IMX585 sensor with long focal length (>2000mm) - Can it work?

A guy local to me has just recently started imaging primarily with a QHY Minicam8 and a Celestron EdgeHD 925… His results are stunning, especially on nights of good/great seeing!

There is a guy in our local club who has that exact combination and he is very happy with it despite that the seeing here is below global standards.

I agree with Tobiasz that this would be a combination for those with an outstanding mount and experience. Otherwise you will be setting yourself up for a tough time.

A C8 at native focal length would require a mount providing you with stable 0.29” RMS in order to avoid bloating and loss of resolution (with the pixel size of the 585). With the reducer you could get away with 0.42” and if you then bin in post you could live with 0.84” which is entirely doable. Maybe this works better with narrowband than it does with broadband though, I’m not sure.

I use my C8 without a reducer and I’ve used two mounts for this. Last season I used my Juwei 17 which didn’t stand a chance really. I got some good images but I could never get the 0.38” RMS needed to sustain full resolution, so I had to use bin2 in post as a minimum. With my AM5 this season things have fared better, but still not doable to get that 0.38” mark on a regular basis - so I’ve still needed to resample several most images in post in order to get satisfying results. The IMX533/571 has faily larger pixels than the IMX585 so in theory it should be a better/easier combo - but it’s still not the easiest.

People are different though, so what I find unacceptable might be what someone else find very acceptable.

If you want good detail (not bloating and loss of resolution) you could get away with the reducer and binning in post, as long as you are comfortable with that limitation. Native focal length would be a huge stretch IMO, unless you have excellent seeing and an excellent mount, not an AM5, AM7 or the likes of it - They are great, but not for this kind of tracking IMO. No amount of Russell Croman’s tools will make bloated images look good. An observatory grade mount however will make imaging like this effortless.

That being said, if you own the equipment already just give it a go and see if the results are enough to make you happy - which is the main thing anyways.

Hi,

0.41”/pix might be considered as oversampled in textbooks but the fact is that deconvolution programmes such as BlurXT better work with (slightly) oversampled data. Even if you seeing is average (let’s say about 2”) you should keep the bin1. I used the 585 sensor on a Edge with focus reducer (around f/7) and it worked fine :

https://www.astrobin.com/adr2vi/C/

I really d’ont see a problem with that, some people here on AB shoot with 0.27”/pix and this can still be interesting but needs of course decent sky conditions.

Kartik Atre · Apr 20, 2026 at 03:13 AM

Hi All,

Not sure if this has been discussed elsewhere. I am thinking of experimenting with using an IMX585 sensor camera with my Celestron C8 at native (or even 0.7x reduced) focal length.

I understand this is grossly oversampled. Has anyone tried this? Or with another OTA of similar or longer focal length?

For context - this will be deep sky work.

Thoughts?

Thanks, Cheers and, CS

Kartik

The IMX585 sensor has 2.9 micron square pixels. If you are in a region with 1” seeing, the optimum pixel size for your C8 is around 4.6 microns so you’d be a bit over sampled. That means that you’d be operating with a lower SNR, which you could make up by simply using more exposure time. On the other hand, if your local seeing is closer to a more common 2”, the optimum pixel size would be 7.6 microns. In that case, you would still be slightly oversampled by using 2×2 binning but you’d pick up a factor of 2 in SNR. You can always oversample but you should understand that the penalty is in the SNR that you achieve and oversampling does not produce a sharper image. It is true that BXT does a bit better with a slightly oversampled image but it’s not a good idea to use that fact to justify extreme oversampling. SNR is ultimately more important than sampling when it comes to producing a high quality image. A good middle ground is to slightly oversample and to make up for the loss in SNR with a bit more exposure time. I think that with your telescope, you are going to find that you’ll get the best performance by binning 2×2 with that sensor.

Here are the charts showing the optimum sampling rate based on MTF analysis that I did a few years ago.

📷 image.png

John