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

CMOS/CCD sensor grades from manufacturers

  • Please log in to reply
19 replies to this topic

#1 mAnKiNd

mAnKiNd

    Mariner 2

  • *****
  • topic starter
  • Posts: 253
  • Joined: 20 Sep 2016
  • Loc: NY

Posted 17 May 2018 - 04:36 PM

Hello and apologies if this has been answered elsewhere.

I kindly invite the public here on CN and even camera manufacturers like QHY, ZWO, SBIG, ATIK, QSI, Altair Astro, RisingCam etc.. to shed more light on which CMOS/CCD sensor grades they purchase from sensor manufacturers i.e. Sony, Panasonic, Kodak etc...to install in their cameras?

What do the different sensor grades mean to advertised sensor performance and furthermore what price differences are we looking at i.e. is buying a sensor from Sony at grade 1 vs grade 2 or 3 worth the extra charge? Why and how?

Are all camera manufacturers purchasing grade 1 from their supply? Or is this why we see the same sensor appear at different price points from different manufacturers (I know these price differrences also entail costs in manufacturing different electronics, features etc..). For example, the QHY163, ASI1600 PRO and Atik Horizon all have the same sensor from Panasonic, but different price points - is this all down to the other costs, or is the Atik Horizon more expensive than the other two because it purchased higher grade sensors? I don't know, I'm asking. Or even look at the good old Kodak 8300 sensor with all of it's different renditions in all those different cameras with different priced. Same sensor grade? Does it even really matter?

Are we to understand that all competing astrocameras sold to us contain the same sensor grade? Certainly the examples of different grades in electronic products coming out of a production line are ubiquitous enough to stop and ask, what grade sensor am I paying for and am I missing out on anything, or not?

I kindly request further clarification
and transparency regarding what "practical" differences, if any, amateur and professional (i.e. photometry) astro imaging can expect from spending the extra money on a higher grade sensor.

Many thanks, good health and clear skies to you all.
Minos


Edited by mAnKiNd, 17 May 2018 - 05:23 PM.


#2 jhayes_tucson

jhayes_tucson

    Fly Me to the Moon

  • *****
  • Posts: 7,388
  • Joined: 26 Aug 2012
  • Loc: Bend, OR

Posted 17 May 2018 - 05:54 PM

Perhaps there is a dark little secret that you should understand about the semiconductor manufacturing process:  It isn't perfect.   That's right, not only is the production process almost hopelessly complicated, but all of those newly produced wafers will have chips that actually have defects or that won't work at all.  Right up front you have to realize that the process needed to produce these sensors makes the difficulties that we encounter with astrophotography look like a day at the beach.  Still, the semiconductor production process is refined all the time in an effort to improve yield.  Yield turns out to be a key parameter that manufactures look at to monitor the stability of the production process.  In any manufacturing run, the process might yield maybe half of the chips that work at all.  From there, the working chips still aren't perfect.  In any large scale wafer production process, the manufacturers gain enough statistical information to predict how many good chips will be produced and how many with known classifications of defects will be produced.  Ultimately, they wind up with a lot of sensors at the bottom of that scale and very few that are perfect--and that's how the prices are ultimately computed.  

 

Then it's up to the customer to decide on how good the sensor needs to be relative to the application and the price.  You can look up the sensor grades on the manufacturer sites to see what they mean.  For most amateur AP applications, you don't need a very high grade sensor. Reliability and stability are probably more important parameters than say the number of bad columns.  Still there are many of us who are driven to perfection and many folks who can afford it, buy the highest quality sensor possible.  Why?  Most because they can.  The higher quality sensor definitely have tighter specs and that also appeals to many folks.  Just recognize that there can be a significant price premium applied to higher grade sensors.  Why?  Well, that goes back to how many do they get out of the process...and what will the customer pay?

 

None of the camera manufacturers are very likely to show up here to rationalize any of this for you.  You need to decide on what you need relative to what you want to accomplish.

 

 

John


  • lambermo, Jon Rista and mAnKiNd like this

#3 mAnKiNd

mAnKiNd

    Mariner 2

  • *****
  • topic starter
  • Posts: 253
  • Joined: 20 Sep 2016
  • Loc: NY

Posted 17 May 2018 - 06:06 PM

John, thank you for your comprehensive reply, much appreciated. I had a vague idea of what you wrote beforehand, but you've shed a lot more light on it for me, thank you.

 

You need to decide on what you need relative to what you want to accomplish.

 

 

John

 

Indeed, in order to make that decision, personally, I'd like a little more info on what grade sensor i'm getting for my money and this information is not published by the camera manufacturers.

 

To go back to a previous example of the aforementioned Panasonic sensor, is one paying more for the Atik Horizon because there is a higher grade sensor in it vs the one in the ZWO/QHY alternative? If that were the case, then I'd be more inclined to reach deeper into my pocket for it, knowing that as you say, perhaps that sensor has less bad columns and tighter specs. More "peace of mind" if you will..

 

I do however share the same gut feeling as you said, that for amateur astrophotography this is not really paramount, as long as the sensor is stable and reliable, like you said - unless one wishes to, and can afford to be at the top of their game..

 

For the rest of us mortals, this is just academic discussion.

 

Thank you again for your time.

 

Minos


Edited by mAnKiNd, 17 May 2018 - 06:12 PM.


#4 jhayes_tucson

jhayes_tucson

    Fly Me to the Moon

  • *****
  • Posts: 7,388
  • Joined: 26 Aug 2012
  • Loc: Bend, OR

Posted 17 May 2018 - 06:40 PM

It's hard to say what sensor grade is in any given camera unless it's in the specs (FLI provides such specs) or you ask the manufacturer directly.  Price differences are not necessarily a key to understanding sensor grade in any give camera.  There are a lot of issues that can drive prices.  The Chinese cameras (ZWO, QYH, and others) operate in an environment of very low manufacturing costs so that's one reason that they tend to be so price competitive.  There may be other difference as well but I'm not going to speculate.  If you want to know more about a particular camera, you are going to have to study specs more carefully and get on the phone to ask specific questions.

 

John


  • mAnKiNd likes this

#5 mAnKiNd

mAnKiNd

    Mariner 2

  • *****
  • topic starter
  • Posts: 253
  • Joined: 20 Sep 2016
  • Loc: NY

Posted 17 May 2018 - 06:42 PM

I honestly thought as much, but many thanks again for the clarifications.

All the best
Minos

#6 maadscientist

maadscientist

    Apollo

  • *****
  • Posts: 1,070
  • Joined: 08 May 2013
  • Loc: Atlanta, Georgia and Deerlick Astronomy Village

Posted 17 May 2018 - 07:25 PM


To go back to a previous example of the aforementioned Panasonic sensor, is one paying more for the Atik Horizon because there is a higher grade sensor in it vs the one in the ZWO/QHY alternative?

You have to also remember that a camera is an execution of many parts other than the sensor. The electronics are important, as sub standard analog to digital converters can degrade the final image. Isolation of the power circuits, and proper cooling mechanisms can affect the final result. Spacing of the chip to the optical window can cause dewing issues. New cmos chips have different settings which the camera manufacturer may or may not be able to take advantage of. All of this has to seamlessly work in the camera drivers with compatibility with image acquisition software.

 

I had a great Kodak 1001e chip a long time ago that SBIG could not integrate properly in their camera (It was an SBIG camera I bought with the 1001e in it). I found out later that they didn't know how to wire it in and time it correctly. It gave stars that had big black dots in the center of them. I went round and round with the owner, who I will not mention. He insisted that was NORMAL.....no....it wasn't.

 

My recommendation is to buy it new while it is under warranty. I ate $3500 on that SBIG.


  • mAnKiNd likes this

#7 mAnKiNd

mAnKiNd

    Mariner 2

  • *****
  • topic starter
  • Posts: 253
  • Joined: 20 Sep 2016
  • Loc: NY

Posted 17 May 2018 - 07:34 PM

Hi madscientist, thanks for your reply.

I made a short mention in the OP that I recognize that there is a differing expense in the circuit designs implemented by each camera manufacturer, I can truly appreciate that. I was simply trying to hone into whether or not some of the price differences in cameras from different companies using the same sensor can, in a way, be used as a gauge to appreciate the certain grade of the sensor used, since it is not openly shared by them what grade sensor is implemented into their designs.

I now understand, that like most things, it is not a simple yes or no answer.

I'm sorry to read about your woes with your old SBIG camera, I would have not expected that from them, as I regard them to be a top-notch camera manufacturer, which I still believe them to be. Black spots in the middle of stars are certainly not normal afaik!

Thank you for your thoughts and insight.
Minos


Edited by mAnKiNd, 17 May 2018 - 07:36 PM.


#8 Shiraz

Shiraz

    Viking 1

  • -----
  • Posts: 617
  • Joined: 15 Oct 2010
  • Loc: South Australia

Posted 17 May 2018 - 08:06 PM

Hard to know for sure with all makers, but according to the Atik website, "The Sony sensors (with the exception of engineering samples) are not graded, so all the Sony sensors we use are the highest and only grade commercially available". Atik do not mention sensor grade for their new Horizon camera, so one could guess that the Panasonic chip is not graded either. From my (very limited) experience of Aptina and Panasonic CMOS sensors, commercial grade CMOS chips can be made with no major faults apart from a smattering of hot pixels and some insignificant amp glow in the Panasonic. Maybe the relatively low cost of CMOS sensors is due to high yields of good quality chips. 

 

In addition, I understand that the new CMOS sensors are essentially the whole camera on a chip - the camera makers really only attach a few wires to make them work. Assuming that the chip is programmed properly, I would think that the differences in quality are not down to the sensor itself, but rather the window material, mechanical design and cooling. And of course software support is a major factor. Cheers Ray


Edited by Shiraz, 17 May 2018 - 08:37 PM.

  • ccs_hello, Jon Rista, mAnKiNd and 1 other like this

#9 maadscientist

maadscientist

    Apollo

  • *****
  • Posts: 1,070
  • Joined: 08 May 2013
  • Loc: Atlanta, Georgia and Deerlick Astronomy Village

Posted 17 May 2018 - 08:11 PM

I'm sorry to read about your woes with your old SBIG camera, I would have not expected that from them, as I regard them to be a top-notch camera manufacturer, which I still believe them to be.

I can add another horror story for you. I owned a Yankee Robotics camera with a great Kodak 6300? ( I think) chip in it. The camera had superior electronics. However, to save a bit of money, it was water cooled only (no peltier). A bit of a issue having a water bucket under the scope, but you could really cool it down with ice in the water.

 

Well, eventually the seals failed and the water started leaking into the CCD chamber, which had no cover slip on it. The prongs corroded over time and one night it got dewed up on the chip before completely failing. Yankee Robotics had gone out of business a month earlier......after I realized I had a giant paperweight, I took it apart to see what happened, ......that burned me between 3,000 and 4,000 dollars, I can't exactly remember....

 

I am fortunate my wife is very understanding.....



#10 mAnKiNd

mAnKiNd

    Mariner 2

  • *****
  • topic starter
  • Posts: 253
  • Joined: 20 Sep 2016
  • Loc: NY

Posted 17 May 2018 - 08:11 PM

Shiraz, that's really interesting, thank you for sharing that!

 

If indeed all CMOS sensors commercially sold by Sony (and by proxy perhaps Panasonic and Aptina like you said) that are found in astro cameras are of the same quality grade, then I suppose my question is moot for these CMOS manufacturers..

 

Thanks again

Minos



#11 CharlesW

CharlesW

    Long time member

  • ***--
  • Posts: 3,659
  • Joined: 01 Nov 2012
  • Loc: Chula Vista, CA

Posted 17 May 2018 - 08:12 PM

This document describes how sensors are graded, https://www.photomet...ccd_grading.pdf  

 

Grade zero and one sensors are substantially more expensive than grade two and I believe that most of the cameras that we are using have grade two chips. 


  • mAnKiNd likes this

#12 mAnKiNd

mAnKiNd

    Mariner 2

  • *****
  • topic starter
  • Posts: 253
  • Joined: 20 Sep 2016
  • Loc: NY

Posted 17 May 2018 - 08:15 PM

I can add another horror story for you. I owned a Yankee Robotics camera with a great Kodak 6300? ( I think) chip in it. The camera had superior electronics. However, to save a bit of money, it was water cooled only (no peltier). A bit of a issue having a water bucket under the scope, but you could really cool it down with ice in the water.

 

Well, eventually the seals failed and the water started leaking into the CCD chamber, which had no cover slip on it. The prongs corroded over time and one night it got dewed up on the chip before completely failing. Yankee Robotics had gone out of business a month earlier......after I realized I had a giant paperweight, I took it apart to see what happened, ......that burned me between 3,000 and 4,000 dollars, I can't exactly remember....

 

I am fortunate my wife is very understanding.....

Omg man!



#13 mAnKiNd

mAnKiNd

    Mariner 2

  • *****
  • topic starter
  • Posts: 253
  • Joined: 20 Sep 2016
  • Loc: NY

Posted 17 May 2018 - 08:17 PM

This document describes how sensors are graded, https://www.photomet...ccd_grading.pdf

 

Grade zero and one sensors are substantially more expensive than grade two and I believe that most of the cameras that we are using have grade two chips. 

CharlesW, thank you kindly for sharing this document, ill take a close read! 

 

From what I've understood so far, we can divulge that CCD sensors are graded but CMOS sensors are not. Would this be accurate?

 

Cheers

Minos


Edited by mAnKiNd, 17 May 2018 - 08:18 PM.


#14 AgilityGuy

AgilityGuy

    Viking 1

  • -----
  • Posts: 892
  • Joined: 20 Jun 2015
  • Loc: Northern CA

Posted 17 May 2018 - 08:20 PM

I’ve thought about the same topic recently since I have a Nikon D800E with a Sony IMX094 sensor.  The camera was released in 2012 and the same sensor was used in the D800 (makes sense) and the D810/D810A cameras.  Now in 2018 ZWO is releasing a cooled OSC camera using the same Sony IMX094 sensor for essentially $4000.  This was enough to prompt me to send mine off for Ha modification and use it at least a bit longer.  I bought the camera in excellent shape from a friend for $1000 more than a year ago and the Ha modification cost $350.  I’ve got a scope with a huge imaging circle so I can easily use the entire sensor.  I understand the benefits of cooling and maybe what’s been learned about sensors and electronics but to spend 4x the money for a camera with the same sensor wasn’t in the cards for me. It will be interesting to see where the cameras and prices go in the next couple years


  • mAnKiNd likes this

#15 mAnKiNd

mAnKiNd

    Mariner 2

  • *****
  • topic starter
  • Posts: 253
  • Joined: 20 Sep 2016
  • Loc: NY

Posted 17 May 2018 - 08:26 PM

Something that really piqued my interest about this was a Facebook post by QHY a couple of weeks back, got me thinking about all this. See attached:

 

This is a CMOS manufacturer other than Sony, Panasonic or Aptina. So I'm a bit confounded as to Atik's claim about Sony sensors commercially sold being all of the highest quality grade. Perhaps (speculation) this is just Sony QC bar being set higher than smaller manufacturers.

Attached Thumbnails

  • 05_17_18_09_22_52_Screenshot_20180517-212125.png
  • 1526606859460.png

Edited by mAnKiNd, 17 May 2018 - 08:28 PM.


#16 mAnKiNd

mAnKiNd

    Mariner 2

  • *****
  • topic starter
  • Posts: 253
  • Joined: 20 Sep 2016
  • Loc: NY

Posted 17 May 2018 - 08:30 PM

I’ve thought about the same topic recently since I have a Nikon D800E with a Sony IMX094 sensor.  The camera was released in 2012 and the same sensor was used in the D800 (makes sense) and the D810/D810A cameras.  Now in 2018 ZWO is releasing a cooled OSC camera using the same Sony IMX094 sensor for essentially $4000.  This was enough to prompt me to send mine off for Ha modification and use it at least a bit longer.  I bought the camera in excellent shape from a friend for $1000 more than a year ago and the Ha modification cost $350.  I’ve got a scope with a huge imaging circle so I can easily use the entire sensor.  I understand the benefits of cooling and maybe what’s been learned about sensors and electronics but to spend 4x the money for a camera with the same sensor wasn’t in the cards for me. It will be interesting to see where the cameras and prices go in the next couple years

Interesting....where is all that extra money going into? 



#17 spokeshave

spokeshave

    Mercury-Atlas

  • *****
  • Posts: 2,616
  • Joined: 08 Apr 2015

Posted 17 May 2018 - 08:53 PM

CMOS sensors are not generally graded. Because of their architecture, they generally either work or they don't. CCDs are graded, typically by the number of column defects (which CMOS sensors do not have because of the different readout architecture).

To add to what John said, yield is also related to die size. The larger the sensor, the more likely it is to have a defect somewhere. Additionally, the larger the sensor, the fewer dies there physically are on a wafer. A single defect on a 36x24mm die will ruin it. That same wafer area will contain, for example, 9 IMX174 dies, and that single defect will only take out one of them. That's why sensor cost goes up exponentially with size.

Tim
  • Alfredo Beltran, jhayes_tucson, Jon Rista and 1 other like this

#18 mAnKiNd

mAnKiNd

    Mariner 2

  • *****
  • topic starter
  • Posts: 253
  • Joined: 20 Sep 2016
  • Loc: NY

Posted 17 May 2018 - 08:55 PM

Thanks Tim

#19 AgilityGuy

AgilityGuy

    Viking 1

  • -----
  • Posts: 892
  • Joined: 20 Jun 2015
  • Loc: Northern CA

Posted 18 May 2018 - 12:31 AM

Interesting....where is all that extra money going into? 

Good question.  That's not the only example.  The Sony IMX071 sensor that was used in the D5100 (released April 2011) and D7000 (released September 2010) bodies is also being used in the ZWO ASI071MC Pro color camera.  Neither of the Nikon cameras sold for near the $1700 price point of the ZWO camera.  All of the cameras are essentially photon collectors but dollar for dollar the sensor fitted in the Nikons likely provided the user with a bit more utility.  I do not claim to know how CMOS sensor technology has changed from 2010/2011 to 2018 but I have to believe the advancement in sensor technology would make a more recently developed CMOS sensor more desirable in a camera of any type.


Edited by AgilityGuy, 18 May 2018 - 12:46 AM.

  • mAnKiNd likes this

#20 ecloud

ecloud

    Mariner 2

  • -----
  • Posts: 222
  • Joined: 17 Oct 2013
  • Loc: Dalian, China

Posted 19 May 2018 - 11:27 AM

Good question.  That's not the only example.  The Sony IMX071 sensor that was used in the D5100 (released April 2011) and D7000 (released September 2010) bodies is also being used in the ZWO ASI071MC Pro color camera.  Neither of the Nikon cameras sold for near the $1700 price point of the ZWO camera.  All of the cameras are essentially photon collectors but dollar for dollar the sensor fitted in the Nikons likely provided the user with a bit more utility.  I do not claim to know how CMOS sensor technology has changed from 2010/2011 to 2018 but I have to believe the advancement in sensor technology would make a more recently developed CMOS sensor more desirable in a camera of any type.

The industrial product cost nearly zero theoretically.

The key is the volume.

Nikon can eat a huge volume of the chip, so they can get the price of the chip 1/2 or even 2/3 off than other small customers.

If Nikon, Pentax and Fuji had no plan to use IMX071 in their camera, do you think Sony will produce this chip? Do you think Sony can reach the balance just by the poor volume purchased by itself and ZWO? That will not happen!

 

So, the price of any industrial product is mainly based on the volume. The more you buy, the lower you pay (per unit).

This can also answer the question of the lower price of ZWO/QHY than Atik.

 

China is a huge industrial country. Some productions are being built in an awesome volume here in China, steal, aluminum, plastic, cooler, fan, mainboard, etc. So the price is low.


  • AgilityGuy and mAnKiNd like 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