Canon or Sony or Nikon

[MaJiC79]

I've been doing my research and I'm wondering if there are any practical differences between Sony, Canon and Nikon DSLR's/mirrorless cameras?

Canon looks like the well travelled path for IR filter removal, with Nikon coming next, but I don't see so much with the Sony equipment - yet it's Sony sensors in use elsewhere.

[MaJiC79]

Well, I think I can answer my own question. Canon and Nikon have some entry level equipment, but Canon has a much greater range of entry level stuff. Sony seems to start a bit higher up the range, with all three brands competing for the top spaces. As I'll be tinkering, there's a certain amount of risk involved, taking out filters etc, and that lends itself to the cheaper end of the market. So Canon wins out for me

[ollypenrice]

The case for starting with a daytime camera for astrophotography has never been weaker than it is today. Just be sure you really want to go that way rather than for a cooled CMOS astro camera. Dual use may be very important to you but, if it isn't, CMOS astro cameras are more affordable than ever.

Olly

[ONIKKINEN]

If technical details interest you, have a look at this site: https://www.photonstophotos.net/. You can filter the cameras you want to compare in the various tests done there. But generally Canon cameras are quite weak from a raw sensor specs standpoint and Nikon cameras are better. Sony cameras are also good but here you have whats called a "star eater problem" where there is noise removal in the raw frame that can remove small stars from the image. That feature cannot be removed so some research should take place before buying a sony camera.

Also, have to agree with Olly above. Dedicated astro cameras are MUCH better than DSLRs and the price differences are getting smaller by the day. I would say that get the DSLR if its less than 500e (preferably much less) but otherwise dont bother and jump straight to the dedicated astro camera train.

[MaJiC79]

I'm looking at the used market, so I can trim my budget back a bit (Christmas coming up, I need to make sure the little kids are taken care of before this big kid), and I'm no stranger to fixing things (usually because I broke them) and making tweaks.

Actually, I want both, but I want to try a mirrorless camera (canon m10/m100/m200 as these are reasonably priced used, for quite large sensors) before I look at a dedicated camera (probably mono)

[900SL]

If a DSLR, I've had excellent results with a cropped sensor Nikon. The D5600 has a tilt screen (essential), a ISO invariant sensor, and pretty good dark noise/low light performance.

You can find them used no problem.  I have full frame and cropped sensor and there are advantages shooting cropped (narrower FOV, less undersampled, lenses are cheaper, less demanding on image circle and field flattening etc..

[MaJiC79]

I'm leaning more towards mirrorless right now than DSLR, but I'll take a look at the cropped sensors, thanks for the info

[MaJiC79]

7 hours ago, 900SL said:

If a DSLR, I've had excellent results with a cropped sensor Nikon. The D5600 has a tilt screen (essential), a ISO invariant sensor, and pretty good dark noise/low light performance.

You can find them used no problem.  I have full frame and cropped sensor and there are advantages shooting cropped (narrower FOV, less undersampled, lenses are cheaper, less demanding on image circle and field flattening etc..

So as far as I can tell, the APS-C sized sensor, is a "Cropped" sensor - is that what you mean? and in reality, all "Full sized" lens means that you get a cropped image?

You can tell I'm not a photographer, as I'm learning a lot more here about the history of these things!

[nfotis]

2 hours ago, MaJiC79 said:

So as far as I can tell, the APS-C sized sensor, is a "Cropped" sensor - is that what you mean? and in reality, all "Full sized" lens means that you get a cropped image?

You can tell I'm not a photographer, as I'm learning a lot more here about the history of these things!

This link should help comparing the framing of a cropped (APS-C) sensor like a Canon 60D and a full frame sensor like the Canon 6D (both using the same scope):

https://astronomy.tools/calculators/field_of_view/?fov[]=310||90||1|1|0&fov[]=310||88||1|1|0&messier=31

In general, cropped sensors mean cheaper cameras and a more tight framing.

N.F.

[MaJiC79]

Yes, it's what I've discovered these past few days. I just don't get @900SL's comment about cheaper lenses (although, if I use it for daytime photography, then it applies).

Really I was trying to understand the practical differences of the major brands with entry into astrography - acquisition (breadth of market/models), tooling (software specific to the brands), other technical reasons to prefer one over the others

I've just discovered (new to me, surely not new to some others) a piece of software called MagicLantern that unlocks a host of features on some Canon models - and if I can find a suitable mirrorless and get it working on that, then maybe that's the low cost route (not low effort, but I don't mind spending time on a hobby)

[nfotis]

The remark about cheaper lenses holds, because a lens destined for a cropped sensor (EF-S mount on Canon) can be built with smaller lens elements, hence cheaper.

On telescopes, you just get a larger magnification from the same tube (like getting a 1.6x Barlow, without losing optical speed)

N.F.

 

[happy-kat]

I think magic lantern may only be usable on the original canon m mirrorless, though I haven't researched recently to see if later models are supported. Canon were better supported with various astro software then the others though the third party software support has got better than it was I gather on the others. Sony and Nikon both had star eating issues and longer exposure length limitations linked to it, you'll have to research to see how much this applies now but if buying second hand it would be worth knowing what it means for the models looked at. 

Edited November 30, 2021 by happy-kat
Removed '1' as misleading

[Budding Star Gazer]

I did quite a lot of reading myself before purchasing a camera, i ended up going for a canon EOS 6D, it seems to be a relatively good starting point that didn't break the bank (second hand £350) from WEX.

[ollypenrice]

Be very careful over this 'cropped sensor' business. It produces more nonsense per square inch than any other topic and, in astrophotography, it is best avoided.

In astrophotography the 'real numbers' from a camera are pixel size and chip size. 'Resolution' does not arise from the camera alone and certainly has nothing whatever to do with sensor size or pixel count. Resolution refers to the amount of detail which can be resolved, in theory, by the scope-camera system and is measured in arcseconds per pixel.  This is the key term and is never mentioned, to my knowledge, in the daytime photography world. It defines how much sky lands on each pixel. The more sky per pixel, the less the detail in the image - except that the 'seeing' (the blurring effect of the atmosphere) limits the real resolution to something between 1 arcsecond per pixel and 2 arcseconds per pixel (or worse) depending on location and air conditions.

The term 'crop factor' implies that a smaller sensor 'zooms you in' to a target - but it doesn't. It just captures less sky around it. What matters is pixel size (not pixel count) because that tells you how many pixels you put under the scope's projected image of your target. If you put more pixels under it you'll get a bigger and more resolved image provided the seeing, guiding and focus allow it. They often don't! 

Very few telescopes can cover a full frame (35mm) camera and the few that can are easily recognized by their price!

Olly

[wxsatuser]

29 minutes ago, ollypenrice said:

 

Very few telescopes can cover a full frame (35mm) camera and the few that can are easily recognized by their price!

Olly

There are a very few camera lenses that can do it as well and they are also noticeable by their price.

[MaJiC79]

That's been very good discussion, thank you, it sheds a lot of light (pun not intended) on the matter. I've seen mentions of looking for 1-2 arcsecond resolution in the system, and I've wondered the why of it - and I take it that if I aim for something larger, I'm more likely to have "success" but at the price of a lower detail image. Whereas a higher detail image will require improved performance in the tracking and system, in order to deliver the required performance, and all subject to the conditions on the night.

From what I can tell, MagicLantern is currently only supported on a specific firmware for the Canon EOS M mirrorless, and the M1, M10, M100, M200 are iterations of the same design. I've got some experience with firmware (from small PIC16 devices, up to ARM Cortex 9 MCU) and sufficient motivation that I will at least have a look to see if it's something I can contribute to if I can find a suitably cheap mirrorless to play with. I'd rather go for something a bit newer, but I'll keep my eyes peeled for an opportunity.

I'm pretty much committed to Canon now, mainly because of that third party support and better astro community support - and as I'm not currently invested in another ecosystem, it's a no brainer

[900SL]

Looks like you have it worked out

For beginners I'd not get too hung up on getting the right pixel scale. I generally shoot undersampled (ie shorter focal length and DSLR sensor pixel size) and still get photos that I am happy with. Look in the Samyang 135mm thread by way of example. For widefield this is fine to start with and you can work on the learning curve without needing to get polar alignment, tracking and guiding spot on. I travel to dark sky sites so this suits my modus operandi. 

Learn the ropes, learn how to process, get out there and capture data, and work your way up the ladder. 

I still find cropped sensor a useful term even if it's not technically correct. I have two Nikons, one full frame and one CMOS (cropped). Using the CMOS gives me a 'zoom' by reducing the FOV

D610    Pixels 5.95 micron  say with GT71 and 0.8 reducer so 320mm FL for pixel scale: 3.9 arc sec/pixel (undersampled but wide FOV)  

D5600  pixels 3.89 micron ditto for pixel scale:  2.5 arc sec/pixel 

Bigger pixels capture more signal and usually have a better dynamic range, but lose resolution if pixel peeping. And on full frame there are more demands placed on the lens or scope to deliver a flat image field that is sharp and without major aberration into the corners 

https://skyandtelescope.org/astronomy-blogs/imaging-foundations-richard-wright/astrophotography-pick-your-pixels/

 

 

 

Edited November 30, 2021 by 900SL