OSC CCD Camera in Light Polluted Skies?

[BlueAstra]

Just read the article by our French Dept. on the comparison between OSC and mono CCD cameras (Atik 4000) in this month's Astronomy Now. What a fascinating discussion! Unfortunately, although the analysis is thorough, it can't come to a conclusion in LP environments.

I currently used a 1000D, but I would love to get a dedicated cooled CCD, but I have quite bad LP with Manchester being close by, and don't get out that often with a clear sky. I like the idea of at least getting some data in a short session with a OSC system, and the reduced outlay, but not sure about its performance in a LP environment.

Quite a few people in the USA seem to use OSC, but on the forum it seems to be all Mono. Does anyone have any comments on OSC v. Mono, especially in a LP environment?

[harry page]

Hi

I have a sx m25 OSC and I love it:hello2:

I also live in a town with bad light pollution and the gradients I get are a bit fierce sometimes , but nothing pixinsight does not sort out

I do like being able to capture a complete image with my limited imaging time.

If you ask me is it as sensitive as a mono , no its not but as our french friend did I got both

Harry

[moocher]

You'll need to include an idas or cls filter to your budget. I read the article too but olly has little or no lp!!! I think you'll be pleasantly surprised what you can achieve with a osc...

Alan

[Catanonia]

Mate, I live in Sandbach about 1 mile from jct 17 of the M6.

I needed a Astronomik CLS clip filter in my 1000D to get beyond 5min subs at iso 800.

My QHY9 mono with the 2inch Filter wheel LRGB option doesn't require any LP filters so far at 5 mins.

The filters seem to have the in built LP reductions as they are separate channels and well to be honest so far, are replacing the clip LP filter I used to use with the 1000D

See my pictures in Deep Space (recent ones) to see the impact.

Seriously, consider it and being bias, I know, the QHY9 mono + filter wheel LRGB is a great bit of kit, almost same size chip as the DLSR's and cheap.

You are more than welcome to visit and check out my setup. PM me if you like. Only 30 mins away from you.

[BlueAstra]

Thanks for the offer and the feedback. Its interesting that you get less LP with the separate filter setup. You'd think the luminance filter would have the same approximate passband as a 1000D ('cept in the IR), while the RGB would more or less be equivalent to the luminance but in three separate chunks. I suppose the RGB are not exactly 'top-hat' shaped transmission passbands, and the dips between the RG and GB transmissions may correspond to some LP lines. Will have to investigate the curves to figure that one out.

I assume if you go the Ha, Hb, O3 filter route then you are using really narrow passband filters and LP is laughed at! Do you then still need a luminance filter, letting all the LP back in again?

And how do you focus? Its a long walk from the focussing knob to the laptop! Easy on the back of the 1000D using a mask.

[moocher]

When you say NB I assume you mean HST? In which case the luminance id the Ha data...

Focusing can be tricky with NB filters but its not impossible.

Alan

[moocher]

Actually that's not strickly true but the Ha data is always the strongest. You can image in lp and strong moonlight with NB filters..

Alan

[BlueAstra]

Sorry, what is HST?

[Astroscot2]

My QHY9 mono with the 2inch Filter wheel LRGB option doesn't require any LP filters so far at 5 mins.

The filters seem to have the in built LP reductions as they are separate channels and well to be honest so far, are replacing the clip LP filter I used to use with the 1000D

Cat, im not sure this is true, your QHY9 being more sensitive should pick up the LP just as your Canon did, as your imaging with different filters you must be processing the LP out when balancing the colour and not seeing the LP effects in the final image.

As Alan mentions OSC cameras can be used effectivly with the correct LP filters. The issue with using LP filters and OSC cameras is slightly more complicated with very short focal lenghts.

Mark

[moocher]

Sorry HST is Hubble palette....

Ha is assigned to the red channel

O3 is assigned to the Green channel

S2 is assigned to the Blue channel.

Alan:D

All these filters have a very narrow "band width" and are unaffected by LP or Moon light....

The only time you really can`t use them is the Spring as galaxies don`t respond to this type of imaging. HTH

[narrowbandpaul]

can I just clear up some confusion?

firstly, why get less LP with a CCD? Well it has all to do with the bit depth. Your DSLR is probably 12 bit (4096) levels whereas the CCD will be 16bit (65536) levels. There will be the same amount of light pollution, its just it will only occupy a few thousand levels at most out of 65000, so it seems less significant

secondly, if you do narrowband imaging, you dont need to use an L filter.

thirdly, you shouldnt use the Ha or any other single emission line as luminance. It will supress the other channels if you do that.

fourthly, the HST pallette is a term coined by my friend, who invented the whole technique in 2001 (and thats NOT russ croman btw). Stands for Hubble Space telescope as thats the scheme they use. And alan was not quite correct with his definition of its ordering. It is also called the wavelength ordered pallette, and is:

R= S2

G=Ha

B=O3

the pallette that alan is referring to is the CFHT pallette....Canada-France Hawaii Telescope

just thought i would clarify a few things....

[moocher]

oops...

Alan

[ollypenrice]

Paul as sharp as ever! (FWHM of less than 1...)

I'm sorry I hve no LP but I did confess to this in the article and look forward to following this thread, since I recommended a discussion of just this. (Actually I do have a bit of LP and the Antares project about to begin is right down in what bit I have. We will be using both cameras so I may find something worth passing on.)

Olly

[George]

Ive been toying with the idea of a dedicated CCD for ages and always decided to stay with the DSLR (currently the 1000D) simply because of the bad LP at my location. Rather than spend x000's Im going ahead with a super cooled mod this summer on the 1000D which will cost pennies, if I see a huge improvment on my images i may think again about a dedicated CCD.

[harry page]

Hi

As I said earlier I have a OSC and a mono camera and have tried the OSC and RGB through filters and straight lum images and I find that they are all subject to light pollution

Obliviously depending of the colour of the pollution will have a different on the RGB channels

I always use a IDAS filter whatever I am doing ( other than narrowband )

Harry

P.s. any body coming on here saying they have no light pollution should be banned

[adamsp123]

can I just clear up some confusion?

firstly, why get less LP with a CCD? Well it has all to do with the bit depth. Your DSLR is probably 12 bit (4096) levels whereas the CCD will be 16bit (65536) levels. There will be the same amount of light pollution, its just it will only occupy a few thousand levels at most out of 65000, so it seems less significant

just thought i would clarify a few things....

That explains why when I tried out my new Atik 383 the light pollution was much less obvious, thought something was wrong with the camera! LOL (12 bit canon 1000D I have been using versus 16 bit Atik)

[ollypenrice]

Ive been toying with the idea of a dedicated CCD for ages and always decided to stay with the DSLR (currently the 1000D) simply because of the bad LP at my location. Rather than spend x000's Im going ahead with a super cooled mod this summer on the 1000D which will cost pennies, if I see a huge improvment on my images i may think again about a dedicated CCD.

But George, if you have bad LP wouldn't a mono CCD and narrowbad be just up your street?

Olly

[BlueAstra]

can I just clear up some confusion?

firstly, why get less LP with a CCD? Well it has all to do with the bit depth. Your DSLR is probably 12 bit (4096) levels whereas the CCD will be 16bit (65536) levels. There will be the same amount of light pollution, its just it will only occupy a few thousand levels at most out of 65000, so it seems less significant

....

I have a query.

I presume LP can be expressed as signal/noise ratio. So if I have a DSLR and measure 10 units of signal and 2 unit of noise (LP) I would have a S/N ratio of 5. If I could now increase the bit depth (amplitude resolution) of my DSLR, but not my sensitivity, my measurement units would get smaller and I may measure 100 units of signal, but I would also measure 20 units of noise, so my S/N would still be 5. If I increased the sensitivity of my DSLR by a factor of 2 I would measure 200 units of signal and 40 units of noise, still a S/N ratio of 5.

Surely the only way to reduce the effect of LP is to selectively reject the LP noise signal by means of a LP filter, or narrowband filters, which affects the LP noise but not the signal? Doesn't the bit depth affect the resolution of the measurement, not the discrimination between signal & LP, since there's no way the bit depth can tell the difference?

I apologize in advance if I've missed something obvious since I'm just writing this as I think it

[Catanonia]

Well i am no expert with ccd, just beginning. But I know that my 10m subs are as good as if not better in terms of lp compared to my 1000d with a astronomik lp clip filter. This is the same setup but with qhy9 and the standard 2inch filter set.

Now maybe it is my interpretation of it due to seperate rgb channels but my pictures are much better on this setup with the clip lp in place.

Or perhaps it is just that the qhy9 kicks butt over the eos.

[George]

But George, if you have bad LP wouldn't a mono CCD and narrowbad be just up your street?

Olly

I like my images in colour

Seriously, I have given this a lot of thought Olly, not only would it be advantagous for LP but also lunar interference.....but I like my colour and I dont have a budget at the moment for replacing my imaging kit.

So I have to make do with what I have

[narrowbandpaul]

cheers olly!

Blue astra to answer your query:

its maybe not obvious but it is of fundamental importance when comparing situations to use electrons rather than DN (or ADU) to express amounts of signal and noise. To do this you need the overall system gain, which is roughly Gain=Full Well (e-)/max number of DN. Say you have a 16bit CCD with a quoted full well of 32,500e-, the gain is 32500/2^16 = 0.5e-/DN

that means that every ADU you see in your image corresponds to 0.5e-.

say then that in one image the background from LP is 1800DN. If we multiply this the gain, this corresponds to 900e-. The statistics of photon counting are such that N(e-) will generate sqrt(N) noise. So 900e- corresponds to 30e- of noise. We could define an SNR of 30.

now take an SLR with a full well of 20,000e- and 12bit depth. The gain would be 20000/4096, 5e-/DN. Assuming you use the same telescope and exposure time, and that the DSLR has the same quantum efficiency as the CCD, then you must recieve the same number of LP photons. And since there is a linear relation between photon number and electrons in the pixel, you MUST recieve 900e- of LP signal. In the camera though you will see a signal of 900/5 DN....180DN. Photon statistics dont apply to DN counts, so we must use the electron figure to determine the noise....

well thats just the same sqrt(900)...=30e-. Both camera will give you the same noise from LP! independent of the bit depth.

the SNR of the DSLR is also 30!

both cameras are equivalent. why isnt the SNR important?...well you can remove the average value of the background during post processing, so only the noise impact is important. Both cameras have the same noise.

thats why expressing things in DN units is irrelevant and meaningless, electrons must be used. Looking at the noise in DN from both cameras might lead you to believe that the DSLR suffers less (as it only has 180DN average and only 6DN noise) but if expressed in real units like electrons they have the same value.

so yes, you need an LP filter to reduce the noise from the background. A 24bit camera vs 12 bit camera will have no effect at all on image quality

hope that helps

paul

[BlueAstra]

OK, if I’m reading this right, as far as the underlying sensor goes (DSLR, Mono CCD, OSC CCD) the LP affects each the same and non has any fundamental advantage in LP rejection or less sensitivity to LP. The sensitivity to LP therefore depends on the optical filtering in place in front of the CCD, whether it be a LP filter, discrete LRGB filters, discrete HST filters or the colour filter mask on the OSC (maybe with additional LP filters).

I would think that discrete HST filters with a mono CCD (and possibly with much lower sensitivity, a OSC CCD) would have the best LP rejection, since they have the narrowest filter band passes.

Referring to the original question of whether OSC performs as well in LP environments compared with mono CCD using LRGB, I guess it depends on the filter responses of the discrete RGB filters compared with the response of the ‘RGB’ mask in the OSC. I have no idea what they are, so a practical experiment is needed! Perhaps one of you mono imagers out there in heavy LP locations could persuade a friendly company to lend the OSC version for a direct comparison.

My last thought is that a LP filter may be the dominant effect in LP reduction, and may be used on either setup to advantage if you are in a heavy LP environment. Either that or buy PixInsight with its DBR function!