CCD reality check

[johnrt]

I'm a little intrigued by the CMOS bandwagon, I know it is possible to take hundreds and hundreds of very short exposures with them, but the images I have mostly seen in practise use 5 minute subs, which is comparable to the sub lengths I use with my CCD (when I shoot RGB), so for me as an owner already of a CCD this is no contest, why would I swap?

If I were gong out to buy my first ever camera today, perhaps I would have a more difficult choice?

[ollypenrice]

I think it's true that CCD is stagnating in research terms. Sony are not going to keep on making them, they say.

Olly

[frugal]

9 minutes ago, ollypenrice said:

I think it's true that CCD is stagnating in research terms. Sony are not going to keep on making them, they say.

Olly

I was just looking into that. I remembered a couple of years ago that Sony said they were going to stop production of CCDs in 2017, but shortly after that they changed to 2026 (http://www.baslerweb.com/en/news-press/news/new-information-about-sonys-ccd-sensor-discontinuation-plan/40656 - May 2015). It does seem that they are no longer doing any development on CCD technology and have moved over to purely CMOS. I guess they believe that by 2026 they will have improved CMOS to the point that it is better than CCD in all aspects. It certainly seems that from reading around the subject that the main driver for CCD/CMOS these days is fast exposure times (5000fps fast), rather than long exposure times (http://www.stemmer-imaging.co.uk/media/uploads/cameras/avt/12/120483-Allied_Vision_White_Paper_CCD_vs_CMOS.pdf - March 2016). So even with CMOS, astrophotography will still be a niche within a niche.

It will be interesting to see how things develop once users have got to grips with the new CMOS cameras and learn how to get the best out of them over the next year or so (gain, offset, noise, exposure time etc), rather than the current images which are mostly "First light" tests.

[zicklurky]

On 3/23/2017 at 08:28, ollypenrice said:

Apologies for posting an image previously seen in a different context. (It was in an effort to show how fast mono and filters are.) However, as interest in the ASI 1600 understandably grows, I think that the 'slowness' of the old school CCD is becoming overstated. Our supposedly insensitive and out-dated Atik 11000 CCD needed one hour (2X30) to obtain this Ha result.

I wouldn't, in all honesty, call that slow...

Olly

When I got my CCD I compared the specs and ended up getting a Sony 690 chip, as I thought the Kodak chips looked slow. But let's compare one of Olly's setups with a rig I was using a while back, imaging at the same resolution. 11000 chip with FSQ 106 (106mm aperture), vs my Sony 690 + Canon 200mm stopped down to f/4 (50mm aperture). Both setups around about 3.5 arcsec/pixel resolution. Because of the larger pixels of the 11000 and larger aperture of the telescope, each of Olly's pixels receives about 4.5 times the amount of light than my pixels do. Yes, the 690 has very efficient pixels, but the 11000 receives so much more light. 

I still love my 690, and I couldn't afford going down the large sensor route anyway. It's just something to think about.

[wimvb]

I've been glancing at cooled cmos to replace my box brownie, and will probably invest in a new camera this summer. But lately, I've started to also consider an Atik 460ex. Reasons: excellent service by the manufacturer, mature technology, stable software, and 16 bit data.

I'm not convinced that difference in QE (if there is a difference) is all-important.

The one thing that cmos really excels at is real estate / price.

The argument made for short exposures making it easier to dodge clouds & planes, is also a valid one for me.

One problem with cmos is of course that by the time you can use your new toy for 1st light, there will be a replacement on the market already.

[frugal]

20 minutes ago, wimvb said:

I've been glancing at cooled cmos to replace my box brownie, and will probably invest in a new camera this summer. But lately, I've started to also consider an Atik 460ex. Reasons: excellent service by the manufacturer, mature technology, stable software, and 16 bit data.

I'm not convinced that difference in QE (if there is a difference) is all-important.

The one thing that cmos really excels at is real estate / price.

The argument made for short exposures making it easier to dodge clouds & planes, is also a valid one for me.

One problem with cmos is of course that by the time you can use your new toy for 1st light, there will be a replacement on the market already.

 

Atik 460: 

ZWO 1600MM:

So if the graphs are to be believed, the ASI1600 is about 10% more sensitive (in absolute terms) than the Atik across most of the wavelengths. So at the Ha band, for every 10 photons that hit the cameras, the ASI1600 will record 8 of them and the Atik will record 7 (and my Canon will record 1 ;(  ). Obviously Pixel size will affect this.

[Filroden]

2 hours ago, frugal said:

So if the graphs are to be believed, the ASI1600 is about 10% more sensitive (in absolute terms) than the Atik across most of the wavelengths. So at the Ha band, for every 10 photons that hit the cameras, the ASI1600 will record 8 of them and the Atik will record 7 (and my Canon will record 1 ;(  ). Obviously Pixel size will affect this.

Your graphs use different scales. One is absolute and the other is relative, but gives no reference to what. Therefore without knowing the absolute QE at the peak around 500nm there is no way to compare the two chips without a controlled field test. 

[xanthic]

I was just looking at the same graphs. The ASI1600 graph is relative - that camera has a peak QE of around 60% (according to ZWOs manual), so at 656 nm, you have about 77% of 60%, so around 46.

[frugal]

Just now, Filroden said:

Your graphs use different scales. One is absolute and the other is relative, but gives no reference to what. Therefore without knowing the absolute QE at the peak around 500nm there is no way to compare the two chips without a controlled field test. 

Doh!

[wimvb]

I think that zwo cameras have about the same sensitivity as ccd's (asi1600 is often compared to kaf8300), but somewhat lower read noise. But these numbers may not be most important. Sensor area, driver stability and manufacturer's service can be at least as important. During the few cloudless nights we have, I don't want to tinker with the camera, it just has to work. I also don't want to correct a camera's deficiencies during processing. I'm already doing too much of that now. But new technology is quickly rewriting the rules of the game. Atm, I'm saving up to join the game next imaging season.

[LightBucket]

Anybody have any feedback on the new ASI071mc-c compared to CCD,  as that camera is 14bit I believe and has APS-C size sensor but is  OSC....read a few reviews on CN, but not seen it mentioned on here..

[wimvb]

Osc in ccd isn't as popular as nb or lrgb imaging. But I think that in a situation with low light pollution (= reasonably dark skies) and only occasional gaps in the clouds, it can be a real advantage to have a sensitive, low noise cmos. Some members here are getting fabulous results with low f-number scopes and dslrs. A cooled, sensitive, low noise osc camera should be capable of excellent images, under the right conditions.