BM 6K FF FPN/banding

Just did some SUPER low light testing at 3200 and banding is definitely bad. But it’s not a scenario I think I’d ever use this particular camera. The original Canon R6 had similar banding in video mode and C-log 3.

That was my observation too with the Canon.
If you need to shoot in the dark, get a Sony A7S series. It can see colours in the shadow under moonlight!

Just did some SUPER low light testing at 3200 and banding is definitely bad. But it’s not a scenario I think I’d ever use this particular camera. The original Canon R6 had similar banding in video mode and C-log 3.

Would you mind posting your test? 3200 is bad but it appears in lower ISOs in my tests.

Just did some SUPER low light testing at 3200 and banding is definitely bad. But it’s not a scenario I think I’d ever use this particular camera. The original Canon R6 had similar banding in video mode and C-log 3.

Would you mind posting your test? 3200 is bad but it appears in lower ISOs in my tests.

Will do, in the next day or so.

EDIT: All my footage features my kids and house, so I'm opting not to post. But see my post further down the thread which essentially sums up that I think banding is not a real-world problem for me with this camera.

I'm also experiencing some bad FPN problemas with the new 6k ff sensor, i also sent an email to BMD support yesterday

I've compared with my 6k pro (which I also bought on it's release) and despite the fact that the 6k pro has som problems, it's a lot worse on the new 6k ff.

here is some footage of it (a quick video, some prints and also the braw file):

https://drive.google.com/drive/folders/1iTEhdyV8pxMIEoSRuv-ed-hsyH__lTXe?usp=sharing

Striping aside, the noise is more prominent in the 6K Pro picture above.

Striping aside, the noise is more prominent in the 6K Pro picture above.

Noise is not the problem, the FPN striping is a LOT worse on my 6k ff

I just shot this sequence, first frame looks normal, but the others are a complete mess

Quick video on this

In motion it's pretty bad

I shot another test with the same lens, same place, on the 6k pro, while it does have some FPN striping, on the 6k FF seems to be worse.

I'm pushing the image quite a bit, but this has never been a problem on the 6k pro, while on the new 6k ff, it seems to happen more and also more visible

Also, if this is normal for this sensor (still waiting on a response from BMD) then this sensor is horrible for shadow quality, you will need to overexpose it for sure :/

Bruno, what was your ISO in that above dimly lit scene?

Noise is not the problem, the FPN striping is a LOT worse on my 6k ff

I just shot this sequence, first frame looks normal, but the others are a complete mess

Quick video on this

In motion it's pretty bad

Ouch.

I'm also experiencing some bad FPN problemas with the new 6k ff sensor, i also sent an email to BMD support yesterday

I've compared with my 6k pro (which I also bought on it's release) and despite the fact that the 6k pro has som problems, it's a lot worse on the new 6k ff.

here is some footage of it (a quick video, some prints and also the braw file):

https://drive.google.com/drive/folders/1iTEhdyV8pxMIEoSRuv-ed-hsyH__lTXe?usp=sharing

Most likely CMOS smearing. If you wave your hand around in front of the lens on the side of the light source the banding will move/disappear. I contacted BM and there response is that it's normal LOL .. That being said this camera produces the nicest images south of the Alexa with enough light. I own an fx6, so I have to decide if this camera's numerous limitations (poor form factor, lack of NDs, bad low light perfomance, etc.) are worth overlooking for its stellar image and codec.

Happen on mine too. If this is the same family of sensor as the Pocket 6k as I've been told on another thread then its obviousely doing something differently on this camera or there is a problem

Bruno, what was your ISO in that above dimly lit scene?

they were at 800 ISO

Okay, some good news (subjectively speaking) for anyone reading this thread and worried about banding on the BMCC6k...

After finally loading a bunch of clips from the last couple weeks into Resolve, I've come to the conclusion that you really have to try to get the sensor to show extreme banding, and those cases would not be routinely replicated in actual regular production work. In my experience so far, banding looks worse on the back of the camera than it does once loaded into Resolve...and it even cleans up a small bit with some chroma noise reduction. I'm not saying this is a perfect sensor...it DOES display banding if you try, and it's possible it may show up in certain rare production circumstances (low-light, poorly-exposed, high-contrast scene with backlit subject), but I wouldn't be as concerned about it as this thread may lead you to believe.

Overall, I'm pretty blown away by the look of this camera. It checks a lot of boxes for me (phenomenal image, large sensor, two xlr inputs, OLPF, open gate, and more). I have loved my three BMPCC4k bodies for the last four years (two of them with Metabones 0.64x speedboosters and Rawlite OLPF, one with MFT lenses), but this BMCC6k is next-level. I really like having my full frame lenses fully utilized again and no extra glass between them and the sensor. I'm VERY impressed with the noise and grain characteristics, even at high ISO settings.

If you're on the fence about this camera because of banding or noise, don't be. If you're on the fence because of framerates, form factor, or any other logistical reason, that's something ultimately only you can decide. But no matter how you cut it, the BMCC6k is an outstanding cinema camera with amazing image quality.

Bruno, what was your ISO in that above dimly lit scene?

they were at 800 ISO

I find 800 noisy. I know this is just a test, but move up to 1250 or down to 400.

I agree it will make a difference.

At one time in the early days of the camera, I used to shoot the BMPCC4K at ISO 3200 but I’ve settled on ISO 1600. 1250 will likely produce the best looking image in the upper range, comparable to ISO 400. Would love to see the BMCC6K for myself, open gate.

You can check how much noise each ISO has by enabling focus peaking and dial up the intensity to say about 80%

Okay, some good news (subjectively speaking) for anyone reading this thread and worried about banding on the BMCC6k...

After finally loading a bunch of clips from the last couple weeks into Resolve, I've come to the conclusion that you really have to try to get the sensor to show extreme banding, and those cases would not be routinely replicated in actual regular production work. In my experience so far, banding looks worse on the back of the camera than it does once loaded into Resolve...and it even cleans up a small bit with some chroma noise reduction. I'm not saying this is a perfect sensor...it DOES display banding if you try, and it's possible it may show up in certain rare production circumstances (low-light, poorly-exposed, high-contrast scene with backlit subject), but I wouldn't be as concerned about it as this thread may lead you to believe.

Overall, I'm pretty blown away by the look of this camera. It checks a lot of boxes for me (phenomenal image, large sensor, two xlr inputs, OLPF, open gate, and more). I have loved my three BMPCC4k bodies for the last four years (two of them with Metabones 0.64x speedboosters and Rawlite OLPF, one with MFT lenses), but this BMCC6k is next-level. I really like having my full frame lenses fully utilized again and no extra glass between them and the sensor. I'm VERY impressed with the noise and grain characteristics, even at high ISO settings.

If you're on the fence about this camera because of banding or noise, don't be. If you're on the fence because of framerates, form factor, or any other logistical reason, that's something ultimately only you can decide. But no matter how you cut it, the BMCC6k is an outstanding cinema camera with amazing image quality.

I agree with you, because with those tests i found out that my 6k pro also shows this type of problem and i've never noticed (in over 2 years using it) so it's not a big problem if you are using this camera with proper light and you know what you are doing.

Still, this new 6k ff sensor is definitely worse in this regard when compared to the 6k pro.

Bruno, what was your ISO in that above dimly lit scene?

they were at 800 ISO

I find 800 noisy. I know this is just a test, but move up to 1250 or down to 400.

moving it down to 400 would only make a difference if I changed other exposure values, if i would only change ISO, then it would make no difference.

But yeah, changing it to 1250 will help on this for sure, but that's not what this test was about

I agree it will make a difference.

At one time in the early days of the camera, I used to shoot the BMPCC4K at ISO 3200 but I’ve settled on ISO 1600. 1250 will likely produce the best looking image in the upper range, comparable to ISO 400. Would love to see the BMCC6K for myself, open gate.

1250 would be the 'same' as exposing to 100 iso because you will prioritize shadows over highlights, that's why it's the cleanest option on the second base ISO, because most of the time, you will be on a low light situation, so shadows are more important.

… moving it down to 400 would only make a difference if I changed other exposure values, if i would only change ISO, then it would make no difference…

Roger that and 10-4.

This camera is like a Production 4K 2.

If you order this camera, I would buy NeatVideo.

It's always a good idea to have NeatVideo in your arsenal.

It's always a good idea to have NeatVideo in your arsenal.

Except it’s the slowest plugin in the world.

I own it. Only use it in VERY specific cases when Resolve doesn’t work.

I'm also experiencing some bad FPN problemas with the new 6k ff sensor, i also sent an email to BMD support yesterday

I've compared with my 6k pro (which I also bought on it's release) and despite the fact that the 6k pro has som problems, it's a lot worse on the new 6k ff.

here is some footage of it (a quick video, some prints and also the braw file):

https://drive.google.com/drive/folders/1iTEhdyV8pxMIEoSRuv-ed-hsyH__lTXe?usp=sharing

Most likely CMOS smearing. If you wave your hand around in front of the lens on the side of the light source the banding will move/disappear. I contacted BM and there response is that it's normal LOL .. That being said this camera produces the nicest images south of the Alexa with enough light. I own an fx6, so I have to decide if this camera's numerous limitations (poor form factor, lack of NDs, bad low light perfomance, etc.) are worth overlooking for its stellar image and codec.

No, it's not normal, i just received confirmation from BlackMagic that my sensor does have a problem and i'll be taking it to authorized dealer here in Brasil for it to repair.

Again, this problem DOES NOT happen on the 6k pro the same way that happens on the new 6k FF.

Interesting. I have the exact same problem. So Blackmagic is acknowledging the issue for some but not others.

Lets hope its a firmware fix and they needed a camera with the problem back for test.

Lets hope its a firmware fix and they needed a camera with the problem back for test.

Let's hope it's not just inconsistent responses to the same problem. I emailed BM support again today. They need to get it together and give us some real answers.

Lets hope its a firmware fix and they needed a camera with the problem back for test.

Let's hope it's not just inconsistent responses to the same problem. I emailed BM support again today. They need to get it together and give us some real answers.

Agreed. While it's not a problem that will plague my production on a consistent basis, I would hate for it to pop up at the worst possible moment.

I can confirm that lenses do not affect the banding. Makes no difference.

I tried:
- sigma 24-70 2.8 art
- canon FD
- canon k35
- lomo illuminas mkii
- Cooke s4i
- speed panchros

Always they're in the shadows in low light. A simple pan from a dark area to well lit.. banding is there in the dark.

For me this cam is usable at iso 100-250 and 1250 when shooting darker scenes. Which is always advisable since you have lots of info in the shadows and therenis enough light ofcourse. But anything above these and the banding is noticable.

So for docu or event work it might be a hard sell because sometimes you gotta shoot in dark places and you have no control. And not even THAT dark. Just a dark corner in a daylight scene is enough (note: if your frame is just that dark corner though! I know.. why would you do that.. but a pan from dark to light or something for a scene intro is a thing).
You can work around this maybe, but for me its just not acceptable for a new camera to have this issue, since no other cam i've had or shot on in 10 years have had this problem (this bad i mean)

Still hope for good news from BM though * cross fingers*

Hi Stefan! Have you heard back from BM? Just upgraded from my BMPCC4K and this issue is terrible.... I just sent an email to support but kinda worried about it.

How can I reproduce this issue? I don't shoot in low light situations, however, this made me concerned that my camera might have the same issue and needs to be replaced/repaired.

The problem apparently doesn’t show in well-lit scenes so you should be fine. However you can always test your camera in ‘poorly’ lit scenes and look for the issue. It may take BMD a few weeks to resolve the issue if it is a firmware solution and then everyone will get a firmware update. I’d start to be seriously worried if there isn’t an official word from BMD by January.

Hi Stefan! Have you heard back from BM? Just upgraded from my BMPCC4K and this issue is terrible.... I just sent an email to support but kinda worried about it.

Sorry for bringing this thread up again but since I've been working on a big project I haven't been able to properly follow up on this issue.

So, after a couple of back and forth mails with BM tech their answer is basically "It's the sensor, can't really do anything about it".
The last "test" I did back then was in a lighting situation that was mostly grey (and blue but no purple) and highlights were touching yellow in false color, so no crazy underlit or high contrast situation, and still had issues with the smearing. Although less noticeable, it was there, and again, also with the lut applied I could see it flickering around in the deep shadows of the image from ISO 320 and up. Yes after compression its less noticeable but it shouldnt be there to start, not like this.

This is the full (final) mail concluding the convo;

"
Thank you for your email and I can appreciate that you are still experiencing the CMOS smear in some under-lit situations.

We are also aware that is can happen sooner on the Cinema Camera 6K than on the Pocket Cinema Camera 6K Pro but please be aware, although they are similar in appearance the two cameras are different internally and are built for different purposes.

The Cinema Camera 6K has different sensor technology in order to implement a full frame sensor and is designed entirely as a Cinema Camera, which may not be comparable in certain situations in comparison to the more versatile Pocket Cinema Cameras that have smaller sensors and multiple codec options."


So to conclude this (for me): Camera is amazing in daylight, normal settings. Once you go for anything moody in a natural setup (aka not controlled with lights etc), I'm not confident to use it... so I wont. It's a great camera for a specific type of shooting I guess. Which is fine, but kind of a bummer.

And the Rolling shutter in Open Gate is the 2nd reason to pass for certain projects. It's really bad lol. The crop mode is great though, but yeah, defeats the purpose of FF.

"The Cinema Camera 6K has different sensor technology in order to implement a full frame sensor and is designed entirely as a Cinema Camera.."

I just had to laugh. So a 'cinema camera' is not just allowed but even expected to perform bad in low-light?! That's a nonsense excuse.

Also cinema and shows turned a lot dimmer/darker in the past decade and shooting keeps trending towards practical/available/natural with lower overall light levels.

Hi all.
I ran tests with my camera. And I found out that this problem is related to the lenses. This defect appeared with the Carl Zeiss 50mm 1.4 and Canon 16-35mm L 4 lenses. The defect did not appear with the Canon 50mm 1.2, Canon 24-105 L 4, and Canon 100mm 2.8 lenses. I carried out tests with cameras BMCC6K FF and BMPC4K everywhere, this defect appeared the same. I even tried the iPhone 14Pro. I'll send scans or videos later. I came to the conclusion that everything was fine with the camera. It is necessary to look for the problem in the lenses.

Vitaly, thanks so much for identifying the apparent source of the issue. Rather unexpected results.

Vitaly, are you using Sigma Mc-21 adapter, or a different one?

Vitaly, thanks for your tests. Maybe it has something to do with the new OLPF they have in front of the sensor? If you dare, it looks like it's pretty easily removable .

Vitaly, are you using Sigma Mc-21 adapter, or a different one?

Yes, I use Sigma Mc-21.

Vitaly, thanks for your tests. Maybe it has something to do with the new OLPF they have in front of the sensor? If you dare, it looks like it's pretty easily removable .

I don't think so. I had the same flares on both cameras: BMCC6K FF and BMPCC4K using the same lenses.

I got rays of light on these lenses:
Carl Zeiss 50 mm 1.4, Canon 16-35mm L 4.
If you use a light source with a flicker frequency that does not match the camera's shutter speed, these rays will appear in stripes.
Test:

pass: 1234

When using these lenses, such rays did not appear.
Lens: Canon 50mm L 1.2, Canon 24-105mm L 4, Canon 100mm L 2.8.
Test:

pass: 1234

I had the same flares on both cameras: BMCC6K FF and BMPCC4K using the same lenses.

I suggest that the guys who have these strips conduct tests with different shutter speeds in the camera and also conduct tests using natural light, for example in a dark room with light from a window.

Or use lights with high frequencies that are usually described as flicker free.

I got rays of light on these lenses:
Carl Zeiss 50 mm 1.4, Canon 16-35mm L 4.
If you use a light source with a flicker frequency that does not match the camera's shutter speed, these rays will appear in stripes.
Test:

pass: 1234

When using these lenses, such rays did not appear.
Lens: Canon 50mm L 1.2, Canon 24-105mm L 4, Canon 100mm L 2.8.
Test:

pass: 1234

I had the same flares on both cameras: BMCC6K FF and BMPCC4K using the same lenses.

I suggest that the guys who have these strips conduct tests with different shutter speeds in the camera and also conduct tests using natural light, for example in a dark room with light from a window.

Am I missing something? You’re showing lens flare in your tests, which have nothing to do with banding or shutter speed.

Not trying to knock your tests, it just seems like you’re testing for something totally different than the CMOS sensor smear at the heart of this thread.

I got rays of light on these lenses:
Carl Zeiss 50 mm 1.4, Canon 16-35mm L 4.
If you use a light source with a flicker frequency that does not match the camera's shutter speed, these rays will appear in stripes.
Test:

pass: 1234

When using these lenses, such rays did not appear.
Lens: Canon 50mm L 1.2, Canon 24-105mm L 4, Canon 100mm L 2.8.
Test:

pass: 1234

I had the same flares on both cameras: BMCC6K FF and BMPCC4K using the same lenses.

I suggest that the guys who have these strips conduct tests with different shutter speeds in the camera and also conduct tests using natural light, for example in a dark room with light from a window.

Am I missing something? You’re showing lens flare in your tests, which have nothing to do with banding or shutter speed.

Not trying to knock your tests, it just seems like you’re testing for something totally different than the CMOS sensor smear at the heart of this thread.

In my case, I wanted to show that if the frequency of the light flickering does not match the shutter speed on the camera, then these rays will be displayed in horizontal stripes. I also wanted to show that when using some lenses, rays appear, but when using others, they do not.
I don’t have these horizontal stripes because the frequency of the light flickering coincides with the camera’s shutter speed. That's why I suggested testing with different flicker frequencies.

P.S.
I couldn't find a flickering light source in my home, so I couldn't do this test. But I conclude that these stripes will appear at the wrong flicker frequency. And different lenses, according to their characteristics and quality, enhance it.

I just bought a BMCC6K and had some of this smearing. It's caused by overflow from bright pixels into dark pixels (bright window versus dark interior will demonstrate it best). It goes away when using first (100) and second (1250) native iso's.

NB don't confuse native iso with the recommended iso's of 400 and 3200 (which are digital gain iso's meant to distribute dynamic range for ease of monitoring shadows and highlights and are not native iso's).

The digital gain applied in camera seems to be the cause of the smearing. So better off avoiding that scenario.

Workflow: try and shoot at native iso and change aperture/shutter speed/lighting to achieve intended exposure.

If you run into an exposure triangle dead end, better to underexpose in camera at iso 100/1250 and boost the gain in post production by the required number of stops (rather than boosting the gain/raising the iso in camera).

This YouTube video provides a good explanation and illustration of the issue and comparison between a number of different cameras, and the solution of native iso.

Am considering upgrading from 6K G2 to the FF soon and this thread has me worried. Any consensus on whether or not these issues are consistent across cameras or if it is limited to a certain batch of cameras? Anyone else upgrade from a 6K Pro or G2 and have any pros/cons?

The digital gain applied in camera seems to be the cause of the smearing. So better off avoiding that scenario.

There is no digital or analog gain applied in camera until ISO6400. The artifact is just buried at lower ISOs.

Good Luck

I got rays of light on these lenses:
Carl Zeiss 50 mm 1.4, Canon 16-35mm L 4.
If you use a light source with a flicker frequency that does not match the camera's shutter speed, these rays will appear in stripes.
Test:

pass: 1234

When using these lenses, such rays did not appear.
Lens: Canon 50mm L 1.2, Canon 24-105mm L 4, Canon 100mm L 2.8.
Test:

pass: 1234

I had the same flares on both cameras: BMCC6K FF and BMPCC4K using the same lenses.

I suggest that the guys who have these strips conduct tests with different shutter speeds in the camera and also conduct tests using natural light, for example in a dark room with light from a window.

Am I missing something? You’re showing lens flare in your tests, which have nothing to do with banding or shutter speed.

Not trying to knock your tests, it just seems like you’re testing for something totally different than the CMOS sensor smear at the heart of this thread.

You're not missing anything. These are just lens flares.

I got rays of light on these lenses:
You're not missing anything. These are just lens flares.

Hi. Agree. It's just glare. But different lenses, even with the same focal lengths, glare differently. For example, Canon 50mm makes natural soft highlights, and Carl Zeiss 50mm makes rough highlights in the form of stripes that look unnatural. This is due to different lens quality. I'll continue. It is important. If you think about it, there will be a flickering light source in the frame. Then stripes appear in the frame. There are shutter speeds at which the bars do not blink quickly. Everyone is used to seeing a standard strobe at the wrong shutter speed, but there are strobe (blinking) that are slow or that stand still in the form of a stripe. I've encountered this. And this is most likely why many people take this as a defect. I suggest you think about this. Most likely this is true. In the tests that were given by other users, it is clearly visible that if they block the light with their hand, the stripes change, then they depend on the light source and shutter speed on the camera and appear in different places in the frame.
For me, for example, lenses that create rough glare look bad and I don’t like it. I didn’t really notice this on set because these are rare conditions. But after conducting this test, I realized that these lenses are not particularly high quality.

It depends completely on the position of strong lights and the kind of lenses you use. Even the same lens can look good in one scene and bad in another.
Oh, and BTW, these days Zeiss, just like Leica, sell their names if somebody is ready to pay for it. My 30 year old Zeiss C/Y lenses have beautiful flares, but old Minoltas are even better. They don't adapt to EF mount unfortunately, with one exception.

Finally, this discussion is pretty much off-topic, this is not what most people would call banding.