Jump to content

  •  

CNers have asked about a donation box for Cloudy Nights over the years, so here you go. Donation is not required by any means, so please enjoy your stay.

Photo

Good news for those looking for smaller pixels and a bigger chip

  • Please log in to reply
12 replies to this topic

#1 jhayes_tucson

jhayes_tucson

    Fly Me to the Moon

  • *****
  • topic starter
  • Posts: 7180
  • Joined: 26 Aug 2012
  • Loc: Bend, OR

Posted 08 August 2017 - 06:16 PM

I am currently at the SPIE meeting in San Diego where I learned that Canon is going to start offering some of their sensor technology to other camera manufacturers.  They were showing a number of new CMOS chips with extended NIR capability out to 1100 mm.  In my view, most of those chips are uninteresting because they have 19 micron pixels.

 

The one new chip that they had on display that was interesting was an APS-H (29.22 mm x 20.20 mm) with 2.2 micron square pixels.  That is a whopping 13,280 x 9,148 pixel, 120 Mpx array!  It has 28 output channels so that it can be read at 9.4 fps.  Wow.  The bin depth is 10,000 e- and dark current is 8.1 e-/sec at 60C.  They couldn't give me a read noise or QE spec.  The chip is available in a color (RGB) or a mono version.

 

And now for the bad news.  The price for chip alone is $10,000.  I explained to them that for the amateur AP market, they'd have to get the price down by a factor of 2-4 before there would be much interest.  To my surprise, they seemed interested to hear that news and expressed some interest in attending the AIC meeting to learn more about the AP market.

 

I don't know if this will go anywhere, but it is interesting to see that Canon is starting  to market their chip technology.  You can view their site along with the spec sheet at:

https://www.usa.cano...owcases/sensors

 

Here's a photo of the chips that they had in the booth.

 

John

Attached Thumbnails

  • image.jpeg

Edited by jhayes_tucson, 08 August 2017 - 06:22 PM.

  • PrestonE, psandelle, rainycityastro and 12 others like this

#2 psandelle

psandelle

    Mercury-Atlas

  • *****
  • Posts: 2941
  • Joined: 18 Jun 2008
  • Loc: Los Angeles

Posted 08 August 2017 - 06:54 PM

Way to represent! That does sound like a chip that my super-fast wide field heart would love.

 

Paul


  • PirateMike and KBALLZZ like this

#3 PirateMike

PirateMike

    Mercury-Atlas

  • -----
  • Posts: 2764
  • Joined: 27 Sep 2013
  • Loc: A Green Dot On A Blue Sea

Posted 08 August 2017 - 07:34 PM

Look on the good side of pricing. Way back when calculators first came out they cost $300-$400 which was a lot of money back then. Now you can get a much better calculator for $5 or so.

 

I'll be getting one year older every year. I wonder how long it will take to get these sensors to a point that I could afford one!



#4 rockstarbill

rockstarbill

    Fly Me to the Moon

  • *****
  • Posts: 6352
  • Joined: 16 Jul 2013
  • Loc: Snohomish, WA

Posted 08 August 2017 - 07:46 PM

My FSQ106 wants that mono chip now.


  • psandelle and TCFenua like this

#5 jfrech14

jfrech14

    Viking 1

  • *****
  • Posts: 910
  • Joined: 17 Jan 2012
  • Loc: Virginia, USA

Posted 08 August 2017 - 08:04 PM

WOOOOOOO that sounds like a price to get observatories interested. I wish they would made a buttable version for future focal plane arrays.  Looks interesting. I wonder how they do their gain? .5,1,2,4,8 times or e/adu and a 10 bit output... sacrificing DR for insane resolution? They mix and match data like 10,000 e- at .5x and then 8e-/s dark current at 8x. Needless to say I look forward to the release to hear more!



#6 Jon Rista

Jon Rista

    ISS

  • *****
  • Posts: 23803
  • Joined: 10 Jan 2014
  • Loc: Colorado

Posted 08 August 2017 - 11:15 PM

Yeah, I read about Canon offering their sensors on the open market on ImageSensorsWorld blog a little while back. I didn't expect them to offer that 120mp APS-H, though. That's pretty interesting!! So, if I am calculating correctly, if the dark current is 8.1e-/s @ 60C, by the time you cooled that sucker to -20C, the dark current would be ~0.03e-/s. That isn't too bad! Considering that Canon's smaller pixel sensors have fairly low read noise, those with 4.3 micron pixels have no more than ~8e- at the minimum gain, and newer ones have much less than that, we should be looking at only a few electrons of read noise. Lets say it's 4e-, that's ~68dB, or 11.33 stops. The ASI1600 at unity is 11.41 stops with 3.8 micron pixels, and has about half the FWC (so shorter exposures). If the pixels have read noise closer to other sensors with similarly small pixels, then it might only be about 2-2.5e-. That would give dynamic range of 12-12.33 stops. 


Edited by Jon Rista, 08 August 2017 - 11:17 PM.

  • TCFenua likes this

#7 jhayes_tucson

jhayes_tucson

    Fly Me to the Moon

  • *****
  • topic starter
  • Posts: 7180
  • Joined: 26 Aug 2012
  • Loc: Bend, OR

Posted 09 August 2017 - 08:46 AM

It would be an awesome HyperStar chip!

 

John


  • wes, psandelle and jfrech14 like this

#8 pfile

pfile

    Fly Me to the Moon

  • -----
  • Posts: 5213
  • Joined: 14 Jun 2009

Posted 09 August 2017 - 08:34 PM

i guess it's another group inside of canon, but almost every time there's canon people at AIC with DSLRs and lenses. so at least canon is not a stranger to AIC/ astrophotography.

 

rob



#9 freestar8n

freestar8n

    Vendor - MetaGuide

  • *****
  • Vendors
  • Posts: 8894
  • Joined: 12 Oct 2007

Posted 09 August 2017 - 09:33 PM

I think they regularly attend neaf also. And years ago they offered the 20da for amateur Astro.

If these pixels kept getting smaller but the read noise stayed around 10e the smaller pixels would be bad for Astro. But if they keep read noise down and dark current manageable - the offerings may just get better and better as the prices go down.

Frank

#10 t_image

t_image

    Gemini

  • -----
  • Posts: 3287
  • Joined: 22 Jul 2015

Posted 09 August 2017 - 10:00 PM

Can someone explain to me the end game of Canon developing such high-end sensors,

and the long delay until they make it to market?

Are they just looking for a high-end gov't or corporate client to sell just a few to make a profit?

The manufacturing setup of such a high-end sensor surely costs $$$$$$$?

 

The 120MP sensor above was announced by Canon in 2010:

http://cpn.canon-eur..._cmos_sensor.do

 

They announced a super high ISO 19micron video sensor in 2013:

http://global.canon/...013/mar04e.html

 

Then finally released a product with it in 2015:

https://petapixel.co...4000000-camera/

 

So maybe the big news in all of this is Canon will let other camera companies have a go at putting these guinness book novelties to actual use!!!!!

 

I won't hold my breath that they will ever care towards a non-university sized budget.....

 

More than likely I'd bet another sensor manufacturer will be putting something of these specs in cell phones for a reasonable price before many on CN will have Canon's novelty items....



#11 jhayes_tucson

jhayes_tucson

    Fly Me to the Moon

  • *****
  • topic starter
  • Posts: 7180
  • Joined: 26 Aug 2012
  • Loc: Bend, OR

Posted 09 August 2017 - 10:48 PM

i guess it's another group inside of canon, but almost every time there's canon people at AIC with DSLRs and lenses. so at least canon is not a stranger to AIC/ astrophotography.

 

rob

 

Yes, I did explain to the folks at SPIE that Canon is at AIC (showing cameras and lenses) but this is a different group.  Canon is big enough that not everyone is connected to what everyone else is doing.  These guys really want to talk to other camera manufacturers--not so much to those of us who buy the cameras as end users.

 

John



#12 jhayes_tucson

jhayes_tucson

    Fly Me to the Moon

  • *****
  • topic starter
  • Posts: 7180
  • Joined: 26 Aug 2012
  • Loc: Bend, OR

Posted 09 August 2017 - 11:00 PM

Can someone explain to me the end game of Canon developing such high-end sensors,

and the long delay until they make it to market?

Are they just looking for a high-end gov't or corporate client to sell just a few to make a profit?

The manufacturing setup of such a high-end sensor surely costs $$$$$$$?

 

The 120MP sensor above was announced by Canon in 2010:

http://cpn.canon-eur..._cmos_sensor.do

 

They announced a super high ISO 19micron video sensor in 2013:

http://global.canon/...013/mar04e.html

 

Then finally released a product with it in 2015:

https://petapixel.co...4000000-camera/

 

So maybe the big news in all of this is Canon will let other camera companies have a go at putting these guinness book novelties to actual use!!!!!

 

I won't hold my breath that they will ever care towards a non-university sized budget.....

 

More than likely I'd bet another sensor manufacturer will be putting something of these specs in cell phones for a reasonable price before many on CN will have Canon's novelty items....

 

My guess is that the big problem with this kind of sensor is yield and that's what drives cost and availability.  It wouldn't surprise me at all if they announced this sensor in 2010 after they got one barely working.  It may have taken this long just to get the bugs worked out and to get the production process figured out.  At 10k$ per sensor, they are looking for customers in the industrial markets first and that makes sense.  To some degree, this is a solution in search of a problem.  Prices will eventually come down but there is little doubt that this sensor was (and still is) ahead of it's time--depending a bit on how well it works.  I agree that at this point, it falls into the "novelty" category.  It would sure chew through disk space!  It's a good thing that you can now buy a 4TB external drive at Costco for $110.

 

John



#13 rkayakr

rkayakr

    Apollo

  • *****
  • Posts: 1497
  • Joined: 27 Oct 2010
  • Loc: Northeast Ohio

Posted 10 August 2017 - 07:25 AM

NASA has defined a set of nine technology readiness levels. The first demonstration of feasibility is TRL 5 and know informally as:

  hand assembled by PHDs, breaks easily

the next level up is know informally as

  safe to have management demo

 

I suspect that Canon makes an announcement when initial feasibility is shown, which is a long way from production.


  • psandelle likes this


CNers have asked about a donation box for Cloudy Nights over the years, so here you go. Donation is not required by any means, so please enjoy your stay.


Recent Topics






Cloudy Nights LLC
Cloudy Nights Sponsor: Astronomics