The Big Shootout of 2016: Phase/Leaf/Pentax/Canon/Sony

[dougpeterson]

I'm surprised that the 50 1.2L was used for the Canon. If has known focus drift.

Trust me, having done tests routinely for public consumption for nearly a decade now, no matter what choices you make there are several "I can't believe you didn't" or "you should have" responses. They aren't wrong; they just show that testing is a rorschach test; you'll approach it with a bias toward whatever shooting style, technical needs, and personal preferences.

If I had used an Otus on the Canon someone would have said "you should have used a Canon brand lens since that's what most people shoot with, and the Otus isn't even autofocus" etc etc. If I had used the 50/1.4 people would have said "why use the cheaper/lower-end 50". If I used used the sigma, etc etc.

Given that we can't run every sub-test with every camera with every lens (with every kind of subject matter with every kind of lighting), the results will always exclude potentially useful combinations. I think comparative tests are always very limited in nature; real world photography isn't a function of controlling every variable or matching focal lengths, it is about getting the most out of a particular set of gear, and picking the right tool for the job, and learning the gear well enough (and having gear with the appropriate features and interface) that the gear is the last thing you're worried about at the time you want to create an image.

We provide these test because we are so frequently asked for them by our customers. But ultimately I think they are far less valuable than using the gear yourself in situations similar (or identical) to your actual intended use. That's why we offer rental-toward purchase, open houses, appointment-based free use of our studio, and attend workshops and trade shows (like WPPI coming up in a bit, or the Breed Workshop we were at in LA this weekend. An hour with a camera in your hand will tell you far more than an afternoon spent looking at these files regarding whether it will suit you.

All that said, I will gladly provide my time and my company will gladly open its studio and equipment vault (in LA and NYC) to anyone who wishes to make a different comparison.

 

[MrSmith]

I don't really see how that will affect the outcome of this test.

It will just reaffirm how mediocre the 1.2 lens is when stopped down. It would as useful as using the old mamiya 1.9 80mm.
Sensible choice would have been the 45mm, sigma 50, 50mm 1.4 but that might have skewed things in the 5ds's favour. :thumbup:

 

[JeRuFo]

I was ofcourse talking about the focus issues of the lens. But you're right, for sheer resolution there are better options. Since the 24mm Zeiss was tested it would be interesting to use the 24mm t/s-e as well to compare to the wider Rodies that will no doubt be in the test, best Canon lens that side of a 50mm I've ever used and a favorite on the a7 for many as well.
But still, this test is about ball park figures, not whether certain lenses have 5% less CA etc. A final decision should still be made by the buyer based on testing it in their own workflow.

 

[dougpeterson]

Sony A7RII, Pentax 645Z, and Canon 5Ds R have now been added.

https://digitaltransitions.com/massive-still-life-shootout/

Over 30gb of raw files available for free download.

 

[aztwang]

Doug,

First off thanks for doing this test, I know it had to be very time consuming. As these tests usually provide results that are very close, picking a winner may be personal taste at times.
So Doug, after completing this test can you provide us with a opinion of what you saw? Maybe specific comments that surprised you.

Thanks

 

[Ken_R]

I agree that no test will satisfy everyone so that is why one must do it's own testing if possible but even if a test is not tailor made to one's needs or quite scientific with a proper device (Like Lens Rental's Roger Cicala's Optical Bench Lens tests) one can get a lot of info out of it. Enough to know that the differences between sensors might be so minor that one looks at other factors for differentiation and choosing criteria between models.

In the case of the Phase One backs there are several great choices now. The sensor is just one component of a system.

IMHO for landscapes the sweetspot nowadays is the IQ160 and the IQ260 (only if you must have LE capability), specially for tech camera use. Even Yunli (which has seemingly endless DR and LE needs) has produced some stunning images with the 260. Money no object the IQ3 100mp is a no brainer whether its for SLR or Tech camera use (w/ Rodenstock HR-W's) and the 80MP backs are close behind if you do not need the extra 1.5 stops of DR and extra LE capability of the 100mp model. For SLR work the 50mp CMOS model is awesome, specially for higher volume photography.

Doug thanks for your contribution. It's a lot of work and it's appreciated.

 

[jagsiva]

the IQ3 100mp is a no brainer whether its for SLR or Tech camera use (w/ Rodenstock HR-W's) and the 80MP backs are close behind if you do not need the extra 1.5 stops of DR and extra LE capability of the 100mp model.

Not so sure. I was just playing with the 23HR and 32HR files posted by CI. There is still quite a bit of cast left over after the new 9.0.3 LCC is applied.

Now, the CI files were quite underexposed, so will need to play with more samples.