Sensor problem?

Hello,

I bought a URSA mini pro G2 4.6K in 2022 and so far loving working with it.

This seems like a sensor problem to me, I wonder if any of you has been having the same issue?

The camera doesn't detect it as a "drop frame" so recording is not stopped automatically. It happens both in BRAW and PRORES. I record on Angelbird C-fast cards

It can happen randomly 1 to 5 times in a minute and then stop for a month or it can happen 10 times the same day.

Thanks for any ideas/solutions

Difficult to say .. Using an Ursa Mini Pro G1 , several years ago , recording in .Dng , I sometimes had some strange frames . It was rather rare and unpredictable ... Beautiful colors , but not exactly what I was expecting ... Twenty four frames per second ... twenty three correct frames , and one defective ... rarely .. Here it is :

DefectiveFrame01.jpg (258.6 KiB) Viewed 539 times

I was using Lexar CFast cards . I told to myself maybe it had something to do with my media formatting and File Allocation table ... the way datas are recorded in the memory cells .. I was using Exfat .. and I began using HFS plus .. and did not see it again ..

For those using Windows , a free software exists .. " HFS Explorer " , very simple and it works very well for me ..

Again .. not an easy explanation . Maybe the Sensor Specialists at BMD have an Idea ?

Good Luck , Regards / Mike ..

Thanks for your answer

It really looks like it, 23 frames ok and 1 defective, very random. I'm on Mac but not using ExFat

Would really love to hear from BM specialist, since Black Magic only proposes to change the sensor for around 3000$ without really being sure that that's the problem... rather disappointed with them.

The examples look quite different. The first is interesting in that the band is an exact repeat of the area of the subject’s mouth whereas the other examples are meaningful in terms of the image being recorded. If you have more than one CFast card, does changing to another card help? If a second card also has the error, it may be a problem in the image generation pipeline or the sensor. If different cards are good, then the original card is failing.

Yes . It may come from the sensor , but it may also come from the Image Processing Chips behind it ..
As we don't have the structure , the diagram , nor the schematics .. difficult to track the problem .. It's a Pipeline , from the sensor to the memory Card ..

A Camera is a complex device .. In Electronics , sometimes a Disturbance can come from another part of the circuitry .. auto-oscillations , spikes , via the power rails .. Even some regulators may auto-oscillate .. Not writing about " Tin Whiskers " .. if you don't know them , you'll find informations on SItes of the Nasa .. Maybe incredible , but true ..

With .Dng , as every frame is like a distinct Still Picture , it's easy to replace the defective frame with an adjacent one , just before or just after .. It's so fast it won't be visible . With Braw or Prores , I do not know , it should be doable within Da Vinci .. Anyway , it's annoying and not normal , it's always some more work and some more troubles ..

Regards / Mike ..

Looks to me as corrupted files during transfer from the card to the computer, especially if it doesn't occur when played back in camera.

The OP didn't write about that point. Actually, I remember to have seen that once with a bad transfer by a faulty reader. Something to be checked before dumping the camera.

Can also be bad media.