URSA Mini 4.6k The Original™ Magenta Cast Issue

I made a comparative test of the Ursa Mini 4.6K versus the Cinema Camera 2.5K, both with the Canon EF 35mm f/1.4 L-series lens at f16. I did not use any filters.


Not only does the Ursa Mini 4.6K exhibit a magenta cast, but more problematic is that the cast is non-uniform and cannot be removed through standard tint correction in Resolve.


Here’s a still from the Ursa Mini 4.6K with the BMD 4.6K Film to REC709v3 LUT applied in Resolve 12.5 and adjustment made to the tint controls in the raw tab of Resolve to try and remove the magenta cast:

UM46_CanonL35mm_BMD_46KtoREC709v3-LUT.jpg

Copying the left side of the image over to the right side for comparison, you can clearly see that left side is balanced while the right side is magenta. The color cast is obviously not uniform and would require painstaking use of power windows to remove.

UM46_CanonL35mm_BMD_46KtoREC709v3-LUT-COMPARISON.jpg

Compare that to this image from the Blackmagic Cinema Camera 2.5K using the same lens and f-stop:

BMCC_CanonL35mm_f16.jpg

The BMCC is completely even all the way across the image.

If the cheaper, older BMD Cinema Camera can produce a balanced image from that lovely L-Series lens, my brand new much more expensive 4.6K Ursa Mini certainly should be able to the same.

Here are the original raw DNG files from both cameras to download and test:

https://drive.google.com/open?id=0B1bnfGXKHqpcWElqWjdHQWhqTHM

Jamie, did you do a test with your Canon 35mm 1.4 wide open or stopped down to no more than f2.8? Many people are saying it gets worse the more the lens is stopped down. This will be my next test to shoot at wide apertures and see if the cast decreases.

I'm past my return window. The issue in seeing is beyond any issues I've ever seen with a camera, but I seem to be free of an uneven cast. I just hope there is a fix, or I'm stuck with an expensive paper weight that can shoot some pretty pictures... When it feels like it.


Sent from my iPhone using Tapatalk

www.ryanhamblin.com
www.brainstem.tv

I've had my UM4.6k PL a week now.

Did a big shoot using B4 mount with Fujinon HA17x7.8 lens on Wednesday. All good, no problems with magenta corners or cast.

I've also done a couple of tests at home using GL Optics 11-16mm and CineAlta primes. Also no magenta corners.

Here's a couple of DNG's shot open gate with no lens. Camera pointing out a shaded window during day, set to 5600K, 11.25 degree shutter angle, and not over-exposed.

1 - https://www.dropbox.com/s/7sd4j7bwbu72k ... 3.dng?dl=0
2 - https://www.dropbox.com/s/qbrp7wdoxaa7p ... 2.dng?dl=0

As you can see - if you push the file to extremes - there's very slight shading, but the colour is even. The first DNG is pointing horizontal out the window. The second DNG is tilting the camera up 45 degrees.

I'm very happy with the camera's performance. Image and sounds captured are great.

Please can someone send ungraded files with magenta problems to check in my pc?

Now BMD says, that the UM 4.6K is working as it was intended, and the magenta is the look of the camera
..hmm, if there is a camera out there, that does not show a magenta cast, BMD has to say, that that one is defective...

I've had my UM4.6k PL a week now.

Did a big shoot using B4 mount with Fujinon HA17x7.8 lens on Wednesday. All good, no problems with magenta corners or cast.

Hi Tristan can you share with us dngs shoot using the b4 mount and lens, it will be very helpful for me because i want to use it as eng camera and in live events. Thanks a lot

Jamie, did you do a test with your Canon 35mm 1.4 wide open or stopped down to no more than f2.8? Many people are saying it gets worse the more the lens is stopped down. This will be my next test to shoot at wide apertures and see if the cast decreases.

Hi Benton. You're a hero for following up on all this.
As my test was in daylight and I didn't want to introduce any variables beyond the sensor and lens (so no ND filters) I could only go down to f11 even with the shutter at 11.5 degrees and not clip the clouds. I saw no difference in the uneven magenta cast between f16 and f11. That's enough to know the camera isn't functioning properly. Whether it's less at 2.8 is a moot point for me. It should work properly at all apertures.

Also, I think there's something we're all forgetting here while getting too caught up in details. The main issue is that there is magenta cast across the image that simply shouldn't be there, regardless of whether it is even or uneven or how lenses respond to it. The 4.6K should be able to deliver a balanced image simply by setting the proper temperature in camera.
None of the raw sample footage released by BMD had a magenta cast. I paid for my camera based on the dynamic range and great color balance I saw in those raw frames in Resolve (as I'm sure you did as well). The sensors in all the 4.6K cameras BMD sells should deliver the same balanced color exhibited in their sample raw footage. Any that don't should clearly be returned.

....Spot on! I can hardly understand, why people would like to spend 5000$ to become beta-testers...

What's your read on the magenta vignette cause?

Simply Put - That it shouldn't be there !

I am not in the business of producing cameras, so I would be guessing...

Given the loud silence of BMD on this issue - I take that they too are in the guessing contest.
So if those who designed the camera have no clue as to why things are as they are, my guessing wouldn't count

Check this video from 9m50s

Jamie, did you do a test with your Canon 35mm 1.4 wide open or stopped down to no more than f2.8? Many people are saying it gets worse the more the lens is stopped down. This will be my next test to shoot at wide apertures and see if the cast decreases.

Hi Benton. You're a hero for following up on all this.
As my test was in daylight and I didn't want to introduce any variables beyond the sensor and lens (so no ND filters) I could only go down to f11 even with the shutter at 11.5 degrees and not clip the clouds. I saw no difference in the uneven magenta cast between f16 and f11. That's enough to know the camera isn't functioning properly. Whether it's less at 2.8 is a moot point for me. It should work properly at all apertures.

Thanks for your kind words Jamie, I'm just turning over all the stones. I want a clean working camera too.

I agree that the camera should and ultimately needs to deliver a clean image at all apertures, but if the cast shows up less with wider apertures, this I believe would tell us that we're dealing with a cast that's caused by the angle of light striking the sensor. When a smaller aperture is set, the light emanates more from a single point within the lens and the light rays are directed toward the sensor at more of an angle than they would if the lens was wide open. In telecentric designs, the light is redirected back in more or less parallel rays as it strikes the sensor thus avoiding the ray angle problem.

I believe the problem will be shown that it is lens cast as BMD says, but this reason alone doesn't give them a free pass to say it's normal, because ALL manufacturers have to deal with this problem and usually work it out with sensor calibration and sensor shading. Also the fact that so far most of the heavier cast is showing up on the right side to me also means something it not right with the calibration or something else is affecting the sensor. If it was just lens cast alone, we should be seeing casting on both sides of the sensor equally, but we don't.

Not sure if anyone has asked the question, but is this something that only affects the UM4.6K EF model? Or are both PL & EF models affected?

I think someone recently posted a problem with their Mini 4.6K PL camera. May have been Jamie.


Sent from my iPad using Tapatalk

Below is a review/test of Ursa mini 4.6k EF for magenta issues. Rick did you test yours?

http://cinescopophilia.com/yet-another- ... ssue-test/

Daniel peters 4.6k review tips and trick ..plus free luts for the 4.6k

https://danieljohnpeters.com/2016/03/18/3731/

What's your read on the magenta vignette cause?

Simply Put - That it shouldn't be there !

I am not in the business of producing cameras, so I would be guessing...

Given the loud silence of BMD on this issue - I take that they too are in the guessing contest.
So if those who designed the camera have no clue as to why things are as they are, my guessing wouldn't count

...guessing too: they already know the source of the magenta cast, but are not able to deliver flawless cameras (maybe sensors) yet ...

Donnell, I'm still waiting for the Titla LS-T03 lens support. Put a Canadian order in last week but if it takes too long to ship, then on Monday I'll quickly order the Tilta LS-T05 from B&H Photo (free shipping to Canada) since they have it in stock. The LS-T05 cradle is better quality but it has such a short adjustment (0.875") I'd dedicate it to the Fujinon Cine zoom with an outside diameter of 110mm. For the SLR Magic APO primes with a 95mm outside diameter, I'll need the longer vertical adjustment that the LS-T03 provides.


Sent from my iPad using Tapatalk

Oh wow, ok gotcha Rick.It must be torture having this camera and waiting to get all that stuff together to finally shoot.

I think someone recently posted a problem with their Mini 4.6K PL camera. May have been Jamie.


Sent from my iPad using Tapatalk

Mine is EF.
But you may be on to something Rick. PL versus EF could be a relevant difference. All of the BMD sample footage is from a PL 4.6k. All of JB cameras are PL mount. Also, Marco Solorio has reported that his cameras is flawless, it's also PL mount.

I'm very interested to know what the image from yours looks like.

I think someone recently posted a problem with their Mini 4.6K PL camera. May have been Jamie.


Sent from my iPad using Tapatalk

Mine is EF.
But you may be on to something Rick. PL versus EF could be a relevant difference. All of the BMD sample footage is from a PL 4.6k. All of JB cameras are PL mount. Also, Marco Solorio has reported that his cameras is flawless, it's also PL mount.

I'm very interested to know what the image from yours looks like.

I asked the question as mine's a PL and is not exhibiting any problems. My DNG's are posted earlier in the thread.

I think someone recently posted a problem with their Mini 4.6K PL camera. May have been Jamie.


Sent from my iPad using Tapatalk

Mine is EF.
But you may be on to something Rick. PL versus EF could be a relevant difference. All of the BMD sample footage is from a PL 4.6k. All of JB cameras are PL mount. Also, Marco Solorio has reported that his cameras is flawless, it's also PL mount.

I'm very interested to know what the image from yours looks like.

I asked the question as mine's a PL and is not exhibiting any problems. My DNG's are posted earlier in the thread.

Please excuse my mistake Tristan. That's what I get for skimming the thread. I'm glad to see your camera is working well. So, we have a number of confirmed PL mount 4.6K cameras that work well.

Does anyone out there have a magenta free EF mount 4.6K and some DNGs to share?

I really dont think there are any free Magenta EF cameras in existence sadly enough

Is this a case of an incorrect flange focal distance on the EF mount, come to think... the magenta reminds me of the lens whacking technique, as if light is passing between the lens and the mount...

Is this a case of an incorrect flange focal distance on the EF mount, come to think... the magenta reminds me of the lens whacking technique, as if light is passing between the lens and the mount...

I said that previously. Maybe you guys should try putting some tape around the lens/camera contact, see what happens ?

arri Alexa vs Ursa mini 4.6k review and tests

https://m.youtube.com/watch?v=Rk7ZFx7j8f8

arri Alexa vs Ursa mini 4.6k review and tests

https://m.youtube.com/watch?v=Rk7ZFx7j8f8

Seems they got an EF mount 4.6K with a good sensor. Interesting that at 14:55 the DP warns prospective buyers about QC issues and recounts experiences of having to send cameras back multiple times before getting a working unit.

Unfortunately I've tested a PL version and it did have the same magenta corners. I have uploaded photos here in this thread, several days ago.

I think someone recently posted a problem with their Mini 4.6K PL camera. May have been Jamie.


Sent from my iPad using Tapatalk

Mine is EF.
But you may be on to something Rick. PL versus EF could be a relevant difference. All of the BMD sample footage is from a PL 4.6k. All of JB cameras are PL mount. Also, Marco Solorio has reported that his cameras is flawless, it's also PL mount.

Is someone with a magenta corner camera brave enough to experiment with the light baffle in the mount?
If you screw it out, does the corners improve? How far can it get screwed in? Maybe it is in the wrong position and casting the magenta?

Is someone with a magenta corner camera brave enough to experiment with the light baffle in the mount?
If you screw it out, does the corners improve? How far can it get screwed in? Maybe it is in the wrong position and casting the magenta?

It's easy to remove. Can be unscrewed with your fingers. Removed mine before. It's silver behind the baffled tho. So will probably cause internal reflections.

I passed my 30 days and according to BMD my camera is Normal even though it exhibits Magenta corners... What do I do now?
I guess I am doomed to be shooting windowed sensor at 4k DCI raw, Anamorphic mode when that comes out, and Ultra HD in prores version since I can't produce a clean image with the lenses I have in full frame raw.

That really blows... I dropped about 10k on this package
thanks so much.

Eli, you might also get decent results with the raw 2.4:1 4608x1920 window when that is enabled in June. That's my hope! If corners are the only problem, may be tolerable. But if the entire right side is magenta as I've frequently seen on other cameras, then the 4096 and 3072 frames will have to be used. Arrrgghhh! When the problem is solved, BMD will probably do the right thing and issue a recall or provide a mechanism to correct existing flawed cameras. That will be outside any normal period to return a camera for a refund. Eventually I would expect any flawed camera to be repaired, not refunded.


Sent from my iPad using Tapatalk

I passed my 30 days and according to BMD my camera is Normal even though it exhibits Magenta corners... What do I do now?
I guess I am doomed to be shooting windowed sensor at 4k DCI raw, Anamorphic mode when that comes out, and Ultra HD in prores version since I can't produce a clean image with the lenses I have in full frame raw.

That really blows... I dropped about 10k on this package
thanks so much.

Eli, under the warranty, you have the right to get a clean working camera. As long as good cameras exist that show no magenta corners using the same lenses that have been shown to produce it, BMD needs to replace your camera. The problem right now is they don't fully know what the cause is, so it's a matter of testing a replacement camera before they ship it out to you. Just make sure that they test it using a lens. The open sensor test reveals nothing if the camera will produce magenta corners.

Unfortunately I've tested a PL version and it did have the same magenta corners. I have uploaded photos here in this thread, several days ago.

I think someone recently posted a problem with their Mini 4.6K PL camera. May have been Jamie.


Sent from my iPad using Tapatalk

Mine is EF.
But you may be on to something Rick. PL versus EF could be a relevant difference. All of the BMD sample footage is from a PL 4.6k. All of JB cameras are PL mount. Also, Marco Solorio has reported that his cameras is flawless, it's also PL mount.

Emilian, apologies to you. It was your post to which I was referring when I guessed it was Jamie that had poor results with a PL camera.

arri Alexa vs Ursa mini 4.6k review and tests

https://m.youtube.com/watch?v=Rk7ZFx7j8f8

That camera looks outstanding. So obviously there are good EF cameras out there.
I find this whole issue extremely confusing. There seems to be lots of agendas going around here. And overly simplify general statement that are just not proven true, just being stated for posturing or just frustration.
I won't judge how effective all of that maneuvering actually is. I will say that it is definitely confusing the heck out of me. We have budgeted for six of these cameras, four for studio use, and I just unable to make a decision about it.

arri Alexa vs Ursa mini 4.6k review and tests

https://m.youtube.com/watch?v=Rk7ZFx7j8f8

That camera looks outstanding. So obviously there are good EF cameras out there.
I find this whole issue extremely confusing. There seems to be lots of agendas going around here. And overly simplify general statement that are just not proven true, just being stated for posturing or just frustration.
I won't judge how effective all of that maneuvering actually is. I will say that it is definitely confusing the heck out of me. We have budgeted for six of these cameras, four for studio use, and I just unable to make a decision about it.

Similar situation with me. I have booked a Raven, still wanted to pick up a UM46. I can't take a decision as BMD RMA procedures are convoluted here in India.

arri Alexa vs Ursa mini 4.6k review and tests

https://m.youtube.com/watch?v=Rk7ZFx7j8f8

That camera looks outstanding. So obviously there are good EF cameras out there.
I find this whole issue extremely confusing. There seems to be lots of agendas going around here. And overly simplify general statement that are just not proven true, just being stated for posturing or just frustration.
I won't judge how effective all of that maneuvering actually is. I will say that it is definitely confusing the heck out of me. We have budgeted for six of these cameras, four for studio use, and I just unable to make a decision about it.

Carlos, There are many theories put forth here as to what may be causing the magenta corners issue. Some or all of them could be completely wrong, but that's really beside the point. The fact remains that at least certain copies of this camera have been shown to have a magenta corners issue that goes far beyond nitpicky pixel peeping and essentially ruins a scene that is struck with it (you can see many examples of this in this forum). A fix in post has also never been shown to exist.

I'm unaware of any agendas (at least from the forum side) other than trying to get clean working cameras that are free from the magenta corners issue. But as a potential buyer you should at least be aware that the issue is real and you should proceed accordingly to make sure you get clean working cameras. I completely agree that the image from this camera is outstanding, most of the time. I wouldn't be waisting my time on all this magenta mess if it wasn't.

So how do the return policies work?

Let's consider a retailer like B&H, can you keep returning defective units till you get a normally functioning camera within 30 days? Does that include (if you are not based close to the NY Superstore) shipping time to send the cameras back, or it is not inclusive of shipping times?

Or, do the policies change from individual re-seller to re-seller?

Or, the returns polices decided by the manufacturer for their respective product?

How can i get an RMA if according to BMD my camera exhibits normal Behavior that is not out of the ordinary?

... BMD RMA procedures are convoluted here in India.

Similar situation in Canada I think since I'd have to deal with shipping to California. International shipments are a pain when you are shipping your own goods for service and will have your own goods returned. I've done it and there are custom declaration forms that handle it, but can be a hassle. I've no idea how customs would react if I shipped an item with a different serial number than the machine that was shipped back to me in the event it is replaced instead of repaired.

Hoping I never have to do that of course. Tristan's experience gives me hope, but Emilian's experience is a sobering reminder that it's a crap shoot at this point.

Very sad that another amazingly great camera gets off to a late and rocky start that sullies BMD's reputation. In time they'll all be made good. There really are only two options for BMD: make everything right when you can or get out of the camera line of business. My bet is they only see it as one option: make everything right.

Is someone with a magenta corner camera brave enough to experiment with the light baffle in the mount?
If you screw it out, does the corners improve? How far can it get screwed in? Maybe it is in the wrong position and casting the magenta?

It's easy to remove. Can be unscrewed with your fingers. Removed mine before. It's silver behind the baffled tho. So will probably cause internal reflections.

It might cause some reflections but otherwise it would be interesting to examine if the baffle has any influence on the magenta corners so we can systematically remove on part of the equation.

How can i get an RMA if according to BMD my camera exhibits normal Behavior that is not out of the ordinary?

You as the customer needs to be satisfied and only you are the judge of that, not BMD. Just because they say this is "normal" behavior, doesn't make it so. Large sensor cameras have been successfully dealing with lens cast for many years now (if that is the problem as they state). Just because their new sensor doesn't handle lens cast well, doesn't make it OK or normal. They need to fix the problem if they want this camera to be successful in the marketplace.

I just emailed them asking to RMA this unit for a replacement. I have bought and shot successfully with all of their line up cams: 2.5k, pocket, production. I sure hope they will do the right thing by my and US as loyal users and replace/fix this problem.
10k is a lot of money for me to spend on a camera that I can't use.

Hoping I never have to do that of course. Tristan's experience gives me hope, but Emilian's experience is a sobering reminder that it's a crap shoot at this point.

Very sad that another amazingly great camera gets off to a late and rocky start that sullies BMD's reputation. In time they'll all be made good. There really are only two options for BMD: make everything right when you can or get out of the camera line of business. My bet is they only see it as one option: make everything right.

Rick, I have had my experience with BM4k. Eventually I got a good camera but my experience did leave a bitter taste in my mouth. BMD is great in post production and I am amazed how they churn out amazing stuffs. Last year they introduced Resolve 12 which was top notch and this year they are unbeatable with their Resolve 12.5.
And after one year of announcing UM46 we are still struggling to decide whether to get this camera or not.

They should get out of this camera business. Or acquire a camera company and start building up on that (like they did with Resolve and Fusion) instead of starting from scratch.

I just emailed them asking to RMA this unit for a replacement. I have bought and shot successfully with all of their line up cams: 2.5k, pocket, production. I sure hope they will do the right thing by my and US as loyal users and replace/fix this problem.
10k is a lot of money for me to spend on a camera that I can't use.

The camera is under warranty and its defective they should returned, talk to your dealer

Is someone with a magenta corner camera brave enough to experiment with the light baffle in the mount?
If you screw it out, does the corners improve? How far can it get screwed in? Maybe it is in the wrong position and casting the magenta?

It's easy to remove. Can be unscrewed with your fingers. Removed mine before. It's silver behind the baffled tho. So will probably cause internal reflections.

It might cause some reflections but otherwise it would be interesting to examine if the baffle has any influence on the magenta corners so we can systematically remove on part of the equation.

Assuming everything about the physical body of the 4K and 4.6K Ursa Mini is identical and the only difference is the actual sensor itself, and that both sensors are the same size since they're both S35, the issue is actually the sensor and not anything to do with the lens mount or baffle. If it was that, you'd see magenta corners on 4K models as well. but since it's only happening with 4.6K EF and PL models, then it's the sensor and not the mount, baffle, or anything else inside the camera.

Mike, the sensor in the BMPC4K and the original URSA and the URSA Mini 4K camera has a resolution of 4000x2160 and is very close in size to what is called the Academy 35 size. Any early references on BMD's webpages about that sensor being Super 35 was an exaggeration at best by BMD marketing. Several manufactures loosely throw around the Super 35 term including BMD when it is in reality smaller than Canon APS-C! If memory serves me correctly, BMD now refers to that sensor as 35mm and that's fine.

The 4.6K sensor is 4608x2592 and even slightly exceeds the Super 35 film dimensions but is a true Super 35 sensor. So it is possible that part of the cause of the dark magenta corners is due to or exacerbated by the increased size of the 4.6K sensor. But there are lovely flawless 4.6K sensors in the wild so we know that size is not the sole cause.


Sent from my iPad using Tapatalk

Mike, the sensor in the BMPC4K and the original URSA and the URSA Mini 4K camera has a resolution of 4000x2160 and is very close in size to what is called the Academy 35 size. Any early references on BMD's webpages about that sensor being Super 35 was an exaggeration at best by BMD marketing. Several manufactures loosely throw around the Super 35 term including BMD when it is in reality smaller than Canon APS-C! If memory serves me correctly, BMD now refers to that sensor as 35mm and that's fine.

The 4.6K sensor is 4608x2592 and even slightly exceeds the Super 35 film dimensions but is a true Super 35 sensor. So it is possible that part of the cause of the dark magenta corners is due to or exacerbated by the increased size of the 4.6K sensor. But there are lovely flawless 4.K sensors in the wild so we know that size is not the sole cause.


Sent from my iPad using Tapatalk

You can see in many of the magenta corner examples that the uneven magenta cast coming in from the corners and sides exceeds the roughly 10% larger size of the 4.6K. They may not be exactly the same size, but you'll still get "magenta corners" on some of these 4.6K sensors when windowed to 4K (which would be greater than a 10% difference too). That means a 4K sensor would exhibit the same issues if it's a design issue with the camera body.

All I'm saying is that it's not a physical issue with the camera. Maybe the vignetting is, but it's already been demonstrated that some other cameras can also vignette in the corners slightly, even without a lens. And the slight vignetting is generally not even noticeable. The baffle is not going to cause a magenta cast unless it's reflecting something (it should not be reflective) or is somehow breaking up the color wavelengths in the light (unlikely). If it was the baffle, the issue would be even and uniform all the way around, but it's not. Unless the baffle isn't installed straight, which is highly unlikely.

Therefore, it's purely a sensor issue, and probably something to do with coating as was suggested previously. BMD has had issues with their sensors in the past. Everyone knows that. This shouldn't be a surprise... heck, with BMD it's almost expected at this point. That's their reputation now.

Also, resolution has nothing to do with sensor size. You need to look at the actual dimensions of the sensor, not it's resolution.

... Or acquire a camera company and start building up on that (like they did with Resolve and Fusion) instead of starting from scratch.

Yes, interesting idea. Perhaps acquire Sigma? Not such a great body company, but they appear to know how to make lenses. Maybe Blackmagic could do something with the foveon sensor?

Mike, the two sensors under discussion (4K and 4.6K) share the same size of photosites (5.5 microns), so their respective resolutions define their sensor size open gate.


Sent from my iPad using Tapatalk

Mike, the two sensors under discussion (4K and 4.6K) share the same size of photosites (5.5 microns), so their respective resolutions define their sensor size open gate.


Sent from my iPad using Tapatalk

That doesn't change anything. The magenta corners still exceeds further into the middle of the sensor than a 4K window in some tests. Yes, the magenta corners issue isn't as apparent, but it is still there, and has not been reported on the 4K sensor. They are different sensor designs. So still, like I said before, it's not an issue with the design of the camera. It's a sensor issue, whether in the design of the sensor or an issue with manufacturing that can maybe be fixed (such as coating). Coating is more likely the cause because it would explain the variation of the issue across cameras. This bodes well, as it makes it possible for there to be good cameras out there (which apparently is the case) now and a possible permanent fix once they figure out who to be consistent with the coating application.

That doesn't change anything. The magenta corners still exceeds further into the middle of the sensor than a 4K window in some tests. Yes, the magenta corners issue isn't as apparent, but it is still there, and has not been reported on the 4K sensor. They are different sensor designs. So still, like I said before, it's not an issue with the design of the camera. It's a sensor issue, whether in the design of the sensor or an issue with manufacturing that can maybe be fixed (such as coating). Coating is more likely the cause because it would explain the variation of the issue across cameras. This bodes well, as it makes it possible for there to be good cameras out there (which apparently is the case) now and a possible permanent fix once they figure out who to be consistent with the coating application.

They can add to the production line of the sensors QC test for the magenta issue but it will add up on the camera price, this is how it work