No announcement yet.

4.6k Ursa Mini magenta issues?

  • Filter
  • Time
  • Show
Clear All
new posts

  • I tried a Zeiss Contax C/Y 50mm at different apertures down to f11 - no problem.

    A 28mm Minolta CLE rangefinder lens had corner problems.


    • I want to thank you all for updating this thread.

      I had read all pages of this thread and the magenta thread on official bmd forums as well.
      I will try other lenses, but what is confusing is that some owners reported a lot less magenta vigneting than others using the same lens.
      To the benefice of BMD and all owners, BMD would need to really find a way to communicate the list of compatible lenses once for all.

      Nyc did you successfully bought your UM46 at 6k$?



      • Hi Harold, exactly. nope. I have still never found a UM4.6k that doesn't exhibit the issue.

        The first run with serial numbers up to around 29xxxxxx had it really bad, and it was worse on the right side. After the pause in availability, the cameras after that were somewhat better, and no longer asymmetrically affected. So no more right side issue. FW 3.3 did also make a noticeable change in it, so now it's like only 33% as bad as it used to be. Some people say FW 4 made additional improvements in this issue, but I didnt see any evidence of that.

        That is not to say that this isnt a remarkable camera. It really is, and most people don't shoot higher than f/4 anyway, so it's not really a big problem for most users. And for sure the image the camera makes is really really good.

        With any camera, you learn how to work with it to get the best image. Some like overexposure, in others you need to protect highlights. Some look better with one LOG profile than another etc etc. So for this camera, if you shoot with non telecentric lenses, which is most of them out there, you will get some noticeable magenta vignetting above f/4 certainly above f/5.6 at focal lengths 24mm to 70mm or so.

        The cameras do vary as well, some are better than others in this respect, but I have yet to see a 4.6k DNG of an evenly lit white wall, 35mm and f/8 that didn't have the issue to some degree. And yes, if it is a rare but necessary setting for you, you can address the coloration in post as well with a power window.


        • Originally posted by nyccomposer View Post
          The Xenon lenses are telecentric, most full frame lenses arent and the magenta corners are there when you use non telecentric lenses, even if they are full frame.
          The Xenon lenses are not telecentric. You're confusing them with the more expensive Xenar lenses.


          • I have never used Xenon lenses, but this brochure calls them "near telecentric"



            • Originally posted by nyccomposer View Post
              I have never used Xenon lenses, but this brochure calls them "near telecentric"

              And the Xenar's are listed as being telecentric.

              It's like saying, as good as we can make it without being telecentric.

              Lens shading and sensor shading are going to give you darker corners and colour changes in those corners. One way to minimise that is to use a telecentric lens design, which is very difficult in a legacy 135 format (I hate the phrase FF) and the telecentric deisgn is one of the main deisgn goals of new formats like MFT for example.

              "Lenses designed for mirrorless camera systems such as Nikon 1 or Micro Four Thirds often use image-space telecentric lens designs,[10] which reduce shading and therefore light loss and blurring at the microlenses of the image sensor."


              and there's a very cool animation you can look at here that describes shading from non-telecentric designs.


              This is also only one piece of this puzzle. There's also sensor shading and the interaction with the microlens array. But they all interact with each other.

              Originally posted by Denny Smith View Post
              There is no real advantage to stopping a modern (and most vintage lenses too) down past f4 or S/5.6, do somon modern digital sensors will only degrade the image and cause artificacts like this, along width reduced resolution due to refraction.
              Denny, I presume you mean "diffraction" or more specifically lens diffraction ?


              Also note what happens as you stop down in the linked animation....



              • Thank, Yes I was John. I was referring to lens diffraction effect from stopping down to small apertures on most lenses. I was trying not to get to technical and keep my comment more generalized. But diffraction is more than just the lens (mostly found in non telecentric lenses), it is also the imtersction'if the lens image projection and the sensor, due to the micro lenses on the sensor. So while you could use f/32 on a film camera (diffraction was not problematic on film), the same lens on a digital sensor will soften (less sharp) the image due to lens diffraction.
                Last edited by Denny Smith; 12-19-2016, 12:23 PM.


                • John, thank you. I am no expert on optics, and I always love learning new information, the more the better. That point about MFT is also very illuminating, thank you.

                  I do still see the remaining magenta corners under those specific conditions, and I also see it has gotten a lot better than it was too. I hope they continue to work on the issue and refine it. Maybe it's just an issue with this sensor design, and the next sensor from BM will be an evolutionary step and not have the issue at all. Maybe that's what you meant about sensor shading and the microlens array. I guess the larger the sensor, the more oblique the angle of the light hitting the outside of the sensor, and the more oblique the angle, the more of the colored microlens the light goes through, and becomes attenuated. Maybe because there are two, or maybe just a property of the green dye, it seems the oblique light is being attenuates in the debayered image more in the green, and the result is that magenta vignette?

                  At any rate, I do appreciate the information. I am no expert, quite the contrary, I'm just trying to understand, so all information is helpful. I don't mean to dog the camera at all, but I do want to keep the discussion on track, so that's why I jumped in when it was suggested this was just regular lens vignetting.

                  At any rate, DSLR lenses that are meant for "FF" still cause the magenta corners at 35mm and f/8 etc. so like John says, it is an interaction of a number of factors, and some of them have to do with the sensor.
                  Last edited by nyccomposer; 12-19-2016, 01:15 PM.