atik 414ex mono/filters to ZWO ASI1600MC Pro Cooled Colour 4/3" CMOS USB 3.0 Deep Sky Imaging Camera

[iwols]

Am i mad going down this road mono/filter to colour ccd,my main reason for this is the complex processing (i know your all going to shout yes but)and the imaging time it takes ,using a ed80 at the minute on a permanent set up mount,any comments appreciated  ?

[RichLD]

I have had great results from OSC but find that dark skies are a must. A bright target like M31 gave me a good return after 10 X 600s using a SW ED80 with its dedicated FF/FR and a QHY8L.

Not sure how this will work with a fixed mount, unless you're planning high gain 'lucky imaging'.

I find processing OSC to be just as time consuming as mono, although calibration time is of course reduced due to the lower number of subs.

HTH

Rich

[iwols]

hi rich by fixed mount i mean permanent set up,sorry confused at the minute? not sure how dark my skies are here 

[Ibbo!]

Mono will allow you to take advantage of NB filters when the moon is arround and also helps with LP.

[Scott]

Olly... Olly... Where for art thou Olly.

 

 

 

[iwols]

how would a ha filter be on a osc 

[Ibbo!]

about a 1/4 as effective as mono

[tooth_dr]

Ha data from a colour camera is definitely doable. And I’ve found it pretty acceptable. But its 1/4 resolution compared to its mono. 

[Adam J]

1 hour ago, Ibbo! said:

about a 1/4 as effective as mono

No idea where that figure keeps coming from. Not its not 1/4 as effective by any maths I have seen.

[Adam J]

58 minutes ago, tooth_dr said:

Ha data from a colour camera is definitely doable. And I’ve found it pretty acceptable. But its 1/4 resolution compared to its mono. 

I think you will find its 1/4 the number of pixels but 1/2 the resolution at least in terms of Arcsec / pix.

[tooth_dr]

1 minute ago, Adam J said:

I think you will find its 1/4 the number of pixels but 1/2 the resolution.

You have a way with words Adam.  But better than I thought then ?

[spillage]

Is my maths awful or would that make it 1/8th as effective as mono if its only 1/2 a 1/4?

[Ibbo!]

My reasoning being the RGGB bayer matrix

Could you explain how my reasoning is wrong?

[RayD]

21 minutes ago, Ibbo! said:

My reasoning being the RGGB bayer matrix

Could you explain how my reasoning is wrong?

I'm intrigued and often wondered if you can shoot Ha with a colour CCD.  Is the loss of resolution on Ha because there is only one red lens on the matrix, so only 1/4 of the Ha passing?

[Ibbo!]

Just now, RayD said:

I'm intrigued and often wondered if you can shoot Ha with a colour CCD.  Is the loss of resolution on Ha because there is only one red lens on the matrix, so only 1/4 of the Ha passing?

Thats what I thought.

[RayD]

Just now, Ibbo! said:

Thats what I thought.

Ah ok, thanks.  I have no idea so it's not a leading question btw.  Seems to make sense. 

[tooth_dr]

42 minutes ago, RayD said:

I'm intrigued and often wondered if you can shoot Ha with a colour CCD.  Is the loss of resolution on Ha because there is only one red lens on the matrix, so only 1/4 of the Ha passing?

Lots of us shoot ha with our DSLRs quite successfully. There is also a member in here recently shooting Ha with a colour CCD. It might be Gorann?

[RayD]

12 minutes ago, tooth_dr said:

Lots of us shoot ha with our DSLRs quite successfully. There is also a member in here recently shooting Ha with a colour CCD. It might be Gorann?

That's really interesting, thanks.  I guess you just need more data then (compared to mono) due to the matrix?

[david_taurus83]

I shot the Veil with my 600d and an Astronomik 12nm Ha filter. Got 10 hours worth of 600s subs back in August. Split the channels after stacking and kept the red. It looks quite noisy, sensor was reading 20/25°C for the most, but I tried to reduce the star field using Morphological Transformation in Pixinsight without using effective masks. A more skilfull processor could probably maintain the nebulosity and improbe the background.

[Adam J]

2 hours ago, Ibbo! said:

My reasoning being the RGGB bayer matrix

Could you explain how my reasoning is wrong?

Happy to explain. Although it does depend on your definition of 4 x more effective.

Shooting via the RGGB matrix will indeed mean that you are collecting about 1/4 the number of photons assuming that the red filter is 100% transmitting at Ha. As it happens its usually a little less than 100% but not by a significant margin (sensor brand depending).

So that means that you get 1/2 the resolution in terms of Arc seconds / pixel. Or 1/4 the number of pixels. So if your measure of "effectiveness" is simply resolution based then you could make the argument that it is 1/4 as effective although i would say 1/2.

However in terms of image quality you are looking at a per pixel signal to noise ratio delivered at sufficient resolution (usually >1'' and less than 3'').

So lets assume that you get about 1.2'' per pixel from a OSC (think ASI1600mc vs mm on my 130PDS). In terms of Ha using only the red pixels you will in fact have 2.4'' / pixel resolution once you use super pixel debayering.

Now 2.4'' per pixel is not too bad and hence your not being overly limited in terms of resolution via the RGGB matrix in the case of Ha. So lets discard that as a factor measuring effectiveness.

So that leaves signal to noise. If we consider a Mono with the same pixel size as the OSC above then it will also give 1.2'' per pixel. So on a per pixel basis assuming that the red filter is 100% transmitting at ha you dont actually lose anything in terms of per Pixel signal to noise ratio. Where this changes is if you bin the mono 2x2 to match the effective resolution of the RGGB image at 2.4'' / pixel. Now you get an improvement of about 2.5x the signal to noise ration for a CCD and 2x for a CMOS (that gets more complex as to why you dont get the fill 4x with CCD hardware binning and its to do with where noise is added in the process, no time to explain this in detail). 

So in the end once you get down to it and you include the little bit poorer transmission via the red filter you end actually end up with about 3x the sensitivity in the mono camera...but only if you bin 2x2 to the same resolution as the OSC Ha image.

People tend to look at the jump from DSLR to dedicated camera and attribute too much of that to it being mono when infact lots of the difference is in the cooling at least in terms of Ha.

Now, moving on to OIII you might think its better because you are using GGB to pick up the OIII with OIII laying between blue and green, so 3/4 of the pixels. On the simple basis of your original statement you would think that you might get  3/4 the performance of the Mono from the OSC... but you dont because with OIII falling between the green and blue means that you are only getting 40% of peek transmission in the blue and 60% in the green if you are lucky, although some of the modern back side illuminated CMOS OSCs are doing much better than that these days. So in the end its OIII that really suffers and Ha is surprisingly good unless you compare it to binned mono data.

Take Davids example above and imagine how much improved it would be with the sensor at -20c instead of +20c (vast reduction in dark current).....I dare say better than only 1/4 of an equivalent mono sensor.

Adam

 

 

 

[Ibbo!]

Thank you.

Seems to make sense but the 4 pane image below took me about 5 hours with 2 filters ( but longer than that to get it to stitch together and my limited processing skills ) ?

I will stick with my mono ccd and filters.

[don4l]

I started out using a dslr.  I wouldn't dream of going back to colour.  I find processing mono data much easier.

I don't really understand all the maths, but I do remember all the problems.

This isn't as good as the examples above, but it is just three subs. 10m Ha, 20m SII, and 20m OIII - total exposure = 1 hour..  I've just banged it together while reading this thread.  You can tell from the hot pixels that I haven't even calibrated the subs.  I could not dream of doing this with a one shot colour.

 

 

[geordie85]

I admit I like the "convenience" of OSC as I don't have to worry about filter changes etc and can leave my set up imaging whilst I nap. No worries of filter wheel failure, aligning stacked images correctly etc. 

But I do miss mono imaging and would return to it in a heartbeat if funds allowed. 

[tooth_dr]

Heres my data comparison with same scope:

DSLR (OSC) 40d and 12nm Ha filter 19 x 600s

CCD (mono) 7nm Ha filter 32 x 600s.

 

[Adam J]

4 minutes ago, tooth_dr said:

Heres my data comparison with same scope:

DSLR (OSC) 40d and 12nm Ha filter 19 x 600s

CCD (mono) 7nm Ha filter 32 x 600s.

 

1/3 more interrogation in the mono with a filter thats almost half the width?