Having a weird issue with the B0527 camera. The live preview looks perfectly fine, but when I capture a full-resolution image through the Python script, the saved image comes out heavily underexposed — like way darker than what I’m seeing on screen.
Also, after taking a full-res shot, the camera just shuts off completely. The live screen goes black and nothing brings it back except physically unplugging and re-plugging the USB cable. And when I reconnect, the Python script weirdly opens the webcam twice before the USB camera finally opens again.
Yeah, I tried toggling the exposure time already — no effect on the saved image at all. I even tried shining a bright light directly on the sensor before capturing, and the saved image still looks just as dark. It’s like the capture isn’t respecting any of the exposure settings that the live preview is using.
Here’s a comparison shot showing what I mean — left is what the live view shows, right is what actually gets saved:
I’m on Windows 10, Python 3.9, direct USB 3.0 port on the motherboard. And yes, the shutdown happens on every single full-res capture — 100% reproducible on my end.
Thanks for the screenshot — that really clarifies the severity of the underexposure. And the fact that neither exposure toggling nor adding external light makes any difference is a strong clue.
So here’s what’s going on with Issue 1 (the underexposed captures):
This is a known hardware/firmware limitation with the B0527. The ISP (Image Signal Processor) functions — specifically gamma correction and white balance control — are not built into the camera’s onboard processing pipeline. The live preview stream goes through a different processing path than the full-resolution capture, which is why you’re seeing that mismatch. The preview gets some host-side processing, but the raw full-res frame doesn’t benefit from those same adjustments.
Unfortunately there’s no immediate fix I can give you for this. Our engineering team is actively working on it and we’re hoping to have it sorted out as soon as possible.
As for Issue 2 (the shutdown after capture) — we haven’t been able to reproduce that behavior on our side yet. Your Windows 10 + Python 3.9 setup might be a factor. Could you try recording a short video of what happens on screen when the shutdown occurs? That would help us narrow down whether it’s a USB enumeration problem, a driver crash, or something on the firmware side.
Appreciate the honest update on Issue 1 — good to know it’s not just me doing something wrong. I’ll keep an eye out for any firmware updates.
For the shutdown thing — yeah, I can grab a video of it. Should I capture the whole screen including the Python terminal output, or just the camera preview window? The terminal does show some messages when it crashes out, might be useful.
Definitely capture the whole screen with the terminal visible — those console messages will be really helpful. If you can also note the exact resolution you’re capturing at when it happens, that would help too (I’m assuming it’s the sensor’s max resolution based on your description).
So to summarize where we’re at:
Issue 1 (underexposed saves): Confirmed ISP/firmware limitation — no immediate fix, engineering is working on it.
Issue 2 (shutdown after capture): Not yet reproduced internally — waiting on video evidence from your setup so we can diagnose further.
Once you’ve got that recording, post it here and we’ll dig into the shutdown behavior next.
