for local adjustments, but something I usually don't use globally):
- Grain has to be uniform. I have one picture which is made in good light, not at some crazy ISO, but the main subject is in relatively dark area, where I've lifted the shadows considerably. So most of the picture is almost noise-free, but the lifted dark area shows strong noise artifacts, making it rather unpleasant. see patch three from left on the first picture. To be processed from scratch.
- We looked at the test prints both in ambient light and under gallery-like spotlights. There is one picture in my series that is very dark, making it very dependent on lighting. It's beautiful under spotlight, the details come through very subtly. Now when looking at the same print under ambient lighting, most of it is just pitch black. So I have to choose whether to make it lighter and viewable under not-so-good light (and too bright for spotlight) or leave it darker - perfect for spotlight, but pointless under any other lighting. Another hard choice because there're no spotlights at Photoville.
All in all, it was a very informative experience.
PS: I feel strong urge to mention I hate discussing camera makes and models and I have avoided it for years already. The reason has been simple - the camera is by far not the weakest link in my work. Cameras just haven't gotten into my way. This post is an exception and this blog will not become place for bashing gear... Then again, why not say out aloud what I think - this is my blog after all, describing what I'm working on and struggling with.