Cocoon Nebula 24 July 2010 00:00-01:15

[The Thing]

LX90 8" UHTC Baader Alan Gee II Telecompressor, Canon 350Da, Astronomik CLS-CCD clip filter, Lakeside Focuser, Bahtinov Mask, Meade 8x50 Finderscope as Guidescope with SPC900NC rehoused and LX modded with mono chip as guide cam, PHD Guiding, DSLR Shutter, DSS Live to monitor subs as they are saved as well as providing lots of useful stats.

48/60 x 1minute subs (needs a lot more but this was to try the loads of short subs approach). Processed in DSS and Irfanview for a quick result. I'm quite happy with this (this is a crop). The guiding was sub-pixel (I was amazed after all the grief I've had) but the collimation (after the scope flocking) was not as good as it should have been.

[Mac]

Its faint, but its there.
Be nice to see the difference when you add the short subs
How many you looking to do and are you going to do a long sub comparison.

Mac.

[MarkS]

The Cocoon is a tough object - it's actually very faint.
 
Your sky background will have been high - especially with the moonlight.
The subs are short - probably far too short.

If you try this again from somewhere dark and using longer subs you'll see a big difference.

Mark

[RobertM]

That's quite a bold attempt at a very difficult object especially with the moon glow.  It's one of those objects where you can never have enough subs.

Looks like there's something nasty going on with those star shapes so yes definitely get that collimation spot on - use a star (artificial or otherwise) to verify if you've used a laser.

Robert

[Fay]

Persevere, keep going Duncan, getting there!!

[The Thing]

This needs a hundred more subs at 1 minute, I think two minutes would be a better short exposure to use but the limiting factor is the skyglow/moonglow washing out everything. Samir Karusi gives specific instructions on how to determine the exposure length to use in a particular situation (as posted by Robert http://www.samirkharusi.net/sub-exposures.html).

Despite the issues with collimation, insufficient subs and it being a difficult dim target, this is still one of the best DSO's I have imaged to date. Roll on clear nights!

[MarkS]

Samir's approach is interesting but slightly over the top.  It is based on the premise that in each sub-exposure the sky fog (to be more precise - the sky fog plus fog from the dark current) should swamp the CCD read noise.  This is certainly a sufficient condition to get the best possible overall signal-to-noise ratio but it is not actually necessary.  It will err on the side of longer than necessary exposures.

In addition I don't necessarily agree with some of the simplifying assumptions in the maths contained in the links his site points to.

The read noise of the Canon 350D is very low (5 electrons RMS).   As a rule of thumb, using a CLS filter round here in light polluted Sidcup, using an aperture of F6 a 5 minute sub-exposure is ideal - 1 minute is far too low - you'll notice a big difference between the two. If you're using a narrowband filter or if you're at a darker site such as Rother Valley or Lydd then you could stretch to 10 minutes but it would only give a marginal improvement over 5 minutes.  If you're in the very dark skies of the Alps on a freezing night then it is possibly worth going up to 15 minutes.

What about a sensor such as the Kodak KAF-8300?  This is much noisier than the Canon (approx 10 electrons RMS) and therefore requires much longer exposure times before it reaches the optimum. In fact, even though the read noise is only double that of the Canon, it requires more than double the sub-exposure length  - 3x longer is a reasonable estimate.

If you're using a slower F-ratio e.g. F10 you'll need longer exposures, a faster F-ratio e.g. F4 means you can shorten exposures.

The opposite problem is that you could end up using a exposure time so long that it saturates your object of interest.  For DSLR users, selecting the right ISO and the right sub length is a very interesting problem.


When I'm less busy I'll have to produce a spreadsheet to do the calculation (without erroneous simplifying assumptions!)

Mark