Best DSLR for astro modification?

[Nikodemuzz]

I'm considering a DSLR for an interim camera before eventually investing in a (mono) CCD. The idea is to modify the camera, and perform the modification myself. While I'm quite confident that I have the tools and skills to perform the modification with the help of online documentation, I would rather not do it on a new, expensive camera. There are, however, a lot of different models which would be reasonably priced on the used market. I can pretty well judge how those cameras would work in the normal photography scenario, but in astro imaging with modifications? Not so much. Are there some models that are particularly well suited for this function?

I was looking at used Canon models that could be had with reasonable prices used (Canon because I don't know anything about Nikon cameras, and I already have EOS adapters for my scope. I could still consider Nikon if there is a benefit to it):

- 600D

- 2000D

- 60D

- 5D Mk2

- 6D (prices start to get high, but I could consider it if there is enough return in terms of quality)

Older models than these are obviously more affordable, and I would be happy to go with one of those if they offer better value.  The way I see it, the main differences will be sensor performance, modification difficulty (if there are differences) and price. The cameras will be used via laptop so I don't see much in terms of features that could make a big difference. In the models above there are both full frame and APS-C models. I could work with either field of view. Primarily my imaging will be with the WO SpaceCat, in the future perhaps also with a larger refractor. I might also have a go with my C11 + reducer, but that is of secondary importance.

Any thoughts?

[Spaced Out]

For info I read somewhere that 6Ds can be a bit tricky to mod (easier to mess up), so I asked Juan at cheap astrophotography if he would do mine but he was reluctant. I did get it done in the end by Andy at Astronomiser, but I'm glad I didn't risk trying it myself !

A lot of people using 450Ds and 600Ds it seems.   

[Nikodemuzz]

Thanks for the heads up regarding the 6D! Gary Honis has the instructions for modifying it, but of course that doesn't mean that there would'nt be mines in that field... Used 6D's are expensive enough to cause sadness if they are turned to paperweights by a ham-fisted hobbyist. 

Edited February 18, 2020 by Nikodemuzz
Typo

[michael8554]

I did a 6DMk1  myself, but that was after practice on 5 or so cheaper models (450D through to 600D) over the years, with a digital depth gauge rig.

Very sensitive camera if the large pixels suit your focal length.

600D and newer cameras have Floating Sensors.  Tricky to get the sensor back without tilt, good luck with Gary's method, best left to the commercial modders.

Practice on a 450D, you can always sell it on if you're successful !

Michael

 

[Spaced Out]

1 hour ago, Nikodemuzz said:

- 6D (prices start to get high, but I could consider it if there is enough return in terms of quality)

 

6D really is a great camera. I've had a lot of Canon DSLRs but for certainly for single shot nightscapes, wide angle landscapey milky way type stuff I always default to using my old 6D just for its image quality and low light performance. I've had it modified recently so I am looking forward to picking up a bit of nebulosity in my images now too and maybe chucking it on a mount for widefield stuff. 👍

[david_taurus83]

I've found my 6D to be more sensitive than expected. I dare say even more sensitive than the ASI1600 mono I had. Seems to pick up a lot of faint signal. Pixels are almost twice as big mind.

[Alien 13]

There are a few Canon cameras that have plenty of Ha response out of the box.

Alan

[Nikodemuzz]

8 hours ago, david_taurus83 said:

I've found my 6D to be more sensitive than expected. I dare say even more sensitive than the ASI1600 mono I had. Seems to pick up a lot of faint signal. Pixels are almost twice as big mind.

You mean sensitivity in general? I would expect the Ha sensitivity to be behind a dedicated astro camera like that, especially without modification. Still, interesting to know! Not all targets are that dependent on Ha.

[Nikodemuzz]

8 hours ago, Alien 13 said:

There are a few Canon cameras that have plenty of Ha response out of the box.

Alan

Which models in particular? No need for modification would be a welcome prospect. Actually, my Fuji X-T3 should be a pretty good performer in that regard. Unfortunately the thorough lack of support from any astro utilities makes any attempts at proper deep sky photography a very laborious process.  Which is why I am now looking at an additional camera to do that job.

[happy-kat]

@The Admiral uses a fuji and I guess has some work arounds.

[Nikodemuzz]

9 hours ago, michael8554 said:

I did a 6DMk1  myself, but that was after practice on 5 or so cheaper models (450D through to 600D) over the years, with a digital depth gauge rig.

Very sensitive camera if the large pixels suit your focal length.

600D and newer cameras have Floating Sensors.  Tricky to get the sensor back without tilt, good luck with Gary's method, best left to the commercial modders.

Practice on a 450D, you can always sell it on if you're successful !

Michael

 

Thanks Michael, it is always useful to hear from others who have walked the path. 😃

The sensor alignment process did jump out as a potential source of headache when I read the modification instructions. The methods of recording the original sensor position that Gary mentions don't seem very precise. Especially if one would be using fast optics, the alignment probably has to be correct within some tens of microns or less. Shouldn't be a problem if I can measure the correct distances with that much accuracy, but I'll have to consider how I can make that happen.

[michael8554]

2 minutes ago, Nikodemuzz said:

The methods of recording the original sensor position that Gary mentions don't seem very precise.

The positioning only has to be accurate to one turn of the screw, might be obvious with Gary's method when the screw is one turn too loose or too tight, I wouldn't risk it. 

Michael 

[Nikodemuzz]

2 minutes ago, michael8554 said:

The positioning only has to be accurate to one turn of the screw, might be obvious with Gary's method when the screw is one turn too loose or too tight, I wouldn't risk it. 

Michael 

Are you sure? One full turn seems like an enormous room for error, although the thread may be very fine. I was thinking that I could measure the screw positions with a digital caliper, but thought it couldn't be trusted to be precise enough.

[Nikodemuzz]

19 minutes ago, happy-kat said:

@The Admiral uses a fuji and I guess has some work arounds.

I would love to hear more about that! Thanks!

[AbsolutelyN]

I recently had my 6d modified. The weather has not been very co-operative but I managed this test image last week - 8 one min exposures through an Esprit 100 @ 1600 in Bortle 5. I was quite impressed for just 8 mins of exposure and looking forward to trying more integration time. Really bought it for travel but it may get used more than I thought. Biggest issue is calibration, you can't easily calibrate like you can with a set point cooled camera.  Please ignore the dust spot.   

Edited February 19, 2020 by AbsolutelyN

[Nikodemuzz]

18 minutes ago, AbsolutelyN said:

I recently had my 6d modified. The weather has not been very co-operative but I managed this test image last week - 8 one min exposures through an Esprit 100 @ 1600 in Bortle 5. I was quite impressed for just 8 mins of exposure and looking forward to trying more integration time. Really bought it for travel but it may get used more than I thought. Biggest issue is calibration, you can't easily calibrate like you can with a set point cooled camera.  Please ignore the dust spot.   

Thanks for the message! Very impressive result for such a short integration time, congratulations! I'm also in a Bortle 5 area, so this is quite informative. Where did you have the modification made, if I may ask?

About calibration, I assume you mean dark frames? I have been under the impression that with the uncooled cameras there might not be much benefit from using dark frames. Dithering should help, if I'm not mistaken (a manual process with the X-T3...)

[AbsolutelyN]

It was modified by Astronomizer - http://www.astronomiser.co.uk/. Fantastic service, he quotes up to a few weeks for the mod but it was sent back modified within days. 

I don't know enough about DSLR calibration at the moment, perhaps others can chime in? The image above had darks, bias and flats applied. I might re-process without at somepoint and see how that looks.  

Here is a much less processed version of same image that does not have noise reduction which shows how noisy the image is. Again this is only 8 mins, I'd hope much longer integration will do much better.

[david_taurus83]

2 hours ago, Nikodemuzz said:

You mean sensitivity in general? I would expect the Ha sensitivity to be behind a dedicated astro camera like that, especially without modification. Still, interesting to know! Not all targets are that dependent on Ha.

I should have said, I bought mine modded. The 6D is supposedly 49% QE before you remove the filter so after modding it will be fairly close to the ASI1600 in terms of photon to electron conversion. I'm using the same focal length as I was with the ASI but I've gone from 2.4" per pixel to 4" per pixel so a lot more light landing on each pixel will add to its intensity value. The ASI is excellent with an Ha filter, always gave impressive results but I always struggled with faint non emission stuff. The 6D seems to fare better for me in this regard. See below, a work in progress as I'm trying to learn to process in photoshop. The colour is all wrong but theres loads of dusty stuff available. Stack of 240 x 60s at ISO1600. Scope is an Altair 70EDQ-R.

[michael8554]

3 hours ago, Nikodemuzz said:

Are you sure? One full turn seems like an enormous room for error, although the thread may be very fine. I was thinking that I could measure the screw positions with a digital caliper, but thought it couldn't be trusted to be precise enough.

Precisely! That's why one turn too tight, spot on, or one turn too loose, might be decernable with the cardboard height gauges. 

(You make adjacent Sharpie marks on the screw heads and the chassis before removing) 

The screws are M1.6 - 0.5.

Michael 

2 hours ago, AbsolutelyN said:

The image above had darks, bias and flats applied.

You might try Dithering 12 DSLR pixels, no Darks, use the Bias as Darks. 

Michael 

[AbsolutelyN]

1 minute ago, michael8554 said:

You might try Dithering 12 DSLR pixels, no Darks, use the Bias as Darks. 

Michael 

It was dithered too .... set to medium in SGP. Not sure how many pixels that is but will reprocess as you say as an experiment. Thanks  

[Nikodemuzz]

9 minutes ago, michael8554 said:

Precisely! That's why one turn too tight, spot on, or one turn too loose, might be decernable with the cardboard height gauges. 

(You make adjacent Sharpie marks on the screw heads and the chassis before removing) 

The screws are M1.6 - 0.5.

Michael

Thank you, Michael! That certainly increases my confidence level about my ability to complete the mod. 

[Nikodemuzz]

1 hour ago, david_taurus83 said:

I should have said, I bought mine modded. The 6D is supposedly 49% QE before you remove the filter so after modding it will be fairly close to the ASI1600 in terms of photon to electron conversion. I'm using the same focal length as I was with the ASI but I've gone from 2.4" per pixel to 4" per pixel so a lot more light landing on each pixel will add to its intensity value. The ASI is excellent with an Ha filter, always gave impressive results but I always struggled with faint non emission stuff. The 6D seems to fare better for me in this regard. See below, a work in progress as I'm trying to learn to process in photoshop. The colour is all wrong but theres loads of dusty stuff available. Stack of 240 x 60s at ISO1600. Scope is an Altair 70EDQ-R.

 

Thanks for sharing. I can certainly see what you mean about being able to pick up faint stuff! How dark is your sky, and did you use filters?

[Nikodemuzz]

13 hours ago, Alien 13 said:

There are a few Canon cameras that have plenty of Ha response out of the box.

Alan

I found this article (which is no doubt familiar to many of you) :https://clarkvision.com/articles/do_you_need_a_modified_camera_for_astrophotography/

He mentions the same thing as you, Alan, about the Ha responses of stock cameras, and there is also a further link that contains transmission curves for a bunch of cameras: http://kolarivision.com/articles/internal-cut-filter-transmission/

The differing responsiveness to Ha is quite interesting and significant when choosing camera model, but I also found the writer's argument about post processing very intriguing. Do you guys agree with him? If he is correct, then what is the appropriate processing method that does not suppress as he describes?

This is quite interesting, there seem to be arguments both for and against DSLR modification, with compelling photographic evidence backing either side. 😃 

I certainly wouldn't mind if the modification wasn't necessary at all. It would also make me all the more frustrated about the lack of support for my Fuji. 😃

[AbsolutelyN]

21 minutes ago, AbsolutelyN said:

It was dithered too .... set to medium in SGP. Not sure how many pixels that is but will reprocess as you say as an experiment. Thanks  

Thanks Michael .... just ran it through BPP with bias as darks and have to say I can't tell any different to the one with matching darks. 

[Nikodemuzz]

5 minutes ago, AbsolutelyN said:

Thanks Michael .... just ran it through BPP with bias as darks and have to say I can't tell any different to the one with matching darks. 

 

Looking quickly it seems that there is less color noise in the latter, "darkless" image, which is interesting. Hot pixels are still there in both images, should they not have been removed during the stacking/calibration? Well, I am by no means an expert here, I'm sure the more experienced users will be able to help.