DannyBurkPhoto
Offline
• • •
Upload & Sell: Off
|
 p.3 #16 · Nikon/Canon/Sony Image Quality vs Panasonic's Multi-Shot High-Res Mode | |
I'll add a bit to the conversation, since it looks like I'm the only one here that has actually tried it, however briefly. (More on that in a moment.)
I got an S1R several days ago and have spent most of the time since then, literally, going through the manual and setting its options. Last November, I got a GFX 50S that basically replaced my A7RIII, but I was still in need of a full frame camera for certain lenses and situations. My frustration with the Sony meant that I wasn't willing to use it anymore, and after a lot of online study, I decided to try the S1R. It will be almost exclusively used with adapted manual lenses (same as the Sony), although I did buy the 24-105 for casual walkaround use since there were a lot of reports saying that it's surprisingly good.
Anyway, it's finally set up and I was anxious to try the pixel shift feature yesterday. Between the reviews and samples from LC, DPR, and Northlight, it looked promising enough to have real value for landscape, given the ability to use it with some ambient motion. I tried Sony's version when it came out, and it was next to useless for anything but product shots and the like; it reduced moire on fabric, but with no actual increase in pixel dimensions, and using it for anything other than fabric resulted in a nearly invisible improvement that only provided slight improvement when interpolating the file.
Conditions weren't good yesterday for the test; there was a lot of wind and constantly changing light. I'm hoping for better sometime this week. I composed a test that included the back of the house (brick) since that part wouldn't move, plus a couple of small trees, various plants around the house, and woods in the background. I tested both "Mode 1" (for unmoving objects) and "Mode 2" (with moving objects) to see what I'd get. Mode 1 wasn't usable at all, due to the large amount of moving foliage, which appeared as a blurred mess. Mode 2 shows a lot of promise, but does have a number of artifacts when viewed at 100%. Most of those can be cloned out, although I also tried downsampling and then most became invisible while still providing a very large file. My impression is that the less movement, the better; the 10-15 mph wind yesterday did it no favors.
I did the tests using a couple of very high quality lenses (Milvus 25 and Otus 55), which worked perfectly with the Sigma adapter. The resulting file is RAW and can be handled in ACR just like any other RAW file, and the camera can be set to shoot a normal file along with the composite. I didn't try it, but the normal file could be used for cloning, if necessary, after interpolating up to the size of the composite.
I don't feel that the tests are ready to post at this point, due to the bad weather and light conditions. I'll work on this soon and will post examples when I have more to show.
|
