We're using your Oak-D Pro camera with IR illumination in a pitch dark environment. We set the sensor resolution to THE_400_P, the IR illumination to maximum strength, and have a 400x300 autoexposure region in the middle of the image. Sometimes we get striped grayscale images that lead to distorted depth images: [upl-image-preview uuid=4426e614-118c-4f64-87d4-e45e07d7d80f url=https://discuss.luxonis.com/assets/files/2026-06-17/1781712057-975774-image.png alt={TEXT?}] In the depth image red means very close by, blue means further away. Another less extreme example: [upl-image-preview uuid=d092f4c0-79a4-43ab-9bc5-0777d8492aaa url=https://discuss.luxonis.com/assets/files/2026-06-17/1781714914-625515-image.png alt={TEXT?}] We're wondering what causes this striping. Could it be interference with IR illumination pulses? Does the IR shines continuously or in pulses? Or any other hints what it could be? Thanks in advance!

Another example with point cloud added: [upl-image-preview uuid=3acac994-0543-433d-9cb4-d29c1a5a452e url=https://discuss.luxonis.com/assets/files/2026-06-18/1781781362-518082-image.png alt={TEXT?}]

Hi @"WouterOddBot"#p34027 Apologies for the late reply. Can you provide more information on: - ir projector brightness - resolution used? - exposure time which is set by the AE - ISO You say this happens sometimes? Have you tried changing the IR intensity when that happens? It looks like an artifact/ some weird combination of image scaling + IR banding. Thanks, Jaka

@"jakaskerl"#1112 Any ideas still?

Hi Wouter, Unfortunately, we haven't been able to reproduce this on our side yet, so at the moment I can only suggest a couple of things to help narrow it down. 1. Could you log the exposure time and ISO/gain when the striping occurs? It would be useful to compare those values between "good" and "bad" frames to see if the artifact coincides with an AE adjustment. 2. If possible, could you run one overnight test with manual exposure and gain (disabling AE)? Since the issue appears intermittently, this would help determine whether the artifact is related to the auto-exposure algorithm or something else in the image acquisition pipeline. If you're able to collect a few striped mono frames together with the corresponding exposure/gain values, we'd be happy to take another look.

Hi @"SasoTamse"#p34087 , Thanks for your response and for trying to reproduce. 1. Yes I'll add logging for exposure time and ISO/gain. It may take some though before we get the logs. This striping happens one our production robots in the field and we'll have to wait until our next release. 2. Would indeed be nice if we could help you with that, but sorry. The software runs on our production robots that work day and night. So we can't turn off autoexposure. We have testing robots too but we can't reproduce the striping on those robots. So hopefully I'll come back here in a few weeks with some images and expose/gain logs.

Striped grayscale images

WouterOddBot

We're using your Oak-D Pro camera with IR illumination in a pitch dark environment. We set the sensor resolution to THE_400_P, the IR illumination to maximum strength, and have a 400x300 autoexposure region in the middle of the image.

Sometimes we get striped grayscale images that lead to distorted depth images:

In the depth image red means very close by, blue means further away. Another less extreme example:

We're wondering what causes this striping. Could it be interference with IR illumination pulses? Does the IR shines continuously or in pulses? Or any other hints what it could be?

Thanks in advance!

WouterOddBot

Another example with point cloud added:

WouterOddBot

jakaskerl

Hi WouterOddBot
Apologies for the late reply.

Can you provide more information on:

ir projector brightness
resolution used?
exposure time which is set by the AE
ISO

You say this happens sometimes? Have you tried changing the IR intensity when that happens? It looks like an artifact/ some weird combination of image scaling + IR banding.

Thanks,
Jaka

WouterOddBot

Hi @jakaskerl ,

Thanks for your response. The IR projector brightness and resolution:

We set the sensor resolution to THE_400_P, the IR illumination to maximum strength

We don't log the AE's exposure time and ISO. But the camera runs at 60Hz, so the exposure time must be smaller than 1/60 seconds.

You say this happens sometimes?

Yeah actually we just got some new footage where the camera was working just fine and then suddenly the striping occurs. We have around 40 robots and this striping happens occasionally on several robots.

Have you tried changing the IR intensity when that happens?

No we didn't try changing IR intensity, as the striping happens in the middle of the night, and we also need max intensity to get proper depth images. And so far we cannot reproduce the striping, so we can't play with parameters to see what causes striping.

some weird combination of image scaling + IR banding

We looked into enabling anti banding mode, but we can't tell whether this has something to do with banding. We know what banding means, but can you explain a bit more about IR banding specifically for the OAK-D?

Thanks in advance!

WouterOddBot

@jakaskerl Any ideas still?

SasoTamse

Hi Wouter,

Unfortunately, we haven't been able to reproduce this on our side yet, so at the moment I can only suggest a couple of things to help narrow it down.

Could you log the exposure time and ISO/gain when the striping occurs? It would be useful to compare those values between "good" and "bad" frames to see if the artifact coincides with an AE adjustment.
If possible, could you run one overnight test with manual exposure and gain (disabling AE)? Since the issue appears intermittently, this would help determine whether the artifact is related to the auto-exposure algorithm or something else in the image acquisition pipeline.

If you're able to collect a few striped mono frames together with the corresponding exposure/gain values, we'd be happy to take another look.

WouterOddBot

Hi SasoTamse ,

Thanks for your response and for trying to reproduce.

Yes I'll add logging for exposure time and ISO/gain. It may take some though before we get the logs. This striping happens one our production robots in the field and we'll have to wait until our next release.
Would indeed be nice if we could help you with that, but sorry. The software runs on our production robots that work day and night. So we can't turn off autoexposure. We have testing robots too but we can't reproduce the striping on those robots.

So hopefully I'll come back here in a few weeks with some images and expose/gain logs.