# ZED 2i: Severe magenta/purple image corruption after bright sunlight exposure

**URL:** <https://community.stereolabs.com/t/zed-2i-severe-magenta-purple-image-corruption-after-bright-sunlight-exposure/11388>\
**Category:** ZED2\
**Created:** [June 2, 2026, 8:57am UTC](https://community.stereolabs.com/t/zed-2i-severe-magenta-purple-image-corruption-after-bright-sunlight-exposure/11388 "2026-06-02T08:57:05Z")\
**Posts on this page:** 4\
**Page:** 1

<div class="post-metadata">

**Author:** ![dweinlader](https://community.stereolabs.com/letter_avatar_proxy/v4/letter/d/c6cbf5/32.png) [@dweinlader](https://community.stereolabs.com/u/dweinlader)\
**Post date:** [June 2, 2026, 8:57am UTC](https://community.stereolabs.com/t/zed-2i-severe-magenta-purple-image-corruption-after-bright-sunlight-exposure/11388/1 "2026-06-02T08:57:05Z")

</div>

Key observations:

- The issue has been observed on a large number of cameras. We have nearly 200 ZED 2i units deployed, so this does not appear to be a failure of a single camera.

- I have observed this issue using both SDK 4.x and zed-open-capture.

- A previous employee discussed this with [Stereolabs](https://www.stereolabs.com/) support and was reportedly told that the camera must be power-cycled when the issue occurs.

- Power-cycling is not an acceptable solution for our application because we cannot experience random interruptions during operation.

- The issue has only been observed outdoors in bright sunlight. I have not yet been able to reproduce it indoors.

- The camera may operate normally for extended periods (for example, more than an hour) before the issue occurs or it can occur within seconds of turning on.

- In at least one case, the issue appeared immediately after the camera was turned toward the sun while being moved/stored.

- The corrupted image is not simply overexposed. Bright regions become white, but many highlight transitions and reflective surfaces develop strong magenta/purple coloration.

- The issue can affect large portions of the image and produces unusable data for computer vision.

Current configuration:

- Using zed-open-capture.

- Camera hardware AEC/AGC disabled.

- Running a custom software exposure/gain control loop.

- Camera white balance control disabled.

- Disabling white balance may have reduced the frequency of the issue but has not eliminated it.

Additional observations:

- The open-capture source appears to indicate an OV580 ISP/bridge and OV4689 image sensors.

- The project includes logging of OV580 ISP registers and OV4689 exposure/gain registers, suggesting there may be ISP state related to this issue.

- Because the problem has been observed with both the SDK and open-capture, I suspect the issue may exist below the application layer.

Questions:

1. Is this a known issue with the OV580/OV4689 image pipeline used in the ZED 2i?

2. What internal camera state is becoming corrupted when this occurs?

3. Is there a documented recovery method that does not require a USB power cycle?

4. Are there specific ISP, AEC/AGC, AWB, or OV580 configuration registers that should be disabled or initialized differently to prevent this condition?

5. Has [Stereolabs](https://www.stereolabs.com/) identified the root cause or a mitigation for deployments operating continuously in bright outdoor environments?

Any guidance would be appreciated.

 ![zed_2i_noise](https://community.stereolabs.com/uploads/default/original/2X/d/d629475f74c54aa8023fc1f9fbc055f671de4151.jpeg)

 ![zed_2i](https://community.stereolabs.com/uploads/default/original/2X/2/2c692c237264597ed55558ebff142d65f8b4a9a2.jpeg)

---

<div class="post-metadata">

**Author:** ![Myzhar](https://community.stereolabs.com/user_avatar/community.stereolabs.com/myzhar/32/11_2.png) [@Myzhar](https://community.stereolabs.com/u/Myzhar)\
**Post date:** [June 3, 2026, 8:03am UTC](https://community.stereolabs.com/t/zed-2i-severe-magenta-purple-image-corruption-after-bright-sunlight-exposure/11388/2 "2026-06-03T08:03:20Z")

</div>

Hi @dweinlader  
Welcome to the [StereoLabs](https://www.stereolabs.com/) community.

What you described is a rare problem caused by a rare combination of factors (mostly outdoors) that cause the overflow of one of the registers of the OV580.  
Regrettably, the only working solution is to call the `reboot` function to force a reset of the camera registers because a simple rewrite of the affected register does not fix the issue.

The `reboot` command is sent to the MCU of the ZED 2i, which forces a reset of the OV580. This operation is the same as disconnecting and reconnecting the USB plug.

> **[Camera Class Reference | API Reference | Stereolabs](https://www.stereolabs.com/docs/api/classsl_1_1Camera.html#a334a36797b49df7dc290360acf5cba2a)**
>
> This class serves as the primary interface between the camera and the various features provided by the SDK. More...

> <https://github.com/stereolabs/zed-open-capture/blob/93739154903a66c112ee52c3c6fbfd83b7218617/src/sensorcapture.cpp#L673>

---

<div class="post-metadata">

**Author:** ![dweinlader](https://community.stereolabs.com/letter_avatar_proxy/v4/letter/d/c6cbf5/32.png) [@dweinlader](https://community.stereolabs.com/u/dweinlader)\
**Post date:** [June 3, 2026, 12:06pm UTC](https://community.stereolabs.com/t/zed-2i-severe-magenta-purple-image-corruption-after-bright-sunlight-exposure/11388/3 "2026-06-03T12:06:57Z")

</div>

> [@Myzhar](#):
>
> What you described is a rare problem caused by a rare combination of factors (mostly outdoors) that cause the overflow of one of the registers of the OV580.

Thank you. I will try that. Is this something you might consider fixing with a firmware update for the camera?

---

<div class="post-metadata">

**Author:** ![Myzhar](https://community.stereolabs.com/user_avatar/community.stereolabs.com/myzhar/32/11_2.png) [@Myzhar](https://community.stereolabs.com/u/Myzhar)\
**Post date:** [June 3, 2026, 12:50pm UTC](https://community.stereolabs.com/t/zed-2i-severe-magenta-purple-image-corruption-after-bright-sunlight-exposure/11388/4 "2026-06-03T12:50:00Z")

</div>

> [@dweinlader](#):
>
> Is this something you might consider fixing with a firmware update for the camera?

Yes, the hardware team is analyzing and trying to fix it.
