Canon EOS 500D for astrophotography?

[Vebjorn]

The title says it all to be honest. I'm thinking of buying Canon EOS 500D for AP But i cannot find any article or review on how the camera performs for this kind of usage. Can anyone assist me? As i am a rookie when it comes to astrophotography.

I have taken a lot of a-focal pictures though, but i have no experience with AP on a GoTo mount with the camera attached to my telescope! Well, back to the subject. If you have used the 500D for ap, I't would be great if you could "enlighten" me as i intend to buy this camera.

Sincerely,

Vebjorn

[MishMich]

I use a Canon EOS1000D. The 500D is similar, but has more pixels so is capable of higher resolution and can capture movies. The 1000D only ranges between 100 & 1600 ISO, so if the 500D supports ISO 3200 that might be useful - however, noise does increase with higher ISO numbers, so I am trying to stick to ISO 800 or lower if I can. The 1000D Seems fine for DSOs. S@N reviewed these a while back, and rated the 1000D as being preferable to the 450D (also higher resolution) - but you'd have to read the article as to the reviewer's reasoning, and decide whether the same logic applies to the 500D; essentially they suggested it wasn't worth the extra for the 450D. I have used the 1000 with a movie hack that captures the video stream to PC, and that can produce video for processing in Registax, and my initial results were OK. I've not seen/heard yet what the product of the 500 is like in native video mode.

M.

[yesyes]

I haven't got a DSLR yet but I read about the subject to narrow down the choices.

At the moment the EOS 500D would be my favourite, then 450D and 1000D. The 500D has the advantage of much less noise and the ability to record video, which can then be stacked in Registax for example.

For the noise comparison, see this thread:

http://stargazerslounge.com/imaging-discussion/92512-noise-comparison-canon-dslrs.html

I would buy a 500D here and now, if only the budget (pronounced: wife) would allow it.

[Matthew.Blake]

Wife and I bought a 500D for Christmas. So far I've photographed the Moon and it's been lovely. I've had a go at M42 for fun but can't figure out how to get Registax 5 to load the RAW file correctly, I'm not sure if it's Registax or the camera yet.

[MishMich]

Try Deep Sky Stacker for M42.

M.

[Takahashi]

I've been trying DSS with RAWs from my 500D, but do I need to de-bayer the image first, and if so, how so? Every time I've stacked a bunch of RAWs they've lost all colour, and don't even look as good as a single JPG frame, LP or sensor-noise notwithstanding.

[MishMich]

I just load the RAW file into DSS. There are two options you can select, one for if the image appears too grey, and the other for if it is hard to fix colour in post-processing. TBH, I have tried the DSS post-processing, but I don't like it, and what is saved seems much the same as the initial stack. What I get in DSS looks poor - kind of washed out. When I save this and then open it in Gimp or PSE7 it looks very different - dark with little detail. However, the detail is there, and processing in either package can bring it out.

I've tried processing DSOs in Registax, but can never get them to line up. I did find my JPGs looked better than my RAWs, but that is because JPEG applies stuff to make the image look visually OK - with RAW you can get to do all that yourself. I started using DSS stacking my JPGs, because the immediate result 'looks' better - but if you look closely there is not the clarity/cleanness/definition in stacked JPGs than RAWs. I tend to use RAW files in my 'normal' photographs, because you can change many of the camera settings after the image has been taken - particularly useful with nigh-time shots, or snow-shots.

M.

[Matthew.Blake]

I've been trying DSS with RAWs from my 500D, but do I need to de-bayer the image first, and if so, how so? Every time I've stacked a bunch of RAWs they've lost all colour, and don't even look as good as a single JPG frame, LP or sensor-noise notwithstanding.

That's exactly the problem I have. I just get red everything in Registax. It displays them properly when previewing them in the select screen though!

[MishMich]

I decided to have a look and see if I could figure out what is going on here.

I stacked all my JPEGs of M42 from over the past month in DSS, and TBH, they look better than my RAWs.

Before I had tried stacking RAWs & darks from one night (5x3, 5x2, 5x1 & 5x30 seconds), as well as two nights (loadsasubs), and even all the RAWs taken over the past month - and the results were little better than one night's worth of 10 - 30 second subs, and didn't appear as good as the above.

Then I tried stacking 5 x 3 minute subs & 2 x 3 minute darks - wouldn't stack at all. So I tried 5x2 minute subs and associated darks, which stacked OK, but was pretty much the same as the stack for the whole night, except for one thing - the stack for the whole night was not so highly exposed around the trapezium, showing some detail, while the set of 2 minute subs was blown out in that area, and no tweaking could alter that.

Then I tried Registax on one night's worth - the RAW files all loaded correctly (not red), but when I went to stack them, only about 8 out of the 20 were selected, and they didn't align properly. This is why I avoid Registax for stellar/deep space work - I've never found any way of getting it to align using stars in the way DSS does.

However, when I look at the stacks of the RAW files from the latest night's imaging, they do show better resolution when viewing at actual pixel size. So, whilst there is not the wealth of nebulosity apparent compared to the JPEG above, there is more finer detail in the nebulosity that is there - it is has higher definition.

The feeling I get is that in stacking long and short exposures, the method of stacking does not work in favour of the long exposures, but diminishes them in relation to the shorter exposures. Similarly, if a significant number of the longer exposures are being rejected (the processing doesn't make this clear if this is what is happening), then that could explain why a sequence of 20 images ranging from 30 seconds to 3 minutes looks 'darker' and less detailed than a sequence of 5 images at 2 minutes.

I can't help out with the question about why everything is red in Registax, Matthew, as I've not had this problem, and when I try loading RAW files and stack them I don't see this happening. Sorry.

M.

[Matthew.Blake]

Well some semi good news there. I've found some of the software that comes with the camera can batch convert my RAWs to 16bit/ch TIFFs which registax loads correctly. When I get home tonight (only found it just before I went to bed last night and being a lazy bum I didn't stay up late) I'll be registaxing away to see what I can get. I also think DSS had been interpreting the RAWs incorrectly as well as when I went looked at the image properties in DSS it thought they were 16bit grey scale not the 14bit/ch the 500D RAWs really are!

[MishMich]

There is a place for setting RAW file properties under Options/Settings/Raw...

I checked that bilinear interpolation was checked, and selected .Use Camera White Balance'.

Star detection threshold I set to 5%

Under Stacking Settings:

Result tab:

Mosaic Mode

Align RGB Channels in final image

Light:

Kappa-Sigma clipping (Kappa=2, Iterations=5)

RGB Channels Background Calibration

Dark:

Median

Hot Pixel detection and removal

Dark optimization

Alignment:

Automatic

Intermediate and Final Format:

TIFF

Cosmetic:

Detect and clean remaining hot pixels (set to 1 pixel)

Detect and clean remaining cold pixels (set to 1 pixel)

Replace with the median

Output:

(unchanged from default).

The properties do list Depth as 'Gray 16 bit'

It appears that only 18 of appx 48 frames were stacked because DSS didn't detect amy stars in at least ten of them, and very few stars in the others. Presumably these were those most affected by tracking error.

M.