# Frame Streaming

**URL:** <https://forums.openmv.io/t/frame-streaming/10743>\
**Category:** OpenMV Boards\
**Created:** [March 19, 2025, 6:27am UTC](https://forums.openmv.io/t/frame-streaming/10743 "2025-03-19T06:27:55Z")\
**Posts on this page:** 2\
**Page:** 1

<div class="post-metadata">

**Author:** ![Santin](https://avatars.discourse-cdn.com/v4/letter/s/858c86/32.png) [@Santin](https://forums.openmv.io/u/Santin)\
**Post date:** [March 19, 2025, 6:27am UTC](https://forums.openmv.io/t/frame-streaming/10743/1 "2025-03-19T06:27:55Z")

</div>

Hello.

I have an application in which we’re streaming and displaying it on an industrial HMI.  
The display control is very basic; it’s not a web browser.

We’re sending the image with the example program, but we do see that modifying the image compression affects the viewer: “cframe = frame.compress(quality=x)”

With compression of 35, the image is sometimes cut out, with 50 the image is flashy and with 90-100 it is displayed correctly.

The problem we’re having right now is that when I send the black and white image to adjust the thresholds, if it goes completely black or white, the control freezes and won’t refresh again.

I think it could be due to compression. I don’t know if there’s any way to improve compatibility with very basic browsers like the one we’re using.

I have been trying with “cframe = frame.to\_jpeg(quality=100, subsampling=image.JPEG\_SUBSAMPLING\_AUTO)” but we are not able to get the image stability.

Thank you and greetings.

---

<div class="post-metadata">

**Author:** ![kwagyeman](https://yyz1.discourse-cdn.com/flex029/user_avatar/forums.openmv.io/kwagyeman/32/4_2.png) [@kwagyeman](https://forums.openmv.io/u/kwagyeman)\
**Post date:** [March 19, 2025, 8:20am UTC](https://forums.openmv.io/t/frame-streaming/10743/2 "2025-03-19T08:20:21Z")

</div>

Hi, you might need to manually select the right subsampling: [image — machine vision — MicroPython 1.24 documentation](https://docs.openmv.io/library/omv.image.html#image.JPEG_SUBSAMPLING_444)

The auto algorithm changes between 444, 422, and 420 at 60% and 35% quality.
