Moravian Direct CMOS camera control

[ProtonImager]

Direct control of the new C1,C3 and the medium format C5 series of Moravian cameras is lacking.

Would the developers need a camera to work with to develop this or is it something to do with Moravian?

[Stefan B]

Hi,

a native integration into the application requires us to have a camera physically available as well as the vendor providing the sdk with the documentation.
Without a camera, implementing the SDK is not feasible as it needs to be tested thoroughly and also long term maintenance requires to have a camera available for future sdk upgrades.

[ProtonImager]

I spoke to Patel at Moravian. In 2023 they will reach out to NINA developers to send test cams. They do have an SDK. Thank you

[Stefan B]

They can reach out to me via [e-mail]

(I’m the creator and maintainer)

[Dale Ghent]

I started on a GXCCD native driver some a year or so ago, but put the work aside when it became clear that developing it without an actual camera on hand to test with would not be a satisfactory time investment, and there was no NINA user that I knew of that had GXCCD camera. I think I might have since cleared that branch of work during a cleanup

But it is very nice for Patel to offer that resource. I take it that the ASCOM driver is sufficient for now? Are there any camera features that would be good to manage from within NINA that the ASCOM driver does not make available, or is wanting a native driver more of a consideration for frame download speed?

[ProtonImager]

Frame download speed is the issue and isolated freezing.

‌

Moravian uses a SIPS software package which is what they recommend. The camera is lightning fast on it when I used it but it is NOT NINA!

It’s slow with the ASCOM driver probably from all the ASCOM overhead.

‌

I’ll pass on the email address.

[Dale Ghent]

The “ASCOM overhead” is the overhead of the Windows .NET SafeArray, combined with possible image row reordering - SafeArrays are slow to start with (due to their flexibility) but memory intensive, and so is any image row reordering that might need to happen. Not a problem when camera sensors were small, but very obvious once you start pushing pixel counts like what the IMX455 and 411/461 sensors have. Avoiding the SafeArray is the primary reason for having a native driver for fast CMOS cameras; another large reason being access to controllable camera features that aren’t covered by ASCOM.

[Stefan B]

Duplicate of #864.