Monochrome VS. Color CCD Cameras

[Alveprinsen]

It seems to me all the serious amateurs, semi-pro's and professionals out there use Monochrome CCD cameras and filters for their astro imaging.

Having looked at a few CCD cameras, I've come to notice that the color versions are only marginally more expensive, but have double or nearly triple the resolution.

Take the Orion Starshoot III Monochrome CCD camera for example, with its 1392x1040 resolution, and then the:

Orion Starshoot Pro DSCI v2.0 Color CCD Camera with its 3032x2016.

So whats the deal?

Alveprinsen

[Stu_2011]

It seems to me all the serious amateurs, semi-pro's and professionals out there use Monochrome CCD cameras and filters for their astro imaging.

Having looked at a few CCD cameras, I've come to notice that the color versions are only marginally more expensive, but have double or nearly triple the resolution.

Take the Orion Starshoot III Monochrome CCD camera for example, with its 1392x1040 resolution, and then the:

Orion Starshoot Pro DSCI v2.0 Color CCD Camera with its 3032x2016.

So whats the deal?

Alveprinsen

It's a can if worms and everyone has a view on mono vs colour I found.

I did lots of research and came to the conclusion that mono is the best if you have time and good weather on your hands. I went for an ATIK one shot colour camera and I'm very happy with it. My time is fairly limited coupled with awful British weather so the OSC suits me, not everyone will agree.......

The resolution issue is a bit misleading sometimes on the colour cameras due the Bayer matrix thing. This means that 4 pixels (red, green and 2 blue in a square array I think) are combined to produce 1 colour pixel so the actual resolution of the final image is less. I think this is right, I'm sure I will be corrected if not.

Thanks

Sent from my iPad using Tapatalk

[Stu_2011]

It's a can if worms and everyone has a view on mono vs colour I found.

I did lots of research and came to the conclusion that mono is the best if you have time and good weather on your hands. I went for an ATIK one shot colour camera and I'm very happy with it. My time is fairly limited coupled with awful British weather so the OSC suits me, not everyone will agree.......

The resolution issue is a bit misleading sometimes on the colour cameras due the Bayer matrix thing. This means that 4 pixels (red, green and 2 blue in a square array I think) are combined to produce 1 colour pixel so the actual resolution of the final image is less. I think this is right, I'm sure I will be corrected if not.

Thanks

Sent from my iPad using Tapatalk

It's a can of worms.......not can if worms, more coffee required...

Sent from my iPad using Tapatalk

[Gina]

I was under the impression that most of the astro CCD cameras were available in both mono and colour versions at the same price. That certainly applies to Atik cameras. The resolution is the same. The difference is that the colour cameras have a Bayer colour matrix on top of the sensor thus the real resolution is actually a quarter of that of the mono version. Each group of 4 pixels have 1 red, 2 green and 1 blue pixel. One group of 4 mono pixels with filters form a single colour pixel.

[pete_l]

Whichever technology you decide to choose, there are trade-offs.

If you go for One Shot Colour cameras (OSC) you benefit from a simple imaging process. Every shot you take can be used as it comes out of the camera and it can simply be stacked with all the other OSC images you took. The camera interpolates the various single-colour pixels for you and (as people found out when they were developing colour TV) the eye is less sensitive to low-resolution colour than it is to low-res. luminance, so the "loss" of resolution isnt really an issue.

On the down-side, you're limited to full-spectrum images - you can't really use an OSC camera for Halpha or other narrow band imaging, and you can't "bin" colour images (apart from a few, sepcialised models) to deeper into the image.

For Mono cameras, you get higher resolution Luminance images, so they can appear sharper. You can also take narrow band images at some astronomically significant wavelengths : Ha, S2, O3 to give striking results. You could also argue that everything you see visually through a telescope is seen as either dark or light: no colour, so a B&W image is closer to "visual reality", and some of the world's most impressive photo-journalism images are monochrome, too. A mono camera also gives you the option of imaging inn the I.R - outside the visual range.

However, there are some drawbacks. If you want colour, you need to spend a lot longer taking images as you need at least 3 sets of exposures (plus Luminance, if you're being fussy ). You also need the added complication and expense of a filter wheel and a set of filters. You'll also find that you almost certainly need to tweak your telescope's focus every time you change filters. Finally, the post-processing of LRGB (and possibly additional narrow band) sub-images is more complex. NOt only do you have more to stack, but you need to make some artistic judgements regarding the colour balance. This is easy to mess up and is a matter for your own personal choice, as there are rarely any universally correct answers,

[Velvet]

Again it comes to preference. Mono you will get more detail I think but requires a little more time and effort.

Most people are happy with detail from OSC but mono will give more, its the 4 pixels to 1 that makes the difference

I have changed from OSC to mono and i know some folks that have gone from mono to OSC only to return to mono in the end.

Its personal preference I think both have pros and cons,

Velvet

[Stu_2011]

I was under the impression that most of the astro CCD cameras were available in both mono and colour versions at the same price. That certainly applies to Atik cameras. The resolution is the same. The difference is that the colour cameras have a Bayer colour matrix on top of the sensor thus the real resolution is actually a quarter of that of the mono version. Each group of 4 pixels have 1 red, 2 green and 1 blue pixel. One group of 4 mono pixels with filters form a single colour pixel.

Slightly off topic but I get confused with the whole resolution thing, is it 1/4 or 1/2? Its a quarter in area but half in either x or y, I think the x and y is more relevant. Also, I'm sure I read somewhere that with modern OSC and good image processing the actual resolution drop of OSC compared with mono is more like 30%.

I'm happy with my OSC, but it is a new ATIK 428.

Thanks

Sent from my iPad using Tapatalk

[GlassWalker]

1/2 linear resolution is same as 1/4 area resolution. Just different way of expressing it.

If the figures are representative or not depends in part on the subject. If you're imaging a Ha rich emission nebula with OSC, then the 1/4 area resolution is quite appropriate. If you're imaging a broad spectrum galaxy, then the penalty is much smaller.

[Gina]

Slightly off topic but I get confused with the whole resolution thing, is it 1/4 or 1/2? Its a quarter in area but half in either x or y, I think the x and y is more relevant. Also, I'm sure I read somewhere that with modern OSC and good image processing the actual resolution drop of OSC compared with mono is more like 30%.

I'm happy with my OSC, but it is a new ATIK 428.

Thanks

Sent from my iPad using Tapatalk

That's fine then Great camera And I agree that with the weather in this country an OSC camera makes a lot of sense. It all depends on your interests, the amount of light pollution and other environmental aspects of your location. I'm very interested in narrow band AP so I've bought a mono CCD, filter wheel and filters. That works out a lot dearer than an OSC camera of similar spec. I have DSLR for OSC at present but will be adding LRGB filters in time for my mono system. The DSLR gives me OSC at high resolution for a modest price but the quality has been improved by DIY mods.

[jabberwocky]

A certain author of a very popular AP book who is a member on here was telling us at his lecture at SGL7 that colour cameras were the way ahead. If they are good enough for him I'd say they are good enough for anyone.

You watch he will now post saying mono is the way ahead, lol.

[ollypenrice]

There are one or two things above with which I disagree. I've had mono and OSC versions of the Atik 4000 for some time, having written a shoot out article for Astronomy Now a couple of years ago.

Both are good cameras and it does come down to preference. However...

OSC imaging is, without any doubt at all, slower than imaging in a mono camera. In a mono you shoot luminance without obscuring two thirds (roughly) of the spectrum with colour filters. You then add colour which can be de-noised without loss of resolution (the luminance reinstates this) so you can get away with not so much. You can also bin the colour 2x2, so saving time. But, and here's the rub, if you are cut off in your prime by the good old English weather you can end up with LRGCloud, which is not a lot of good. I don't want to speak for Steve but when we've chatted about this in the past he's given this as a major reason for his liking OSC. I can quite understand that. My own reason is that I do this for a living and can be running round between mounts like a headless chicken, so having at least one simple capture going on is rather nice!

Resolution; my understanding is that OSC cameras are not as simple minded as some people think! The software interpolates the data from the three colours to generate a synthetic luminance layer. So imagine a curved, sharp edged Ha nebula, ie a red one. Its curved edge will be represented, in a mono camera, as a series of little one-pixel sized steps. Now in the OSC only the red pixels will pick up this edge, so you might think it can only be represented in the larger steps of the RGGB matrix of 4 pixels. But no, if my undertanding is correct the software will create a pixel by pixel interpolation of the curved edge in synthetic luminosity.

There is a loss of colour resolution, but does that matter? I doubt it. If you convert an RGB image to Lab colour mode in Photoshop and blurr the two colour channels like crazy (ie the l and b channels) you will probably see no loss of resolution when the channels are recombined.

You can shoot Ha quite successfully in an OSC camera but clearly a mono will do the job a lot better. Some objects will never sing without an Ha layer but on others it really doesn't make any odds.

Capturing OSC is easier than capturing mono and having a single flat rather than five to do is pure heaven!

Caturing the colours simultaeously means that even if the seeing is bad it is bad for all channels so you stars will at least be the same size in each.

Shooting low in a mono you can be crafty and shoot the red at the lowest elevation (it scatters least), the bue at the highest (it scatters most) and the green whenever.

Processing an OSC image is not easier. It is best to process the synthetic L layer in a different way from the colour layer, just as it is in mono processing.

Not all camera chips respond well to the Bayer matrix. One I would absolutely never consider is the OSC version of the Kodak 8300. I've seen these in action (if that's the right word to use for nothing happening!) and they are just too slow, slower than a modded DSLR. The Atk 4000 OSC, however, is only marginally slower than the mono.

Usually you get your OSC colour filters for free, as Gina says, and this is quite a saving.

Most but not all of the finest imagers use mono. Greg Parker/Noel Carboni work in OSC and so did Dietmar Hager for many of his best results.

Ah well, I've now sold my Atik 4000 OSC but not because I didn't like it. I liked it a lot, but I'm forking out for a full frame Atik 11000 and need the cash!

Olly

[GlassWalker]

Resolution; my understanding is that OSC cameras are not as simple minded as some people think! The software interpolates the data from the three colours to generate a synthetic luminance layer. So imagine a curved, sharp edged Ha nebula, ie a red one. Its curved edge will be represented, in a mono camera, as a series of little one-pixel sized steps. Now in the OSC only the red pixels will pick up this edge, so you might think it can only be represented in the larger steps of the RGGB matrix of 4 pixels. But no, if my undertanding is correct the software will create a pixel by pixel interpolation of the curved edge in synthetic luminosity.

There is a loss of colour resolution, but does that matter? I doubt it. If you convert an RGB image to Lab colour mode in Photoshop and blurr the two colour channels like crazy (ie the l and b channels) you will probably see no loss of resolution when the channels are recombined.

At the end of the day, you're still interpolating data you never got in the first place. You could arguably interpolate mono data too, so the performance gap between them will remain. So all else not limiting, having bayer pattern would mean loss of resolution compared to mono, of both luminance and colour data.

In general photographic terms, on a bayer matrix the green detectors are used as luminance data, which still needs interpolation. Red/blue provide chroma data. There's occasional talk about with introducing "L" elements into the filter array, but they never caught on. Currently Fuji have an unusual variation, which is very heavy on green detectors and even sparser red/blue. It seems to work well for regular photography in low light conditions.

Olly, hope someone else picks up that '4000 mono before I get tempted!

[Coco]

Again it comes to preference. Mono you will get more detail I think but requires a little more time and effort.

Most people are happy with detail from OSC but mono will give more, its the 4 pixels to 1 that makes the difference

I have changed from OSC to mono and i know some folks that have gone from mono to OSC only to return to mono in the end.

Its personal preference I think both have pros and cons,

Velvet

Me too... twice!!!

[r3i]

So far I've only used OSC cameras - up to now I've have had to set up and tear down my gear at each session and adding to that the rarity of clear skies in the UK, the convenience of the OSC has definitely suited me. However, you cannot ignore the added value you get with mono CCDs from things like narrowband and the sharpness achievable with the luminosity subs. When my obs finally get completed and I'm a bit further up the imaging learning curve, some kind of mono setup will surely beckon.

[Gina]

I've been following this debate with great interest. One thing that occurs to me is that if you take a particular mono sensor and add a Bayer matrix you are going to make it less sensitive since the Bayer cuts out light. So I'm thinking that for faint DSOs the mono camera won't require any more imaging time overall than the OSC and if binning 2x2 for the colours it might actually be less.

[ollypenrice]

I've been following this debate with great interest. One thing that occurs to me is that if you take a particular mono sensor and add a Bayer matrix you are going to make it less sensitive since the Bayer cuts out light. So I'm thinking that for faint DSOs the mono camera won't require any more imaging time overall than the OSC and if binning 2x2 for the colours it might actually be less.

Gina, trust me, OSC is always slower than mono for a given quality. I've done the same object in both several times. With the Atiks, on a bright object it made little difference. On a faint one it could make quite a lot, maybe 20%. The reason, as you say, is that Lum shooting is full spectrum so about three times the light. Another way to think of it is that in an OSC you are wasting time shooting for colour information that you don't need. What you need is detail (luminance) or the particular response of an emission line, usually Ha. Some say that you need equal doses of L,R,G and B but I don't find this to be so. I find I can use, say, an hour per colour and 3 to 4 hours of Lum. The colour may look washed out at first but there are ways around that in processing. You can end up with fully adequate saturation. On the French forum I'm quite often told that my colour is over saturated.

In an Ha dominated object you can shoot truly minimal colour (as little as 30 mins per channel at F4) and this will give you stunning star colour. (It doesn't burn out the cores.) The Ha will do the rest.

Olly

[Gina]

Gina, trust me, OSC is always slower than mono for a given quality. I've done the same object in both several times. With the Atiks, on a bright object it made little difference. On a faint one it could make quite a lot, maybe 20%. The reason, as you say, is that Lum shooting is full spectrum so about three times the light. Another way to think of it is that in an OSC you are wasting time shooting for colour information that you don't need. What you need is detail (luminance) or the particular response of an emission line, usually Ha. Some say that you need equal doses of L,R,G and B but I don't find this to be so. I find I can use, say, an hour per colour and 3 to 4 hours of Lum. The colour may look washed out at first but there are ways around that in processing. You can end up with fully adequate saturation. On the French forum I'm quite often told that my colour is over saturated.

In an Ha dominated object you can shoot truly minimal colour (as little as 30 mins per channel at F4) and this will give you stunning star colour. (It doesn't burn out the cores.) The Ha will do the rest.

Olly

Thank you Olly That was what I thought and I should get much better results from a mono camera in a given time even with wide spectrum objects like galaxies.

[kirkster501]

When I get a CCD I am going to get a OSC. A personal thing as everyone says, but from lately in the UK in accordance to my weather diary we are averaging one night per fortnight of clear skies. My reading is that it's just not practical having to take four set of lights. But then, what would a numpty newbie like me know!!!

Happy Xmas, Steve

Sent from my iPhone using Tapatalk

[Stu_2011]

When I get a CCD I am going to get a OSC. A personal thing as everyone says, but from lately in the UK in accordance to my weather diary we are averaging one night per fortnight of clear skies. My reading is that it's just not practical having to take four set of lights. But then, what would a numpty newbie like me know!!!

Happy Xmas, Steve

Sent from my iPhone using Tapatalk

That's exactly the reason I went OSC, lack of time and the lack of good weather in the UK dominated the decision.

I might go mono a few years down the line but for now I'm happy with my ATIK 428 OSC, its a great camera.

Thanks

Sent from my iPad using Tapatalk

[kirkster501]

Mono is for me when I have got an observatory, not studying for a Masters part time and kids are older. *Then* I will go for a mono perhaps but at the moment I just haven't got the time to pull it off. "Know your limitations" a wise man once said......

[Gina]

Nothing ventured, nothing gained - I say

[ollypenrice]

Yes, these are good reasons in favour of OSC. It might also be worth sayng that I think OSC will show its best side in fast optical systems.

Olly

[steppenwolf]

Like Olly, I have used both mono and filters and OSC CCD cameras and enjoy both. For me, here in the UK, with such limited imaging opportunities because of the weather, an OSC device fulfils most of my needs for broadband imaging where my main interest is nebulae but I also enjoy narrowband imaging so use a mono CCD for this purpose. If I lived somewhere where I had better skies I'd be more inclined to work a little more with a mono CCD camera.

At least with a OSC CCD camera I always get a full colour image at every session!

[Black Knight]

At least with a OSC CCD camera I always get a full colour image at every session!

I think that's a very fair point, given the weather in the UK over the last year or so.

[ollypenrice]

I think that's a very fair point, given the weather in the UK over the last year or so.

Yes, it is indeed a perfectly valid point. My dream setup would probably be parallel scopes with OSC in one and mono in the other, the mono catching luminance or narrowband. I think Steve has this on the go?

Olly