The GetDPI Photography Forum

Great to see you here. Join our insightful photographic forum today and start tapping into a huge wealth of photographic knowledge. Completing our simple registration process will allow you to gain access to exclusive content, add your own topics and posts, share your work and connect with other members through your own private inbox! And don’t forget to say hi!

Hmm… DPreview has added the GFX 50S to their studio test scene…

ErikKaffehr

Well-known member
Hi,

DPreview has added the GFX 50S to their studio test scene and they conveniently matched up it against the Pentax 645Z, Canon EOS 5DsR and the Sony A7rII.

The first question I have asked, how good is the Fujinon 63 mm lens? Let's check the corners:

Screen Shot 2017-03-14 at 07.02.03.jpg

So, what do I see? Well, the best lens of the bunch may very well be the Canon 85/1.8, at least in the corners. Pentax is great, but has some lateral color. Seems like Fuji has some astigmatism and so has the Sony lens. Winner? All are quite good, should I rank them it would be Canon, Pentax, Fuji and Sony.

So, what about on axis?
Screen Shot 2017-03-14 at 19.20.39.jpg

Well, all lenses outresolve the sensor leaving some colour artefacts. Pentax may be best balanced, Canon still great. My ranking would be Pentax, Canon, Fuji and Sony.

So, the lenses are good! So we expect some moiré? Yes!
Screen Shot 2017-03-14 at 07.03.06.jpg

This is a tough test and all fail. No ranking…

Let's look at high ISO noise, theory says that the 44x33 mm sensors should have similar noise at 1360 ISO as the smaller ones at 800 ISO. Let's check 44x33mm at 1600 ISO and 24x36mm at 800 ISO. Let's not forget that both 24x36mm cameras are tested with f/1.8 lenses…
Screen Shot 2017-03-14 at 07.04.45.jpg

The Sony A7rII plays a small trick it has in it hand. Sony uses "dual gain conversion" a technology it has cross-licensed from Aptina. The trick is that full well capacity is reduced at a certain ISO, in this case ISO 640. Thus darks are rendered cleaner. The cost is that an additional transistor is needed, so some surface area is given up for improved high ISO performance.

The great surprise for me is that the small difference in resolution between the 5DsR and the Sony matters. It may be that the Canon has a much better lens.

It is not possible to judge the performance of a system from a single exposure using a single lens, but I still find that these four images give a lot of interesting info…

Best regards
Erik
 

Attachments

Paul2660

Well-known member
Hi Eric, thanks for posting.

Yes the Fuji glass has some astigmatism, I see it mainly in the corners, but not excessive. Sharpness has totally impressed me with both the 32-64 and 120mm.

On the noise front I am willing to bet Dpreview used LR/ACR for their conversion. LR IMO current is leaving quite a bit of room in the higher iso's from 1250 and up. There is another raw converter I am using that does exceptional high ISO on the GFX, both in regards to detail, and color. PM me if you are interested.

The deal killer for me on the A7rII, was both high ISO past 1600 (which I found very noisy on the camera I tried) and long exposure noise which rendered exposures of 1 minute or longer worthless due to excessive noise (there was a ton written about this), to which Sony's answer was "use long exposure noise reduction" which did not work for me. It takes up way too much battery life, heats up the sensor and places gaps in stacks.

The DR of the 50Mp sensor is very impressive, and in the lower ranges, ISO 100 to 400, 2.5 stops is a very possible push if needed with a lot of details remaining. But with the right raw converter, ISO 1250 to 3200 can be pushed as much as 1 stop maybe 1.5 stops.

Paul Caldwell
 
35mm format is the "APS-H" of 44x33 format, and 44x33 format is the "APS-H" of 645 format. The difference would be marginal when the number of pixels are close. Thus when 35mm format advances towards 70MP, almost all current digital backs (except the 100MP) could become "obsolete" in terms of ease of use.
 

ErikKaffehr

Well-known member
Hi Paul,

I may have been expecting a miracle...

In recent time I have been playing a lot with the idea of buying a Canon camera. I have three Canon lenses, the 16-35/4L, 24/3.5 TSELII and the 24-105/4L and I am happy with all of them, although I have some reservations with the 24/3.5 TSE LII and also the 24-105/4 L. For long telephoto I am still Sony, but I am quite attracted to the Canon alternatives. So, a Canon 7D may be in my future… But I have just realised that I have a very good Sony A77 for those images, so that body is not any longer for sale.

I have done some long exposure test on the Sony, and I got a lot of hot pixels without long exposure noise reduction. In that area the P45+ may be a champ, but I don't do it very often.

I feel that the Sony A7rII is flexible enough for my needs. Not the best tool for everything, but it can do most things quite well and that may be "good enough". It is easier carry one gear instead of two and focusing on one kit of gear instead of two may be beneficial. I have the dark perception that I missed a good number of shooting opportunities by using MFD instead DSLR. Having a workflow that works well may have benefits over two workflows that don't work so well…

Best regards
Erik



Hi Eric, thanks for posting.

Yes the Fuji glass has some astigmatism, I see it mainly in the corners, but not excessive. Sharpness has totally impressed me with both the 32-64 and 120mm.

On the noise front I am willing to bet Dpreview used LR/ACR for their conversion. LR IMO current is leaving quite a bit of room in the higher iso's from 1250 and up. There is another raw converter I am using that does exceptional high ISO on the GFX, both in regards to detail, and color. PM me if you are interested.

The deal killer for me on the A7rII, was both high ISO past 1600 (which I found very noisy on the camera I tried) and long exposure noise which rendered exposures of 1 minute or longer worthless due to excessive noise (there was a ton written about this), to which Sony's answer was "use long exposure noise reduction" which did not work for me. It takes up way too much battery life, heats up the sensor and places gaps in stacks.

The DR of the 50Mp sensor is very impressive, and in the lower ranges, ISO 100 to 400, 2.5 stops is a very possible push if needed with a lot of details remaining. But with the right raw converter, ISO 1250 to 3200 can be pushed as much as 1 stop maybe 1.5 stops.

Paul Caldwell
 
Top