Wouldn't have thought this possible...

[namreg]

Came across this on Flickr yesterday:

Horsehead nebula (untracked) | Flickr - Photo Sharing!

Nothing special to look at, but how he's taken this image is quite amazing (not to mention time consuming and camera destroying). 785 subs @ 2-3 seconds! I wouldn't have thought this possible until I saw the iso he'd used.

Out of idle interest, anyone got any idea what the equivalent exposure would be for, say, iso 1600? Do you just multiply by 8 i.e. 16-24 seconds, or is it not as simple as that?

[Zakalwe]

I don't understand. Why would it be a "camera destroyer" or time consuming? Surely it is no more harmful than a smaller number of longer subs, unless you are referring to the mechanical wear on the shutter? It was taken on a 50D, which has a shutter with an expected life of 100K operations (source: here)

As for the time, his exposure length was 39 minutes....just taken with a lot of very short subs.

I think that it is a very innovative way of doing things, and shows what modern kit is capable of.

[namreg]

As for the time, his exposure length was 39 minutes....just taken with a lot of very short subs

And countless realignments of his fixed tripod no doubt.

Well it impressed me!

[Zakalwe]

And countless realignments of his fixed tripod no doubt.

Well it impressed me!

Hey...I am mightily impressed too. it goes to show that decent images can be had without going for fancy mounts, autoguiders etc.

I am guessing that the final image was cropped loads, so he mightn't have needed to re-align all that much. Just cropped different areas as the image drifted across the FOV?

[namreg]

Yes, looking at it again you can get the whole belt and M42 in the FOV at 200mm, so it is severely cropped. An achievement nonetheless. Can't imagine going for an image like that without some form of tracking but it shows what can be achieved with a little imagination.

[pete_l]

Out of idle interest, anyone got any idea what the equivalent exposure would be for, say, iso 1600? Do you just multiply by 8 i.e. 16-24 seconds, or is it not as simple as that?

The photo says it was taken at ISO 12800 That's What the spec calls "H2 expansion". Hopefully someone with the camera will correct me, but that sounds to me like the camera electronics just amplifies (i.e. doesn't add any more signal) the top, ISO3200 setting by a factor of 4. If that's the case, then maybe you'd get the same results at ISO3200 if you stretched the image after taking it?

I expect the guy had a remote control that he just programmed in to snap so many hundred exposures (I wonder how many were discarded), rather than having to take each one by hand

[Shibby]

Very impressive, yes it's at f2.8, but this just goes to show the power of stacking.

I don't understand. Why would it be a "camera destroyer" or time consuming? Surely it is no more harmful than a smaller number of longer subs, unless you are referring to the mechanical wear on the shutter? It was taken on a 50D, which has a shutter with an expected life of 100K operations

You do realise, that at this rate, that's only 127 images before the shutter reaches the end of its expected life!

[Zakalwe]

Very impressive, yes it's at f2.8, but this just goes to show the power of stacking.

You do realise, that at this rate, that's only 127 images before the shutter reaches the end of its expected life!

Huh?

I said 100K shutter operation...that is 100,000

[Shibby]

Yeah I know... 100,000/785 = ?

[G2EWS]

The photo says it was taken at ISO 12800 That's What the spec calls "H2 expansion". Hopefully someone with the camera will correct me, but that sounds to me like the camera electronics just amplifies (i.e. doesn't add any more signal) the top, ISO3200 setting by a factor of 4. If that's the case, then maybe you'd get the same results at ISO3200 if you stretched the image after taking it

Both Nikon and Canon admit that the extended range of their cameras is only their as a marketing excercise. I am sure they would not admit that in print by the way!

Once you go outside the 'normal' range you just introduce unecessary noise.

Best regards

Chris

[shaunster]

Huh?

I said 100K shutter operation...that is 100,000

lol

Yes its a good result for 2 or 3 sec exposures !

[namreg]

Yes its a good result for 2 or 3 sec exposures

It's an amazing result. If, as discussed above, the 'real' iso top end as regards data collection is 3200, then this is the equivalent of about 5 second exposures at iso 1600, or 10 secs at 800 (assuming that's the way it works). Wouldn't have thought that sufficient to collect any signal at all from the horsehead. Stacking just reduces noise, it doesn't increase signal, or at least it's not supposed to.

[roundycat]

The signal in a succession of light frames is repeating and regular. The noise in the same frames is random. So the noise tends to cancel out and the signal reinforces.

Dennis

[namreg]

The signal in a succession of light frames is repeating and regular. The noise in the same frames is random. So the noise tends to cancel out and the signal reinforces

Yes, which means there has to be some signal there in the first place to be reinforced. The signal picked up here seems to be amazing for such incredibly short exposures.

[Zakalwe]

Yeah I know... 100,000/785 = ?

Ahh, right, I see what you mean:icon_salut:

[dph1nm]

Yes, which means there has to be some signal there in the first place to be reinforced. The signal picked up here seems to be amazing for such incredibly short exposures.

Providing you are at an ISO with unity-gain or better (i.e. you only need one photon to trigger an ADU), then the signal will always be picked up, no matter how short each exposure, once you add lots of them together.

The noise, however, is a different issue!

NigelM

[ollypenrice]

I'd guess he was using the Canon EF200L lens, which is a beauty and can run wide open as shown. It does cost more than an ED80, though... It is one of the best bits of glass I've ever used.

Olly