A while ago, I posted a detailed article about a very defined pattern of red dots / artifacts that I saw on the Fuji X-series cameras when shooting against the sun. This was the first time I encountered such a problem, so without fully researching the issue and understanding the real cause, I wrongfully blamed the Fuji X-trans system for creating those patterns (my sincere apologies to all the Fuji fans!). A couple of our readers pointed me to some other links on the Internet that show a similar issue on different camera systems from Sony, Panasonic, Olympus and a number of others. The pattern indeed seemed to be quite similar between those and what I saw on Fuji cameras. I then decided to take my Olympus OM-D E-M5 camera for a side-by-side comparison and see if I could reproduce the issue on it as well. Now that I have done enough research to understand the root cause of this problem, I will not only explain the red dot phenomenon in detail, but also show image samples from two different mirrorless systems to illustrate the point.
The red dot patterns can be quite frustrating to see in images. Although this particular phenomenon only happens when the light source is very intense and the lens aperture is small, one would still probably wonder what causes it to happen and how one could minimize or even eliminate it. Before I talk about those things, let me first demonstrate that the red dot flare issue is not related to a particular camera or a lens. When shot in the same conditions, pretty much every modern mirrorless camera will show this and even our DSLRs are potentially prone to the same problems, as discussed below. Take a look at the following image taken by the Fuji X-Pro1 camera and the Fujinon XF 14mm f/2.8 lens at f/22:
And now take a look at an image captured by the Olympus EM5 with the Olympus 12mm f/2.0 lens at f/22:
The two images are taken 2 minutes apart, with about the same light intensity from the sun. As you can clearly see, the red dot pattern is visible in both shots, at slightly different intensity levels. But it is certainly there. And if I took every other APS-C mirrorless camera, whether it was Canon, Nikon, Sony, Panasonic, Pentax or Samsung, all of them would produce similar results. The effect can look worse at different angles and sun intensity levels, as demonstrated in my original article.
Occurrence
Before I explain why the red dot flare issue happens, let me first explain the conditions when you will be seeing this problem. The red dot issue only occurs when all of the below conditions are met:
- The camera is pointed at a very bright source of light (i.e. the Sun)
- The source of light is very intense
- The lens is stopped down to a small aperture (typically f/11 and smaller, but can be visible at f/8 on smaller systems)
This is not seen at large apertures, as clearly seen in the below image (Fuji X-Pro1 + 16-50mm @ f/3.8):
While the effect can be seen on pretty much any digital camera (yes, including DSLRs), it is highly amplified on mirrorless cameras with short flange distance – see more on this below.
Cause
So what causes the red dot flare issue? Basically, this has largely to do with the reflective nature of the sensor surface. As light rays enter the lens, they get squeezed into a very small aperture. At this point, each internal reflection in the form of flare is already part of the image. The light rays reach the sensor and immediately bounce back to the rear element of the lens. In essence, each pixel on the sensor that gets hit with the bright source of light reflects some of the light back to the rear element – that’s what creates the grid pattern. The reason why we see the red dots so large, is because of the flange distance. The small pixels from the sensor become larger when they first reach the rear lens element, then when the light reflects back from the rear element to the sensor, they are even bigger in size!
The reason why the effect is amplified on small mirrorless cameras has to do with the shorter flange distance. All of that back and forth reflection madness is happening because the intensity of reflections is higher at such short distances. If the flange distance is doubled like on DSLRs, those pixel reflections get too big to cause much trouble. This does not, however, mean that DSLRs are immune to this particular issue. If the lens is stopped down enough, DSLRs can produce different patterns too.
Hi Nassim,
do you know if a faster sensor readout speed could mitigate this problem?
Maybe it’s a silly question. But could the Nikon Z8 show less of this pattern, because it reads the sensor faster than the Z7 (II)?
It’s an optical artifact that, based on the description given, doesn’t depend on readout speed, only on the geometry and optics of the sensor, the aperture, and the rear element of the lens.
@Ircut thank you! :)
Nasim, In the article you state
“Although this particular phenomenon only happens when the light source is very intense and the lens aperture is small,”
I didn’t find this happening exclusively at small apertures. However, if the aperture is small, the possible area of reflecting surfaces becomes a bit bigger. So it makes sense – but not only .
In my findings, the red/green dots happen more often with wide-angle lenses. I faced it first with the Sigma dp0 quattro (14 mm fixed lens).
The situation always needs a dark shape (pointing at the sun through a window- or door frame, or through trees) to become so massive that I need to trash the picture.
Nikon is not better than others – in my findings I didn’t experience this kind of flaw as often as with the Z-series with, say, Fuji.
A friend and Sony owner showed me the rearside of his Tamron, Voigtländer, Laowa and Sony 200-600 lenses. The mount-Ø of Sony is much smaller than on Z-mount. That’s one reason for less red dots. And the other is, all this companies do much more to prevent reflections than Nikon does. Light catching steps, velvet, matt laque, rectangular masks. Try for yourself to provoke the red/green dots on a Sony.
Hey Nasim, thank you for all the data. I appreciate a well-thought out discussion. I had actually been avoiding the X-Trans system specifically because of this, in spite of really wanting to try it! This is the first time I’ve seen a reasonable explanation to the problem.
Interestingly, I’ve only ever used adapted vintage 35mm lenses on my current mirrorless camera, and I often shoot into the sun. But with my aperture as wide-open as possible. Now I know why I’ve never noted this issue! (I had previously thought it was owing to the Bayer filter design.)
Cheers!
Hi
I have just found this phenomenon on my Nikon Z7. Very disappointing. And yes, it is less pronounced when using the FTZ Adapter with an F-mount lens on it.
Best regards
Tobias
Tobias, every mirrorless camera has it and the Z7 is one of the least offenders in this department. DSLR cameras show it too, but the dots are much bigger in size, making them harder to see.
This is not just a small camera problem, I have encountered this on large cinema cameras (RED EPIC).
Make sure to use NDs!!! they will help you expose correctly at a more open F stop.
I suggest having an ND9 for most situations, although the variable ND filters are really handy because you can dial them in exactly and work in a variety of lighting conditions, though keep in mind variable nds are actually 2 pieces of polarized glass so they will slightly affect your overall look.
If using NDs of specific strength, make sure you are using a non IR ND, as some cheap ones available are IR and have a red tint to them that will definitely be in your image. This can be removed but its a pain and does decrease your color information.
Hi Nasim!
Thanks a lot for your suggestion sharing info.
After reading more than 1000 reviews, I bought Nikon D5500 with 18-140 mm lens in the same box last month and shoot more than 1000 good quality photos. First time i shoot at auto only. After suggestion from the nikon sales person i now generally shoot at P mode. My main concerns are family, park, zoo, travel etc. I have a few questions.
(1) In which mode and settings for which situation, i can take excellent quality pictures and my child’s family video.?
(2) I’m just using one Nikon original lens hood. No UV filter installed yet. Is there any need to buy
one UV Filter? If yes the brand Nikon or Hoya or something else and its size.
(3) Suggestion for lens cleaner. To avoid moisture and fungus can i keep few silica gel?
(4) Is there any need to buy one 50mm/f1.8G for portrait only or should i stick to this versatile lens.
Thanks so much for this post. I was in a panic after I bought a camera I love EXCEPT for the mysterious red dot glare in photos of a sunrise, and this is the only place offering an explanation!!
The rainbow colours seen on the right of the second image are due to diffraction caused by the sensor as per Wikipedia:
Diffraction artifact in digital cameras
One form of flare is specific to digital cameras. With the sun shining on an unprotected lens, a group of small rainbows appears. This artifact is formed by internal diffraction on the image sensor, which acts like a diffraction grating. Unlike true lens flare, this artifact is not visible in the eyepiece of a digital SLR camera, making it more difficult to avoid.
Unfortunately the explanation is totally wrong. The dots you see are actually polygons as the pattern is caused by light reflecting from the diaphragm in the lens. As you stop down the diaphragm gets bigger so more reflecting area. Light reflecting from the diaphragm gets multiply reflected from lens elements and hence the numerous reflections in a geometric pattern. The better the lens coating the less of a problem you get. Prime lenses have fewer elements and hence less reflections than zoom lenses. Good lenses have coatings on ALL surfaces. You can easily see the effect of coating all surfaces if you look at a pair of cheap binoculars and then an expensive pair with the same optical size where all elements are coated. The good coated pair will be far brighter. If a coating reflects only 3% of the light multiply that by the number of elements and the light loss is very significant – easily 1 stop (15 elements * 3%).
If you ant a better explanation look at: www.cambridgeincolour.com/tutor…-flare.htm
Light reflected from the sensor produces ghost images around bright light.
I turned the af off and more dots thanks
no more dots that’s what I wanted to say