M42 - single 6min exposure

[malc-c]

Took advantage of the conditions last night to grab a few more subs of M42, and to check the tracking took a single 6 min sub. I've attached the un-processed sub, and one with the curves tweaked.

I also took around ten 200s subs with 3 x 200s darks, but as yet can't seem to get a decent stacked result (still learning the software).

I'm pleased with the amount of data recorded in the single capture, and makes me wonder why we resort to lots of stacking of multiple images. Would welcome comments.

Hummm... why is it that when you upload images to the forum they never look like the ones on the PC... the processed image looks over cooked and very orange !!

EDIT = corrected image uploaded

Added a 3rd image - processed in PaintNET

[Earl]

Nice start Malcolm, the advantage of stacking is to improve the Signal to Noise ratio, if is clear tonight and im feeling upto it, (had a cold all week) ill have a bash myself.

[RogerTheDodger]

you have a running man hiding in the data at the top...

[malc-c]

you have a running man hiding in the data at the top...

Ooo... better get started on stacking the 10 x 200s subs then

[RogerTheDodger]

had a little fiddle with the image, there he is

[Ags]

In the old film days no-one did any stacking. I think if you can do long exposures like 6 mins you have a good signal to noise ratio already so stacking has less of an obvious effect.

[malc-c]

In the old film days no-one did any stacking. I think if you can do long exposures like 6 mins you have a good signal to noise ratio already so stacking has less of an obvious effect.

Good point... I guess in the good O'l days with film there was no choice but to get a good site, good alignment and spend hours manually guiding. The other thing I've noticed with modern digital cameras and the processing that goes with it is that you can make the image look however you like. My two post processed images are a good example, one warm with yellows and orange, the other blue and red....

Still I guess as long as the image looks good to the eye that's what counts. I might try taking a 20m sub next time and see if that comes out as good - most of my attempts as stacking result in a final image that has less detail than the original

[Ags]

Good point... Still I guess as long as the image looks good to the eye that's what counts. I might try taking a 20m sub next time and see if that comes out as good - most of my attempts as stacking result in a final image that has less detail than the original

Ha ha! When I get to image with my dipsomaniac SE mount the subs are 30 seconds long and all contain random tracking errors. Stacking is essential to kill the noise and get the stars round.

Another reason why your image looks good: it is M42, uber-bright, so of course signal to noise is high. I think you might find stacking more essential on a fainter target like the Veil of the NAN.

[Revs]

Looking good Malc. Stacking should reduce the noise and enhance the detail, but collimation and tracking look good. Nicely done:)

[Explorer 190 MN]

Just to for fun this is m42 stacked to make 6.4 mins exposure time.

8" Newt fl 1000mm

Light pollution 5-7

Canon 550D

7 x 55 sec s lights ISO 800

25 darks,

25 flats

Deep sky stacker and then CS5

I think I will try either 45 sec subs at ISO 800 next time or 120 sec at ISO 400 next time.

[Explorer 190 MN]

I am only learning myself but as far as I can work out short exposure times for bright objects For my camera I find ISO 800 and for bright targets 45 - 55 sec subs and for fainter objects ISO 800 and 2 min subs but thats for my sky fog and light pollution levels. They say you need to gather at least 26 mins of data to get the best S/N ratio. Due to noise with most system cameras exposures of more than 5 mins is not recommended due to signal getting lost with noise. I tried M45 with 20 something mins exposure ISO 800 2 min subs, darks,flats,offset bias and then I tried ISO 400 and 2x5 min Lights and 2 Darks. I got a lot more information and better contrast with ISO 400.. I read somewhere the best way to messure skyfog is to use your normal camera lens set to the same f number as your OTA then try diffrent exposure times until you get your histogram on the back of your camera to 20% from the left hand side.

I try to stretch as much information out of my image using the controls in deep sky stacker before before gently adding color to the image I then adjust the color channels then back to middtone brightness and contrast then highlight btightness contrast then back to color channels and then saturation. I then save it as a 16 bit tiff open it in adobe CS5 and adjust it.

[ollypenrice]

In the days of film the best photographers did much longer exposures (40 mins etc) and they did stack. David Malin used different film emulsions (relatively more sensitive to red, green and blue) on successive nights.

The virtue of deep stacking really is a no brainer. The more data you have the more you can stretch the faint stuff without drowning in grainy noise. Just try it, it honestly is dead obvious. If you have 2 hours the faint stuff looks like sandpaper. If you have 12 hours it looks gorgeous. No free lunches in this game, which is why it's such a nice game! If you come back to the same target over several years you can get a tremendous depth of signal. I love doing this.

Olly

[malc-c]

OK, followed a tutorial and stacked 9 x 200s @800ISO subs with 3 x darks of the same exposure.....

Here is the result...

Now in theory, this equates to 30 min exposure, but IMO doesn't sow as much detail as a single 6 min sub.

If someone wants to have a go at processing the raw images I can zip them up and stick them somewhere for a download, but would like to have a blow by blow account of how they get the results.

[Ags]

Yup, that's about right I think. DSS Help includes an animation of the effects of stacking - the image is no brighter after stacking, regardless of the number of subs, but you can see the noise go down. So a 6 min sub will be brighter than any number of 3.3 min subs.

To go really deep you need both exposure and stacking - the long exposure to catch the faint stuff, and the stacking to cut away the noise. Once the noise is gone you can stretch the image etc etc.

[malc-c]

To go really deep you need both exposure and stacking - the long exposure to catch the faint stuff, and the stacking to cut away the noise. Once the noise is gone you can stretch the image etc etc.

So basically if I was to take say 5 x 6min subs and then stack them I would get better data than 10 x 3min subs... right, (insert sound of penny dropping )

[malc-c]

Just out of curiosity - why take x number of subs of the same duration etc. Can you not simply take say one 6 min sub and then stack that same image 10 times to reduce the noise levels and get the same result ?

[Ags]

The point is that noise is noisy - so it does not repeat from sub to sub. Signal repeats from sub to sub. This difference between signal and noise enables stacking software to distinguish between good and bad data and strip away the bad.

If you stacked 10 copies of the same image the noise pixels would repeat like perfect signal, and would be stacked by the software and added to the final result

[Explorer 190 MN]

OK, followed a tutorial and stacked 9 x 200s @800ISO subs with 3 x darks of the same exposure.....

Here is the result...

Now in theory, this equates to 30 min exposure, but IMO doesn't sow as much detail as a single 6 min sub.

If someone wants to have a go at processing the raw images I can zip them up and stick them somewhere for a download, but would like to have a blow by blow account of how they get the results.

200s is too long for the core of m42 you blow out all the detials. You can take some long exposure to catch the outside edge detail stack and process them and then shorter exposure to catch the center of m43 and then stack the to final images in photoshop. Try again with 55s subs

[austen4]

From this Im going to start doing longer exposures, I never new how much more detail you get. Its a great image though, well done.

[malc-c]

Thanks for the reply guys... whilst you were replying, just for an experiment I stacked 6 copies of the 6min sub, then saved it as a greyscale tiff then tweaked in PS and saved the output in RGB colour

Here's the results

I note the comments about noise being increased... maybe this would explain the fact that pulling the levels I get a similar amount of detail in the 5 x 6min subs as I do when pulling the single image, as the ratio between S and N remains the same. But couldn't this also happen when taking multiple subs. Wouldn't there be a chance that subsequent subs could have a worse S/N ratio as the CCD would be hotter and less sensitive ??

[Earl]

So basically if I was to take say 5 x 6min subs and then stack them I would get better data than 10 x 3min subs... right, (insert sound of penny dropping )

In reality i found this to be wrong, I tested it on M33 using 5 min subs and 10 min subs, but with the same total length for each, the images looked identical.

[Ags]

Earl I think there is a balance between noise reduction, number of subs and length of subs. I think the balance point depends on the brightness of the target and how you process it. It may be that M33 was bright enough after 5 mins for your processing, so going to 10 mins only increased the noise in return for only a slight improvement in signal.