Whats your personal FWHM cut off point?

[White Rabbit]

Just curious what the maximum FWHM is before you chuck an image.

I've notice with my own images that they can start out in the early evening at about 2.15 and as the night goes on it can get as high as 5.6.
I normally chuck anything above 4.5, just wonder what you do?

Cheers

[Shiraz]

below about 3 arc sec, priority is luminance or Ha data - anything much worse and I do colour channels only, in the hope that I might get better lum at some later date. Good conditions around here are about 2.5 - occasionally better.
Anything above 5.5 and I go to bed.

[Lee]

Good conditions here probably more like 3.0.... although it does improve through the wee hours, but these days I'm in bed by then!
I make the most of the clear nights, when pretty bad bin 2x2 and chase asteroids... :)

[Merlin66]

I suppose another benefit of doing spectroscopy, is you can handle any light pollution and any seeing conditions!!
The processing of the spectrum "drops" out light pollution, and adjustment of the entrance slit looks after the FWHM issues. Majic!!

[multiweb]

Looks like you might get out of focus as temperature drops. 2.5/3 is good. Anything over 4 is not good.

[PRejto]

'

Are you asking because the weather in Sydney has been shocking in terms of seeing? A pity with all these clear days as the FWHM I've been seeing has mostly been 5+. I've just packed it in in frustration! I'm pretty happy when it gets below 2.5-3...

Peter

[Logieberra]

Peter, how do our TheSkyX HalfFluxDiameter HFD figures correspond to FWHM? The same? If so, in the burbs of Canberra, I never see below 3. Evenings start at mid 5's and get marginally better over time.

[rat156]

Surely the numbers here are subjective, don't they depend on your image scale? Or does your software take that into account? Are the numbers here expressed in arc sec or pixels?

Cheers
Stuart

[Lee]

You're right - depends on scale, also really depends on the field you are measuring.

[PRejto]

Hi Logan,

I normally don't pay much attention to the HFD in TSX as I use FocusMax. Generally I find the HFD after using FocusMax is higher than the FWHM values I see when inspecting the images in CCDInspector which of course uses actual scope resolution in the calculations.

For example after focusing Ha the average HFD from FocusMax = 3.99
The range of FWHM over 7 images 10 min guided CCDInspt = 1.96-2.13

So, perhaps a 50% reduction. I'm not sure how the HFD value is calculated in FocusMax. Perhaps I have not set something up in the software properly but I have not found anywhere to set in the imaging scale.

Peter

[Logieberra]

Thanks Peter. Found a little something from Mr Ron Wodasky on this, which answers my Q:

FWHM is what you would need to be concerned with if image sharpness is what you are looking at.

Half-flux diameter is used on out-of-focus star images to aid in focusing away from best focus. It is a good measure for determining exactly how far from focus you are, and thus is very useful for a software product like FocuxMax. It's not very useful for human eyes, however, since we are mostly concerned with knowing when we are at good focus.

FWHM is useful for that - you want the smallest FWHM you can get if you are focusing yourself, or if you are evaluating seeing conditions (better seeing = smaller FWHM). More specifically:

Half flux diameter is the distance where half of the total light flux is inside that diameter, and half is outside.

FWHM is the diameter of a star image at half the height of the maximum brightness level.

[multiweb]

Software accounts for this. I image from 0.5 arc second per pixel up to 11 asp sometimes. So you have to account for various FL. I rarely got under 2. Maybe 1.8 on an exceptional night. Routinely between 2 and 3. 4 and over is noticeably blurry and bloated up close.

[rmuhlack]

Does spot size and pixel size have any role here?

[gregbradley]

I am not convinced these FWHM numbers are always meaningful.

I think it may vary with software. I would just look at the image and is it sharp or not rather than a number.

With long focal length bad seeing becomes obvious because its very hard to achieve a focus. It looks out of focus when its not.

Greg.

[White Rabbit]

Thanks guys.

I use CCD inspector to monitor my imaging folder live. As soon as a new image gets deposited it measures it for various attributes FWHM being one of them.

I suspect that my ever degrading FWHM has to do with three things, temprature, seeing and the angle of the object relative to the horizon. I notice that the closer it gets to the horizon the worse my FWHM gets. I'll need to add a few more focusing runs in my sessions.

I just wish to god that Software Bisque would hurry up and develop the camera add on a bit more to allow automated focusing runs between filters selection and temperature change. I'm kinda wondering why I paid so much for it when you consider programs like Astro art and the likes.

[gregbradley]

CCDinspector is a handy program. I'll have to get that.

In CCDstack if you click on several different stars you'll get several wildly different FWHMs. Hence my scepticism. Larger stars inevitably give a large FWHM and the lowest FWHM will be a small tight star near the centre of the image.

Greg.

[Peter.M]

The thing here is FWHM is supposed to be for you to measure the relative quality of each frame. The result it gives you is an average of the particular images FWHM. This means that you can compare each images focus and guiding against another image. This is how I use FWHM, because you are right different software will report different absolute numbers, but as long as you are using one piece of software you can gauge relative quality.

[Shiraz]

For FWHM to be useful as a quality measure within a given system, it does not matter what the units are. However, if you use it to compare with the results others get - which was implicit in the OP question - then the system FWHM in pixels must be converted to arc secs via the plate scale.

I just did a comparo to see how much measures varied with software - Pixinsight, Iris and Nebulosity all reported ~ the same results (well within 10%) for a few chosen stars (which must be unsaturated of course) in a recent image. Pixinsight gave almost identical FWHM to Iris when forced to use Gaussian rather than Moffat profiles. Neb uses HFR (half flux radius) which is similar to FWHM, but being a radius, needs to be doubled to a diameter to be comparable - results were very similar to the other two when this was done.

FWHM (or HFD) in arc seconds is a useful absolute measure of seeing for any scope larger than about 100mm, provided the pixel sampling is reasonably close to the Nyquist optimum.

[SkyViking]

Interesting thread. I usually experience between 2" and 3". I have seen below 2" on occasion, but it's rare. If it's over 3" I switch to taking colour, binned 2x2. I rarely see it higher than 3.5-4" though (only on a few nights per year).

[White Rabbit]

Yes it is very interesting and little over my head. I have much to learn, things are never as simple as they seem with this hobby. I love the challenge though.

[alistairsam]

Hi Rolf,

I'm a bit curious about this. what software did you use for the above measurements?
I'm struggling with constant above 4 FWHM as reported by Maxim, ccdstack and ccdinspector but I thought it could be my guiding/mount performance/scope rigidity all of which could smear the stars over a few pixels increasing the overall fwhm. I use an OAG with the lodestar and guiding is pretty flat.
I constantly get blurred images, with my F4 I get pretty good details, but compared to others, its blurred.
I even tried with a refractor, but ended up with 3.9.
would this mean that this is basically due to poor seeing?
if it is, then I'm moving to the hills.

Cheers
Alistair

[SkyViking]

Hi Alistair,

Here is an example of a good night at my site, FWHM ~2":
http://www.rolfolsenastrophotography.com/CCD-Camera-Testing/i-bNVbbNZ/X3
That is measured with PixInsight. I get similar results in Maxim for the same frame, so expect it to be reasonably accurate.

First of all I'd suspect that your image scale is set incorrectly so FWHM is not reported properly, but if you say images are also blurred compared to others then maybe not. Just something to double-check.
And secondly, collimation could be off.

I don't know if your site has particularly poor seeing. If it's always 4 or above then that sounds odd, but I'm really just guessing :shrug:
What is the experience of others in your area?

Cheers,
Rolf

[Geoff45]

I throw out everything over 3" fwhm in luminance, over 3.5" in colour.
Geoff.

[SkyViking]

Tonight's seeing is pretty good here, below 2" in both axes: FWHMx 1.9" and FWHMy 1.7" :thumbsup:
I'm currently making the most of it but my target will go behind trees in about 10 mins. And the Moon is also coming up soon... I think it's time to go to bed :lol:

[alistairsam]

Officially jealous...:D;)

[White Rabbit]

Yip, me to. Forget about the quality of seeing, I'd settle for a clear sky. We've had clouds for weeks now. Next week looks promising though.

[Gem]

With the ATIK and 80mm, I can get down to 1.6 late at night in Canberra on a good night. No hope earlier in the evening or with the SCT. Usually around 2-3 earlier in the evenings on good nights.

[Logieberra]

Funny you say that, Grant. I ran an experiment shortly after my post, above, last September. I found that my Kendrick dew controller heater strap was ruining my focus! Bumped the heater settings way down and saw instant results, with HFDs in high 1s and 2s!

[Gem]

No coincidence that my SCT usually uses a dew heater then! :)