RX1 cross frame colour caste test results

[tashley]

Here's a link to a gallery that has shots from f2 through f16 using a Phase One LCC sheet held close to the lens mount and pointed at a mildly hazy sky. The frames were made RAW and then developed in LR 4.3 with Adobe Standard camera calibration profile and no lens corrections. I took a white balance from dead centre of the f5.6 frame and then applied it to all the other frames. Otherwise, nothing: no sharpening, NR, lens corrections, exposure adjustments.

The shots were then exported to 50% sized 92% quality JPEG in pro photo - so if you really want to get the most accurate view you must download the originals and look at them in a colour space aware application on a wide gamut monitor that is calibrated.

I will say nothing at this point about what I think - it is more interesting to test the range of reactions from a wider group of people...


ps there's one obvious mistake that came from testing too many things on too many cameras today: I shot with Auto ISO so as the aperture got smaller the ISO got higher. I doubt that it makes much difference but if people feel otherwise I will try to repeat...

 

[Mike Hatam]

Tim - thanks for posting these images - it's very helpful to see them.

Here's my initial reaction after studying them...

The red cast on the right is a non-issue for me. It's very minor, and essentially gone by f4.

The cyan cast on the left is more noticable and problematic. It's strongest at f2 (probably more visible due to vignetting), but reduced significantly by f2.8, and nearly gone by f4.

Even with the left-side cyan cast at f2, I don't see this as a show-stopper or serious issue in general use. In general use, the color cast will be essentially undetectable in most situations.

I'm taking delivery of an RX1 tomorrow, and looking forward to putting it to use over the holidays.

 

[pophoto]

Tim, it is what it is, I can either look at these white shots all day or real world images.

Thanks for these tests, your review is enjoyable and informative.

 

[tashley]

Both eminently sane reactions IMHO. I regard it as a non issue for all but the sorts of specialist pro tasks this camera would never be used for-but there's been a lot of fuss so I thought I'd try to lay it to rest...

 

[Mark Gowin]

Tim, thank you for sharing your test results. The cyan transition on the left is more abrupt than the magenta on the right. However, it is doubtful either will be noticeable in the types of photos which I expect to use the RX1.

I plan to test my RX1 to verify it doesn't have any extreme color cast issues that could be problematic. If my camera is no worse than yours, then it should be OK.

 

[m_driscoll]

Tim,

Thanks for posting this. This is no big deal to me. Every camera that I've had has some issue. I'm more upset about not being able to buy an EVF and the kind of Seattle weather that makes you suicidal!

Keep posting and reviewing.

Cheers, Matt

 

[douglasf13]

It's tough to say, but this might bother me. Even using the ZM 35/2 on the NEX-5N bothered me, and most don't consider that combo to have issues. These shifts can be particularly noticeable in concrete, which, living in L.A., is often a factor.

All of that being said, I may be ok with it, now that LR4 has the capabilities to fix it, although I haven't used this new feature, so I don't know how well that it works. The color shift is light enough that I'd guess that CornerFix and C1 would fix it easily.

It's pretty surprising to me that these shifts are present, considering this is a fixed lens 35mm. You'd think that Sony would correct it in firmware at the very least.

 

[barjohn]

That is about the same amount I see in my test shots and it doesn't seem to cause any real problems with normal photos. It would be nice, nonetheless for Sony to add an in camera fix which they should be able to do.

 

[Godfrey]

Looks like someone could update CornerFix to do the appropriate corrections.

These look much like what comes out of the M9 when fitted with the Color Skopar 28mm lens. CornerFix cleans up these effects very nicely:

(Each row (two M9 lens codes and no code) shows the out of camera image paired with the CornerFixed version. The columns show the apertures.)

 

[tashley]

I've been testing all this in detail for an article I've written and there's a lot to learn and a lot that's interesting around various options for dealing with it. Now I've written it all up I understand it in some detail and it's somewhat more complicated than it looks but very interesting!

 

[barjohn]

So, Tim, where is the article?

 

[Mark Gowin]

I look forward to reading your article Tim.

 

[barjohn]

Some of you may recall the discussion we had about this back in December. Imaging Resource labs encountered the issue too and discovered the cause and solution. It seems that it is caused by the vignetting correction in the camera. It applies it to both JPG and RAW files. If you turn it from AUTO to OFF no more color shading from left to right. However, you now need to apply vignetting correction in PP. LR4 fixes it easily as does C1. If you don't want any color cast, turn off the shading feature until SOny provides a firmware fix. No need to use corner fix or go through difficult gyrations. Latest from the IR Lab: A perplexing Sony RX1 mystery solved and a sneak peek at ongoing camera tests - Imaging Resource

 

[Mark Gowin]

John,

Thank you pointing this out. Sounds like an easy solution to simply turn off the in camera vignetting correction and add vignetting correction to a LR import preset.

Here is an example the color cast on my RX1.

Converting to B&W solves the problem in this case, but obviously that is not a preferred solution.

 

[Shashin]

Hmm, B&W, that would do it. LOL.

Well, I reported the problem to Sony and after a few rounds of "can you send me pictures," they have asked me to return the camera for "repair." I sent my camera last Thursday and I will let you know what happens when I get it back.

 

[barjohn]

This should be fixable with a firmware upgrade. IR said they have contacted Sony and given them the information. I have confirmed this in my own testing and now have that turned off.

 

[Shashin]

This should be fixable with a firmware upgrade. IR said they have contacted Sony and given them the information. I have confirmed this in my own testing and now have that turned off.

Apparently, Sony wants the camera. I contacted Sony in December and they knew what the problem was then.

 

[Mark Gowin]

Apparently, Sony wants the camera. I contacted Sony in December and they knew what the problem was then.

That is interesting. Please let us know you find out from Sony. I am curious. I am going to give the no in-camera auto correction approach a try.

 

[Steen]

So, Tim, where is the article?

I look forward to reading your article Tim.

John and Mark, I think this is the link you were looking for:

Tim Ashley Photography | Sony RX1: Colour Shift & Calibration, Aperture Series, ETTR & more

 

[D&A]

John,

Thank you pointing this out. Sounds like an easy solution to simply turn off the in camera vignetting correction and add vignetting correction to a LR import preset.

Here is an example the color cast on my RX1.

Converting to B&W solves the problem in this case, but obviously that is not a preferred solution.

Mark, a little off topic, but Sparks Lane (in Tenn) for image above? I first noticed some other images of yours (in the other RX-1 thread) and thought Great Smokies for the images? Just curious and love the scene in the winter time and was tempted to make the trip during this time of year. Are you there now? Thanks!

P.S. *** Just noticed the "hhttp" says "Sparks Lane"...so that answers my question about location.

Hope to hear how Sony resolves this color cast issue as I am thinking about aquiring this camera.

Dave (D&A)