Dither interferes with Download of Image - results in capture of dither motion

[Former user]

When PHD2 and sky conditions are a little unstable and settle time is taking longer than normal, NINA appears to be listing the shot as "downloading" but in reality still has the chip open and capturing the dither motion - once the dither timeout is reached the camera is closed and the shot loaded in the image window = doubled stars. sequence take with no dither have no delay on closing the shot. It appears to be a synchronous blocking call on dither settle that does not allow the calls to end the current image.

Version Beta 007 reproducable : yes - run under condition that force PHD2 to struggle with settle - turn off dither and download is instant

no logs unfortunately

[Yannick Dutertre]

Your observations may be correct, but I think your conclusion is wrong

More likely NINA has started capturing while dithering from the previous shot was not over

Dither and download will indeed occur in parallel to save time

But the shutter is already closed by then

As soon as download is over and dither is over (as per PHD2) OR the NINA dither timeout is reached, the next exposure will start

So if PHD2 had not recovered by then, then you will have issues.

The "Downloading" thing is purely display - because of all the parallel stuff going on, it is difficult to update accurately

[Stefan B]

Hi,

which camera are you using? It might be an issue for a specific camera implementation.

[JAMIE FLINN]

Thanks for looking at this - Camera is QHY183M (native - not ASCOM)

I was able to confirm that using other tools connected to the same PHD2 instance (with NINA closed down) were able to sequence with dither without interfering with the download - it was always fast and finished before the dither started. as a work around I think I can set the settle to < 3 pixels and it will wrap up the download - last nights PHD operation was “messy” due to local seeing. All the same settings worked perfectly night before with stead seeing

[Stefan B]

Duplicate of #527.