When is a CMOS/CCD good enough for DSO?

[iapa]

I was wondering if the more experienced have guidelines as to determining whether a CCD/CMOS camera is useful for DSO imaging.

Some cameras are identified as planetary or guide cams, yet I often see very good results for them when used for DSOs.

Image resolution (arcsec/pixel) based on focal length and sensor pixel size seems to need to be around 2"/pixel for UK skies.

A ratio of image scales for guider:imaging <4 also seems to be a preferred value.

It seems to me that you could choose your imaging and guide tubes and plug all into a spreadsheet and come out with the necessary camera specs. - and I have done that  around 8um pixel size on an 8" SCT, f5 400mm fl guide scope, 3.75um guide camera.

Where does sensor size/pixel count come in?

If I hang around the 8um pixel size from above, if I end up with a sensor that's 1920 x 1080 (c2M pixel), how would that compete with a full size sensor at 4500 x 3000 (c13M pixel).

I'm thinking that the 2M 8um sensor would only 'see' 40% of the area of the diameter of the light path - assuming the 42mm T2.

It seems I am looking for a full frame, 13M pixel CMOS/CCD camera, wouldn't it be great if there was a comprehensive list of cameras with pixel size, sensor size (either in pixels, or mm) and price.

But, back to my original question

I have seen some very good stuff from 1M pixel cameras with much smaller pixels.

What makes a camera 'good enough' for DSO.

Just add cooling? Delta or set point?

 

[Davey-T]

We're really spoilt nowadays as regards imaging cameras as almost any CCD / CMOS camera can produce some sort of reasonable image of DSO's

If you do a bit of research you will be amazed at the primitive equipment that original imagers had to use and the comparative cost of it.

Dave

[iapa]

I'm neither archaeologist nor historian but fully appreciate and am more than willing to acknowledge their efforts.

Even more willing to build on the backs of and efforts of those early pioneers. LOL.

I came from valve radios/transmitters - I've no intention of going back when I have VOD I'm my internet connection

i want tomorrow and I want it today ??

[ronin]

Quote

 

Some cameras are identified as planetary or guide cams, yet I often see very good results for them when used for DSOs.

Likely they started out as a planetary camera and have undergone updates to enable DSO imaging, Simplest is new software and module components that enable a long duration exposure. Think this was the method by which the Philips webcam became useful for DSO, people reflashed the software and were able to get long exposures. The camera manufacturers realised the market and got in on the act.

Planetary cameras may also tend to have an arcsec/pixel size of 0.5" (and less) not the 2" you give for DSO's. Find the SBIG site they explain why somewhere in their information. Equally SBIG also suggest a ratio of 1" not 2". So as in many aspects you may find more then one "rule" and all different. Another area where I have read 1, 2 and 4 as the optimum.

If a planetary camera has a smaller pixel size then you also may have to take into account history. The WO Star 71 is a good imaging scope. It's small focal length means small pixels can be used (planetary camers size?) and not that many years ago it did not exist. People were using longer focal length scopes which meant bigger pixels. Smaller better scopes mean that sensors with smaller pixels are now useable for DSO imaging, whereas they probably used to fall into the planetary imaging realm.

There is also a reason for pixel size it has to match to stars size, something to maintain a "roundish" star when the image is expanded. Was told it, did write it down, cannot recall what it was.

One other aspect is that the DSLR that many use was never designed around doing astrophotography. It is aimed at landscapes, wildlife, portraits and holiday pictures. So there is a camera even less "suited" to DSO imaging that is very commonly used for DSO imaging.

[MattJenko]

Main distinction between them is cooling and sensor size. A lot of the high framerate planetary cmos chips are small and are uncooled. Lack of cooling makes them noisy, which is not so much of a problem for high res planetary imaging or guiding, but does impact DSO imaging. You can use them, just that the results are easier to process with a cooled sensor and more targets are available with a larger chip size.

[ollypenrice]

Chip size doesn't matter for planetary but it certainly does for DSOs. The simplest thing is to model the field of view on a planetarium software package. I use SkyMap Pro but many offer this facility. In a nutshell you just see what will fit on the chip.

How crtical is sampling rate? Not very, in my book.

0.66"PP:

3.5"PP:

Horses for courses. They both take pictures.

For long exposures you need cooling and for noisy chips like the Kodaks it is best to have set point. It matters less with Sony.

Regarding spec lists, what drives me potty is the fact that it is so hard to find the chip size in mm along both sides. This is so obviously necessary information yet it is rarely provided.

Olly

[MartinB]

Modern DSLRs are truly amazing with big strides in reducing noise, increases in sensitivity with the aid of microlensing and a generous amount of chip real estate.  With their video capabilities they can make great planetary and lunar cameras.  They aren't cooled and need the IR filter removing to be properly effective for imaging Ha emission nebulae.  Cooling makes a big difference in the summer and less so when imaging in sub zero winter conditions.   However, one aspect which often seems overlooked is bit depth.  A DSLR will have bit depth up to 14 whereas a cooled CCD is likely to be 16bit.  This is of no consequence when imaging bright solar system objects but is when imaging dim DSOs.  The capacity to stretch the histogram that 16 bit gives you shouldn't be underestimated.  

[Zakalwe]

6 hours ago, MartinB said:

 The capacity to stretch the histogram that 16 bit gives you shouldn't be underestimated.  

Its less important now with the rise of CMOS. Stacking many subs (hundreds or even thousands) gets the dynamic range back.

http://www.cloudynights.com/topic/549417-asi1600-experiments-with-high-gain-nb-imaging/?p=7424465

CTB-1, 4 hours integration, 120 second sub exposures. 12 bit ADC

[ollypenrice]

7 minutes ago, Zakalwe said:

Its less important now with the rise of CMOS. Stacking many subs (hundreds or even thousands) gets the dynamic range back.

http://www.cloudynights.com/topic/549417-asi1600-experiments-with-high-gain-nb-imaging/?p=7424465

CTB-1, 4 hours integration, 120 second sub exposures. 12 bit ADC

That has to be impressive.

Olly

[MartinB]

11 minutes ago, Zakalwe said:

Its less important now with the rise of CMOS. Stacking many subs (hundreds or even thousands) gets the dynamic range back.

http://www.cloudynights.com/topic/549417-asi1600-experiments-with-high-gain-nb-imaging/?p=7424465

CTB-1, 4 hours integration, 120 second sub exposures. 12 bit ADC

Perhaps it's best not to go into discusssions on noise, gain and dynamic range, these threads always end in tears!  The camera used for that shot has less than half the read noise of a Canon 5d Mkiii at 6400iso (which is very good in it's own right).  So that's very exciting and does away with the need for long subs and high end mounts.  So a cheap, high performance camera which can be used with a relatively cheap mount.  I think this is the future but I would still prefer the camera to be 16 bit!

[ollypenrice]

The acid test is simple: is the image good? In the example above, the image is good...

Oh hell, I hate new technology!!! I'm too old for it.

Olly