Sigma Sony Comparo Part 2
Sigma Sony Comparo Part 2
8 months ago
There are 3 sets of raw images on dropbox here:
Each set has 6 images in sets of 2 apertures f1.8 and f5.6 at 3 focal lengths of 18mm 24mm and 35mm.
Nomenclature is as follows:
|sdim/dsc|9999|l(eaves)/r(palmfronds)/p(pink flowered bush)|focal|aperture
so sdim0001l18f56 is sdim or dsc0001 18mm leaves f5.6
When shooting I set the Sigma to Exp -1 so some of the images will need to be bumped in exposure and there is little more noise than there should be. Most of the images have gray cards in the image for white balance. I forgot a few times. Af was used for all images. Sony af is much better in low light than Sigma's.
I used Iridient Digital for processing the raws.
I haven't looked close enough yet to form any opinions other than they are both surprisingly good. I've been working on this for a couple of days refining the techniques to equalize the images.
If for some reason an image won't load it may be because it is not fully downloaded on dropbox yet so try it again later.
Samples to follow.
sigma sd1M with canon mount vs Sony A7r using the same sigma 18-35mm f1.8 lens.
The first part is here: http://www.dpreview.com/forums/thread/3588219
for those not knowing. What did you compare? or give us a link to part1 pls.-- hide signature --
I've been looking through the images and have settled on just one to post that shows both how good each camera is and that the Sigma seems to me to keep the crown.
Here are the images, both with the Sigma 18-35 at 35mm and f5.6 of the leaves on grass. First the sigma
Next the equivalent Sony image taken at the same settings:
sigma on the left sony on the right:
Looking at these two crops the grass just behind the gray card across the image seems to be subtly sharper in the sigma image than in the sony. Some of the fallen leaves appear slightly better on on or the other partly due to the light/colors of the leaves but the grass in that area seems to be in excellent focus for both cameras (the dof borders are a bit different between the two images but the area being looked at seems to be well focused in both images) and to my eye the sigma has an edge.
Sony seems to produce cleaner and less noisy images. The Sigma was shot at -1ev while the Sony was 0. The sigma was bumped up by 1ev in editing and that may have contributed to the noise but on the whole the sigma does seem to be more noisy than the sony under similar conditons.
Sigma retains its edge in sharpness. I have long wondered how much of the sharpness/accutance of sigma images was due to the lack of an aa filter and how much to the 3 layer technology. I think we have a little more info now. The sigma, even with very similar pixel pitch remains sharper. That may be due to the 3 layer technology. There is much more difference in sharpness between similar images from the sony A7 and A7r or the Nikon D800 and D800E then between the Sony A7r and the Sigma SD1.
The sigma wb was set to cloudy in the camera even though it was being shot in late day daylight. When the dropper was used in Iridium Developer it changed little if at all meaning the camera got the wb almost perfect according to the gray card. The sony was off considerably; it was set to auto. To my eye the sony was more pleasant to view but I adjusted it with the gray card and it was very close to what the sigma gave out of the camera.
One more thing I wanted to mention. The sony has no issues with focus on the 18-35. In low light it takes some time, maybe 2-3 seconds at the most but it nails the focus every time. I'm not familiar with the different kinds of focus and why Sigma seems to think it has focus when it clearly doesn't on that lens. Maybe Sigma needs to look at Sony to see how they can stop the camera from saying it's focused when it clearly isn't.
For sharpness-test i would shoot a newspaper-kiosk showing newspaper-titel outside. same subject: 1. center 2. edges with different apertures. precise focus needed with loupe.
yes if its sunny there could be a problem testing speedy apertures 1.4, 2.0 when camera goes only to 1/4000 with iso 100.