@alex_ai OK, the GX series and the VEYE series use completely different ISP solutions, so their parameters do not correspond one-to-one. I look forward to your test results.

@zoldaten
As long as you follow the steps and place our driver and DTS files into the Orange Pi source tree, then compile the image according to the official Orange Pi instructions and flash this image onto the board, the VEYE camera driver should appear correctly.

Essentially, this operation just adds the driver to the original standard Orange Pi system. No additional steps are required after flashing the image. I suspect that you may not have fully applied the modifications in all the necessary parts of the source code.

@sheko Sorry, I just got back from the Chinese New Year holiday. Regarding using the RAW-MIPI-SC132M on the Raspberry Pi 5, I suggest you primarily refer to this document. It covers driver installation, how to use the V4L2 mode, and methods for previewing using OpenCV.
https://wiki.veye.cc/index.php/Mv_series_camera_appnotes_4_rpi

@nkos

Sorry, regarding the attempt to port the camera to the Radxa Zero 3 + 6.x, we tried to set up the standard build environment but were unable to compile successfully. I posted a question on the Radxa forum, but have not yet received a follow-up response.

As for the method you mentioned to configure our registers, we recommend that you do not use the driver. Instead, please use our mv_mipi_i2c.sh script.

https://wiki.veye.cc/index.php/Mv_mipi_i2c.sh_user_guide

@veye_xumm
Thanks for the answer, have a happy holiday!

@veye_xumm

That's right. It worked with the lens. Thank you very much for your technical support.

@xlwvu7mx You are welcome.

@d_101 Sorry for the late reply. Based on the current situation, I can only suspect a hardware issue that is causing the I2C path to be unavailable. This could be a problem with your mainboard or with the camera itself. Please return the camera to us, and we will perform testing and inspection.

@ffgemanof
The veye_viewer client currently does not support the mono10 and mono12 formats. If you need to use these formats, please configure the data format using mv_mipi_i2c_new.sh, and then acquire the data using v4l2-ctl or the V4L2 API.

Hi @veye_xumm, thank you again for all your help. I found that the VEYE camera you made available with I2C address 0x22 does work on the J401B. And now, after a long series of follow-ups with Seeed Tech Support, I found that i2cdump generates I2C transfer timeouts at address 0x3b (but only 0x3b!).

Then Seeed noted that applying the VEYE "compatibility fix" updating the USB Hub firmware to 08C3 removed the timeouts. So I tried that with the VEYE camera and it works! (This with Jetpack 6.1).

A weird puzzle remains, I was sure the J401B had USB Hub Firmware version 08C3 as supplied, although with one of the "devices" giving different details to 08C3 on the J401. Hence I did not expect to need the "compatibility fix", and had some concerns that reapplying it could cause a problem. But reapplying the update just now the firmware version was initially 08B3!?

It has updated to 08C3, and has fixed camera access to 0x3b.

This has been time consuming, but I might now test JetPack 6.2

@veye_xumm
Sorry for the late reply.
Thank you for the new code!
It seems to work partially in the rpi5 standard OS. I was able to compile the read and write files, and some commands work while others don't.

The new mgain is stuck at zero, so if i use new exp mode i can set the shutter speed with new mshutter, but not the new mgain, despite changing it returns zero when i probe with the -r parameter. so using the new exp mode i only get a black image.

the old mshutter works and i was able go get up to 1 second exposure per frame in streaming mode but the capture mode and capture command don't seem to work.

Also i would like to have it work in the rpi archlinux OS that i use for astronomy. The veyecam2m driver works and i get video in the archlinux OS, but i wasn't able to compile the i2c read and write in that system.

@veye_xumm it is resolved now, the image is shown. Before with the steps on the forum the .dtbo file was also missing, I copied it from another folder and it didn't work. The right and supported way of installing drivers is very confusing, I tried installing by wiki guide and then compiling locally (github), since it didn't work. Thank you for your patience.

@nkos Cheers!

@veye_xumm Hi,
Firmware update does not solve the problem, after software reset auto exposure mode works normal

@nimadibaj
You have implemented the function of writing 0xff to 0x000b. Similarly, proceed to write the subsequent registers as well.
The method for writing the subsequent registers is the same as that for writing the 0x0b register, but they must be written in sequence.
Addr--->data
0x10 0xDE
0x11 0xC2
0x12 0x00
0x13 0x00

@veye_xumm

Hello my friend.
We are sending I2C commands via a FPGA.

1_ Writing 0xff to 0x000b solved our problem. Thank you. Please give us a command to save this change.

2_ We are trying to change video frame rate (now is 30fps and we need 25fps).
by these commands we cannot do this :
0x10 0xDE
0x11 0xC2
0x12 0x00
0x13 0x00
(sending Two byte, then sending next two byte and ... ).

Please guide me. Thanks a lot.

@deepcamera

Only ISP-related parameters can be persistently saved after power-off through the paramsave command. Therefore, I recommend that you reconfigure your desired parameters in the startup script during program design.

As for which commands can be saved, you can refer to the veye_mipi_i2c.sh source code. If the corresponding read/write functions contain operations like './i2c_write $I2C_DEV $I2C_ADDR 0x10 0xDA', it indicates that the function's register belongs to the ISP category and can be saved using paramsave.

@lpcq2000 抱歉现在MCU bypass模式已经不再被支持了。我们的资料有一些迟滞，回头会更新到上面。

@siderat I think this one will work.

@barbanevosa
Please refer to part8:
https://wiki.veye.cc/index.php/Mv_series_camera_appnotes_4_jetson/