which ZWO camera for 18" f/4 Newt ?

[jonstarrysky]

About to pull the trigger on a camera for an Alt-Az driven 18" dob (with Starizona reducer f/3, FL=1375mm) for live EAA viewing using a telescope-mounted Windows 11 tablet (Fusion5 FWIN232 PRO N5-8GB DDR5 RAM, 256GB SSD Storage +512GB micro SD, N100 Quad-Core CPU) to be running 64-bit SharpCap Pro. Maybe remoted to a laptop indoors or HDMI cable to big monitor. Just a matter of which camera to get and to check does the plan adds up technically....Choices of camera were 294MC, 533MC or 2600MC. The aims are very much for simplicity, at least to start out. Folks with similar setups (big dob + ZWO camera) are reporting nice results without the added complication of flats, darks etc (obviously suboptimal but still looking great).

Looking through the SharpCap forum folks are having issues with ZWO cameras. By the looks of it the issues are often to do with required updates and have simple solutions. Or are there real caveats ?? Hopefully the new Windows 11 tablet, with fresh installation of 64-bit SharpCap pro & drivers will avoid issues. Camera feed would go to directly to the USB3 port of the tablet and the camera will have separate 12v DC power. Due to the small pixel size the plan was to do hardware binning at least 2x2 as default (will still be limited by seeing). Maybe 3x3, 4x4 or more for standard viewing. This will ease the pressure on the tablet ?

The 2600MC would be an investment to get both the wider FOV + advanced CMOS sensor. Wide FoV gives access to larger DSOs and facilitates plate solving (Push-to Assistant). And smaller objects would use ROI in SharpCap to increase effective magnification and reduce file size. Possibly ROI would be in use most of the time. But full sensor view available for the bigger objects and Push-to Assistant....

Otherwise with the 294MC non-cooled gives a cheaper solution with wide FoV but older sensor. And 533MC smaller FoV whilst getting the newer sensor technology.... I am assuming there are no real technical downsides to just going for the 2600MC to get both size and new sensor ? Re telescope ergonomics etc, weight is a non-issue (the UTA can handle double this weight).

Any comments/ thoughts appreciated
Jon

[admin]

Hi Jon,

I don't think that there are any significant worries with ZWO cameras regarding issues or reliability - they are by some way the most used cameras with SharpCap, so more people will post about them here just because more people are using them. Like any other brand there will be some bad samples that are either defective initially or go bad over time. Overall though, the ZWO cameras seem to work reliably in SharpCap.

You have pretty much covered all the pros/cons of the 294/533/2600 in your post. 294 is older technology while the 533 and 2600 are newer and similar capabilities with different sensor sizes. Having used all three, I would definitely recommend either the 533 or 2600. The lack of any appreciable fixed pattern noise (amp glow) on these sensors makes them much more forgiving of simple imaging workflows (no darks, hot/cold pixel correction, etc). In particular, having an uncooled 294 sensor will make it harder to dark subtract out amp glow, as the temperature cannot be controlled.

So, in the end I would suggest that the choice comes down to 533 or 2600 based on cost and field of view... The FOV of the 533 will be approx 0.5 degrees square, so should plate solve fairly reliably with the extended SharpSolve database, but the 2600 will give you about 1x0.67 degrees, which opens up more targets as well as making plate solving easier.

The only downside of the 2600 compared to the 533 is that the 2600 is likely to have lower frame rates at the same ROI as the 533, which might have an impact if you try to do planetary imaging. That said the 2600 will hit 28fps at 640x480 in 16 bit mode, so will probably hit 50fps or so in 8 bit mode at that resolution, so not too bad.

cheers,

Robin

[jonstarrysky]

Thanks Robin. Useful the feedback away from the 294, as a hardened visual observer I want the EAA experience to be as direct as possible (not remotely straying into astrophotography territory) and minimizing the need for complexity (flats etc). Good point the volume of use of a camera make will dictate the apparent problems (a camera which no one uses has no reported issues....).

Looking into the binning issue this attracts differing views and confusion (maybe due to a lack of definitive info from ZWO). The pixel size of the 533 or 2600 is several fold lower than the seeing limit on a scope of >1000mm FL. I am cool with the theoretical 2X SNR gain for CMOS 2x2 binning (accepted, not as good as 4x SNR for CCD analogue binning). But from what I read ZWO are recommending against on-camera binning, and there is even the implication that their cameras do this by taking 1 of the pixels from a 2x2 array and literally bin (no pun intended) the other 3 pixels. No averaging or summing up, literally binning (as in trash can). I assume this is incorrect and they'd not consider that lazy approach. In which case on-camera binning sounds like a huge waste (3/4 of the area of the sensor), and maybe the smaller chip (533) or the bigger pixel size (294) is favored if one wants to avoid gigantic files from the 2600 1x1 images....

[oopfan]

Hi Jon,

My opinion of uncooled cameras is that they are great (i.e. inexpensive) when imaging bright objects, however, deep-sky requires longer exposures which necessitates meticulous dark frames management. I've used a 290-based sensor with good results but I found myself having to take dark frames every single session due to ambient temperature differences.

Purchase a cooled camera, that way you can take dark frames once or twice a year from the comfort of your home. Of course, then you'll have to decide on the sensor temperature and exposure. I'd suggest using Robin's Sky Background Calculator here: https://tools.sharpcap.co.uk/ and then plug that number into a simple formula. Let's work through an example:

Bortle Number: 5.0 (use your specific value for your location)
F Ratio: 4.0
Pixel Size: 3.76 (or 4.63um if ASI294)
Quantum Efficiency: 75% (use the average QE across the band, not the peak)
Type: Color (you said you want a color camera)
Selected Filter: Luminance
OUTPUT: Sky Electron Rate: 2.66 e/pixel/s (or 4.03 e/pixel/s if ASI294)

Then plug the Sky Electron Rate into this simple formula:
Minimum exposure = 10 * (Read Noise * Read Noise / Sky Electron Rate)

ASI2600: pixel size 3.76um, read noise 1.4e-, therefore minimum exposure = 7.4 seconds
ASI533: pixel size 3.76um, read noise 1.5e-, therefore minimum exposure = 8.5 seconds
ASI294: pixel size 4.63um, read noise 1.8e-, therefore minimum exposure = 8.0 seconds

(Please check my math, and use your own Bortle class. If your sky is brighter than Bortle 5 then the minimum exposure will be shorter.)

Which camera you purchase is up to you. Here is another calculator for determining Field Of View:
https://astronomy.tools/calculators/field_of_view/
Be sure to click "Imaging Mode"

Brian

[oopfan]

Jon,

One thing I did notice with the ASI2600 and ASI533 is that you will be over-sampling. This means that your pixel size is too small compared to your focal length. There are to ways to fix that: use a reducer or increase the sensor's bin size. The most cost-effective solution is to increase the bin size. Watch what it does to your Minimum Exposure:

Your pixel size goes from 3.76um (bin 1) to 7.52um (bin 2). Plug that number into the Sky Electron Rate calculator. Your value goes from 2.66 e/pixel/s to 10.64 e/pixel/s. Next, plug it into the Minimum Exposure formula. Now, your exposure goes from 7.4 seconds to 1.8 seconds. Magic!

Here is another useful calculator:
https://astronomy.tools/calculators/ccd_suitability

Here is a graphical depiction of how suitable the ASI2600 at bin 1 is for your f/4 scope. You are over-sampling, meaning that the image is less bright than it could be:

Screenshot 2024-06-07 at 10.54.33 AM.png (113.93 KiB) Viewed 3319 times

And here it is at bin 2:

Screenshot 2024-06-07 at 10.55.02 AM.png (93 KiB) Viewed 3319 times

Since you are visual observer, you should be happy with 1.8 seconds.

Brian

[jonstarrysky]

HI Brian, thx for that. I have the Nexus 0.75 reducer so will be working at f/3 and effective FL=1375mm. Yes binning seems very relevant regarding the oversampling especially for the 533 or 2600. The plan was to select bin2 (or more) in Sharpcap for live stacking. I don't plan any post processing work, this is EAA not astrophotography....

However, binning seems controversial for ZWO cameras. A simple additive or averaging binning approach on-camera (happening after A>D conversion for a CMOS sensor) would improve S/N by 2 fold (vs. 4-fold for CCD). Both of these sound good. But there are suggestions that ZWO does the binning in a way that essentially discards pixels (?!)

[oopfan]

Jon,

The negative aspect of using a reducer is that it introduces another optical element. They are not perfect unless you pay big bucks for one.

True, you get a higher SNR with a CCD. Regardless of what the chattering class says about ZWO, I would try it at bin 2 and see what you think. Another advantage is the size of the files are 1/4 of bin 1 size.

Brian

[admin]

Hi,

I haven't seen any suggestion of ZWO binning not being proper binning, so I have obviously missed that discussion. Do you have a link to the source for that? I'd be very surprised as it would surely have been easy to prove/disprove - It's something that is testable using a slightly modified version of SharpCap's sensor analysis, or other sensor analysis routines. Other brands clearly advertise either average or sum binning (or perhaps a choice), so it would be odd for ZWO to not do the same.

On the flip side, it is possible to run the cameras in a 'use 1 pixel out of 4 mode', which can be useful sometimes to get higher frame rates at lower resolutions without sacrificing FOV. It's also possible that it may depend on other settings like 'Hardware binning' or on the model of sensor in the camera.

cheers,

Robin

[oopfan]

Jon,

Right, if CMOS behaved like CCD then you'd pay the price of only one unit of Read Noise per 4 pixels. Unfortunately that's not the case with CMOS. So the Minimum Exposure stays the same at 7.4 seconds regardless of bin 1 or bin 2. Sadly. Perhaps we should look for CMOS cameras with larger pixels and small Read Noise. There is only one that stands out:

ASI2400MC Pro: pixel size 5.94um, read noise 1.5e-, only $2,999, minimum exposure 3.2 seconds
I even looked at uncooled cameras.

I think that the ASI2600MC looks to be the better value. Pray that you've got more LP than Bortle 5

PS: The reducer looks like a good option now.

Brian

[jonstarrysky]

Brian thx for the advice, and calculations. Useful that a defined exposure time can be calculated, so basically I'd be aiming for around 8 sec subs. The AltAz dob mount will allow this, Servo-CAT tracks in an eyepiece field for 10s of minutes, but occasionally there are movements. Hopefully Sharpcap pro can somehow automaticaly regect the lower quality subs. Yes looks like the ideal camera for such a long FL scope doesnt really exist, the market is aligned to the more ideal shorter focal lengths for imaging. Likely I will get the 2600MC (or 533MC) and use bin 2 as you say.

Robin, The comments about ZWO binning method, I read a few posts, but nothing definitive, good to know ZWO binning is likely OK, its not a widely known issue then....