Talk to me about resolution

[swag72]

I have got the SW120ED out again after playing with the Pentax 75SDHF for a number of months. With the Pentax, when I check my average FWHM figure in Maxim I am always under 2.0 with 20 minute narrowband subs. Putting the SW120ED back to work and I am finding that I am struggling to get a FWHM much below 3. So, why is this and should this concern me? I was hoping that I'd be getting the same figure as my Pentax.

Does this mean that none of my images will be as crisp with the 120ED as the FWHM is almost 2x that of the Pentax?

I am using my Atik 314L+ on both scopes. I can not get my head around this massively different figure and what it means, as if I had a FWHM figure over 2 with my Pentax I'd bin the sub, yet here I am now struggling to get below 3.

Can anyone give me some ideas please?

[swag72]

Just added a file of M33 luminance that I have just taken with the 120ED. This is a 600s sub, but to me looks very soft and useless. With the Pentax, I know it would look miles batter.

[JamesF]

Will the FWHM figure not increase anyway as the focal length increases?

James

[ollypenrice]

Oh, what a confusing term this is, resolution! Optics can have their resolution defined easily by the Dawes limit. Aperture wins. Ah, but we are not in space so the atmosphere messes with the incoming beam. Sometimes smaller aperture is less affected. Eek, how do we factor that in? OK OK so we are imagers. What then? We have small pixels or large pixels so small must give more resolution? (Hey, I'm just back from the future with the Canon Gazillion D so your little CCD is gadoomed, earthling...) Uh-oh, you are tryng to resolve details below that which is allowed by the seeing and 2) if your smaller pixels collect less signal they support less software sharpening. AAArrrgghhh. (Damn that real world, how I hate it!)

No, but it is dead simple. Aperture, focal length, sampling rate, resolution. No? No! Obviously not, since the same imaging system over a single sub cannot match its own resolution over the same sytem with a Sigma stack of 500 subs. Ask Mr Peach.

Ditch the term resolution. Or at least give it one consistent definintiton. We all know what you have to do. Get a reasoable sampling rate and then get an unreasonable amount of data. I'd be doing that right now only, whisper ot, it's sort of cloudy!

Olly

Olly

[swag72]

Olly, you need to get out!!!! Look what the clouds do to your brain!!!

So, sampling rate .......... explain that and what I should be looking for / at.

[JamesF]

Just added a file of M33 luminance that I have just taken with the 120ED. This is a 600s sub, but to me looks very soft and useless. With the Pentax, I know it would look miles better.

Isn't the image scale much larger with the Pentax?

James

[swag72]

Isn't the image scale much larger with the Pentax?

The Pentax is 500mm and the 120ED is 900mm, so yes - But can someone explain stuff with regards to stuff looking soft and fuzzy and poor FWHM figures?

[Martin-Devon]

Sara, I'd always assumed (perhaps wrongly) that FWHM was a measure of how sharp the focus was. I don't know how the Maxim FWHM values compare against those for Artemis (capture software for Atik), but I use the latter and for my humble 720mm focal length refractor (with Baader steeltrack focusser) can routinely get FWHM values of at least 1.5 and on a very good night this figure comes down to 1.0 or even less. As James points out I would imagine there is a correlation between focal length and FWHM values, certainly shorter focal lengths are easier to work with and I assume more straightforward to get sharp focus with. Maybe I'm wrong though? On another note - I see that "eggy" stars are creeping back into your subs again, I know you've had issues with this in the past - are you sure that Skywatcher refractor is all bolted down/focusser tight etc?

[swag72]

When I was doing the focus in Artemis, I struggled to get the figure to stay below 2 on a 1.5 second sub. I at least know now that any issues are not my focusing technique as I can focus nicely enough with the Pentax! So the focal length of 900mm, will it naturally give me larger FWHM figures?

Regarding the eggy stars - No excuses I know, but the guiding wasn't set up for the 120ED and I'm still waiting for a new step down flange from Skywatcher, that I'm hoping will assist with my issue. It's only been 3 months waiting so far!!

I also interestingly found that focuing using FWHM values in Artemis was FAR more accurate then using the Bahtinov mask and Bahtinov grabber software, which I always thought was pretty accurate. Mmm, back to the little Pentax me thinks!!!

[Martin-Devon]

Sara - what you say about the Bahtinov mask/grabber software versus the Artemis FWHM is exactly my experience as well. For some time I diligently used the Bahtinov mask (customized for my F ratio) & grabber software and got reasonable sharpness. Then one night having done this, having presumably optimized my focussing, I then switched to looking at the FWHM figure in Artemis and was shocked - it was high. So, the two focussing methods, at least in my experience do not match up. More recently I've stuck with the Artemis FWHM method to focus and I think I have sharper stars/images now. I'm aware that this is contrary to popular belief, in that the Bahtinov mask method is though to be the gold-standard for focussing, maybe I'm doing something wrong (although the grabber software is pretty straightforward and tells you when you a spot-on).

[JamesF]

There's a theoretical figure for FWHM, which is 1.02 x wavelength x focal ratio. If we assume Ha (wavelength 656nm), that's about 4.8um for the Pentax and 5um for the ST120, but I don't think that's what we're talking about here. I assume the figure you're getting is effectively measured off the CCD sensor?

Back in the real world, we don't get perfect seeing and point source stars. We get something that's a bit fuzzy. Let's say for the sake of argument that a star actually appears as 3 arcseconds wide. I don't know how good an approximation that is, but it's easy to work out different values. The apparent star size on the image plane will be actual width x focal length / 206265, which works out as 7.2um for the Pentax and 13um for the ST120. That's pretty close to one pixel for the Pentax and two with the ST120. I can't help wondering if this is what you're actually measuring and might therefore be the expected result?

It's possible that the ST120 image is "softer" because the optics just aren't as good or because of a mechanical problem with the imaging train, but is it also possible that the image just doesn't look less distinct because it's twice as big?

James

[JamesF]

For interest, I'll point out that the formulae I've used above come from "The Handbook of Astronomical Image Processing". It's not a read I'd recommend to everyone

James

[ollypenrice]

Sampling rate is measured in arcseconds per pixel. This is a nicely presented over-view. There is also some spicy controversy in there!

http://starizona.com/acb/ccd/advtheorynyq.aspx

I too am confused as to why the largest scope I use gets the highest FWHM and yet resolves the most detail. This is the case, though. Like Martin I sometimes go below 1 (0.85 is the record, I think) but that is in the Tak 85. The 14 inch (FL just under 2.4 metres) never gets below 2 but that is in Nebulosity, not Artemis, so we are not comparing like with like. However, the TEC (140/980mm) never gets as low as the smaller Tak either but obviously out-resolves it.

So, while I cannot understand it, it seems perfectly normal that your larger scope will give a higher FWHM. I hope that one of the technical people on here will step in.

Personally I would focus in longer subs. I generally use 3 seconds, the reason being that this gives the sub long enough to let the seeing go through its cycle of disturbance to give you an average value. I just find that this gives a FWHM that bounces around less.

And another thing; using FHM (Nebulosity's version of FWHM) on the big scope is a total waste of time. The readings bounce around all over the place and when you have gone half crazy but are certain you can't get them any lower you find you are not in focus anyway! We have found that the best way so far is to focus using the diff spikes on a bright star, getting them first into four singles rather than four split pairs and then getting them as thin as possible. What we really need but haven't yuet sourced is an enormous Bahtinov.

Olly

[ollypenrice]

Oh, and a PS regarding Bahtinov versus Artemis FWHM; I find they correlate well but not perfectly. I can usually get maybe 0.1 off the Bahtinov value for FWHM, though I really don't know which to trust.

Olly

[JamesF]

Do you know what the FWHM figure calculated by Artemis actually means in terms of the data collected, Olly? Or is it just a "magic number" that's desirable to reduce as far as possible?

James

[swag72]

The apparent star size on the image plane will be actual width x focal length / 206265, which works out as 7.2um for the Pentax and 13um for the ST120. That's pretty close to one pixel for the Pentax and two with the ST120. I can't help wondering if this is what you're actually measuring and might therefore be the expected result?..........but is it also possible that the image just doesn't look less distinct because it's twice as big?

I think this is what I am thinking and seeing James, that the star size is probably x2 in the 120ED. So the FWHM figure will be x2 between the Pentax and the 120ED?

Off to do some reading now Olly thanks! And I'll focus using 3s subs, although that is often using a very small star to keep the brightness figure down - Is that right?

[swag72]

In the CCD calc, my image scale with the 120ED is 1.48 arcpixel/sec and for the Pentax it's 2.64 arcsec/pixel.

What does this mean then in the real world and is this what I am fighting against?

[ollypenrice]

Do you know what the FWHM figure calculated by Artemis actually means in terms of the data collected, Olly? Or is it just a "magic number" that's desirable to reduce as far as possible?

James

Well, I assume that it is the measurement of the diameter of the star at half it's maximum brightness. The way to visualize it is look at a bell curve for brightness as a section through the star. At extreme left and extreme right you have the faintest outer edge low down. The highest point is in the middle for the core so the FWHM is the measurement half way up. However, I don't know what the units are and am ashamed to say that I have never thought about it! I wonder if it is the proportion of width to height. Time for a quick Googel, methinks.

Olly

[JamesF]

So the FWHM figure will be x2 between the Pentax and the 120ED?

Without knowing exactly what Artemis means by "FWHM" I can't be certain, but if it's related to the star size as it appears in the image that is the conclusion that seems logical to me, yes. Even more so given Olly's comments about the largest scope giving the highest figures.

James

[JamesF]

In the CCD calc, my image scale with the 120ED is 1.48 arcpixel/sec and for the Pentax it's 2.64 arcsec/pixel.

What does this mean then in the real world and is this what I am fighting against?

That's a measure of "how much sky" is represented by a single pixel on the camera. If the 120ED is 1.48 arcseconds/pixel then you'd expect an object 3 arcseconds wide to cover fractionally over two pixels, whereas on the Pentax it would cover just over one.

Does that make sense?

James

[swag72]

So when imaging a galaxy for example in both scopes - Does that mean I'll get an image out of the larger scope has 1/2 the sharpness of the smaller scope?

[JamesF]

Well, I assume that it is the measurement of the diameter of the star at half it's maximum brightness. The way to visualize it is look at a bell curve for brightness as a section through the star. At extreme left and extreme right you have the faintest outer edge low down. The highest point is in the middle for the core so the FWHM is the measurement half way up. However, I don't know what the units are and am ashamed to say that I have never thought about it! I wonder if it is the proportion of width to height. Time for a quick Googel, methinks.

Yes. I understand what FWHM actually means, but presumably Artemis is measuring something like "apparent FWHM" by looking at a star, or number of stars and measuring the relative intensities of pixels across the star image, perhaps fitting them to an ideal curve and calculating the value from that. If it gives the result in pixels then I think it may well explain the results Sara is getting. (Gah, keep typing Sarah there. Apologies if I misspell your name, Sara. My wife is Sarah so my fingers tend to add the terminal "h" without me noticing

A brief google leaves me none the wiser as to what Artemis is actually doing.

James

[swag72]

If it gives the result in pixels then I think it may well explain the results Sara is getting.

I'm not clear on this James - What does it explain?

No problem on the name - It often gets typed with an 'h'!!

[JamesF]

So when imaging a galaxy for example in both scopes - Does that mean I'll get an image out of the larger scope has 1/2 the sharpness of the smaller scope?

Ah, well, that depends on what you mean by "sharpness", I guess My gut feeling is that we'd interpret it in terms of detail resolved. We need Lord Rayleigh's help here, I think. His formula says that the angular size of the smallest resolved detail is wavelength / aperture (in radians -- multiply by 206265 to get arcseconds). That works out that the ST120 should in theory resolve 1.71 times as much detail as the Pentax.

Now I'm out on a limb here and I've no scientific basis for suggesting this, but it seems to me that if the ST120 can resolve 1.71 times as much detail as the Pentax, but the decrease in image scale is not at least 1.71 relative to the Pentax, the image may well not look as sharp. I'm so far out on this particular limb though that I have very little confidence in this actually being a valid way to consider it, or even one that makes sense in terms of the physics.

James

[JamesF]

I'm not clear on this James - What does it explain?

The FWHM figure for the ST120 being twice that of the Pentax. It might well basically be saying "that's what you should expect" rather than "the focus isn't as good".

James