The X-Pro2 is a camera with lots of positive characteristics but more importantly, it is a relevant upgrade, taking Fujifilm’s mirrorless system to the next level alongside the recently announced X-T2. However, no camera is perfect and sometimes certain flaws can cause more irritation than others.
Update I (October 2016): the X-T2 has the same issue which doesn’t surprise me (same sensor). An example has been included at the end of the article.
Update II (June 2017): my copy of the X-T20 doesn’t seem to suffer from the same issue, despite having the same sensor as the X-Pro2/X-T2. This might suggest that Fujifilm is fixing the problem.
In the past we shared our findings about the sensor ghost flares of the first generation Sony A7. This time I want to talk about a problem related to the X-Pro2 that I already mentioned in my in-depth review: grid artefacts. Since I keep seeing this topic appear in Facebook groups and on forums, I thought that a dedicated article could be of help.
Before proceeding any further, there are three important things I want to highlight:
- Note #1: My aim with this article is to provide a solution (not definitive, but simple and effective) for users who already own an X-Pro2 (or X-T2), have encountered the issue, but don’t know how to deal with it. It can also be of interest to owners who haven’t encountered the problem yet, or to potential customers. My intention isn’t to cause panic, which is something that happens easily on the internet when a flaw is found on a new camera.
- Note #2: There have been reports of X-Pro2 users NOT being affected by these grid artefacts. This might suggest that the problem is limited to only a select number of X-Pro2 cameras. If this is true, Fujifilm might be able to identify the serial numbers if the brand acknowledges the issue later on. Update: since the X-T2 is also affected, I somehow doubt that the issue is limited to select samples.
- Note #3: The grid artefacts are mostly noticeable when you enlarge the image. In this article I included 2:1 crops to better illustrate my examples. This means that even if it is present, the “grid” might not always be a big concern when it comes to the final result.
Identifying the problem: grid artefacts in backlit situations
When shooting in a backlit situation you can encounter a purple veiling flare. This type of flare is not uncommon and I’ve seen it from other mirrorless systems, Micro Four Thirds being the most relevant example. However in the case of the X-Pro2 a grid artefact is also visible in the image. Depending on how strong the purple flare is, the visibility of the grid can increase or decrease.
Invasive purple veiling flare.
Click on the image to open the full res version
Less invasive purple flare.
Click on the image to open the full res version
What is causing the purple flare and grid artefacts?
A flare occurs when a bright source of light reflects and bounces onto the various glass elements inside the barrel. It is a natural characteristic of a photography lens and can appear in different forms (veiling, ghost) and different colours (green, purple, rainbow). Some lenses are affected more than others and this is mainly related to the manufacturer’s ability to reduce flares by using special glass and coatings. The latter has different names depending on the specific brand (T* coating for Zeiss, Nano GI for Fujifilm, etc.).
Veiling flare, which is what you see in the two images above, usually occurs when the source of light is outside your frame but close enough to reach the front element of your lens. A hood can help but depending on its size, the angle of the light source and your shooting position, it might not be enough.
Based on my research, the purple colour of the veiling flare is not caused by the lens itself but by the micro lenses on the sensor. Since all digital cameras have sensors with micro lenses, it is clear that some chips are more prone to reflections than others.
Sensor flares are not uncommon: the light reflects from the image sensor to the lens and back to the sensor again. Basically the camera sees and records a reflection of a portion of its own imaging sensor. Usually this type of flare can be identified with green and red dots appearing in your image (it is also called red dot flare) when a source of light is in the composition.
An example of sensor flare (also known as red dot flare)
So what about this grid artefact pattern? Well, the purple colour tells us that the flare is not just caused by the lens but also the sensor. This helps us identify the problem: it is another type of reflection from the sensor that is different and more subtle than the classic sensor red dot flare we just saw above.
In this case, it is probably related to the unique pattern produced by Fujifilm’s X-Trans technology. So I asked myself: would this happen with the X-T1 and its 16 MP X-Trans II sensor? I ran a side-by-side test using the XF 35mm 1.4 and the difference is noticeable: I had a hard time finding any purple flare with the X-T1 while with the X-Pro2 they were much more prone to appear. It seems that the problem is specifically related to the new 24MP X-Trans III sensor.
However, during my various attempts to find a purple flare using the X-T1, I did come across to a very subtle one (almost blinding myself in the process!). If I enlarge it there are some traces of a similar grid pattern. This seems to confirm that the problem is related to the X-Trans pixel array but in the case of the X-T1, the issue is far less annoying and can be easily filed into the “irrelevant” folder.
Click on the image to open the full res version
The work-around: it’s basic yet effective!
Purple flares are annoying and difficult to minimise afterwards. With a little bit of post-production, you can correct the colour but the veiling effect is still there. Plus in the case of the X-Pro2, correcting the colour in post won’t eliminate the grid artefact.
Purple flare color corrected in Lightroom with the adjustment brush but the grid pattern remains.
Click on the image to open the full res version
The solution is to avoid the purple flare by slightly adjusting your composition (and I mean “slightly” in its purest definition).
Let’s analyse the same example seen at the beginning. You can notice that by varying my composition just a little, which involved tilting down the camera and standing a little higher (a few millimetres really), the purple flare disappears and so does the grid artefact.
Purple flare + grid artefacts
Click on the image to open the full res version
Adjusting the composition slightly got rid of both the purple flare and grid artefacts.
Click on the image to open the full res version
This solution might seem primitive but if you follow my reasoning, you can understand how it can help in many situations. I explained before that the purple colour and grid artefacts are linked because they are both the result of a reflection from the sensor.
This particular purple flare can be easily seen live in the electronic viewfinder while you are composing your image. So if the purple flare appears, use it as a notification alert for the grid artefacts. Actually, in a few situations, I saw the grid pattern inside the EVF itself.
The artefacts will appear in the OOC JPGs too so you can also double check your results straight after taking the shot.
Of course this solution is not a permanent fix because we are talking about a hardware issue but since it involves such little effort from the photographer, I think it is worth explaining and can be useful for someone lacking experience.
Some users suggested that it happens especially with prime lenses whose rear element is close to the mount. However the images shared above were taken with the XF 35mm f/1.4: its rear element is not that close to the mount in comparison to other lenses. Just as a reference, a lens with a rear element that is really close to the mount looks like the Sony/Zeiss FE 55mm 1.8 (on the right in the image below).
The rear element theory doesn’t work also because I detected the grid artefacts with various lenses like the Fujinon 90mm f/2, the Samyang 50mm f/1.2 and even the XF 100-400mm. The last example is the perfect proof since it can house the 1.4x and 2.x TC (teleconverter) on the rear. However I don’t deny the possibility that the problem can be more present with some lenses than others (depending on how strong the veiling flare is).
Purple flare and grid on the right
Click on the image to open the full res version
Crappy picture but it shows the same effect with a telephoto zoom lens
Click on the image to open the full res version
Double check: does the grid artefact appear in “non-purple” zones as well?
I analysed all the images I took in backlit situations and the answer would be no: the only time I can clearly see it is in a purple zone. Actually I have a pretty clear example below: here we have green and purple flares but the grid is mainly visible in the purple area.
Click on the image to open the full res version
The only time I saw the grid artefacts in a non-purple zone was in the image below. You can notice it slightly in the veiling flare at the top right. The flare has a very subtle warm tint but I am not sure I can classify it as a purple flare.
Click on the image to open the full res version
This shows that there isn’t a definitive answer but purple flares are the main reason I’ve encounter so far for the appearance of these grid artefacts. For example, in this other image below, there is veiling flare but no grid artifacts whatsoever.
Click on the image to open the full res version
What about the X-T2? It’s the same thing.
Since publishing this article, I’ve received several questions about the X-T2 and the possibility of it being affected by the same problem. Now that I’m back from Photokina, I’ve finally started to review the X-T2 and I can confirm the presence of the same purple flare/grid artefact, as you can see in the example below. The same suggestions and tips mentioned above for the X-Pro2 also apply to the X-T2.
What about the X-T20? It’s much better, at least on my copy!
We recently purchased the X-T20 to conduct our full review and comparisons. I compared it side-by-side with our X-T2 to see if the infamous purple flare/grid artefacts would appear and to my surprise, they didn’t. The X-T20 doesn’t seem to have any serious purple flare and I could only spot the grid in very limited zones of select images. It looks like Fujifilm is fixing the problem on this camera which hopefully means they are doing the same for the new flagship models that come out of the factory.
Conclusion: should you be concerned?
When a flaw is found on a new camera, the issue can easily be blown out of proportion, sometimes rightfully, sometimes not. The truth is that we all react differently to these problems, mainly because it depends on how much the issue affects our shooting. Personally, I started to investigate the problem after reading a complaint in a Facebook group, which proves that I didn’t notice it at first.
It is clear that the sensor has an issue with backlit images. I’ve seen purple flares from other cameras but not the grid artefact. If backlit shots play a big role in your photography style, I can understand that you might have some doubts about cameras that present a weakness in a specific shooting condition you like so much.
You may also be uncomfortable because the cameras can’t be completely trusted. I would also add that it depends on how much you post-process your images and how large you print. In some cases, the grid won’t catch your eye while in others it can be more distracting. But remember that most of the time, you need to pixel peep to clearly see it. And in this case, there is the effortless solution that I described above and that will work in most cases. It is a matter of assessing the problem, finding a work-around and trying the shot again to see if that work-around is good enough for you or not.
As a final note, I don’t think there can be a software fix since it is a hardware problem. But my positive results taken with the X-T20 suggests that the company is addressing the issue although no official statement has been made.
Same location as the Samyang shot seen before, but I re-positioned the subject
and changed the composition to avoid the flare.
Click on the image to open the full res version
Here is a detailed summary of the purple grid phenomenom in X-Trans III sensors: https://petapixel.com/2017/02/23/fuji-cameras-produce-strange-purple-flaregrid-artifact/
I can’t see it on the rear LCD…just in the EVF.
I’m not sure what those lines are exactly, but if they don’t appear in the final image I wouldn’t worry about. Do they appear on the LCD screen as well.
The lines are mostly very faint and are sort of a light gray color. They appear anywhere in the EVF (depending on the lighting of the scene) in mostly mid tone areas. I took an iPhone photo of my EVF to show an example. Photo in the link. It’s not the best photo of the EVF and I think the lower area I marked with a red box is more exaggerated by the iphone photo because that’s not what I usually see. The upper area marked with the red box is more typical of what I see. Faint gray vertical lines. Also it seems those areas with the lines will sort of ‘shimmer’ as you move the camera slowly across the scene while those lines are visible.
https://photos.smugmug.com/photos/i-4cSNwJk/0/5503fc3b/O/i-4cSNwJk.jpg
I never noticed a vertical line pattern on my X-T2 and X-T20. Is it thick or in a particular colour? Does it appear at the center?
Question. I haven’t yet seen this grid or line pattern in any of my final photos (viewed on the computer), but I DO almost always see a vertical line pattern in the EVF (not rear LCD, or maybe it’s too hard to see on the LCD) when reviewing photos in playback mode or in live view before taking photos, in many different lighting conditions with no flare. So, is it possible that the EVF is simply exaggerating or enhancing my view of this grid/line pattern because I don’t see it in my photos, I just see it in the EVF (mostly as very faint vertical line pattern). Please give me yours thoughts. Just want to be sure my camera is ‘fine’ and working as expected since I see the vertical line (faintly) in the EVF most of the time even when there’s no flare from a strong light source).
Hallo Mathieu,
Did you try with the electronic shutter ? Or with a slower exposure ?
Regards
its not really an issue for me because i dont shoot a lot against the sun. but i’m very curious if fuji will somehow make a statement. a friend of mine has the 16-55 2.8 and 50-140 2.8 which both have thr purple grid like crazy shooting against the sun on the longest end. at shorter focal lenths its not so strong. yesterday a portrait photog told me his 56mm 1.2 has some flare here and there but not the extreme purple grid with his xt2 and he shoots a lot against the sun. its strange indeed.
I’m not sure about a firmware fix because it isn’t a software problem.
and here is my Fujinon 16mm 1.4 test https://uploads.disquscdn.com/images/e144b7d91cf94f20f6c1aa5e8d1dcc73c2b44bcd92647b490986629327b932fb.jpg
Here is one of your example shots edited with PS and the boxblur, purple allmost completely removed with PS (you can take even more purple out of the image) there is some detail loss but thats probably the only way to save images which have the annoying purple grid https://uploads.disquscdn.com/images/ed4ffad94a48e34f5088b52f542ed6094378fdba672c5e37e030aec11a95a21e.jpg
Here is a before and after comparison. edited in PS with the boxblur 1px. i shot it with the fujinon 50-140 2.8 @ 140mm https://uploads.disquscdn.com/images/19df1a7f59bf14d0a0c15c11408398ed4ad6976f0b2f8baa9054b2cb3b0df3c5.jpg
very good article, mathieu look here, the grid can be minimised in post using boxblur filter with only 1px: http://blog.roos.photography/2016/09/purple-flares-and-grid-pattern-on.html
I was lent it for a few hours, so no idea!
I got mine almost 1 month ago as well so it looks like yours is as good as mine 😉
Hi, Thanks for your detailed review. I didn’t realize this problem until I came across your article. So I tested it next day and did find the purple flare. It seems to me very easy to reproduce the problem, but when I zoomed into the purple flare area of the photos to check the details, I could hardly see the grid pattern/texture (just appears in some tiny spots if at all). I’m sure I wouldn’t have noticed those grids even with pixel peeping if I hadn’t known of this problem beforehand. I’m wondering if this means my X-T20 is as good as yours. When did you purchase yours? I purchased mine 1 month ago (early May). It’s been out of stock on most retailers’ websites (I’m living in US) for quite a while. I was very excited when I got mine, but now I’m debating if I should keep it or just return it. Attached is the picture (I cropped it to just show the purple flare area) https://uploads.disquscdn.com/images/48aa69fed567741dd13e57aaa8430d87a1952848b73158a23387175ed0a1eaed.jpg
I don’t think it’s firmware related. When did you buy your X-T20?
Interesting that you didn’t come across them at all. They’re there in many photos where I shot into the sun. Maybe yours has a firmware upgrade that deals with it somehow? No idea, and no way to check if the X-Pro2/X-T2 would’ve looked significantly worse in the same conditions my photos were taken in.
Update: my X-T20 doesn’t suffer from this issue. More details in the article.
Update: my X-T20 copy doesn’t have the same issue, it’s much better (article updated with images and text).
I updated the article with my findings about the X-T20 and it is much better.
I used the X-T20 and can confirm the issue is still there.
It does, first-hand experience.
Also there on the X-T20. Had the chance of using it for a few hours and forgot what I’d read about the grid issue (apparently it was rare so I didn’t bother) and shot into the sun, and sure thing, purple flare and grid.
In Petapixel article it is explained that the issue came with the XTrans II sensor, when they added the phase detect autofocus layer on the sensor.
https://petapixel.com/2017/02/23/fuji-cameras-produce-strange-purple-flaregrid-artifact/
I confirm that I never had this issue with the X-PRO1, which was first XTrans generation, without phase detect autofocus.
I guess Fuji had to make a call here between autofocus speed and this “random” artifact.
Maybe they could produce a X-pro2 S with slow autofocus and perfect image quality in all situation ?
Thanks for this explanation ! Fyi, I read this because I have bought a X100f and…
I can say that the sensor flare is definitely there and much worse than the X-T10. I tend to do a foolish thing and shoot into towards the sun quite a bit. I could deal with the sensor flare from the X-T10, but the sensor flare with the X-Trans III is quite intense.
I’ll double check as soon as my X-T20 arrives.
So is the same problem with x-t20.
My local fuji manager say that they solve de problem with firmware, can I trust them?
I wondered (I am not a professional) about using a Polarizing filter? Could it help reducing the occurrence this purple flare and grid pattern?
I have a XE-1 and shoot only Leica M mount glass on it with an adapter that I’ve had for one month and I get these artifacts every day so it is most definitely not a rare occurrence. They say this is a by product of the mirrorless system and if it isn’t corrected soon it will hurt all mirrorless sales. I never saw this on my M8 so it’s not my lenses.
Damn…I thought it was my lensbaby!! Disappointed…
Thanks for taking a look. I figured out partly what caused this problem. I had opened the file in Lightroom and converted it to a DNG file. Then, when I opened the DNG file in Photoshop, it had that grid of dots. Sounds like you are spot on that it is a compression artifact. I was doing that because I was using CS5 which does not have adobe camera raw profiles for RAF files. If I export it from Lightroom as a JPEG, no dots.
very sensible article. thanks
I see what you mean for the photo on the left but I don’t think it’s the same thing. To be quite honest I am not sure what does grids are, they look more like a bad compression artefact.
I tried different softwares for the RAW file but it can be visible in the JPG as well.
I just tried to open one of my RAF files in Silkypix and it shows the same pattern. I remember I tried different softwares (including Iridient) when preparing the article and each one displays the grid.
As for returning it, do you often shoot in backlight? As I shown above, it’s often a matter of sligthly re-compose the shot.
I recently bought the xpro2 and over the first few days found that at least one of my photos has this grid pattern when I process the RAW file in Lightroom. However, I also found that if I process the same RAW file in Silkypix, then there are no grid dots. Has anyone else found this to be the case too? I am still trying to decide whether to keep my xpro2 or return it. Thanks.
Thanks for the reply Mathieu 🙂 Most of the posts/threads/etc I’ve seen about the issue seem to treat the phenomenon as being random or rare, but I think your “experiment” demonstrates that the flare/blob/grid is a repeatable and predictable property of the sensor. Being a Fuji convert from m43, I guess I’m slightly disappointed the most important part of the camera is (in my opinion) compromised. If we’re discontent with lenses, we can easily sell them and purchase new ones. We can live with clunky menu systems. We can deal with ergonomic issues. But we can’t change the sensor. It’s up to the individual if this undesirable effect is a quirk of the camera that can be tolerated or not. It’s not really something I would want to think about having to work around. I love the Fuji camera system and its stellar jpeg quality (since I shoot jpegs 95% of the time), but I’m afraid that there is a considerable flaw here in their latest sensor.
I will definitely get my hands-on a X-T20 sooner or later so I’ll check it the problem appears again. As for the comparison with the X-A3, that’s a good idea. If I can get that camera too, I will definitely compare them.
I wouldn’t be surprised at all since the sensor is the same, unless Fuji managed to fix the problem. Somehow I doubt it but I could be wrong.
Is it safe to assume the X-T20 will exhibit the same issue?
It will be interesting to see if this problem continues to be evident in the X-T20, the third camera body with the 24mp X-Trans III sensor. While Fuji has had some delays/other issues with the X-A3, it has the 24mp Bayer sensor. Would it be possible to see a comparison between both Fuji 24mp sensors to see if it truly is an X-Trans III issue? The article provides a decent solution, but as the comments below indicate, the photographer doesn’t always have the luxury of time to recompose the shot. Nor does the photographer necessarily want to recompose the shot, as in the case of shooting into the light for a desired effect (for example sun-on-the-horizon shots, backlit portraiture, etc.). I’m a huge Fuji fan and was considering upgrading from the X-T10 (which doesn’t seem to have this issue), but I am hesitant to spend my hard-earned dollars on a camera that has a known hardware-based image quality issue. If the fault is mine, no problem. If it’s the camera, I may skip this generation and wait for the X-Trans IV.
I am so glad I saw this article before purchasing the X-T2. I really love the way the Fuji handles when I tested it in the store, but all of the accounts I have read about the purple grid issues has turned me off. I will stick around and see if Fuji ever decides to address this before I make my next purchase, otherwise I will go from my A7ii to A7iii/riii. Great article!
I have ran into this issue on a somewhat regular basis.. on a shoot last
week I ended up with a bunch of what would otherwise be good shots that
are ruined (it’ looks like they are printed on a canvas surface before
even leaving the computer.. client might not notice soo much, but I sure
do. Disappointing).. crop of image attached. This was with the 90mm f 2
and MILD back lighting (from what I could see NO strong light source
was directly hitting the front of the lens) https://uploads.disquscdn.com/images/1114027cc412a3b89ec8d69219a5ec5d4d487c63ffae4fd955c45d1587a907f1.jpg
I noticed this issue in the summer when I got my X-T2. I had an UV filter on the lens, I thought it’s caused by the filter…
Indeed. However, as a landscape guy who often shoots with the sun just on the horizon or high in my frame I find that I can not eliminate it completely. I just shift where in the frame the sensor flare is occuring. I love the handling of the Fuji bodies and do very much like the output of my XE1. I’ve just never owned a camera that I have to readjust my vision in order to take a shot. Yes, some cameras require certain settings or lenses to get a proper “look” but a camera that actually requires me adjust my “vision”? Not too happy about that.
You can easily see the purple flare in the EVF and most of the time you can get rid of it by changing your composition slightly.
Really wish I would have seen this article before purchasing my XT2. Took it outside this morning to shoot a few test pictures to make sure the camera worked and discovered some rather odd grid artifacts in what I thought was purple flare. Quite obvious at full size and even more so when I convert my image to black and white which is my mainstay style. After searching about it I found your article. Not too pleased with my new camera at this moment. I tend to keep gear for the long haul (still shooting with my XE1 and an old Sony A850). Now… Not sure I want to work with a camera that I have to work around.
I truly wonder why more sites are not making comment about this issue.
I’m sorry, I cannot reproduce this “grid” and I also have an X-T2. I checked 4 pictures (not a lot, I admit), found with heavy backlight at 100 % in Aperture (JPGs only), CaptureOne and Iridient last releases (uncompressed RAW), no grid in the purple areas.
What I miss in your article is something about the RAW-converter – or were the sample pictures OOC JPGs? Was there in-camera denoising involved? And when on earth you ever need 200% enlargement? I mean, except finding grids which very possibly will not make their way out of the nozzles of an inkjet printer?
Andrew Zargoza, what you show as “grid” to me looks very much like incandescent light and electronic shutter. I may be wrong, but it’s very different from the effects Mathieu was showing here.
Every camera has one issue or another. Don’t lose any sleep just adapt.
Cool. I’d be interested in what you think revisiting the X-Pro2 with latest firmware v2.00. I’m sure there will be and has been lots of comparisons to the X-T2 because of this.. seems to basically make them perform essentially the exact same (MINUS the ability to customize AF-C, 4K, and the performance enhancements with the battery grip and the grip’s Boost mode, which is different and not quite as good as the in-camera body Boost mode). I’d be particularly interested to know what you thought of the Corrected AF Frame. I really feel it does not make complete sense the way it works and they had implemented better before the update which was more logical and usable.
Yes, all the lenses I have that have an update have been to the latest.. There’s only really one, which is the 18mm f/2. The 35mm f/2, 23mm f/2 and 56mm f/1.2 I have have no updates.
FYI, I know someone who has the 18mm f/2, did not update the firmware for the lens, but is exhibiting the “old way” the Corrected AF Frame works.. was asking me if I think he should up date it.. I said if he doesn’t like the new update which causes the Corrected AF Frame to perform unsatisfactory the same way I am saying it is, then don’t. He really does not like the change to the way the Corrected AF Frame works when On, as I do.
It really has nothing to do with parallax correction. I totally understand parallax correction. And, ideally, if you’re going to get in that close, you should be switching to the EVF to make sure you got the focus and framing you want. They’ve gotten rid of macro mode so you can focus down to minimal focus distance even though it’s probably not recommended.. which is fine, Fuji leaves the decision to switch to the EVF or not up to the camera user. That’s fine. Again, this is not really about parallax error nor is it about accurately focusing and framing at very close range.
Again, my main issue is NOW, the Corrected AF Frame only shows the infinity box (upper left box) and “recommended value/focus point” (lower right box) which is 50cm (for the X-Pro2). Why 50cm? Why not the complete range from Infinity to the specific lens’ minimal focus distance?
It would be far more helpful to represent the perceptual depth and focusing range of the lens so you have a clear reference to aim your camera at an area anywhere between infinity and minimal focus distance to find and make your green AF confirmation box.
– If you know you’re going to focus closer, then aim closer towards the minimal focus distance box.
– If you know your subject is further out, then aim it closer towards the infinity box.
– And, if you know it’s a mid-range subject, obviously aim it somewhere in the middle between infinity and minimal focus distance.
That seems fairly easy to use and with practice can probably hit a green AF confirmation box exactly where you want to using the OVF and AF shooting.
Why can’t you do that now with after the update? You certainly can, but it’s significantly less intuitive and less easy to mark your intended AF point without a “complete” visual reference.
NOW, you still have infinity box (no problem here), BUT the other box (lower right corner) represents 50cm. Why? anything less than 50cm to your lens’ minimum focus distance is completely open ended. There’s not way to see and know where your minimum focus begins or ends. So, where would you think to begin to aim your camera to obtain AF focus lock?
LET’S COMPOUND THE PROBLEM FURTHER. Each lens has a different minimal focal length, right? Some can focus closer to the lens, some further away from lens.
– If the lens can focus really close, then you can expect if you wanted to focus really close using the OVF, you’d aim passed & away from the 50cm box towards the lower right corner in the OVF, probably as close as you think you can get.
– If the lens cannot focus that close and further away from the lens, then you would expect to focus really close using the OVF, you’d have to aim passed & away from the 50cm box towards the lower right corner in the OVF, as close as you think you can get, but obviously not as close as the example above. BUT, THAT IS THE QUESTION. Just how far towards the lower right corner do you think you can get before you go past the “invisible” minimal focus distance? You could be making several attempts before you finally get a green AF confirmation box.
Do you see how this can be a waste of time and effort because the OVF and Corrected AF Frame no longer tells you where the minimal focus distance is at so you know not to try to aim the camera passed that point?
The margin for error and your search parameter to get the correct intended green AF confirmation box has increased.
LET’S COMPOUND THE PROBLEM EVEN FURTHER. The v2.00 firmware release notes states: “If the minimum focus range of the attached lens is greater than the recommended value, the frame is displayed on the position of the minimum focus range.”
So, if the lens’ minimal focus range is greater than 50cm, then it will display the minimal focus range instead of the 50cm. Am I interpreting this correctly? Then, for those certain lenses, the Corrected AF Frame will display the full range from Infinity (upper left box) to Minimal Focus Distance (lower right box)? So, they will display and represent the full focus range with the Corrected AF Frame with these particular lenses, right?
If so, then why the inconsistency in how the Corrected AF Frame is shown and consequently can be used? Some lenses that have greater minimal focus distance will show the complete focus range, and some lenses that do not have minimal focus distance will not show the complete focus range, only from 50cm to infinity, anything below 50cm has not reference and basically a bigger guessing game.
Why can’t they just represent the complete focus range from infinity to minimal focus distance (or vice versa, same idea) for ALL LENSES? Why a partial visual reference for some lenses? And, I would be highly inclined to believe that the majority of the lenses that you would use with the X-Pro2 and the OVF fall under the category of lenses that get this partial visual reference.
It just makes SO MUCH MORE SENSE if you get the complete focus range from infinity to minimal focus distance for all lenses all the time.
It never used to work this way before.. the 50cm thing… minimal focus distance was shown instead of 50cm. And, you never could nor would get a green AF confirmation box outside the box (the lower right one of the 2 you get from with Corrected AF Frame On). I’m 99.9% certain of this as I use the OVF at least 90% of the time with both my X-Pro1 (when I had it) and X-Pro2.
Obviously, if minimal focus distance was shown instead of 50cm, you could expect the lower right box of the Corrected AF Frame to be fixed closer to the lower right corner of the EVF. Wherever it may be, at least you’d see the minimal focus distance and know where not to search and try the space between the infinity and minimal focus distance boxes ONLY.
As it stands with v2.00 update, you can search between Infinity & 50cm AND between 50cm and god knows where to the lower right corner of the OVF.
Doesn’t this lack of a certain and confined area to find your AF point less efficient and therefore less accurate.. the opposite of what the intent was for the Corrected AF Frame change and supposed “improvement”?
Are my comparisons and differences making any sense to you?
Maybe it’s me?… it’s probably me…. But I don’t get why X-Pro2 (& X-Pro1) + OVF users (who use AF and Corrected AF Frame On) do not understand why the way it works post v2.00 update is not as good as before the update?
Other than the Corrected AF Frame, the v2.00 update is awesome!
What did they miss in the v2.00 firmware update (besides making the Corrected AF Frame worst)?
I would still have liked customizable AE-L & AF-L buttons like the X-T2 can do. Lastly, I think it would be helpful to be able to change the color of the OVF displayed frame lines and information along the edges. Sometimes white is harder to see (no matter how bright you can make it) against a fair number of situations.. at least give me a red or blue.. much like the Leica M240 can change its frame line colors.
.. and if you do start to get too close in focusing with the OVF and the framable area begins to get cropped out, maybe the frame lines should change to yellow to warn you so you could decide right then if you want to keep using the OVF to focus and frame or switch to the EVF.
I should get my hands on an X-Pro2 again soon. Just to double check: did you update the lens firmware as well? Because there is a specific update for the parallax correction.
I have a slightly updated version of my gripe with the Corrected AF Frame. I was right, but wrong about 1 detail regarding the minimal focus distance indicator… It was the way I was stating it, but the manual says otherwise, not the update makes it perform more like what the manual says, but I think the way it was before is actually better, still, and makes more sense.
You may want to make notes, maybe a short video or photos, with your X-Pro2 before updating so you can compare the differences before and after the update.
First, I stand corrected about the 2 boxes for the Corrected AF Frame. The solid box that’s more upper left IS infinity, there’s no disputing that. BUT, the box that’s dashed/not solid that’s more lower right IS NOT minimal focus distance, but the recommended focus distance of the camera(?) is, which is 50cm in the case of the X-Pro2 (80cm in the case of the X-Pro1). So, there never was a box that represented the minimal focus distance in the Corrected AF?.
I don’t see how that’s possible.. I’m almost 99% certain that’s exactly how it was always represented before. You could never go passed the lower right box with a green AF confirmation box.. that was indeed the minimum focus indicator box and the limit.
I don’t like the idea the corrected AF is fixed on the recommended focusing distance OR further away on some lenses. It’s so inconsistent this way, no? Don’t like this at all.
I’m still on this firmware release note: “·ON:The CORRECTED AF FRAME in the OVF is fixed on the recommended value described in the Owner’s manual. If the minimum focus range of the attached lens is greater than the recommended value, the frame is displayed on the position of the minimum focus range. ”
I referenced the X-Pro2 manual, 50cm is the recommended value I believe this is referring to. Fine, but why fixed on this?
It’s like anything beyond this point to infinity can be represented accurately, and any shorter and in front of this point towards minimum focus distance can be anywhere. Why?
It would just make visual sense if the fixed point always includes the minimal focus distance. But it is not always like this and set to the 50cm. The exception to this is if the attached lens’s minimal focus distance is greater than the 50cm (so i assume it will be offset a little more to the upper left). If it can do this and set the minimal focus distance on some lenses, why not all?
It makes way more sense for each lens, if the corrected AF is always fixed, that it should always be at and show the minimal focusing distance and infinity for that particular mounted lens… not some arbitrary midpoint which could greatly vary from lens to lens so you never know just how much room you have to search and lock focus.. minimal focusing distance of each lens is different, so the 50cm (or 80cm on the XP1) can be the same, but in actuality where that is in relation to the minimal focus distance of the lens will vary and move.. it may be about half way, it could be further away from the min. focusing distance of the lens and you will have more space to search for the AF, or it could be closer towards the minimal focusing distance and you will have less space to search for the AF.
To use this new system, it would require you to always know what your lens’s minimal focusing distance is and then figure out how that might visually be imagined in relationship to the 50cm (or 80cm of the XP1) in the OVF.. That’s too much thinking to grasp that concept.
If it’s like the old way [and I now believe it was not represented as the manual(s) originally dictated of 50cm(XP2) or 80cm(XP1) before these updates], as we are describing and you have infinity AND a set minimal focusing distance represented with the corrected AF on all lenses, you will always know where you limits and working range is. It’s way less guessing and mental calculations to imagine the working range and limits of this new update.
Done, that’s it.. it could and was clearer and simpler, but now it isn’t with the latest firmware update(s) to the XP1 & XP2 in regards to the Corrected AF Frame.
.. and why it changes spacing when simply moving the 2 boxes from upper left (looser) to lower right (tighter) still fails to make sense. You are at the same distance and focal plane and not adjusting focus, only moving the point Corrected AF Frame around laterally and vertically, not in and out of focal distance. It should not be changing it’s spacing like this, only if you are focusing further out of in would that remotely make sense to change the 2 box spacing of the Corrected AF Frame.
No prob. On a side, but related subject regarding the X-Pro2. Do you still own or have one? If so, have you updated to the latest firmware released yesterday? v2.00?
Overall, I think there are a lot of improvements and I’d be interested in your opinion.
However…FOR ME, I did find a problem with the firmware update. This of course applies to OVF users, and if you’re using the X-Pro2 and own one, you’re almost certainly a big OVF user.. pretty much the main argument to get one. Ok, here goes, it’s kind of long, but I will try and explain the problem(s) below which I really think Fuji should re-examine and fix.
Let me quickly point out, also, I THINK the distance between the 2 boxes (Corrected AF) varies depending on where you move the AF point. Check it out.. if you move the Corrected AF point towards the upper left corner, it spaces out more, if you move the Corrected AF point towards the lower right corner, the 2 boxes get closer together (in some cases might even overlap). I’m pretty certain it never use to behave like this. Do you think it even makes sense that the 2 boxes should space out more if you’re just moving the Corrected AF point around the frame? It’s all lateral and vertical repositioning. You’re on the same plane and distance because you are not refocusing, just repositioning the Corrected AF point/frame(s). It’s not like you’re actually focusing in or out more.. if you were, then yes, that would make sense. But, just moving the Corrected AF around and the infinity and minimum focus distance spacing between each other changing does make any sense right now. If it should, please explain. Thanks.
I also see how the green AF confirmation box goes to the lower right and outside of the minimum focus distance box.
This doesn’t look right and is not intuitive. THEY DONE F*CKED UP THE CORRECTED AF! If you have the 2 boxes, the green AF confirmation box should never exceed these limits. If it’s out, then what’s the point of having them as a reference to gauge where the “true” AF point actually is?
Isn’t the “improvement” to the corrected AF supposed to make it more accurate? How can it be more accurate if you’re getting a green confirmation box outside/beyond of the minimal focus distance indicator?
It doesn’t seem to happen on the other end going beyond infinity mark (towards the upper left)
Still, you actually do not have to be THAT close to still get the green AF confirmation box outside the minimum focus distance. Even at moderate distances it’s still outside. What’s up with that?!
I can also confirm this same “unintuitive” effect where the green AF confirmation box falls beyond the minimum focus indicator/box (more to the lower right) on the 23mm f/2 as well as the 18mm f/2 (and yes, I did the update on the 18mm f/2). I did not try on my 56mm f/1/2, but I have no reason to think it will be any different.
I’m not sure if Fuji did the right thing here?
It REALLY DOES NOT MAKE SENSE to have the green AF confirmation box FALL OUTSIDE/BEYOND THE MINIMUM FOCUS DISTANCE BOX/INDICATOR OF THE CORRECTED AF FRAME.
I use MF most of the time as it’s actually easier than AF, but there are times where I actually would use AF and this does sort of mess up my flow and harder to guesstimate where the actual AF point is since the Corrected AF Frames (the 2 boxes) that would serve as reference for infinity and minimum focus distance are no longer accurate and less relevant.
Am I wrong and not using the Corrected AF Frame correctly?
I REALLY THINK FUJI NEEDS TO RECONSIDER AND RE-EXAMINE THE CORRECTED AF FRAME FUNCTIONALITY AGAIN TO BE SURE THIS IS THE BEST WAY TO USE IT WHEN USING AF.
Thanks for the feedback, good to know.
Thanks!
Great article
I thought I replied back about this, and so this may be late, but I did find some time to try with ND filters and CPL.. doesn’t really have any effect to reduce or eliminate the artifacts, fyi..
thank you for your reply! this is very sad my friend… there are too many fanboys and few photographers around! 🙁
Thank you… sad, very sad…
I can’t give you a 100% accurate answer, but somehow I doubt it. My X-Pro2 had it before, now my X-T2 does the same. It looks like a sensor hardware issue and my guess is that all cameras are affected but most of the time you won’t notice. Even I didn’t notice it at first until someone else mentioned it on Facebook.
Thanks Mathieu. Is it possible that one camera is more prone to this than another copy of the same camera? Asking since I’ve only found one other case of someone talking about artifacts in their images from their X-T2. I would assume this is across the board on all cameras considering what causes it, but I’m wondering your opinion. Adorama asks that I return it for an exchange but I’m thinking it won’t make a difference if a get a different camera, however I would appreciate your opinion. Thanks for this write up, it seems to be the only one on the entire internet right now lol.
Yes, it’s confirmed, there an example in my article now.
Yes, I updated the article with an X-T2 example at the end of the post.
Can you post an example? It’s easier to understand the issue.
If you shoot a lot in backlight situations I understand it can be a pain. With the speedlights one idea could be to to some tests at home to find out more precisely the angles where the problem doesn’t appear or is minimal so that you then know how to position your flash and at which angle to shoot. Granted during work you don’t always have the freedom or the time to to do exactly when you want.
Thanks for this article, especially since it seems to be the only article in the whole world that even mentions this problem. I can’t believe this isn’t talked about more. My problem is I love using speedlights to back light my subjects and I’m finding that it’s basically impossible to do this with the X-T2. I’m getting horrible purple flares and grid artifacts in almost all my shots. During sunsets and continuous backlight, I see how managing the flare is possible, I’ve done it myself. I had to compose a shot, see the flare, change the shot and try again. However, I use flash a lot. I can’t see the flash/flare before I fire an exposure so I can’t tilt my lens a different way. Any advice here? I loved my Fuji cameras since they never got in the way of my photography. Now things are different and I’m having to stop and worry about whether or not light is going to get into my sensor at a bad angle. Yet everyone is ok with this? I LOVE Fujifilm, but why is this problem not being acknowledged by anyone? I’m discouraged, depressed and a bit lost with my new camera. I’ll wait for everyone to tell me I’m an overreacting pixel peeper =
Unfortunately lol, I have too many. I am so damn dissapointed and people say I’m over reacting. I’m finding that the jpeg straight out of camera shows more purple flare than the raw, but the raw shows more grid artifacts. I’m beyond frustrated with this. I can’t believe hardly anyone talks about it and Fuji has never acknowledged it. sigh.
https://uploads.disquscdn.com/images/3070021edf5dfdc1754dd02bfa6174372ddb313f32a3404d72c1159369db3184.jpg
I also had an issue with x-pro2 flare shape when shooting with mechanical shutter over 1/250s, the light source become longer like a vertical lazer beam.
This issue is not show up if i shoot with electronic shutter or slower than 1/250s in mechanical shutter mode.
Anyone have the same issue too?
Hello, do you confirm the problem on xt2?
Thank you… very bad news! 🙁
Do you have samples?
Unfortunately yes, it has the same grid artifact and purple hue problem. I was pulling my hair out trying to figure out what was going on. Very discouraged, but I’m not giving back my X-T2
OK! thank you so much for your reply
I just got the camera yesterday. I’ll check it out in the follow ing days 😉
Hi Mathieu,
I dug up some more info and I have a post production solution too. Could you post an affected RAF file. Id like to test it on one of yours. They have more distinct patterns than mine. Your smugmug jpegs have another diagonal pattern on them though and its hard to see when the xtrans grid pattern is gone when that one is there too.
Have a look!
http://fujifilmstuff.blogspot.se/2016/09/purple-flares-and-grid-pattern-on.html
Have the brand new X-T2 the same problem?
I was in Central America this (hot) summer, no issues at all with my xpro2. I noticed that all purple effects appear above 4K shutter speed, which I barely use. Maybe that’s the reason
I had the same issue with a few images of my X-Pro 1.
Digilloyd is irrelivant. If you enjoy the camera it’s a mute point.
Seriously? Then go buy something else! The idiosyncratic sensor is what makes the Fuji! Why the heck would they move to a Bayer sensor?
Just to clarify one point….
Not all camera sensors have micro lenses. My Phase One most defibrillator does not.
I haven’t tried with a polarizer, I’ll check it out if I have the chance.
Yes the angle the light comes in makes the difference, that’s why varying the position sligthly gets rid of the veiling flare. Let me know your findings after your next shoot 😉
I would also note that the time of day was between 10:30am – 12:30pm. So, the angle of the sun was rather high by this time of day and so I actually did not get much flaring, and I believe most of the shots were basically taken with a bright background.. not all of them had this grid pattern, luckily, but there was still just enough to make we wonder what was going on.
I suppose it depends on the angle the light comes in, and not necessarily flare related.. perhaps if the background is bright enough or even some side or rim light off subject if strong enough might be enough to create this pattern… maybe.
I should try again without the ND filter… It’s a Hoya variable ND filter and I think for the money I spent on it, it should be a fair performer. I don’t recall having any issues using the ND filter before, but that was on my OM-D E-M1 and now I’m using the Fuji X-Pro2.. perhaps the combo variation might account for the different IQ as far as image artifacts are concerned.
I have another shoot coming up in a couple weeks and will review the results again, without an ND filter.. unless I have opportunity to test sooner.
Thanks.
I don’t think so at this point, it was more of a curiosity.
Yes, actually, I was only using the 56mm f1.2. Why? Does it make a difference?
Yes the trick is to avoid the purple flare, which in many situations will ruin your photo anyway regardless of the grid pattern being present or not.
Which lens were you using for your shoot? 56mm?
Mathieu, the night before i read the article , i lost out for a used x pro -2 on a ebay auction , ((£1000) by £20 , in the morning reading your comments i thought i’ve had a lucky escape , but as you said flare is always a problem , so the issue can arise in different cameras and lenses , as for lightroom causing artifacts , i use the in camera fuji raw converter for great results , then fine tune jpegs in dxo . lightroom is a pain to use , i’ll wait for the prices to drop when the x -T 2 arrives and see if the issue is acknowledged by fuji. yours norman
Mathieu, the night before i read the article , i lost out for a used x pro -2 on a ebay auction , ((£1000) by £20 , in the morning reading your comments i thought i’ve had a lucky escape , but as you said flare is always a problem , so the issue can arise in different cameras and lenses , as for lightroom causing artifacts , i use the in camera fuji raw converter for great results , then fine tune jpegs in dxo . lightroom is a pain to use , i’ll wait for the prices to drop when the x -T 2 arrives and see if the issue is acknowledged by fuji. yours norman
Thanks for this article.
I had a fairly recent photoshoot outdoors at the park and I tend to backlight my subjects with their backs towards the sun and I face them (& the sun) and bring their exposure back up with a flash.
I have to admit I think this was the first time I noticed this grid pattern you described here. I though I was imagining things.. perhaps it was the way smart preview looks like (I had disconnected my hard drive where the raw files are located so that Lightroom would work off of the smart preview only to speed its performance) and/or perhaps because I was using a variable ND filter.. I did not have HSS available with my X-Pro2 at that time and had to get my shutter speed down low enough to flash sync and be able to shoot wide open as well.
Having read your article, I believe what I was seeing in a number of the shots when magnified is the grid pattern.
Interesting, and I’d be lying if I didn’t say I wasn’t a little disappointed the X-Pro2 has such an issue… granted, as you have discovered in your own experience with many cameras, it’s not a problem limited to just the Fuji X cameras.
Reluctantly, I have to agree this looks like a hardware based issue, so a sensor update, if they are going to do one, is probably the most likely course to remedy this situation on the hardware level.. IF Fuji thinks it’s important enough to retrofit certain cameras (if it is only limited to certain production batches) with a new sensor and have customers send them back in for the fix.
At least I know what to look for and how to more carefully take my shots by the method you outlined. I have to admit the “slight” change in angle IS really slight.. I could hardly tell the difference.. which I suppose is a good thing as you do not have to radically rearrange your composition just to compensate for grid patterns.
I think for most viewing purposes, even if this grid pattern shows up, it will most likely be missed and not noticed unless zoomed up very close or you post or print at large resolution. Still, better to avoid this by exercising the techniques you showed than to have it at all, if possible.
I’m not saying this is necessary the silver lining in this story, but I have to say the grid pattern reminds me of old printed film photos..I don’t know if they still have this, but this grid pattern reminds me of the texture you see in vintage (talking like maybe 1960’s-1970’s) photos I have found going through old photo albums from my parents. The texture on the printed side is almost like a very fine canvas texture.. this is what the grid pattern reminds me of… I suppose if it does show up again, I could take it as an unexpected nod to looking like a vintage film photograph.
Overall, I’m very happy with the image quality I’m getting out of my X-Pro2. I rarely experience any IQ issues and I think most photographers using an X camera (or any other type that may exhibit such image artifacts) will get nice clean images on the whole where this issue will probably not be as big an issue as it could potentially get over-hyped and ramped up to be. I don’t think any X-Pro2 owners (or other camera owners who may get this grid pattern) have that much to worry about, but it’s good to know.
Thanks for bringing this issue up and offering a practical solution to resolve it. Looking forward to more news, reviews, tips & insights from you.
The X-Trans actually is a very nice filter array. If one gets used to a different workflow results will be better than bayer (I noticed that when getting into the system)
But yes, a X-T2B bayer version for all the whiners and dxo fanatics please 😉
When I use a Fuji camera on my own, I always find that the sensor gives great result. When I do in-depth comparison or analysis, I admit I find little advantages from the X-Trans array. Actually what I like the most about Fuji cameras is the color rendering with the film simulation modes but those can be replicate on a Bayer sensor like it happens with the X-A2. I guess the primary objective was to get rid of the AA filter 5 years ago but today many Bayer sensor are without low pass filters anyway.
It is disappointing that Fuji makes such nice cameras with such an idiosyncratic colour filter array. I think they would achieve much better mainstream success if they used a Bayer filter, or at least gave us an option. The image quality out of the X-A1 was much nicer than the X-M1 when using good Fuji lenses.
Nikon sells a D810A with different filter stacks for astrophotography. Can we can a Fuji X-T2B with a normal Bayer sensor?
Hi Greg, I noticed the purple flare with many micro four thirds cameras so I am not surprised if it happens with the E-M5 II as well. Most of the time, it is a matter of tilting up or down with the camera just a little to get rid of it.
Hi Mathieu, I recently upgraded from the EM5 to the EM5 mark 2 and I have been noticing the same purple flare on the new model which I didn’t notice on the original EM5. Some shots have been so bad I’ve had to convert to black & white. Really frustrating! I was wondering if the lack of an AA filter on the new model contributes to the problem? Just wondering if you or any of your readers have also experienced the same problem with the EM5 mark 2 ?
I think Diglloyd is one of the only tester to mention the problem. However personally I find it relevant with purple flare especially.
Otherwise I believe what Chambers refers to as fractal-like artefacts are related to the Camera Raw/Ligtroom software not be able to demoisaic the RAW files as well as other softwares like Iridient Developer. But in most situations, I don’t find it to be a concern.
Lloyd Chambers has taken a hard look at the X-Pro2. Nothing he found has dissuaded him from his earlier observation “The X-Trans sensor has its weird peculiarities: the fractal-like artifacts seen in the X-Pro1 an the X-E1 are likely to be there all the same in the X-Pro2, which is a sensor solution in search of a problem”. He confirmed that the problems are indeed still present. Not just where flare is present either. Please have a look at Diglloyd.com.