nyvz
Veteran
I just got my A7S and have started playing around with it so I thought I'd start a new thread with some thoughts as I come up with them.
First of all, I am using the camera with a Commlite EF to NEX adapter which so far seems to work alright for aperture control and IS. It certainly controls the aperture of my lenses, though AF is a bit hit or miss on the lenses I've tried so far. Then again I don't own any Canon-brand lenses, mostly Sigma plus a Tokina 11-16 and Tamron 28-300. The VC (IS) on the Tamron 28-300 seems to work fine in movie mode, but I haven't been able to enable it in the photo modes.
The first test I did was a quick comparison of the picture profiles. I tested to see how the white clip point of each of the gamma curves compares. I found that SLOG2 has the same white clip point at ISO3200 that CINE4 has at ISO1000, CINE3 at ISO800, CINE2 at ISO500, and CINE1 at ISO640. This relationship is identical to the relationship all the cinegammas have to eachother and SLOG2 on the FS700 only that they are all 2/3-stop more sensitive. I wonder if Sony designed it that way because that gain setting gives the best read noise or perhaps there is something otherwise odd about the lower ISOs. I should test further against ISO settings in photo mode as well. Anyway perhaps my initial thoughts were wrong about SLOG2 vs cinegammas, and maybe it is more like the FS700 where they all have similar DR but it is just distributed differently rather than like some Sony cameras where cinegammas clip stops off the top. The only way to know will be with a shadow noise test comparing each of the gamma modes at each of those settings that have the same white clip point. I wish the camera had a gain db option as that might be a bit revealing about where base gain is for each gamma, but instead we're stuck inadvertently changing gain every time we change gamma curves.
I didn't even bother to test the other gamma curves since pretty much all the other ones looked quite bad in overexposure and overall had that high contrast clippy look with bad color shifts in highlights, though REC709 800% may have some promise and I just didn't test it because it didn't have an equivalent since it clips sooner than SLOG2 and cannot go below ISO 3200. I'm beginning to think that this might mean it's much like the FS700 where REC709 800% is bad and clips whites while cinegammas still use the full DR.
I'll see what else I find.
First of all, I am using the camera with a Commlite EF to NEX adapter which so far seems to work alright for aperture control and IS. It certainly controls the aperture of my lenses, though AF is a bit hit or miss on the lenses I've tried so far. Then again I don't own any Canon-brand lenses, mostly Sigma plus a Tokina 11-16 and Tamron 28-300. The VC (IS) on the Tamron 28-300 seems to work fine in movie mode, but I haven't been able to enable it in the photo modes.
The first test I did was a quick comparison of the picture profiles. I tested to see how the white clip point of each of the gamma curves compares. I found that SLOG2 has the same white clip point at ISO3200 that CINE4 has at ISO1000, CINE3 at ISO800, CINE2 at ISO500, and CINE1 at ISO640. This relationship is identical to the relationship all the cinegammas have to eachother and SLOG2 on the FS700 only that they are all 2/3-stop more sensitive. I wonder if Sony designed it that way because that gain setting gives the best read noise or perhaps there is something otherwise odd about the lower ISOs. I should test further against ISO settings in photo mode as well. Anyway perhaps my initial thoughts were wrong about SLOG2 vs cinegammas, and maybe it is more like the FS700 where they all have similar DR but it is just distributed differently rather than like some Sony cameras where cinegammas clip stops off the top. The only way to know will be with a shadow noise test comparing each of the gamma modes at each of those settings that have the same white clip point. I wish the camera had a gain db option as that might be a bit revealing about where base gain is for each gamma, but instead we're stuck inadvertently changing gain every time we change gamma curves.
I didn't even bother to test the other gamma curves since pretty much all the other ones looked quite bad in overexposure and overall had that high contrast clippy look with bad color shifts in highlights, though REC709 800% may have some promise and I just didn't test it because it didn't have an equivalent since it clips sooner than SLOG2 and cannot go below ISO 3200. I'm beginning to think that this might mean it's much like the FS700 where REC709 800% is bad and clips whites while cinegammas still use the full DR.
I'll see what else I find.
