RASA8 - One imaging scope to rule them all?

[ollypenrice]

25 minutes ago, tomato said:

that's why I would chose a RASA11 if I could only have one scope.

As would I.  I'm bored by the slow systems I know so well.  They were a great adventure for many years and a double FSQ106/Full frame CCD seemed like a wild adventure - and was - till I started working with Paul and the RASA 8-CMOS rig. Vlad may tell me that I'm dealing only in impressions, here, and he may be right - but what I feel is that I'm working with data which are different in kind from any I've worked with before. I can make different pictures from these data. I'm excited!  I wasn't exactly a cheapskate on exposure time, either, with the slower systems. 20 hours was routine for a single panel. But when I drop those 0.9"PP 20 hour panels onto 3 hour RASA images to improve resolution, there is nothing there in faint stuff.

We are in this for enjoyment (which can be hard to believe!) and, when it comes to enjoyment, give me the RASA. It's the difference between sitting on a nail and sitting on a sofa.

I was a very reluctant convert but there is no going back.

Olly

[vlaiv]

I think that RASA is very good / excellent in what it does - don't get me wrong.

It is just that I'm not so sure about One imaging scope to rule them all part - for my style, I would not choose RASA - even for imaging alone, and we haven't touched other "disciplines" related to data capture - like solar / lunar / planetary imaging, spectroscopy ....

For me - there is too much lack of sharpness in RASA, and that's coming from RC owner - those scopes have massive secondary obstruction.

[ollypenrice]

3 minutes ago, vlaiv said:

I think that RASA is very good / excellent in what it does - don't get me wrong.

It is just that I'm not so sure about One imaging scope to rule them all part - for my style, I would not choose RASA - even for imaging alone, and we haven't touched other "disciplines" related to data capture - like solar / lunar / planetary imaging, spectroscopy ....

For me - there is too much lack of sharpness in RASA, and that's coming from RC owner - those scopes have massive secondary obstruction.

Where do you see this? Mostly in stars or in nebulosity? RASA stars are not good, straight from the linear stack, but they can be fixed. (I'm becoming an expert at this! )  If I can fix it I don't mind.

Olly

[vlaiv]

1 minute ago, ollypenrice said:

Where do you see this? Mostly in stars or in nebulosity? RASA stars are not good, straight from the linear stack, but they can be fixed. (I'm becoming an expert at this! )  If I can fix it I don't mind.

Olly

What affects the stars equally affects the nebulosity. Star is just "shape of the blur" affecting the whole image. If there were no blur affecting the image - every star in the image would be single point of light (as they are from this distance.

While you can "fix" the stars - using some of the morphological tools - you can't fix the nebulosity that way. Closes thing to fixing the image is sharpening that is performed on the whole image and not stars alone. How much it fixes the image can be seen in stars - if stars look better - then whole image will look better after sharpening.

Morphological tools "reorganize" bright pixels - and do so only with and around stars (like "make stars round" action and alike). They are not sharpening tools.

[andrew s]

If you want the one to rule them all go for an 8 or 10 inch f5 Newtonian. Add quality coma corrector, appo reducer plus barlow/power mate and you have it all. 😊

Regards Andrew 

[andrew s]

If you want the one to rule them all go for an 8 or 10 inch f5 Newtonian. Add quality coma corrector, appo reducer plus barlow/power mate and you have it all. 😊

Regards Andrew 

[ollypenrice]

14 minutes ago, vlaiv said:

What affects the stars equally affects the nebulosity. Star is just "shape of the blur" affecting the whole image.

No, I do not believe this. I cannot believe it. I spend many hours per week working from linear data and every minute of those hours tells me it is incorrect.

Olly

[vlaiv]

1 minute ago, ollypenrice said:

No, I do not believe this. I cannot believe it. I spend many hours per week working from linear data and every minute of those hours tells me it is incorrect.

Olly

I can't really convince you that this is indeed so if I don't start with highly technical stuff and math.

Best I can do to show you that this is in fact true (without going into technical stuff) is to do the following:

I do hope we are in agreement that for all intents and purposes stars in astronomical image are in fact point sources (before the light reaches earth's atmosphere and aperture of telescope).

Imagine following scenario to understand what a blur is - there is only one star in the sky and nothing else - no light pollution, no nebulosity, nothing. Photons from this star always arrive from same angle in the sky - from same point, and in perfect world - they would all be focused into same spot, right?

However, we have atmosphere, we have aperture of telescope and we have aberrations. All of these conspire together against our poor photon so it does not land on a single point but gets thrown randomly around this point. Every photon that comes from this star gets the same treatment - majority of them are thrown off course just a bit, some are thrown off course a bit more. When they all accumulate - we get nice bell shaped distribution of photons with certain FWHM. In some scopes it will be nice gaussian like shape - where seeing dominates over diffraction effects, while in other scopes we won't get nice gaussian type shape but it will still be sort of bell shape but little distorted - this is where aberrations dominate seeing (seeing should be fully random and produce true gaussian shape - others are predictable and produce different shapes and result is combination).

Ok, now that we know how star light behaves - it gets scattered in particular way around that single point - we will call this "blur" as it blurs star into some non point like shape.

Punchline - atmosphere, telescope, sensor - all of them have no idea if that photon came from a star or from nebulosity or from some galaxy. All photons will receive the same treatment. Every point in the image will be blurred regardless of the source of photon - it is just important that photon arrives from same spot in the sky and it will get scattered around in that bell shape.

Do this with every point of the image and you have blurred image - that is how blur works - it performs above operation on every single point of the image.

Again - all the participants that scatter photons around from its true position have no clue if photon is coming from a star or from somewhere else - they will treat it the same.

[ollypenrice]

1 minute ago, vlaiv said:

I can't really convince you that this is indeed so if I don't start with highly technical stuff and math.

Best I can do to show you that this is in fact true (without going into technical stuff) is to do the following:

I do hope we are in agreement that for all intents and purposes stars in astronomical image are in fact point sources (before the light reaches earth's atmosphere and aperture of telescope).

Imagine following scenario to understand what a blur is - there is only one star in the sky and nothing else - no light pollution, no nebulosity, nothing. Photons from this star always arrive from same angle in the sky - from same point, and in perfect world - they would all be focused into same spot, right?

However, we have atmosphere, we have aperture of telescope and we have aberrations. All of these conspire together against our poor photon so it does not land on a single point but gets thrown randomly around this point. Every photon that comes from this star gets the same treatment - majority of them are thrown off course just a bit, some are thrown off course a bit more. When they all accumulate - we get nice bell shaped distribution of photons with certain FWHM. In some scopes it will be nice gaussian like shape - where seeing dominates over diffraction effects, while in other scopes we won't get nice gaussian type shape but it will still be sort of bell shape but little distorted - this is where aberrations dominate seeing (seeing should be fully random and produce true gaussian shape - others are predictable and produce different shapes and result is combination).

Ok, now that we know how star light behaves - it gets scattered in particular way around that single point - we will call this "blur" as it blurs star into some non point like shape.

Punchline - atmosphere, telescope, sensor - all of them have no idea if that photon came from a star or from nebulosity or from some galaxy. All photons will receive the same treatment. Every point in the image will be blurred regardless of the source of photon - it is just important that photon arrives from same spot in the sky and it will get scattered around in that bell shape.

Do this with every point of the image and you have blurred image - that is how blur works - it performs above operation on every single point of the image.

Again - all the participants that scatter photons around from its true position have no clue if photon is coming from a star or from somewhere else - they will treat it the same.

Perhaps we should be talking about camera-telescope systems.

I will take a moment to send you a crop from a RASA 8 image, given a basic stretch, which includes both stars and detailed nebulosity. It will not be at all difficult for me to find such a crop which shows poor stars and good nebulosity. It won't be tonight: I'm old and tired out!!

lly

[andrew s]

@vlaiv and @ollypenrice you are both right. @vlaiv is correct that that abberations are universal and impact both point like and extended sources. 

However, just as dimmer stars show less pronounced diffraction spikes, extended objects like planets tend not to show them. This is what @ollypenrice observes. Due to differences in contrast they seem for all practical purposes to be absent.

A classic example is curved spider vanes compared to straight ones. Both suffer diffraction but the curved blades result in a distributed low contrast result compared to the high contrast focused spikes of the straight ones.

Regards Andrew 

Edited August 26, 2023 by andrew s

[vlaiv]

7 minutes ago, andrew s said:

@vlaiv and @ollypenrice you are both right. @vlaiv is correct that that abberations are universal and impact both point like and extended sources. 

However, just as dimmer stars show less pronounced diffraction spikes, extended objects like planets tend not to show them. This is what @ollypenrice observes. Due to differences in contrast they seem for all practical purposes to be absent.

A classic example is curved spider vanes compared to straight ones. Both suffer diffraction but the curved blades result in a distributed low contrast result compared to the high contrast focused spikes of the straight ones.

Regards Andrew 

I'm not sure diffraction spikes are good comparison point.

Their intensity is always percentage of brightness of the source that is creating them. They are equally present on stars as they are on extended objects like planets. Take image of Jupiter taken with reflector with spider, and stretch it very hard - you will get diffraction spikes from planet as well.

We just don't see them because intensity needed to notice them is very very large and comes from brightest stars only in normal stretch levels.

Here is old planetary capture of mine extremely stretched:

Besides that strange circular feature that I think is refection of a barlow lens used and shows aperture of the telescope (some unfocused light) in top corner - spikes are starting to show

But back on issue of blur - FWHM is the same for star and for nebulosity and nebulosity simply won't show features that are that order of size or smaller - due to blur.

FWHM is important bit. That and how much high frequency components there are (or are expected) in signal itself.

Take uniform light without variation - no matter how much you blur it - it will look the same. Sky in daytime photo for example - if it's clear and there are no clouds - it will look the same in sharp and in blurred image.

If you expect smooth varying nebulosity and that is what you get - you might think it was not affected by blur in the same way stars were - but you'd better check other sources to verify if nebulosity is indeed smooth or not.

[ONIKKINEN]

58 minutes ago, andrew s said:

If you want the one to rule them all go for an 8 or 10 inch f5 Newtonian. Add quality coma corrector, appo reducer plus barlow/power mate and you have it all. 😊

Regards Andrew 

Took the words right out of my mouth. Well, i am biased since i have an 8'' newtonian, but i think this is as close to a jack of all trades scope as can be.

Planetary and lunar is easy with the right corrector/camera combo, as is high-ish resolution seeing limited DSO imaging with a good quality coma corrector. Wider field is plausible with something like the Starizona 0.75x corrector, although dont have myself so maybe shouldn't advocate for it. But mosaics at f/5 are still reasonably fast, although not competing against the RASA.

I think a newtonian still has the best chance of being a scope that can fit all imaging purposes even somewhat.

[andrew s]

3 minutes ago, vlaiv said:

I'm not sure diffraction spikes are good comparison point.

Their intensity is always percentage of brightness of the source that is creating them. They are equally present on stars as they are on extended objects like planets. Take image of Jupiter taken with reflector with spider, and stretch it very hard - you will get diffraction spikes from planet as well.

We just don't see them because intensity needed to notice them is very very large and comes from brightest stars only in normal stretch levels.

Here is old planetary capture of mine extremely stretched:

Besides that strange circular feature that I think is refection of a barlow lens used and shows aperture of the telescope (some unfocused light) in top corner - spikes are starting to show

But back on issue of blur - FWHM is the same for star and for nebulosity and nebulosity simply won't show features that are that order of size or smaller - due to blur.

FWHM is important bit. That and how much high frequency components there are (or are expected) in signal itself.

Take uniform light without variation - no matter how much you blur it - it will look the same. Sky in daytime photo for example - if it's clear and there are no clouds - it will look the same in sharp and in blurred image.

If you expect smooth varying nebulosity and that is what you get - you might think it was not affected by blur in the same way stars were - but you'd better check other sources to verify if nebulosity is indeed smooth or not.

I don't think what you say here contradicts what I said. It's a matter of contrast .

Obviously,  the FWHM gives an estimate of the maximum spacial frequency that can be seen (but it's complex point v edge  v gradient etc.) .

Your examples seem to confirm what I was saying. 

I don't think @ollypenrice is disputing the resolution of the RASA.

Regards Andrew 

 

[vlaiv]

Just now, andrew s said:

I don't think @ollypenrice is disputing the resolution of the RASA.

As far as I understood Olly - he claims that star size don't correlate with background detail in RASA data - that somehow stars are being large / soft (sign of high level of blur - high FWHM) - but that detail is still there in the background.

[andrew s]

17 minutes ago, vlaiv said:

As far as I understood Olly - he claims that star size don't correlate with background detail in RASA data - that somehow stars are being large / soft (sign of high level of blur - high FWHM) - but that detail is still there in the background.

I'll have to let Olly comment on that.

However,  what counts to the eye is the detail it can pick out. A bright star will have an obvious impact over many pixels. A chain of dim stars (with the same FWHM as the bright star) might well be visible as a linear feature just a pixel wide.

This difference between point and linear resolution was well known to well respected visual observers of the past.

It would be easy to assume by looking at the bright star blur seeing the pixel wide feature would be impossible to see.  (I am not saying you are doing this.)

I feel though this effect may be the root of your different positions. 

Regards  Andrew 

Edited August 26, 2023 by andrew s

[vlaiv]

48 minutes ago, andrew s said:

This difference between point and linear resolution was well known to well respected visual observers of the past.

It would be easy to assume by looking at the bright star blur seeing the pixel wide feature would be impossible to see.  (I am not saying you are doing this.)

I feel though this effect may be the root of your different positions. 

Visibility of line features is not directly related to resolution (it is indirectly related, but I think you know that). It is the same as saying - angular size of star is smaller then resolving power of telescope - therefore we should not be able to see it.

Seeing / detecting is one thing - resolution is another. Resolution is tied to resolving stuff - having two stars next to each other and identifying them as separate - having two lines next to each other and resolving them as separate features, or resolving the shape of the feature rather than just saying - hey there is this smudge there .

I think that we perfectly understand each other. Resolution that we speak of will impact even transition from dark to light region - it won't be clear sharp transition any more (much like line won't be line but wider feature and star is no pin point but is circular blob).

As far as different positions, yes, it's best for Olly to clearly state his position and even better - provide example he mentioned when he gets the chance to do so.

[ollypenrice]

11 hours ago, vlaiv said:

As far as I understood Olly - he claims that star size don't correlate with background detail in RASA data - that somehow stars are being large / soft (sign of high level of blur - high FWHM) - but that detail is still there in the background.

Here's a close crop at full size from a work in progress. The stars are lousy but, in my view, Trunk details are good.

I can fix the stars very easily, however, and get them good enough for my own taste. They are still not the best but I can live with them...

 

I'm not concerned by how much of this arises from my processing workflow because it's the workflow I use and it does what I want it to do. What you see above is the working reality of using RASA data. The non-stellar parts of the image neither had nor needed any intervention. RASA stars do need attention, though. They also needed it when I was using an earlier workflow.

Another example,. Before:

After

In this case the difference is slight, seen as a crop, but an extended starfield looks much better with the adjustment.

When using good refractors the stars needed far less attention and I've had to come up with efficient solutions for adjusting RASA stars.

Olly

 

 

Edited August 27, 2023 by ollypenrice
typo

[tomato]

Certainly if you go by the number of scopes being used for DSO imaging, then an 8’ Newtonian must  be close to the top of the list, but there are a lot of 80mm refractors out there also. However, I think they tend to be favoured by nebula imagers, small galaxies, it seems to me, are a less popular group of targets to image.

Perhaps the title of my original post was misleading, I did mean it to apply to DSO imaging (all objects) only.
 

 I tried DSO imaging originally back in the 80’s and 90’s with an 8’” F4 Newtonian, and the frustrations I experienced with that scope persuaded me to buy a refractor when I took up the hobby again 35 years later.

Another great DSO imaging scope which I think deserves a mention as a candidate for a one stop scope is the MN190, but I will stick stick with the RASA11, aperture is good.

[ollypenrice]

16 minutes ago, tomato said:

Certainly if you go by the number of scopes being used for DSO imaging, then an 8’ Newtonian must  be close to the top of the list, but there are a lot of 80mm refractors out there also. However, I think they tend to be favoured by nebula imagers, small galaxies, it seems to me, are a less popular group of targets to image.

Perhaps the title of my original post was misleading, I did mean it to apply to DSO imaging (all objects) only.
 

 I tried DSO imaging originally back in the 80’s and 90’s with an 8’” F4 Newtonian, and the frustrations I experienced with that scope persuaded me to buy a refractor when I took up the hobby again 35 years later.

Another great DSO imaging scope which I think deserves a mention as a candidate for a one stop scope is the MN190, but I will stick stick with the RASA11, aperture is good.

The MN190 deserves a makeover with quality hardware around the optics. 1000mm FL and F5 is right on the money for galaxies and nebulae.

The joy of widefield and a fast system, though, is that you can do something new or rarely seen. Whatever you do on M51 will have been done better by the professionals. That doesn't stop it from being satisfying but getting a new perspective on a region is an incomparable buzz for me.

Olly

[vlaiv]

1 hour ago, ollypenrice said:

The non-stellar parts of the image neither had nor needed any intervention. RASA stars do need attention, though. They also needed it when I was using an earlier workflow.

I don't know - to me it looks all mushy and without distinct detail. I pulled 3 different images here from SGL, did not pay attention to aperture, and sure, some of them are narrowband or enhanced by NB data - but feature should still be there and should be sharp:

[ollypenrice]

7 minutes ago, vlaiv said:

I don't know - to me it looks all mushy and without distinct detail. I pulled 3 different images here from SGL, did not pay attention to aperture, and sure, some of them are narrowband or enhanced by NB data - but feature should still be there and should be sharp:

OK, I've probably zapped that feature in processing but, then again, I can see features in ours not seen in the others. The question marked feature is there in ours but softer - as you'd expect from broadband.

Of those posted, I think ours has the most information. This may be down to different ways of looking, of course.

Olly

[vlaiv]

4 minutes ago, ollypenrice said:

OK, I've probably zapped that feature in processing but, then again, I can see features in ours not seen in the others. The question marked feature is there in ours but softer - as you'd expect from broadband.

Of those posted, I think ours has the most information. This may be down to different ways of looking, of course.

Olly

Well, you can't get around physics, and it says that whatever affects stars, affects the whole image.

Here, look at this:

If I reduce image to 40% - which is equivalent of using 3.76 / 0.4 = 9.4um pixel size - feature marked with arrow in the top part is now comparable across all images and sharpness detail on the edge of trunk looks very similar.

[Elp]

1 hour ago, vlaiv said:

to me it looks all mushy

I don't have access to large aperture out of choice but from my experience the sct usually comes last for definition (refractor then Newtonian then sct), my HS6 is a similar FL to my 60mm refractor, same cameras etc, the 60mm beats it for clarity. If the C6 was imaging at F6.3 however, it usually contains more detail but takes so much longer to get an equivalent signal. I experienced this when doing NGC2359 earlier this year, as a result I didn't use the F6.3 data, used the F2 HS for most of the nebulous regions and supplemented the data with the sharper 60mm data.

An 8/925/11/14 may perform differently to my setup.

Edited August 27, 2023 by Elp

[andrew s]

Putting aside differences in perceived sharpness I think there is possibly a technical reason for a difference between stars and hydrogen emission nebulae. 

Stars are wide band and subject to the full force of atmospheric and chromatic aberration . While the nebula is predominantly narrow band in the red and thus less impacted by the atmosphere and chromatic effects.

It's noticeable that @ollypenrice original stars have blue halos.

I doubt a reconciliation is possible though 😊.

Regards Andrew 

[vlaiv]

10 minutes ago, andrew s said:

Putting aside differences in perceived sharpness I think there is possibly a technical reason for a difference between stars and hydrogen emission nebulae. 

Stars are wide band and subject to the full force of atmospheric and chromatic aberration . While the nebula is predominantly narrow band in the red and thus less impacted by the atmosphere and chromatic effects.

I'm with you on that - if it were not for working resolution. It is ~2"/px and such differences should be visible in higher working resolutions but not so much on 2"/px. In wide field images FWHM per channel is roughly the same with very little difference.

From RASA paper - RASA8 data sheet

4.55um max RMS - equates to 2.34" RMS, which in turn equates to 5.51" FWHM over the field, just from optics.

@ollypenrice can you confirm above for us? Can you take one of your calibrated subs and measure average star FWHM on it? I wonder which value you'll get from measurement.

In any case - 5.5" FWHM requires sampling of 3.43"/px - or at 400mm FL that is equal to 6.65um pixel size - and this is without mount and seeing influence. So telescope alone requires at least that size of pixel.