Driving and Capturing from Two Cameras

[Kitfox]

I am going to be doing some scope comparisons and part of the evaluation will be utilizing Sharpcap to provide images of star fields and various deep sky objects. The plan is to do head-to-head comparison of a pair of astrographs and/or flattened or reduced normal refractors (all scopes in this initial test will be refractors).

I have the option of running this from two NUCs in my warm room, but my intial question is would it be possible for a single computer to drive, control and stack from two cameras simultaneously?

I will be using a pair of ASI2400 Pros, but will almost certainly need to use different exposure lengths as the individual scopes being tested will have different focal ratios in most cases.

Just to give you an idea of scale of this scope test, I will be working through in excess of 20 scopes ranging from a diminutive Vixen FL55ss up to a TV NP127is and Pentax 125 SDP. My intent is to whittle down this group to 4 or 5 "winners" that cover a wide range of focal length and apertures and will remain mounted and ready to image in my observatory.

I will have lots more questions as I progress, but just wanted your advice on whether a single computer is practical so I can begin setting up my hardware in preparation for the first pairing.

[admin]

Hi,

sounds like an interesting project

You should certainly be able to get that to work on a single NUC PC (at least if your NUC has enough resources). CPU is probably not going to be a big issue in this case as you will be deep sky stacking and it's usual for SharpCap to be spending most of the time waiting for the next frame from the camera in that case. Memory will potentially be a problem as you are using large sensor cameras. My guess would be that 16Gb would be fine for two simultaneous live stacks with ASI2400s, 12Gb probably OK, 8Gb dubious and less than 8Gb very little chance. This will depend to some extent on what additional settings you have in the live stacking config (enhancements like noise reduction or sharpening consume more memory and more CPU resources).

You will need to run two copies of SharpCap, one for each camera. If you want to use dithering then that becomes an issue as there is no way to synchronize the two copies so that copy B knows that copy A is trying to dither.

Hope this helps,

Robin

[Kitfox]

Thank you, so much to think about on this project.

One NUC is 32 gb and the other is 64gb, so it sounds practical. This will be interesting to try.

Your point about dithering has me thinking...couldn't even do that with two NUCs unless I use separate mounts. I guess I'll try without dithering, initially.

I'm not even sure how to distinguish between camera one and camera two, since they are the same model. The more I think about this, the more it boggles my mind! I may lose less sleep if I just run both NUCs in parallel!

[Jean-Francois]

Hello,

If you use 2 identical cameras with SharpCap, it should detect it and shows a serial number (if the camera has one ... in case not ... Robin ??)
That is what I see with my QHY174 cameras. Now the question is which camera is "xyz1234" or "xyz5678" ... that you need to find it by yourself.

Regards,
Jean-Francois

[admin]

Hi,

yes, in most cases you will see the two cameras identified by serial number. In some cases where the cameras do not have serial numbers you will just get 'XYZ Camera #1' and #XYZ Camera #2' - if that happens then which one becomes #1 and which #2 might change if you reboot, but I think for recent ZWO cameras you should get serial numbers.

Dithering two cameras on a single mount would require some sort of synchronization between the frames being captured, or, at the very least, the two imaging applications to wait for each other to both be between frames when a dither is needed. If you are running different length subs on the two cameras as you mentioned, it becomes very interesting... Anyway, SharpCap doesn't handle that situation, so best to stick with no dithering for now!

cheers,

Robin

[archer1960]

Or just accept that you will have to throw out an occasional frame from one camera due to dithering the other camera. If you don't dither very frequently, you won't lose a large percentage of your frames. Maybe every 5 or 10 frames of the longer exposures?

[Kitfox]

I am starting to edge rather strongly towards using two separate NUCS; I will have one just sitting there if not, so I see no advantage to trying to keep everything jibing on one CPU. I will just manage the AP1100/1200/1600 and plate solve from one unit, and drive the second camera from the second NUC.

The dithering issue has me stumped, though. Since I am doing head to head comparisons, I think I can get away with not dithering as the result would impact both imaging trains equally (true or false?).

That said, if I did set up dithering for camera/CPU #1 (since it will be controlling the mount), what would the impact be to camera/CPU #2 as it runs simultaneously? Would a Filter setting for a specific FWHM level kick out the affected exposures on camera #2, given I set dithering on the camera with longer exposures? And would stacking on camera #2 be affected? I am well out of my skill set here, for sure.

Even though the primary purpose of this trial is to whittle down my collection with some level of logic, I would like the output images to be the best they can be, while giving a fair representation of what the optic in question is capable of delivering.

My intent is to develop a workflow that I can live with and standardize across the 20+ scopes. Sky conditions and target may require some adjustment to exposures, but I want to try and standardize as much as practical, from flats and darks all the way through capture. Consistency overrides everything else.

Thank you all so far for the input. I am spending a lot of time in the SC Manual trying to make what I read, and then what I write, make sense. But your patience with me is much appreciated. I really do think a combination of SC captures and DPAC comparisons will make this evaluation successful and fair for all "competitors".

[Jean-Francois]

Hello,

If you want to measure and compare 20+ telescopes, then you need to think about building a laboratory setup.
So something with an artificial star (different LEDs, at least 3 colours) ... so you can repeat the measurements without extern changes (weather, seeing, mount and guiding problems).

It is a lot different optical quality to measure ...
- wavefront ... so PSF ... depends on the wavelength (alternative a Foucault / knife-edge test)
- field curvature ... maximal detector size with homogenous focus
- distortion ... maybe less important
- contrast/straylight/ghost
- focus shift with the temperature
- lens alignment
- transmission
- ...

Regards,
Jean-Francois

[Kitfox]

Hello,

If you want to measure and compare 20+ telescopes, then you need to think about building a laboratory setup.
So something with an artificial star (different LEDs, at least 3 colours) ... so you can repeat the measurements without extern changes (weather, seeing, mount and guiding problems).

It is a lot different optical quality to measure ...
- wavefront ... so PSF ... depends on the wavelength (alternative a Foucault / knife-edge test)
- field curvature ... maximal detector size with homogenous focus
- distortion ... maybe less important
- contrast/straylight/ghost
- focus shift with the temperature
- lens alignment
- transmission
- ...

Regards,
Jean-Francois

I will be DPAC'ing these all on the bench, but since I will be judging these as astrographs, I am equally interested in their ability to render images on an FF camera. And what better way to judge a wide, flat field and understand a scope's ability to produce sharp stars to the corners than by putting a camera on it?

[Jean-Francois]

Hello,

Yes, OK ... a full frame image capture is a good "summary".

But how do you want to discern the lens misalignment and the camera sensor tilt ?
If you are located in a desert with full night constant condition ... then I wish you a lot of fun

What you can do is a through focus scan and then analyse it with CCD-Inspector or similar software.
You can generate a virtual image with no tilt of the sensor.

The next "problem" is the correct back focus of the camera to the corrector/flattner lens, if any is mounted.
From my experience ... it can take a long time until you found the right back focus for your camera.

Regards,
Jean-Francois