Jump to content


Photo

CCD vs. CMOS Video (DSO Astronomy)

  • Please log in to reply
33 replies to this topic

#1 mmalik

mmalik

    DSLR camera modifications

  • *****
  • Vendors
  • Posts: 5402
  • Joined: 13 Jan 2012
  • Loc: USA

Posted 07 March 2013 - 01:18 AM

Is there some kind of war/competition brewing between astro video CCD, i.e., SONY and astro video CMOS, i.e., Canon?


Current (high-end) offerings with SONY astro video CCD sensors: (Corrections/additions welcome)
•Mallincam Xtreme Color: SONY ICX418AKL / ICX428AKL
•Astrovid StellaCam3 Mono: SONY ICX418ALL


Prospective/futuristic offering/s with Canon's new video CMOS sensor?
New CMOS Sensor for video...


Few questions:
•How do you foresee Canon's new CMOS chip's potential for DSO video astronomy?

•How do you foresee Canon offering this new chip to other vendors (e.g., Mallincam, StellaCam, etc.) or will it be exclusive to Canon cameras?

•How do you foresee Canon catering to video astronomy with the new chip or would it end up going to other domains, medicine, surveillance, security, etc.?

•In short, how do you see high-end CCD and CMOS paly out in the future as far as DSO video astronomy is concerned?

Note: Focus of my inquisition is futuristic DSO video astronomy (NOT planetary)


I am new to the video forum; your insights welcome! Regards


I come from DSLR background; my DSLR astro work here....
A relevant post here...

#2 wolfman_4_ever

wolfman_4_ever

    Apollo

  • -----
  • Posts: 1245
  • Joined: 15 Jul 2011
  • Loc: El Segundo, Ca, So. Cal

Posted 07 March 2013 - 02:10 AM

:gotpopcorn:

#3 Moromete

Moromete

    Viking 1

  • *****
  • Posts: 677
  • Joined: 15 Feb 2012
  • Loc: Romania

Posted 07 March 2013 - 02:15 AM

Considering the size and specs of the new Canon sensor I think we will not find it in cameras with prices for amateur astronomers (<1500-2000 USD). I hope I'm wrong.

#4 nytecam

nytecam

    Voyager 1

  • *****
  • Posts: 11509
  • Joined: 20 Aug 2005
  • Loc: London UK

Posted 07 March 2013 - 04:07 AM

Mike - after seeing your quality long exposure DSO webpage I'd doubt video will satisfy you unless you want a change-of-direction. The brevity of video exposure realtime imaging can't match the quality of CCD imaging and obligitory post processing. Video is quick, easy and 'dirty' by definition in swapping a dim and solitory eyepiece view into a bright, detailed and colourful instant screen view after a few seconds exposure that can be shared by many. :grin:

IMHO the new camera chips are an interesting but needless distraction :o

#5 mmalik

mmalik

    DSLR camera modifications

  • *****
  • Vendors
  • Posts: 5402
  • Joined: 13 Jan 2012
  • Loc: USA

Posted 08 March 2013 - 02:48 AM

The brevity of video exposure real-time imaging can't match the quality of CCD imaging and obligatory post processing. Video is quick, easy and 'dirty' by definition in swapping a dim and solitary eyepiece view into a bright, detailed and colorful instant screen view after a few seconds exposure that can be shared by many.


Understandable; video doesn’t come close to long exposure still imagery but video is bound to get better. I think Canon is on the right track with the new chip if it translates into something serious for (DSO) video astronomy. Thx

#6 Chris A

Chris A

    Vendor Affiliate - Mallincam

  • -----
  • Vendor Affiliate
  • Posts: 1152
  • Joined: 03 Feb 2007
  • Loc: Toronto, Canada

Posted 08 March 2013 - 11:46 AM

I am sorry but this statement from you "Understandable; video doesn’t come close to long exposure still imagery but video is bound to get better" does not make any sense to me because video in it's very own unique league. As we all know that astrophotograpghy and astrovideo are two different worlds are serve different wants/needs.

Again you use a dedicated astro ccd camera for creating pretty astro pictures or for scientific studies and you use a dedicated astro video camera for live (planets, solar and lunar) or near live (DSO's) observations.

To this day you still cannot compare a CMOS sensor to a ccd sensor even though the CMOS are improving overtime. The large CMOS sensor that Canon announced seems to be sensitive with it's very large pixels but is probably still good for only the brighter DSO's and it takes a lot more then then a sensitive sensor to make it in the very faint DSO video world.

Chris A

#7 mmalik

mmalik

    DSLR camera modifications

  • *****
  • Vendors
  • Posts: 5402
  • Joined: 13 Jan 2012
  • Loc: USA

Posted 08 March 2013 - 12:56 PM

As we all know that astrophotograpghy and astrovideo are two different worlds are serving different wants/needs.


Maybe I didn't say it right; I like your version better. I guess too early to say what Canon's new CMOS has to offer in terms of DSO video, what I like is the initiative on Canon's part.


What I don't like about contemporary video (Mallin/Stella) is the interfaces; although things are changing but slowly. Plus I feel the two vendors may have fallen behind times in terms of underlying technologies regardless of the SONY sensors in them. Imagine SONY making a DSO video cam in competition to Mallin/Stella, things would not be the same, correct? What Canon may bring to the table, "if" they were to offer a DSO video cam, may not be just the sensor per se but underlying infrastructure (electronics, etc.), hardware interfaces (USB, etc.), software (image manipulation, etc.), future innovation, etc. All I mean to say, it’s a good start, let’s hope it becomes a trend.


FYI: My images... were taken with a Canon CMOS; not CCD sensor.

#8 Chris A

Chris A

    Vendor Affiliate - Mallincam

  • -----
  • Vendor Affiliate
  • Posts: 1152
  • Joined: 03 Feb 2007
  • Loc: Toronto, Canada

Posted 08 March 2013 - 02:17 PM

Yes it is a good start on behalf of Canon and I also hope they will venture more into this part of astronomy. Canon did take a big initiative step when they brought out their 20Da for astrophotography and now their 60Da unlike other competitors like Nikon, Sony and Pentax. Lets see what Canon brings us in the near future when it come to live or near live DSO video observation.

Chris A

#9 Lorence

Lorence

    Viking 1

  • -----
  • Posts: 852
  • Joined: 15 Sep 2008

Posted 08 March 2013 - 02:28 PM

From the user standpoint what does it mater if you are using a video or a CCD camera. Live video really only works on very bright objects. Once video camera sense up is enabled the image is near live. Integrated video is in reality the same as long exposure with a CCD.

I use several Mallincam video cameras as well as a Mallincam Universe. All my deep sky viewing is near live with exposure times generally between a few seconds to two minutes. What I see is a constant succession of individual images regardless of what camera I use.

Who care how those images get there as long as they are there.

Those chips or ones very similar will will eventually become available to the public. People like Rock will put them into cameras for people like me that are willing to pay for them. All the rest will be free to speculate to their hearts content.

#10 Mogster

Mogster

    Viking 1

  • -----
  • Posts: 503
  • Joined: 14 Mar 2004
  • Loc: Manchester, England

Posted 08 March 2013 - 03:14 PM

Its good to see that Canon are obviously very aware of the astro imaging potential for this tech.

There's a definite market for high end products.

#11 bwallan

bwallan

    Mariner 2

  • -----
  • Posts: 220
  • Joined: 08 Jul 2010
  • Loc: Alberta, Canada

Posted 09 March 2013 - 02:42 PM

Image quality comes down to one simple law... the more actual photons you capture, the better the image. Image quality is not improved by high gain (the reverse is true).

Capturing enough photons from faint stellar objects to create a good quality image takes a fast lens/telescope and time... What I've seen coming out of the current offering of astro-video systems are not quality images. They are an improvement over the "fuzzies" one sees through an eyepiece BUT not up to even the lowest standards expected by astro-imagers!

#12 ccs_hello

ccs_hello

    Fly Me to the Moon

  • *****
  • Posts: 6548
  • Joined: 03 Jul 2004

Posted 09 March 2013 - 02:55 PM

It's a matter of S/N trade-offs. High-gain imagers are trading off the picture S/N, resolution, etc. with "one term":
previously I can't see it, now I can!

We call it as a different viewing style.
There is no definite perfect answer one way or another.

(Until CANON sells that full-frame sized 19um pitch CMOS image sensor at $200 a pop. Keep on dreaming...)

BTW, even if that sensor is sold in millions in quantity, the fast optics, long focal length, and large image circle that can fully illuminate the full-frame image area will be very costly. Say fl=300mm, f/2 ED (or APO OTA) lens for FX format :) :) !!

Clear Skies!

ccs_hello

#13 Dwight J

Dwight J

    Apollo

  • *****
  • Posts: 1069
  • Joined: 14 May 2009
  • Loc: Lethbridge, Alberta, Canada

Posted 09 March 2013 - 09:24 PM

Image quality comes down to one simple law... the more actual photons you capture, the better the image. Image quality is not improved by high gain (the reverse is true).
Capturing enough photons from faint stellar objects to create a good quality image takes a fast lens/telescope and time... What I've seen coming out of the current offering of astro-video systems are not quality images. They are an improvement over the "fuzzies" one sees through an eyepiece BUT not up to even the lowest standards expected by astro-imagers!


I guess you havent seen the Mallincam Universe camera in action then. Astrovideo is not imaging but observing. If you want to do imaging use a dedicated CCD camera. I find the lower resolution of astrovideo a fair tradeoff for hours of image acquisition and processing. Astrovideo is an evolving technology so there are exciting things to come. Just watch us - pun intended.

#14 Lorence

Lorence

    Viking 1

  • -----
  • Posts: 852
  • Joined: 15 Sep 2008

Posted 10 March 2013 - 03:23 PM

What I've seen coming out of the current offering of astro-video systems are not quality images. They are an improvement over the "fuzzies" one sees through an eyepiece BUT not up to even the lowest standards expected by astro-imagers!


You obviously have to get out and see more. Only a few imageres produce magazine quality work, many can't even do the equal of the casual video user. I don't even consider myself as being an imager but I get better near live results with my Universe that I've seen from many of the pixel butchers out there and I get those results in minutes rather than hours.

#15 ccs_hello

ccs_hello

    Fly Me to the Moon

  • *****
  • Posts: 6548
  • Joined: 03 Jul 2004

Posted 10 March 2013 - 03:49 PM

What I've seen coming out of the current offering of astro-video systems are not quality images. They are an improvement over the "fuzzies" one sees through an eyepiece BUT not up to even the lowest standards expected by astro-imagers!


You obviously have to get out and see more. Only a few imageres produce magazine quality work, many can't even do the equal of the casual video user. I don't even consider myself as being an imager but I get better near live results with my Universe that I've seen from many of the pixel butchers out there and I get those results in minutes rather than hours.


Before some people's blood pressure gets slightly elevated, let me try saying something here...

Purpose-made astroCCD imagers have design targets and intended audiences.

Purpose-made fast-image acquisition/high-gain CCD imagers have design targets and intended audiences.

Both are just tools and certainly have pros and cons. Compromises are the name of the game!
One tool's pro (as design features) can be the other tool's con (more like a hindrance.)

There are overlapping zones and there is no rule saying there must be no cross-overs.

The result of the work is heavily dependent on the tools used, the skill level of the owner-user (or artist), and time & effort made.

Can we stop this silly mine is better than yours (without even willing to go deep into the details) type of chatting, especially in this public forum?

Astro is supposed to be fun, tranquility, and less personality conflicts.

Clear Skies!

ccs_hello

#16 mmalik

mmalik

    DSLR camera modifications

  • *****
  • Vendors
  • Posts: 5402
  • Joined: 13 Jan 2012
  • Loc: USA

Posted 10 March 2013 - 04:56 PM

From the user standpoint what does it matter if you are using a video or a CCD camera. Live video really only works on very bright objects. Once video camera sense up is enabled the image is near live. Integrated video is in reality the same as long exposure with a CCD.


You meant video or still… either being CCD or CMOS. I think @Dwight J put it right, "Astro video is not imaging but observing", and these subtleties do matter regardless of the basics of image capture which may not be much different.


Yes live video may work on bright objects today but that's not to say it is going to stay that way in the near future. While Mallin/Stella may be adequate today as makeshift consumers of CCD chips produced by Sony, real advancement and innovation lies with the chip manufacturers (Sony, Canon [new CMOS chip], Kodak, etc.) and it does matter to us because astronomy has been left at the back burner of the big manufactures. Nikon has yet to acknowledge we exist. Canon wouldn’t have built the new chip let alone mention the word ‘astronomy’ in their communique had it not been for one of their executives being an astronomy enthusiast. So yes, our attitudes do matter! Regards

#17 ccs_hello

ccs_hello

    Fly Me to the Moon

  • *****
  • Posts: 6548
  • Joined: 03 Jul 2004

Posted 10 March 2013 - 05:19 PM

Well, we don't know the details on Canon big pixel Full-Frame CMOS yet (price and if there is even a new technology introduced)...

However, while waiting...
there is a nice article to read (some not applicable to astro), but worthwhile to have another dream :)
http://www.andor.com...physics-worl...

If there is a possibility to get EMCCD sensor cheap :) :) ...

Clear Skies!

ccs_hello

#18 nytecam

nytecam

    Voyager 1

  • *****
  • Posts: 11509
  • Joined: 20 Aug 2005
  • Loc: London UK

Posted 11 March 2013 - 05:24 AM

From the user standpoint what does it matter if you are using a video or a CCD camera. Live video really only works on very bright objects. Once video camera sense up is enabled the image is near live. Integrated video is in reality the same as long exposure with a CCD.

....
Yes live video may work on bright objects today but that's not to say it is going to stay that way in the near future. While Mallin/Stella may be adequate today as makeshift consumers of CCD chips produced by Sony, real advancement and innovation lies with the chip manufacturers (Sony, Canon [new CMOS chip], Kodak, etc.) and it does matter to us because astronomy has been left at the back burner of the big manufactures. Nikon has yet to acknowledge we exist. Canon wouldn’t have built the new chip let alone mention the word ‘astronomy’ in their communique had it not been for one of their executives being an astronomy enthusiast.....

Live-video was once just that but as sensup etc extended exposures 'live' ceased and was seen after-the-event so that video and CCD in brief exposure were the same - only the means of presentation was different via TV or laptop screen.

I find it strange that of all the astro-cam manufacturers only Mallincam persists and even they acknowledge regular CCD work with their Universe cam which noticeably has vanished from these pages after launch. I've spoken to other cam makers and apart from marketing 'planetary' cams with fast acquisition rates to computer, video is not on their radar as redundant technology - whatever you make of that. Of course there remains plenty of security cams for us to latch onto.

As to the Canon sensor with 19um pixels as I recall the first CCD cams from SBIG [Texas Inst chips], Starlight Xpress [Sony chips] etc had big pixels but low QE. Times have changed to small pixels and hi-QE. The Canon CMOS sensor will be favoured for night shots by big-game [camera] hunters and those spectacular rotating stellar skies in TV documentaries - real astro will be way down the list. :grin:

#19 mmalik

mmalik

    DSLR camera modifications

  • *****
  • Vendors
  • Posts: 5402
  • Joined: 13 Jan 2012
  • Loc: USA

Posted 12 March 2013 - 03:58 AM

As to the Canon sensor with 19um pixels as I recall the first CCD cams from SBIG [Texas Inst chips], Starlight Xpress [Sony chips] etc. had big pixels but low QE. Times have changed to small pixels and hi-QE. The Canon CMOS sensor will be favored for night shots by big-game [camera] hunters and those spectacular rotating stellar skies in TV documentaries - real astro will be way down the list. :grin:


Fair assessment... but Canon has a way of surprising astro community, e.g., with their 20Da and 60Da offerings. Let's hope new Canon chip does more for astro video than anticipated. If the new chip were to materialize for astronomy, I would hope Canon offers a full astro video camera solution instead of Mallin/Stella type. Regards

#20 mmalik

mmalik

    DSLR camera modifications

  • *****
  • Vendors
  • Posts: 5402
  • Joined: 13 Jan 2012
  • Loc: USA

Posted 15 March 2013 - 04:54 AM

If there is a possibility to get EMCCD sensor cheap :) :) ...


Some interesting arguments...


"Dark current is a bigger problem with EMCCD technology than it is for standard CCDs because the former technique involves amplifying any electrons – both the photon-generated electrons and the dark electrons alike."
Reference...


"When recording video of astral bodies, while an electron-multiplying CCD, which realizes approximately the same level of perception as the naked eye, can capture magnitude-6 stars, Canon's newly developed CMOS sensor is capable of recording faint stars with a magnitude of 8.5 and above."
Reference...


Question: What is Mallin/Stella sensitivity in terms of star magnitudes?

#21 ccs_hello

ccs_hello

    Fly Me to the Moon

  • *****
  • Posts: 6548
  • Joined: 03 Jul 2004

Posted 15 March 2013 - 06:58 AM

EMCCD is an extreme high gain device on the source end, thus noise control is ultimate important. The TI device was produced a decade ago and TI is not a main stream CCD sensor producer thus some drawbacks are expected.
On the other hand, use what you can get (so long as not overly expensive) is the name of the game.

RE: Canon marketing claim on its sensor's sensitivity...
No technology disclosure was made thus I suggest treat it as a grain of salt.

My personal view: unless there is something new, (e.g., prepare to file patent thus cannot disclose), new sensor chip is just using the larger pixel area to fish for more photons. In that case, the new Canon sensor, just like other astroimaging devices you mentioned (both are using the regular commercial SONY CCDs), are still subject to absolute QE < 1 limitation. I.e., no way to overcome the law of physics.

Clear Skies!

ccs_hello

#22 nytecam

nytecam

    Voyager 1

  • *****
  • Posts: 11509
  • Joined: 20 Aug 2005
  • Loc: London UK

Posted 16 March 2013 - 04:51 AM

My personal view:.....new sensor chip is just using the larger pixel area to fish for more photons. In that case, the new Canon sensor, just like other astroimaging devices you mentioned (both are using the regular commercial SONY CCDs), are still subject to absolute QE < 1 limitation. I.e., no way to overcome the law of physics. Clear Skies! ccs_hello

Well said - so all the "magic circuitry" and other gismos can't create more real electrons than original photons arriving on the sensor because no sensor is 100% efficent [QE=100%] in converting photons to electrons but some folk believe in magic :grin:

#23 mmalik

mmalik

    DSLR camera modifications

  • *****
  • Vendors
  • Posts: 5402
  • Joined: 13 Jan 2012
  • Loc: USA

Posted 17 March 2013 - 04:04 PM

My personal view: unless there is something new, (e.g., prepare to file patent thus cannot disclose), new sensor chip is just using the larger pixel area to fish for more photons. In that case, the new Canon sensor, just like other astroimaging devices you mentioned (both are using the regular commercial SONY CCDs), are still subject to absolute QE < 1 limitation. I.e., no way to overcome the law of physics.



Technology is evolving and science is evolving; who is to say our understanding of quantum efficiency (which is just a calculated ratio...an estimated percentage of photons hitting a devices photoreactive surface) is going to stand the test of time. We are far from even understanding nature of photon itself (wave-particle duality); our standard model... of particle physics is far from being complete, part of it being theoretical and part experimental.


I really don’t know whether Canon's new chip has any value to video astronomy; it is just a good step forward in otherwise stagnant industry. Chips, their designs (large/small pixels alike), their efficiencies, how we measure those efficiencies “today”, etc. are an ever changing dynamics. We are far from scratching the surface of any laws of physics, let alone invoking them with our limited understanding. Regards

#24 TonyBegg

TonyBegg

    Lift Off

  • -----
  • Posts: 19
  • Joined: 13 Aug 2012
  • Loc: Santa Fe, NM, USA

Posted 06 April 2013 - 03:57 PM

My understanding is that early work on lucky imaging at Cambridge University in the UK used EMCCD or E-cubed something CCDs and was definitely not doing just planetary imaging (with big telescopes). I believe the finder scopes on the AAT when it opened in 1974 used modified vidicon vacuum tube TV cameras and showed quite a lot of stars (this was before CCDs). I asked one of the high end astro CCD camera makers ($20,000+ cameras) at last year's PATS whether they intended to use EMCCD and the guy said that Scientific CMOS was better. Here we have a non-cooled CMOS with amazing performance. I would like to see some sort of heads-up device that mixes real time color images of DSOs with the visual experience (something of a registration challenge I fear). Video astronomy seems to me to be stuck and I welcome perhaps a new approach here from Canon.

#25 scout72

scout72

    Apollo

  • *****
  • Posts: 1204
  • Joined: 12 May 2008
  • Loc: SF Bay Area

Posted 19 September 2013 - 08:54 AM

Sample video from new Canon CMOS

Pretty cool...I have a Canon EOS C100 for non-astro film project, there is a firmware update coming in November to allow expanded ISO to 80,000....I will have to see if 1/3second at iso80k is enough for bright DSO and not TOO noisy.






Cloudy Nights LLC
Cloudy Nights Sponsor: Astronomics