Anyone make the move from a recent CMOS to CCD? - Forums

Topic starter

Sep 1, 2021 · Edited: Sep 1, 2021

Hi there, I've been using a range of CMOS cameras over the last 18months or so for deep sky imaging, more recently moving into mono narrowband imaging with the ASI294mm and more recently the ASI2600mm.

I've had some pleasing results, but my seeing here in Brisbane Australia is pretty poor and feel maybe the smaller pixel size of these CMOS cameras esp the 2600 isn't the best with my 780mm FL (APM130/F6). The 294 has larger pixels and I seem to get better results at full FL, but it's smaller fov and more sensitive to precise calibration than the 2600. I've been thinking of getting a longer FL scope as well which might exacerbate pixel size seeing issues.

A lot of the images that really draw my eye on Astrobin I've noticed are with CCD often with pixel sizes from 6 to 9… hard to describe but see these as clean and deep colour… granted this may have more to do with the processing skills of those using CCD on the whole likely being deeper in this game than most, rather than the cameras themselves?.

Just wondering if anyone has and would like to share any experiences of chasing the greener? grass of large pixel CCD. Worth it, any downsidesover CMOS…esp from perspective of narrowband over LRGB?

many thanks, rob

David Nozadze

Sep 1, 2021

I think those large pixel CCDs are coupled with very large telescopes (observatory/scientific grade). Hence the high image quality.

robonrome

Topic starter

Sep 1, 2021

Thanks David, some are, but I'm looking at mamy using similar size and quality of scope (5in apo refractor).to what I'm using and other amateur level scopes

Lynn K

Sep 2, 2021

Hi Rob. I have mainly used Starlight Xpress CCD cameras with Sony chips. I did experience the (Now antiquated) St237a & the clubs ST2000 years ago.
My experience with CMOS is planetary video and DSLR. .

However, researching reviews of the KAF 12600 chip, I cam across a lot of discussion on Cloudy Nights about the new Sony CMOS IMX 455 vs. the CCD Kodak KAF-16803.
ZWO, QHY and now Atik have come out with cameras using the IMX 455. The discussion centers around your question of focal length and pixel size. Some advocate there is no advantage in using smaller pixels with long focal length and poor seeing. The IMX455 supporters argue that the IMX455 can be binned to accommodate long focal length and use shorter exposures.

Here is the link to one of those discussions. It starts off with questions about the KAF-12600 (which I have- Starlight Xpress Trius SX46)
https://www.cloudynights.com/topic/761474-ccd-vs-cmos/ It doesn't get into the IMX455 until the latter pages. But, over all it is an argument about larger CCD pixel VS binned CMOS smaller pixels, and the CMOS ability to use shorter exposures requiring less accurate tracking.

I'm an old CCD guy and use 10min (sometims15 min) narrow band subs. Usually 20 to 30 subs per channel is the limit. I see people with 60 to over 200 2min subs and I am bewildered by why so many. All that disk space. The CMOS and CCD technology seems to be far more different than just read-out method. My CCD cameras can take 5 sec. to download. Not a big deal when your only doing 25-30 subs. That only adds up to 1.5-2 min. overall. Of course there in no gain settings. The cameras are set to accommodate deep sky imaging and incapable of doing planetary or lunar. My Sony chips (ICX694AL & ICX814AL) do not require dark frame subtraction. They are so low noise that calibrating a dark will add noise. I only use bias. Dithering is necessary to eliminate hot pixels and and will also smooth out the pixel bias.

I use MaximDL v6 and having trouble with the SX46 drivers. The camera runs fine but the downloads cause the bright stars to bloom to the top of the chip.
Terry Platt of Starlight Xpress says he uses the SX46 with Maxim with no issues. I have used it with 3 separate laptop/Maxim and all have the same issue. It works very well with Sequence Generator Pro.

I also use a 130mm refractor (Astro Physics AP130GTX at F4.7 with the Trius 694 Pro at 1.53 ArcSec/Pix). When focusing a star I can get 1.5 FWHM, but never obtain that in the final stack. Occasionally I get a stack of 2.5 FWHM, but usually around 3.5 to 5. If I pair it with the SX46 (6nm pix), I get 2.02 and I doubt I will loose any detail. But despite the larger SX46 6nm pix compared to the SX694 4.54nm pix, I thing the sensitive Sony chip will out perform the Kodak S/N.
What the SX46 gives me is almost 4x the field of view.

Hope this helps some,
Lynn K.

Helpful

SemiPro

IOTD Staff

Sep 2, 2021 · Edited: Sep 2, 2021

When it comes to the CMOS vs CCD debate, from the threads I see it usually devolves into people vehemently defending the particular type of camera they personally use in fear of possibly having to accept that they spent a small fortune on something that might not be the most efficient thing ever. Its like the xbox vs playstation debate, or amd vs intel in my opinion. Any facts get lost in the argument. The bottom line is I think people are going to use what the are comfortable with. Someone who has been doing this a long time would probably be using CCD's because that was what was prevalent back in the day. If you are like me and you just go into astrophotography recently, you are probably using a CMOS. If I am being honest, when I sift through the top picks / IOTD's on this website there is nothing that makes you go "aha! This was a CCD!" without actually looking at the technical card. All you see are amazing pictures of space with data acquired and processed with incredible skill, no matter the equipment.

You mentioned pixel scale which I looked into quite a bit recently so this thread might be of interest to you as it was to me: https://www.cloudynights.com/topic/536637-image-scale-and-dawes-limit/

Helpful Insightful Respectful Engaging

robonrome

Topic starter

Sep 2, 2021

Lynn K:
Hi Rob. I have mainly used Starlight Xpress CCD cameras with Sony chips. I did experience the (Now antiquated) St237a & the clubs ST2000 years ago.
My experience with CMOS is planetary video and DSLR. .

However, researching reviews of the KAF 12600 chip, I cam across a lot of discussion on Cloudy Nights about the new Sony CMOS IMX 455 vs. the CCD Kodak KAF-16803.
ZWO, QHY and now Atik have come out with cameras using the IMX 455. The discussion centers around your question of focal length and pixel size. Some advocate there is no advantage in using smaller pixels with long focal length and poor seeing. The IMX455 supporters argue that the IMX455 can be binned to accommodate long focal length and use shorter exposures.

Here is the link to one of those discussions. It starts off with questions about the KAF-12600 (which I have- Starlight Xpress Trius SX46)
https://www.cloudynights.com/topic/761474-ccd-vs-cmos/ It doesn't get into the IMX455 until the latter pages. But, over all it is an argument about larger CCD pixel VS binned CMOS smaller pixels, and the CMOS ability to use shorter exposures requiring less accurate tracking.

I'm an old CCD guy and use 10min (sometims15 min) narrow band subs. Usually 20 to 30 subs per channel is the limit. I see people with 60 to over 200 2min subs and I am bewildered by why so many. All that disk space. The CMOS and CCD technology seems to be far more different than just read-out method. My CCD cameras can take 5 sec. to download. Not a big deal when your only doing 25-30 subs. That only adds up to 1.5-2 min. overall. Of course there in no gain settings. The cameras are set to accommodate deep sky imaging and incapable of doing planetary or lunar. My Sony chips (ICX694AL & ICX814AL) do not require dark frame subtraction. They are so low noise that calibrating a dark will add noise. I only use bias. Dithering is necessary to eliminate hot pixels and and will also smooth out the pixel bias.

I use MaximDL v6 and having trouble with the SX46 drivers. The camera runs fine but the downloads cause the bright stars to bloom to the top of the chip.
Terry Platt of Starlight Xpress says he uses the SX46 with Maxim with no issues. I have used it with 3 separate laptop/Maxim and all have the same issue. It works very well with Sequence Generator Pro.

I also use a 130mm refractor (Astro Physics AP130GTX at F4.7 with the Trius 694 Pro at 1.53 ArcSec/Pix). When focusing a star I can get 1.5 FWHM, but never obtain that in the final stack. Occasionally I get a stack of 2.5 FWHM, but usually around 3.5 to 5. If I pair it with the SX46 (6nm pix), I get 2.02 and I doubt I will loose any detail. But despite the larger SX46 6nm pix compared to the SX694 4.54nm pix, I thing the sensitive Sony chip will out perform the Kodak S/N.
What the SX46 gives me is almost 4x the field of view.

Hope this helps some,
Lynn K.

thanks So Much Lynn, this is a great answer and set of experience to mull over. Very much appreciated!

Respectful Supportive

robonrome

Topic starter

Sep 2, 2021

When it comes to the CMOS vs CCD debate, from the threads I see it usually devolves into people vehemently defending the particular type of camera they personally use in fear of possibly having to accept that they spent a small fortune on something that might not be the most efficient thing ever. Its like the xbox vs playstation debate, or amd vs intel in my opinion. Any facts get lost in the argument. The bottom line is I think people are going to use what the are comfortable with. Someone who has been doing this a long time would probably be using CCD's because that was what was prevalent back in the day. If you are like me and you just go into astrophotography recently, you are probably using a CMOS. If I am being honest, when I sift through the top picks / IOTD's on this website there is nothing that makes you go "aha! This was a CCD!" without actually looking at the technical card. All you see are amazing pictures of space with data acquired and processed with incredible skill, no matter the equipment.

You mentioned pixel scale which I looked into quite a bit recently so this thread might be of interest to you as it was to me: https://www.cloudynights.com/topic/536637-image-scale-and-dawes-limit/

thanks Semipro...i'll check that discussion out

Boyan Kassabov

Sep 2, 2021 · Edited: Sep 2, 2021

Hi Rob.

IMHO the dispute 'should I use CMOS or CCD" is useless. It's not exactly the technology of the censor that makes the difference -maybe it's the suitable settings. The MAIN advantage of CMOS cameras is - it's simply cheaper. Yeah - maybe the modern ones have a little higher QE, less noise etc, but I'm sure if CCD technology develops, it would catch this difference.

More important for you is the relation FL/size of pixel. Guess you know it, but still - the shorter the FL, the smaller the pixel size. I think with your 780 mm, the ASI 2600 with its 3.76 microns matches perfectly. A camera with 9 microns pixel size would give you an undersampling.

I use a QHY 22 CCD and so far I'm happy with it. I wouldn't change it if not forced to.

Helpful

Björn Arnold

Sep 2, 2021 · Edited: Sep 2, 2021

Hi,

The major difference is determined through the market:
CCD technology is not suitable for a lot of consumer electronics and also industrial applications. Therefore, the development of the CMOS as been pushed enormously leading to much smaller pixels, lower read noise and lower price.

For professional (research) astronomy, the choice is clearly CCD. For consumers, you typically will have to follow the trend of the market because available CMOS technology will outpace what you‘ll be able to afford or want to invest if you want a competitive CCD.

I see one technological aspect where the CMOS is better for the amateur astronomer: the smaller pixels. I have scopes from FL 370mm to 2032mm. I adjust the binning according to FL.
Furthermore, I actually can always shoot at native resolution of the sensor and do binning at any time later. For the CCD, I‘d have to make the binning choice while imaging in order to make use of its „binning“ advantage.
So a CMOS gives some more flexibility, IMHO.

Hope that helps!

Björn

PS: if you’re interested in EAA, where a lot of shorter exposures are made, CMOS is the winner.

Helpful Insightful Respectful Engaging

robonrome

Topic starter

Sep 3, 2021

Thanks Boyan and Bjorn, from your and the other helpful replies here I'm moving to the conclusion that my camera is not the issue and that either path is fine if the time and effort is invested in learning and making the most of it especially in the area of processing. As I'm already familiar with cmos and have a couple of excellent cameras think I will leave that ccd green grass on the otherwise of the hill and get on with appreciating my own lawn :-) thanks again all.