Blackmagic Forum

I noticed it before but totally ignored it, I thought maybe some sort of reflection. However, yesterday I did a shooting session and noticed the same issue exactly in the same spot.







You can download BRAW video here: https://we.tl/t-nG3rgo3oBa

Here is a video exported in Rec.709/Gamma 2.4



Update:

viewtopic.php?f=2&t=194564&p=1045169#p1045169

Contact BMD Support while your camera is under warranty. They’ll be more than happy to make things right before the internet thunderstorm that could arise overnight.

This may go away by itself but I can’t predict if it will be gone in a month or a year. In my opinion this is caused by some residue of a coating applied to the sensor cover glass or used in the manufacture of the actual sensor.

I don’t speak for BMD of course so I could be wrong. This is almost ’nothing’ compared to the cloudy magenta casts that afflicted some of the URSA Mini 4.6K cameras around 2016.

Many thank Rick. I’ll contact them right away.

It’s worth mentioning that there were no filters attached to the lens itself. It looks like it has to do with light reflection coming from the window.

From a different clip. Same spot below the color chart.

Light reflection?

Now I’m not so sure as the last screenshot with the colour chart clearly has a blue cast not magenta. I’d suggest the other shot is way too dirty to draw conclusions.

Clearest example that will illustrate a real problem would be to shoot a grey card (with even lighting that isn’t causing reflections) and the colour checker if possible where your only correction is colour balance of temperature and tint. Then pull up the waveform window and place it right below the frame. Perhaps boost the colour saturation. In a perfect balance, the waveform will only show a grey line across its frame. But if it is contaminated, you’ll see a small bump in the red and blue channels. Take a screenshot of the waveform and the video frame.

rick.lang wrote:Now I’m not so sure as the last screenshot with the colour chart clearly has a blue cast not magenta. I’d suggest the other shot is way too dirty to draw conclusions.

Clearest example that will illustrate a real problem would be to shoot a grey card (with even lighting that isn’t causing reflections) and the colour checker if possible where your only correction is colour balance of temperature and tint. Then pull up the waveform window and place it right below the frame. Perhaps boost the colour saturation. In a perfect balance, the waveform will only show a grey line across its frame. But if it is contaminated, you’ll see a small bump in the red and blue channels. Take a screenshot of the waveform and the video frame.

Very informative! Will do so. Thanks again Rick.

You might already have done this but have you checked this with a different lens to make sure it is a sensor issue? I had something similar some years ago shooting stills with a Nikon D800E and a Nikkor 50mm 1.4 lens - a magenta spot visible mainly when shooting in backlight in the centre of the field of view. At first I thought it was a problem with the sensor but using a different lens on the D800E and using the same lens on a different camera showed it was a problem with the lens. The lens was out of warranty by the time I discovered it so I never did anything about and never figured out exactly what the problem was.

Hi Mick. I use the same lens for stills, never had any issue with it. I shoot photos and videos under same conditions more or less.

I'm in contact with BMD's support. They asked me to replicate the same scene and then twist the lens without removing it, if the spot moves then it's a lens coating issue, if it remains in its place, then the camera's sensor is the problem. I did the test, the spot didn't move while twisting the lens. I'm sending them the videos now. Let's see what they say.

OK well that does indicate a sensor issue. Best of luck sorting it out.

As others have said because the spot is moving, it seems to point to some kind of optical interactions.

Your test shots are all backlit or have a very bright field in them and the lens maybe be doing that as an internal reflection. What lens are you using?

JB

John Brawley wrote:As others have said because the spot is moving, it seems to point to some kind of optical interactions.

Your test shots are all backlit or have a very bright field in them and the lens maybe be doing that as an internal reflection. What lens are you using?

JB

Hi John. The spot didn’t move when I twisted the lens. I’m using Sigma 12-24mm F4 DG HSM Art. If it was a lens issue, the same spot would appear in stills, right?

Here are some test videos.

https://we.tl/t-uSe56V5yg5

https://we.tl/t-vEuYYzwdZH

I’m using Sigma 12-24mm F4 DG HSM Art.

This lens is not made for L mount right? That would mean you use an adapter. Do you use an adapter when making stills? If so, is it the same adapter?

Bunk Timmer wrote:

I’m using Sigma 12-24mm F4 DG HSM Art.

This lens is not made for L mount right? That would mean you use an adapter. Do you use an adapter when making stills? If so, is it the same adapter?

It’s an EF mount, I use sigma mc-21 adapter for the BMCC6K. The DSLR camera is canon, no adapter is needed. Maybe I should have twisted the adapter with the lens attached. Maybe the adapter is causing the issue. Good point

I've absolutely had adapters cause annoying optical issues with lenses that work fine without. The Blackmagic F-mount for the Ursa Mini Pro has shiny black baffles that I had to put telescope flocking on because it was always causing weird flares.

Jon Hustead wrote:I've absolutely had adapters cause annoying optical issues with lenses that work fine without. The Blackmagic F-mount for the Ursa Mini Pro has shiny black baffles that I had to put telescope flocking on because it was always causing weird flares.

Even if the adapter has no optics?

Omar Mohammad wrote:

John Brawley wrote:As others have said because the spot is moving, it seems to point to some kind of optical interactions.

Your test shots are all backlit or have a very bright field in them and the lens maybe be doing that as an internal reflection. What lens are you using?

JB

Hi John. The spot didn’t move when I twisted the lens. I’m using Sigma 12-24mm F4 DG HSM Art. If it was a lens issue, the same spot would appear in stills, right?

Yes it can be something that’s interacting with the sensor stack. This is the sensor itself AND the IR (and sometimes OLPF) that’s in front of the sensor. There can be interactions with light and just the right angle that means you see an issue on specific model of camera but not others.

Think of a bright light reflecting off a mirror and then bouncing onto an another wall in your house. The sensor can reflect some light back into the cover glass internally and the sensor then can see this reflection. Of course they put anti reflective coatings on things but sometimes it can be the right combination to make this happen. I’m not saying this is what’s happening specifically with you, I’m just giving a general example. I haven’t looked at your work other than the two stills.

JB

John Brawley wrote:Think of a bright light reflecting off a mirror and then bouncing onto an another wall in your house. The sensor can reflect some light back into the cover glass internally and the sensor then can see this reflection. Of course they put anti reflective coatings on things but sometimes it can be the right combination to make this happen. I’m not saying this is what’s happening specifically with you, I’m just giving a general example. I haven’t looked at your work other than the two stills.

JB

Many thanks for the clarification. It makes sense that the sensor reflecting light as the adapter is just a spacer per se.

The problem appeared so far in two shooting sessions, so I have to replicate it in the same place and time. I did a shooting session today, but I didn’t notice any thing except for normal flare when sunlight hits the lens, other than that, the image is perfect.

It happened, or to be specific, I noticed it again in yesterday’s shooting session. You’ll notice it at 1:03 in the bathroom with sunlight coming through the small window.

Might be worth checking if it is affected by the aperture. I did a search to see if anyone ever reported problems with Nikon 50mm f1.4 lenses as I mentioned yesterday and couldn't find anything. However, there were reports of a blue/purple spot from a number of Nikon 50mm f1.8 D lenses some years ago that was only visible at apertures around or smaller than f11. At the time I had the problem lens, I was shooting mainly landscapes with that lens and often at f11. I had an older copy of the same lens which I started using on the saame camera and that never showed any problems. I guess it was an issue with the coating.

Good idea Mick. When I replicated the problem yesterday, I zoomed in/out, focused in/out too. As other mentioned, I should try twisting the adapter. I’ll check the aperture as well. Thank you!

Omar Mohammad wrote:… I did a shooting session today, but I didn’t notice any thing except for normal flare when sunlight hits the lens, other than that, the image is perfect…

This is good! I realize you still have a serious problem but those results at least seem to show the sensor is likely not the fault.

And you going to forego sending the camera to BMD for them to test?

rick.lang wrote:

Omar Mohammad wrote:… I did a shooting session today, but I didn’t notice any thing except for normal flare when sunlight hits the lens, other than that, the image is perfect…

This is good! I realize you still have a serious problem but those results at least seem to show the sensor is likely not the fault.

And you going to forego sending the camera to BMD for them to test?

I’d send it if I replicate the same problem at the same site following everyone’s suggestions.

This is today’s video. Handheld. Gyro stabilization works as expect, however, a gimbal would be helpful.

I have followed your advice using 4K DCI timeline/export resolution. The video might need some tweaks.

Very nice, thanks for sharing. Only thing I’d do differently is substitute an evening sunset Timelapse, rather than the day, and use it as, or very near, the end of the video.

Omar Mohammad wrote:

Jon Hustead wrote:I've absolutely had adapters cause annoying optical issues with lenses that work fine without. The Blackmagic F-mount for the Ursa Mini Pro has shiny black baffles that I had to put telescope flocking on because it was always causing weird flares.

Even if the adapter has no optics?

Yeah, it's bouncing off the adapter's baffles in the case of the Blackmagic F-Mount. Could be something about the sensor stack bounces light somewhere inside that's causing it too as JB mentioned.

In a similar vein, RED have for a long time had issues with interactions on their OLPF filters showing up as magenta / red dots in certain situations.

https://support.red.com/hc/en-us/articl ... id-Pattern

Again, not saying your issue is this but these kinds of things can’t be eliminated by comparing on another camera.

Adaptors, specific sensor design and internal design can affect this as well as the lens itself. You should try testing around the frame edge or edge boundary and see if you can figure out the conditions that it always can be made to happen. That will help in working out what’s going on.

JB

Rick, thanks to you for your kind words and advice. I did the sunset scene in the precios video. The apartment is next to the office so it was easier to go back and forth.

Jon, thank you for the clarification.

John, if you check the last video I’ve posted, at 1:12 you’ll notice the flare at the same spot more or less as in the photos in the very first post.

IR-cut filters are recommended OLPF sensors? Anyway, I don’t think I could fit one for the lens I use.

Omar Mohammad wrote:

John, if you check the last video I’ve posted, at 1:12 you’ll notice the flare at the same spot more or less as in the photos in the very first post.

This looks like a regular lens flare to me. Actually I think it’s the sun itself that is in shot, it’s just that image circle you’re recording doens't include the sun, but if we imagine the sensor was a little bigger, you’d maybe see the sun in this shot. You’re certainly seeing internal reflections of it on the lens. That’s why it’s a dot and not like a regular flare. I’m also guessing you’re shouting at quite a deep stop of 11 or F16?

I notice too that it moves relative to the little bit of camera wiggle you have in your camera tracking system. That also tells me it’s a flare. If it was a sensor issue it wouldn’t move with a camera bump in the tracking.

I haven’t picked up what lens this actually is, but I presume it’s on the quite wide side of things, and you aren’t using a mattebox or lens hood? This is a hot source (in this case the sun) near but just outside the frame boundary. What focal length is this? Looks like 12-15mm?


This is my guess.

Try shading the lens next time. You can do this with a mattebox or a flag on a stand off-camera. It can be hard on a moving shot though and when it’s wide you tend to see the flag as well, and the shadow it might cast on the ground.

Omar Mohammad wrote:
IR-cut filters are recommended OLPF sensors? Anyway, I don’t think I could fit one for the lens I use.

The IR OLPF recipe in the new 6K is a totally new filter and it’s a little more aggressive than with previous BMD cameras Is this what you’re asking?

JB

John, thank you for your detailed message.

Yes I agree it's normal lens flare. This shot was taken at 12mm f9.5. It might not look as wide because I zoomed in to fill black bars - 4K DCI timeline -, also Gyro stabilization zooms in.

When I was looking into BMD cameras, I came across some videos recommending IR-Cut filters but I just recalled it was meant for pocket 6K not the new one I have. Moreover, it is about colors contamination, so it is of no use to me.

I actually had similar thing happen with ALEXA recently, though scene is moody and darker so it doesn’t
Bother me much, probably the reason why I didn't noticed it at the first place.

You can see strong purple spot in the scene around 02:05 - 02:15 as well as 03:03 - 03:17 in the same spot but bit less pronounced cos of bluish background.
This was shot with UST projected background and mixed HSL lights with honeycomb filter from the both side…

I guess larger monitor on the set would eliminate this issues having to see better what you are getting in real time…

Canon's Speedbooster together with the C70 can cause a purple circle. But that adapter has glass in between. I am not sure if what you posted aren't just regular, purple lens flares that crept in, as happened in the last video you pointed to.

twist the lens without removing it, if the spot moves then it's a lens coating issue, if it remains in its place, then the camera's sensor is the problem. I did the test, the spot didn't move while twisting the lens.

Did you just twist or also wiggled it around a bit?

Michel, that was normal flare but it appeared at the same spot maybe because of the same angle of shooting.

Here are all the videos I sent to BMD’s support.

https://we.tl/t-vEuYYzwdZH

https://we.tl/t-hIjZndHBnl

https://we.tl/t-uSe56V5yg5

https://we.tl/t-nr1sQPYpDd

I will do more tests on Monday.

Yes, just a flare that went unnoticed, I actually had mattebox the most of the shoot, but when I changed from 35/2 lens to 24/2.8 I took matte-box off mainly cos of time constraint…not good idea…

Like it was already mentioned by John, this could have had been easily prevented and was nothing to do with camera…lesson learnt.

Darko Djerich wrote:I actually had similar thing happen with ALEXA recently, though scene is moody and darker so it doesn’t
Bother me much, probably the reason why I didn't noticed it at the first place.

You can see strong purple spot in the scene around 02:05 - 02:15 as well as 03:03 - 03:17 in the same spot but bit less pronounced cos of bluish background.
This was shot with UST projected background and mixed HSL lights with honeycomb filter from the both side…

I guess larger monitor on the set would eliminate this issues having to see better what you are getting in real time…

I noticed the spot, not obvious unless you focus well.

I did another test today, managed to capture the magenta spot on the wide angle lens. I tried 50mm lens, nothing out of the ordinary appeared, although I tried to let sunlight directly hit the lens.

What is the curvature like on the surface of the last lens in the rear?
I have observed quite a few such effects with relatively flat ones.

Uli Plank wrote:What is the curvature like on the surface of the last lens in the rear?
I have observed quite a few such effects with relatively flat ones.

It looks slightly concaved towards camera’s sensor. Really very little.

Th rear element moves back and forth when zooming.

Edit:

Verdict: BMD support has confirmed it is lens’s coating issue, as they tried replicating the same behavior using similar lens and adapter.

Update:

I have contacted Sigma Spain and provided them with videos, here’s their answer:

Good morning,
We do not believe that it is a defect in the coating of the lens, in fact from what you send us is not a problem of natural light but backlight, in backlight is easy to produce this type of effect.
If the coating was wrong, you could detect it by viewing the lens in backlight.
Best regards,

It reminds me of taking long exposure shots in DSLR, viewfinder should be covered to prevent light entering from there hitting lens’s rear element. If light enters, the image will be contaminated with similar effect. So as JB said, light is bouncing from camera’s sensor back to lens.

'Flare occurs more often with digital cameras than film cameras due to the use of Optical Low Pass Filters in front of the image sensor which can reflect light back into the lens optics before being finally recorded on the image sensor.'

One of my concerns with the OLPF.

Even without an OLPF you’d still have a more reflective surface than a film emulsion.

The spot is only visible on the gray card. When I move away from it the spot disappears.

Video and stills https://www.transfernow.net/dl/202405097939I5iH

Omar Mohammad wrote:The spot is only visible on the gray card. When I move away from it the spot disappears.

That’s an illusion. In the first shot the grey card is in the middle hence you can see the spot easy. In the second shot the grey card is no longer in the middle of your frame. If you look close, you can see the spot now on the wall slightly above the grey card.

Bunk’s got a keen eye, but it is there faintly. My guess is a single drop of whatever wash is applied to the sensor or sensor filter during manufacturing. It may go away by itself from evaporation (sublimation) but that will take a long time so I’d return the camera.

Bunk, it’s barely visible. I’ll increase magenta’s saturation to examine the difference. But still, why is it strong when the background is dark?

Rick, I’ll contact BMD support again and ask for a replacement.

Thank you for your input.

Omar Mohammad wrote:Bunk, it’s barely visible. I’ll increase magenta’s saturation to examine the difference. But still, why is it strong when the background is dark?

I think you can compare it with shining a weak flashlight on a white bright wall, in a well sun lit room and do the same with the curtains closed. I was wondering did you ever test it with a comparable L-mount lens? I still suspect the adapter.

Bunk Timmer wrote:

Omar Mohammad wrote:Bunk, it’s barely visible. I’ll increase magenta’s saturation to examine the difference. But still, why is it strong when the background is dark?

I think you can compare it with shining a weak flashlight on a white bright wall, in a well sun lit room and do the same with the curtains closed. I was wondering did you ever test it with a comparable L-mount lens? I still suspect the adapter.

Makes sense.

No, I haven’t got the chance to test a L-mount lens, both lenses I have are EF. I’ll try to replicate the same issue with the 50mm, if I get similar results, I’ll send it for revision.

Yes, I tried it with curtains: viewtopic.php?f=2&t=194564&p=1045419#p1011638

As far as I recall it happens when light is coming through glass. When window is open it doesn’t happen. I’m not sure though.

The adaptor is only a spacer per se. It has no optics, however, maybe the sensor is reflecting light to the lens. I use the same lens for photography with canon 6D-II, I’ve never had similar problem.

Thank you for your input, Bunk.

Omar Mohammad wrote:Yes, I tried it with curtains: viewtopic.php?f=2&t=194564&p=1045419#p1011638

Ok so that was not a good analogy.
What I meant is that a weak light builds up (flashlight in the analogy) somewhere in between your lens and the camera. When the light reaches a grey or dark area it becomes visible ( curtains closed in the analogy), when it falls on a bright part it will be hardly visible ( well lit room in the analogy).
It’s clear that the phenomenon only shows up when there is enough contrast. My house has two windows facing south. I have a couple of different mft adapters. A brand-less one with optics, some well-known adapters with optics and one without optics. Both the brand-less one and the one without optics show the phenomenon you mentioned, when I point the camera on the wall in between the windows (contrast). All well-known adapters survive the test (to a reasonable point).

[snip] maybe the sensor is reflecting light to the lens.

I think it’s the adapter.

I don’t remember trying it with a flash light, but I will do, with both lenses.

Both adapters with/out optics showed the same problem?

Thank you for the detailed explanation