How much stretching is too much?

[5haan_A]

Hi Guys,

The answer seems obvious, well it is to be fair- stretch until you have enough detail from nebulosity and stop when it gets too noisy. But, is there a more methodical/scientific approach to this. I use DSS and photoshop, the autosaved stacked file out of DSS is always very dark so a lot of stretching ensues. On my most recent project, C9 cave nebula, I have found stretching to be a real fine line because I need to stretch a lot to draw out the detail, but when I start to stretch too much I begin to lose the contrast between the lovely dark areas of the image, and then there's the noise element to contend with as well. 

I have attached a screenshot of a single stacked frame of H alpha in photoshop. It's not anywhere near done by any means, but you can see on there the histogram and what seems like millions of level stretches in the history tab. It would be good to know when enough is enough. Or a workflow that you use that works. At the moment I work purely by eye, I levels stretch by say 5-10% a couple of times then move the black point across by 5-10 points then rinse and repeat until I have something that looks visually ok. 

 

Thanks

 

[Xilman]

How long is a piece of string?

Seriously: stretch until you can see what you need to see within the limitations of your data and then stop.

See https://britastro.org/observations/observation.php?id=20200805_212500_dc7c69e9f34bb551

and

http://www.astropalma.com/Projects/Satellites/caliban.html

for extreme examples of mine.  The former  completely destroyed the appearance everything except the target of interest. AFAIK, this is the only amateur image of Thebe ever taken, and I have yet to see a professional one not taken by a probe in Jovian orbit.

[5haan_A]

Yeah I think knowing the limitations of your data is an important point. Having just got started on the O stack I can safely say that for this particular target, given my set up and that the majority of imaging time was done under a fairly full moon no amount of stretching is going to bring out the type of detail I desire. Shame as I have 8 or so hours of what now seems to be fairly impotent O data. 

 

Best,

[5haan_A]

Speaking of which attached is a single stacked frame of Oiii data. There are 46x10 min subs in there with all the dark and dark flats necessary. Nearly 8 hours of capture time on a eq6 mount through a Esprit 100. I know the Cave Nebula is a hard target but I feel that I may have been overly ambitious thinking I could capture the required nebulosity while the target was relatively low on the horizon and there being a fullish moon. 

Can anyone squeeze any juice out of this one, or am I flogging a dead horse.

 

Best,

Autosave002.tif

[vlaiv]

You seem to have issues with calibration in that image:

OIII data is very faint and your flat calibration has issues that are higher intensity than intensity of data itself.

[5haan_A]

On 02/03/2021 at 13:34, vlaiv said:

You seem to have issues with calibration in that image:

OIII data is very faint and your flat calibration has issues that are higher intensity than intensity of data itself.

There is almost no OIII data there, its faint. 

The point about the flats, is it perhaps something that can be solved by taking a shorter exposure time making the flats darker. I use APT CCD flats helper. I select a ADU of 25000 and then put it to work. Maybe I will turn it down and see if that makes a difference.