View Full Version : FWHM...do you trust these numbers when judging seeing...?

2008-Jun-03, 09:49 PM
...I do not.........not entirely.

I am not sure if I missed a previous discussion on this subject:
is there a consensus on judging seeing with this FWHM methode...?
any standard???
I am talking about the life-view-window when focusing on a star...

which stars are allowed to be used, to have this FWHM as a standard-
- hight above horizon?
- stellar magnitude?
- spectral class of star?
- luminance filter or none?
- type of CCD?
- software used for evaluating seeing via FWHM (life-view)?
- etc...?

it is my humble but e x p l i c i t experience over the years that:

- there is a huge difference in those values if you use an OSC or mono
CCD (I am talking about the focus window - so life-view)
- if you take another star you will get different FWHM values
- if you take another software it will also differ significantly - it
should not, but it does, that is a fact!
- pickering suggests something else than FWHM tells you about seeing.
- people use this values as a 'pseudo-standard' obviously, not
realizing that different software was used, and therefor they compare
incompareable things...(?)

there seems to be no standard on this subject,...?
so, I do not trust these numbers.

I rather got with pickering's scale, even if this is very much
dependant on personal experience and of course this scale was made
with a scope, only a couple of HR deep sky imagers use: a refractor...
{[(which quite frankly in my eyes is the very best dedicated telescope
for any kind of high resolved photography... *GGG* - don't take that
too seriously, I am only joking)]}

honestly: any thought on this?
I'd really appreciate clarification - if there is such - cos I am
afraid by posting these numbers we are comparing apples and pares, as
we say here in austria...

if there is no standard, we could think of one, shouldn't we ?

best regards,

BTW: FYI I have added a crop on one of the luminance layers of the M57
image I recently posted, that is supposed to illustrate what was in
the data I acquired. DDP+ sharpen, pos constr. deco in CCD stack,
wavelets in registax and smart sharp as well HP filter was globally
applied to that crop)

2008-Jun-04, 01:46 AM
I've found the measurements I get with CCDSoft and The Sky to be very reliable. Stars need to be in the linear range of a NABG chip. Not using OSC I don't know how that would alter things. But when I stay in the linear range the results I get are very consistent. Bright to dim stars vary very little on the same shot. As I get toward the edges the readings grow as field curvature starts to take its due from the image. But within most of the image bright to dim vary only about 0.1" to 0.2". Bright stars above about 27000ADU are no longer in my linear range and tend to give larger readings. Now this is comparing 5 to 10 minute shots. Shorter exposures like a minute down to a few seconds or less give a very wide spread though on any one image the results are pretty consistent. Seeing hasn't had a chance to even out in short exposures so gives me widely varying readings (up to 1" different") on nights when in a 5 minute or longer frame seeing would run about 3.2" or more. On nights (rare for me) to 2.2" then I see less than a 0.5" variance between short exposures.

Comparing my images to a friend in another state imaging with an ST-10 rather than my STL-2000 and binned 1x1 in his shorter focal length scope I find his readings (different software as well) very consistent to mine. That is if my FWHM on the same object is the same as his the resulting resolution of our photos will be very similar but since he is at over 3000 meters compared to my 350 meters his FWHM is usually better than mine and his images show it. On nights mine is better than his I get the better resolution and by about what the FWHM readings would indicate.

I do get slightly smaller FWHM readings when using 1x1 binning than 2x2 with 3x3 showing slightly larger yet. Difference is about 0.2" between each. Since pixel size is 0.5", 1" and 1.5" I put most of that down to resolution limitations of readings smaller than a pixel. The larger the pixel the more room for jitter as the star falls in a different pattern on fewer pixels.