How to start? Color or mono

[ollypenrice]

31 minutes ago, iapa said:

I wish there were equivalent mono and OSC cameras produced by a company that could be used for a side by side comparison to put this to bed  

 

There are, aplenty. I wrote a side by side shoot out for Astronomy Now a good few years ago, comparing mono and OSC Atik 4000 cameras. I shot M42 in both cameras over the same exposure time and found the images to be very similar. Indeed I was so taken by the OSC that I bought it for use here. Longer aquaintance, though, showed it to be at much more of a disadvantage on fainter targets. When I decided that we needed larger format cameras I went for two monos instead. I'm not anti OSC, lots of people like them, but I'm keen to argue that they are not quicker and, in my view, nor are they easier. It's debatable, of course, which is easier.

Olly

[tomato]

I bought a OSC CCD to move from a DSLR, and although I can always produce a 'complete' image from a single session, they will never be as good IMHO as the equivalent (in total exposure time) LRGB image.  I think the factual information on how the Bayer matrix works supports this.

I plan to move to mono + filters very soon, but unfortunately with a different chip, or else I would have posted some back to back comparisons.

[LightBucket]

Any thought on the SXVR M25c OSC camera, I was told they are very good for a OSC, and they have a 25 x 15mm sensor....

[frugal]

I would definitely go mono, for no other reason than this: With OSC you can only do RGB. With mono you can do: mono, RGB, LRGB, HaRGB, Ha, Narrowband or any other combination of the above.

[Adam J]

On 08/04/2017 at 11:09, frugal said:

 would definitely go mono, for no other reason than this: With OSC you can only do RGB. With mono you can do: mono, RGB, LRGB, HaRGB, Ha, Narrowband or any other combination of the above.

Yes I would also go for a mono 100% but I also feel the need to correct you as its just not true to say that a OSC can only do RGB.

Here is the Rosetta taken with a Canon 550D DSLR and 2 inch H-a and OIII filters.

Its cropped because I did not have my CC spacing correct so the quality of the outer edges of the image was not really usable, same problem caused the artifacts on the edges of the stars as I fiddled with the spacing for the OIII data and then they did not quite line up as the stars were a different shape from the H-a image. All fixed now.

Now would I try that with a large pixel CCD OSC....maybe not depending, you get away with the loss in resolution more with a DSLR that has 18mp.

This was about 16x600s Ha and 16x600s OIII at ISO800 if memory serves. By the way SII works just fine too.

Now if this was a mono CCD would it be much better...yes you can bin, the resolution may be higher in some cases and its more sensitive. My point is just that this is a Bi-color narrow band image taken with a OSC camera be it a DSLR.

[jACK101]

I have read this thread in bemusement. I have also read a few books on Astrophotography and none of them go into this level of detail. Can you recommend a book which goes step by step through the theory you have been describing, or do you have to gain this experience by making possibly expensive mistakes?

 

Thanks

[ollypenrice]

10 hours ago, Adam J said:

Yes I would also go for a mono 100% but I also feel the need to correct you as its just not true to say that a OSC can only do RGB.

Here is the Rosetta taken with a Canon 550D DSLR and 2 inch H-a and OIII filters.

Its cropped because I did not have my CC spacing correct so the quality of the outer edges of the image was not really usable, same problem caused the artifacts on the edges of the stars as I fiddled with the spacing for the OIII data and then they did not quite line up as the stars were a different shape from the H-a image. All fixed now.

Now would I try that with a large pixel CCD OSC....maybe not depending, you get away with the loss in resolution more with a DSLR that has 18mp.

This was about 16x600s Ha and 16x600s OIII at ISO800 if memory serves. By the way SII works just fine too.

Now if this was a mono CCD would it be much better...yes you can bin, the resolution may be higher in some cases and its more sensitive. My point is just that this is a Bi-color narrow band image taken with a OSC camera be it a DSLR.

That's a nice Rosette but I don't know that I agree with your concerns over large pixels. An OSC camera will deliver good resolution, provided it is debayered by a good algorithm which effectively derives what is going on in each colour between that colour's filters on the matrix. In theory a mono camera will give better resolution because it is actually measuring pixel by pixel while an OSC is interpolating - but in reality I never found much difference. My objection to OSC is speed, especially in narrowband. Craig Stark has written about debayering algorithms somewhere in net land!

I would only discuss resolution in units of arcsecond per pixel or pixel dimensions. Pixel counts alone can be misleading.

Olly

[iapa]

On ‎07‎/‎04‎/‎2017 at 22:49, ollypenrice said:

There are, aplenty. I wrote a side by side shoot out for Astronomy Now a good few years ago, comparing mono and OSC Atik 4000 cameras. I shot M42 in both cameras over the same exposure time and found the images to be very similar. Indeed I was so taken by the OSC that I bought it for use here. Longer aquaintance, though, showed it to be at much more of a disadvantage on fainter targets. When I decided that we needed larger format cameras I went for two monos instead. I'm not anti OSC, lots of people like them, but I'm keen to argue that they are not quicker and, in my view, nor are they easier. It's debatable, of course, which is easier.

Olly

Ever get that 'oh, sod it' moment, give up on the research and discussion and just bite the bullet, and see for yourself??

Cooled mono 8 position filter, LRGB and NB filters, and equivalent cooled colour. If I have a massive preference, sell the one I don't want.

Unless I change my mind in the next couple of weeks LOL.

 

[Adam J]

11 hours ago, ollypenrice said:

That's a nice Rosette but I don't know that I agree with your concerns over large pixels. An OSC camera will deliver good resolution, provided it is debayered by a good algorithm which effectively derives what is going on in each colour between that colour's filters on the matrix. In theory a mono camera will give better resolution because it is actually measuring pixel by pixel while an OSC is interpolating - but in reality I never found much difference. My objection to OSC is speed, especially in narrowband. Craig Stark has written about debayering algorithms somewhere in net land!

I would only discuss resolution in units of arcsecond per pixel or pixel dimensions. Pixel counts alone can be misleading.

Olly

Huum not sure on that one, I would agree that you can regain resolution via cleaver debayering in an RGB image but I would not agree that the same applies to narrow band imaging on an OSC which was the point i making as I have always found that you get best results by removing the unused pixels and then using superpixel debayering, which is in fact reassuringly also the mode recommended in DSS.

While you do regain most of the resolution via debayering in an OSC when you shoot standard RGB, partly as there is overlap between the RGB filters in an OSC camera meaning that information is almost always contained in all the pixels (unlike the separate filters used in a mono), the problem is that you dont have that effect in H-a as you process the image to remove the green and blue pixels (that just contain noise anyway) leaving only the red. So there is nothing really left to interpolate with. Yes you can in effect blend between pixels but that will just make the image look softer at high zoom exactly the same as you would see by up-scaling it post debayering, it wont gain you information like you gain in an RGB image. But maybe I am missing something.

The way I understand it is that all that leaves you with half the resolution in terms of arcseconds per pixel as you would get with a mono with the same sized pixels when imaging in H-a. As most OSC CCD do tend to have larger pixels than DSLR that will cause you to under-sample if you try h-a on something like a QHY8L on a ED80 or 130PDS. I know because I tried in on my old 1000D (5.7nm pixels) and found the resulting image blocky where as I dont have that trouble with my 550D (4.3nm) due to its smaller pixels, in fact that was the main reason I changed to the 550D. So if you think that a QHY8L is 7.9um for about 5 arcseconds per pixel when you use Ha with it you can see the issue. Of course there are other OSC with smaller pixels and different scopes available, but to be honest that is why I said "maybe not depending" as I was just considering the general rule that CCD pixels are larger. Bottom line if you half your sampling then your always going to have less head room before you end up under sampling.

Yep mono is both higher resolution and faster. I was only correcting the statement by FRUGAL that you could only do RGB with an OSC and not narrow band as that is demonstrably not true.

[ollypenrice]

10 hours ago, Adam J said:

Huum not sure on that one, I would agree that you can regain resolution via cleaver debayering in an RGB image but I would not agree that the same applies to narrow band imaging on an OSC which was the point i making as I have always found that you get best results by removing the unused pixels and then using superpixel debayering, which is in fact reassuringly also the mode recommended in DSS.

While you do regain most of the resolution via debayering in an OSC when you shoot standard RGB, partly as there is overlap between the RGB filters in an OSC camera meaning that information is almost always contained in all the pixels (unlike the separate filters used in a mono), the problem is that you dont have that effect in H-a as you process the image to remove the green and blue pixels (that just contain noise anyway) leaving only the red. So there is nothing really left to interpolate with. Yes you can in effect blend between pixels but that will just make the image look softer at high zoom exactly the same as you would see by up-scaling it post debayering, it wont gain you information like you gain in an RGB image. But maybe I am missing something.

The way I understand it is that all that leaves you with half the resolution in terms of arcseconds per pixel as you would get with a mono with the same sized pixels when imaging in H-a. As most OSC CCD do tend to have larger pixels than DSLR that will cause you to under-sample if you try h-a on something like a QHY8L on a ED80 or 130PDS. I know because I tried in on my old 1000D (5.7nm pixels) and found the resulting image blocky where as I dont have that trouble with my 550D (4.3nm) due to its smaller pixels, in fact that was the main reason I changed to the 550D. So if you think that a QHY8L is 7.9um for about 5 arcseconds per pixel when you use Ha with it you can see the issue. Of course there are other OSC with smaller pixels and different scopes available, but to be honest that is why I said "maybe not depending" as I was just considering the general rule that CCD pixels are larger. Bottom line if you half your sampling then your always going to have less head room before you end up under sampling.

Yep mono is both higher resolution and faster. I was only correcting the statement by FRUGAL that you could only do RGB with an OSC and not narrow band as that is demonstrably not true.

You're quite right. There can be no interpolating with NB filters over an RGGB matrix.

Olly