ISO on a EOS 5D MkII

[Swiso]

Hello everyone,

I am new to this site and a novice in regard to photography, right now I am doing a photography course (here at our local photo club) that I like it very much !

I own a Canon EOS 5D MkII and I just purchased two lenses...the Sigma Art 35mm f/1.4 and the Tamron SP15-30 f/2.8 zoom.

This summer, me and my wife, we plan to go to Azores for a couple of weeks of vacation and I would love to have a shot at taking Milky Way pictures.

I would like to ask Canon owners here about how high you suggest I can go with ISO settings, based on your experience...

I know that actual models (MkIII and MkIV) can easily go up to very high ISO and they could use that pictures right away without much work...

Since I never tried night sky photography pictures, and since my camera is already 7 years old ( purchased in the 2011) I would like some suggestions.

Thank you very much !

 

[1DSmII]

ISO 1600

Loading…

dslr-astrophotography.com

 

[rnclark]

Hello everyone,

I am new to this site and a novice in regard to photography, right now I am doing a photography course (here at our local photo club) that I like it very much !

I own a Canon EOS 5D MkII and I just purchased two lenses...the Sigma Art 35mm f/1.4 and the Tamron SP15-30 f/2.8 zoom.

This summer, me and my wife, we plan to go to Azores for a couple of weeks of vacation and I would love to have a shot at taking Milky Way pictures.

I would like to ask Canon owners here about how high you suggest I can go with ISO settings, based on your experience...

I know that actual models (MkIII and MkIV) can easily go up to very high ISO and they could use that pictures right away without much work...

Since I never tried night sky photography pictures, and since my camera is already 7 years old ( purchased in the 2011) I would like some suggestions.

Thank you very much !

Normally, I would recommend ISO 1600, but the 5d2 suffers from excessive banding problems. See: http://www.clarkvision.com/reviews/evaluation-canon-5dii/

See the images in tables 2a and 2b.

Night images I have made with the 5d2 at iso1600 show too much banding (in my opinion), so I suggest ISO3200. On a fixed tripod with a 35 mm f/1.4, 8 seconds is the maximum I would do to keep stars round. Then make 3 to 4 eight second images and average them together. With the Earth's rotation, this will reduce the banding problem some and gather enough light to make spectacular images.

Roger

 

[bobsimage]

I have had good success with my 5DII and Sigma 35 Art lens with 90 sec exposures at ISO 1600. You need to stop down the lens to f/2.8 or more to eliminate the distortion in the corners. A 90 sec exposure also requires tracking to obtain pinpoint stars.

This image of the Milky Way at Cygnus was taken with my 5DII at ISO 1600, 90 sec exposure and Sigma Art 35mm at f/2.5. The camera was mounted on a Celestron CG-4 mount ...

Bob Brunck

 

[Swiso]

Thank you very much for the suggestions.

I will try it before going on vacation.

@rnclark : when you say to take 3 to 4 eight seconds pictures and then average them together, you mean using PTGui ?

@bobsimage: WOW impressive picture....that's what I would like to obtain ! Nice !

I remember very vividly some years ago in Bolivia, when travelling back to Chile, one morning we wake up at 4 am to depart for the border, and when I get out of the house we were sleeping, I was impressed by the mighty view of the Milky Way above us....was an impressive view !

I will never forget that view !!

Thanks a lot for the suggestions.

 

[Swiso]

Tonight I will go for my first test with the Sigma 35mm 1.4.

One question:

Should I use it at f 1.4 or better going at 2.0 or 2.8 ? Thus to be sure not to have coma ?

And, if I go to 2.8, is 8 seconds still the max value to use ?

Thanks !

 

[kiwi2]

Tonight I will go for my first test with the Sigma 35mm 1.4.

One question:

Should I use it at f 1.4 or better going at 2.0 or 2.8 ? Thus to be sure not to have coma ?

And, if I go to 2.8, is 8 seconds still the max value to use ?

Thanks !

According to this and this , that lens can be used wide open at 1.4 without too much coma.

 

[Swiso]

Thank you all for the suggestions.

This is the result of my first try.

Used the Sigma 35mm open at 1.4, eight seconds shot.

ISO 3200.

 

[kiwi2]

Thank you all for the suggestions.

This is the result of my first try.

Used the Sigma 35mm open at 1.4, eight seconds shot.

ISO 3200.

Nice! That's certainly looks coma free enough to me.

--

The Colour of Night

Explore this photo album by Pete Oliver on Flickr!

www.flickr.com

 

[Swiso]

Thanks !

Yes, just a little vignetting at the edges....but it is to be expected right ?

 

[t_wade]

Swiso,

I completely agree with Roger. I would avoid ISO 1600 with the Canon 5D2. It does very well at ISO 800 and 3200.

Wade

 

[sharkmelley]

Swiso,

I completely agree with Roger. I would avoid ISO 1600 with the Canon 5D2. It does very well at ISO 800 and 3200.

Wade

So you think 800 is preferable to 1600? Why do you think so?

Mark

--
Takahashi Epsilon 180ED
H-alpha modified Sony A7S
http://www.markshelley.co.uk/Astronomy/

 

[t_wade]

Mark,

So you think 800 is preferable to 1600? Why do you think so?

The fixed pattern noise of the 5D2 is at its worst at ISO 1600. Even though it is "fixed", most people who use the camera for astrophotography have no way of cooling the camera to a specific temperature so the fixed pattern noise is difficult to get rid of, even with an excellent master bias frame. I found the only way to consistently remove it is through dithering. For those who can't dither their frames, it is best to avoid ISO 1600. For guided images, I use ISO 800. For fixed tripod (i.e. no equatorial mount), I use ISO 3200. The read noise for ISO 800 and ISO 3200 are quite similar, 5.1 and 4.9 respectively. ISO 800 has more dynamic range too.

Wade

 

[Astrophotographer 10]

I have seen some great nightscapes with the Canon 5D2 so take heart.

Most of those seemed to use ISO3200.

I used a 5D2 once briefly but was disappointed in how much colour noise it had at relatively low ISOs (by todays standard).

As Roger said you're best option is take several images and stack them. I usually take 6 exposures when I stack but if you are only imaging one final image and not a mosaic then the more the merrier.

I use the free program Sequator to stack images and it works really well.

You can also use Photoshop for aligning and stacking images but its more unreliable and a bit crude really. A bit of black mark for Adobe when there are plenty of excellent stacking programs out there and their expensive software has a pretty ordinary stacking algoritihm. A bit 3rd world really.

Luckily you have some excellent lenses there. I have seen quite a few excellent Art 35 1.4 images which seems to be the lens of choice for many nightscapers.

Greg.

 

[Astrophotographer 10]

Mark,

So you think 800 is preferable to 1600? Why do you think so?

The fixed pattern noise of the 5D2 is at its worst at ISO 1600. Even though it is "fixed", most people who use the camera for astrophotography have no way of cooling the camera to a specific temperature so the fixed pattern noise is difficult to get rid of, even with an excellent master bias frame. I found the only way to consistently remove it is through dithering. For those who can't dither their frames, it is best to avoid ISO 1600. For guided images, I use ISO 800. For fixed tripod (i.e. no equatorial mount), I use ISO 3200. The read noise for ISO 800 and ISO 3200 are quite similar, 5.1 and 4.9 respectively. ISO 800 has more dynamic range too.

Wade

Also that would be a large dither ie. move the camera between successive images by a reasonable (but fairly small) amount so none of the same values of pixels line up when stacking).

FPN would be several pixels wide more than likely. On the other hand ISO3200 is quite noisy on this camera but several stacked and dithered images should remove a lot of that.

Greg.

 

[Astrophotographer 10]

Lovely.

Greg.

 

[Astrophotographer 10]

Thank you very much for the suggestions.

I will try it before going on vacation.

@rnclark : when you say to take 3 to 4 eight seconds pictures and then average them together, you mean using PTGui ?

I'll answer for him. No. PT Gui does not stack images. He probably used Deep Sky Stacker. You can use Sequator which is free and is very good.

@bobsimage: WOW impressive picture....that's what I would like to obtain ! Nice !

I remember very vividly some years ago in Bolivia, when travelling back to Chile, one morning we wake up at 4 am to depart for the border, and when I get out of the house we were sleeping, I was impressed by the mighty view of the Milky Way above us....was an impressive view !

Yes I get that here in Australia from about July to Sept hence my interest in nightscapes it is indeed an impressive sight.

I will never forget that view !!

Thanks a lot for the suggestions.

3-4 would be the minimum. Unless you are doing a large mosaic I would go more as 5D2 is pretty noisy. I use 6 x 30 seconds with my clean Sony A7r3. The coma and corner aberrations of the Sigma Art 35 1.4 weren't that bad in the copy I had. You could crop out the worst corners or do a 2 or 4 panel mosaic if you feel adventurous.

6 x 8 second images doesn't take long. Why not do 30? Its still only takes a few minutes.

Take a 100! Imagine how nice that image would be.

Deep sky stacker can stack hundreds of Canon images I see that all the time from Canon DSLR deep sky imagers. 300 stacked images is not uncommon. You may need an intervalometer to do that though. They are about $20 on ebay and plug into your camera and now you can program it to take as many exposures as you want.

I intend to get another copy. I had to return mine as it would not adapt to the Sony A7r3 and was unreliable. They now have a Sony mount version so its on my list.

Greg.

 

[Astrophotographer 10]

Also to be clear coma is not a function of how long you expose. Its there at any exposure length. Its an optical aberration.

Greg

 

[Bjrich]

Is your 5d ii modified at all?

 

[Swiso]

Is your 5d ii modified at all?

Are you asking to me ?

In this case, no...it's not modified.