Need help ! Blackmagic support not wanting to help...

Hello guys !

It's been almost two years that I'm discussing with blackmagic support about an issue I have with my blackmagic micro cinema camera. I have a quite visible grid pattern effect whenever I apply sharpening on solid colors, especially on skin which is very annoying. I contacted blackmagic from the moment I noticed the issue. At first they told me that they couldn't see the problem. Then after sending a lot of examples (prores files, dngs...) they finally admitted they see something wrong on the export I sent them. But they told me they were unable to replicate the problem with the files I gave them...

I think they're being dishonnest here and I'm very disappointed by the treatment I got from Blackmagic, a brand I still consider greatly. To get the kind of results I have with my blackmagic footage, one just needs to apply some sharpening on lumetri (or an unsharp mask for instance in After effects...).

Anyway, here's a jpeg where I bumped the sharpening to enhance the problem of the sensor (watch the jpeg at 200%). (the jpeg is too big to be uploaded directly on bm website)

https://ibb.co/gmsVwT

I have the original files but they're quite heavy. Can please someone honnest at blackmagic take my problem seriously. I don't know if it's a sample / quality control issue but it is definitely there. I also have a BMPCC and it doesn't show that kind of issue (well at least not in those kind of proportions). I want to love this little camera and this brand, but the pill is hard to swallow when someone from support is acting a bit dishonnest...

Thank you for your time, hope to hear from someone at blackmagic.

This sounds more like an issue with your NLE, which program are you using? Is the an issue in all cases, or only in some images? What is the lighting and exposure on the problematic images?
A little more information will help here.
Cheers

Looks like the same crosshatch problem that hit the 4.6k, for that it was caused by an inbalanced green channel.

viewtopic.php?f=2&t=51936&hilit=crosshatch

This sounds more like an issue with your NLE, which program are you using? Is the an issue in all cases, or only in some images? What is the lighting and exposure on the problematic images?
A little more information will help here.
Cheers

Hi Denny !

I'm using Adobe Suite, Premiere and After Effects as well. It doesn't come from the NLE. I also use da vinci resolve and whenever a video is resized in Da Vinci, the grid is ultra apparent (that's actually what caught my attention first) ! I created a topic where I described the problem a little more a while ago on this very forum.

The problem appears whenever I over expose, under expose, expose spot on... I usually over expose by a stop to help reducing the noise by bringing down the signal afterward. But the grid is there no matter what... It is more visible on solid brownish or pinkish colors (furnitures, skin...) which is even more problematic.

Looks like the same crosshatch problem that hit the 4.6k, for that it was caused by an inbalanced green channel.

viewtopic.php?f=2&t=51936&hilit=crosshatch

Hey David ! Thanks ! That's what I think it is (when invastigating on the forum). It most likely looks like it. I just need someone at blackmagic acknowledging it (which they did partially) and process with a recalibration / replacement of my camera. That's all I'm asking And it's been almost two years that I'm regularly writing to bmd support with no success.

I have looked at your jpeg in Mozilla and am 99% sure that the grid is a viewer artifact, not a camera defect. I am so sure because the grid is visible in zooms near 90-110% and period of grid is changing when I change the zoom. And I see no grid at 200%.

Also the image is slightly noisy but it's expected.

In GIMP, I see lots of noise and nothing similar to grid in 100, 110, 200, 400 and 800%.

Further to Anatoly’s remarks, the only time I saw very annoying grid patterns was in my early days of using the URSA Mini 4.6K camera and it was indeed a viewer artifact at some resolutions using an earlier DaVinci Resolve, not present now in my viewers and not present on deliverables.

I’m also on an iMac 27” screen that supports 10 bit DCI P3 and 5120x2880 resolution.

You also must rely on the image you get from the Delivery tab at the proper viewing resolution and properly graded beginning with a decent (spot on not required) exposure. If I pump up the grade to extremes and magnify the image, it can be ugly.


Sent from my iPhone using Tapatalk

I have looked at your jpeg in Mozilla and am 99% sure that the grid is a viewer artifact, not a camera defect. I am so sure because the grid is visible in zooms near 90-110% and period of grid is changing when I change the zoom. And I see no grid at 200%.

Also the image is slightly noisy but it's expected.

In GIMP, I see lots of noise and nothing similar to grid in 100, 110, 200, 400 and 800%.

Hey ! There are two kinds of grid you can see in the jpeg

As you mentionned, when scaled down, the grid is all over the place ! But when zoomed in at 200% you can see parts of grids on her skin and on the chairs... The general grid effect when downscaled is concerning, but it is actually the small grid on her face that is more annoying.

Btw, the image is noisy because it hasn't been denoised and I added a fair amount of sharpening to make my point. It's just a prores file with a rec709 lut and sharpening, but it is actually shot overexposed by a stop.

I just re-read the thread about crosshatch on Ursa mini 4.6k, and there was this same debate at first, like it is a scaling issue. It is not. The scaling enlightens the grid patern, but no other camera footage (or jpeg extract) would have this "scaling issue". If I export a jpeg of any photo in raw, or any footage from a sony FS7 for instance, I will never encounter a scaling issue, using the very same software.

In the end, everybody acknowledged the fact that there is a cross hatching issue on certain cameras. When I look at the jpegs people posted in the thread, it's the exact same kind of grid I have with the bmmcc. I'm not as a technical expert as David is (he's the one that found out the green channel was the problem) but I know enough to know when it's a scaling issue and when it's within the footage.

But thank you for your contribution !

Further to Anatoly’s remarks, the only time I saw very annoying grid patterns was in my early days of using the URSA Mini 4.6K camera and it was indeed a viewer artifact at some resolutions using an earlier DaVinci Resolve, not present now in my viewers and not present on deliverables.

I’m also on an iMac 27” screen that supports 10 bit DCI P3 and 5120x2880 resolution.

You also must rely on the image you get from the Delivery tab at the proper viewing resolution and properly graded beginning with a decent (spot on not required) exposure. If I pump up the grade to extremes and magnify the image, it can be ugly.


Sent from my iPhone using Tapatalk

Hello Rick !

I totally agree with you, you can always break appart an image if you want to... But here I just applied some sharpening and watched the content on a projector and it was there. I mean, if you go 200% on the image you see the grid on her face, it's not acceptable. Well not by my standards for professional work. And once again, I'm not talking about the general grid appearing all over the image (although i'm sure both are related) but the small grids here and there, on her face, on the back of the chairs...

I do not know if this test is relevant in this case.

See if you can see this pattern on the sensor.

Cap the camera and push to maximum gain or ISO.

Adjust the Brightness and Contrast on the monitor you use to see if the pattern is being generated by your sensor.