Forgot Password
Pentax Camera Forums Home
 

Reply
Show Printable Version Search this Thread
10-16-2015, 02:51 AM   #1
Site Supporter




Join Date: Nov 2014
Posts: 211
Sensor Pixel density, and resolving power?

I recently had the chance to try out a Phase One IQ3 80MP at a 'hands-on' day in Birmingham - an amazing piece of equipment , if you happen to have 32,000 to spare.....or sell your used Mercedes AMG.


After looking at the specs I realised that pixels per square millimetre on sensors varies widely depending on sensor size , and the league table looks like this:


1 Canon 5DS - 578 pixels per sq mm


2 Nikon D810/800 - 418 pixels per sq mm


3 Phase One IQ3 80MP - 369 pixels per sq mm


4 Pentax 645Z - 344 pixels per sq mm


It seems to follow that the glass in front of the sensor plays an increasingly important role in determining the OVERALL image quality , regardless of definition/sharpness however that is measured......


What interested me most , was Phase One's claim that the Schneider-Kreuznach lenses developed for the 80MP camera , are 'capable of resolving beyond 100MP'.... I'd be interested to hear comments on this, as most opinions I've read seem to suggest that a lens cannot 'out-resolve' a sensor ??

10-16-2015, 03:01 AM   #2
Pentaxian
Digitalis's Avatar

Join Date: Mar 2009
Location: Adelaide.
Posts: 8,795
QuoteOriginally posted by BostonUKshooter Quote
as most opinions I've read seem to suggest that a lens cannot 'out-resolve' a sensor ?
Many lenses can, and many do indeed out resolve the sensor. The important thing to note that lens performance is limited by diffraction itself. Coincidentally, commercial mass produced imaging sensors are also limited by diffraction due to the photo lithographic process by which they are made*. Theoretically a diffraction limited, optically perfect 50mm f/1.0 lens can resolve approximately 1400 lpp/mm - there isn't a imaging sensor on the planet that can record that level of detail, However truth be told there isn't anyone who has attempted to construct a lens like that...after all what would be the point?

* Electron beam lithography is a way of making smaller devices, however it is a slow process.
10-16-2015, 03:31 AM   #3
Veteran Member
Kolor-Pikker's Avatar

Join Date: Dec 2014
Posts: 341
QuoteOriginally posted by BostonUKshooter Quote
What interested me most , was Phase One's claim that the Schneider-Kreuznach lenses developed for the 80MP camera , are 'capable of resolving beyond 100MP'.... I'd be interested to hear comments on this, as most opinions I've read seem to suggest that a lens cannot 'out-resolve' a sensor ??
The missing key here is - at which level of contrast does the lens out-resolve the sensor? Many old lenses are technically capable of resolving very high levels of detail, but due to primitive coatings and other processes, most of them have fairly poor contrast, which leads to less perceived detail.
This wasn't too big of a problem on film, because as you near the resolution limit of film, it's ability to record contrast gradually degrades, whereas sensors maintain 100% contrast right up to the resolution limit. The effect of this is that any discrepancy in the ability of a lens to render really fine details at high contrast is immediately visible.
Anyway, there is no such thing as having too few pixels, as long as cramming more of them in doesn't come at a cost of noise or dynamic range.
10-16-2015, 05:48 AM   #4
Pentaxian




Join Date: Feb 2015
Photos: Albums
Posts: 3,527
The above being said, I'd add that lenses do not have a very sharp drop of contrast to 0.0 beyond their 50% cutoff frequency, that means that higher sensor resolution gather micro-contrast info, even if the sensor outresolve the lens (that's why you still get a usable image out of a 16Mpixel Point and shot camera, the contrast for high frequency components of the image is just not as good as with a larger sensor).

QuoteOriginally posted by Digitalis Quote
Coincidentally, commercial mass produced imaging sensors are also limited by diffraction due to the photo lithographic process by which they are made*.
That's a good point. However, todays tech node is at 14nm, basically 142 times better than the pixel pitch of a Pentax Q sensor. Image sensor can be made with techno that is far behind the best available silicon techno.

10-16-2015, 07:10 AM   #5
Pentaxian
Digitalis's Avatar

Join Date: Mar 2009
Location: Adelaide.
Posts: 8,795
QuoteOriginally posted by biz-engineer Quote
todays tech node is at 14nm, basically 142 times better than the pixel pitch of a Pentax Q sensor.
The Best is sometimes a very different thing from what is Optimal. The problem is with that kind of pixel pitch the photosites would be smaller than any of the component frequencies of visible light. Off the top of my head I can't think if any reason someone would want to photograph subjects illuminated by monochomatic 14nm light (extreme vacuum UV, bordering on X-rays) - you wouldn't capture anything due to the absorption of such high frequencies by the air itself.

Larger photosites are superior when it comes to low light imaging, larger sensors have more area for the photosites to be distributed. It is possible for small photosites to be merged into a larger one - achieving greater charge capacities than they would alone, however this approach reduces overall resolution, it makes the chip design considerably more complex, increases electrical cross talk, and makes spurious effects like noise and dark current more problematic and the manufacture of such chips is inherently more failure prone.

Last edited by Digitalis; 10-16-2015 at 07:20 AM.
10-16-2015, 08:01 AM   #6
Pentaxian




Join Date: Feb 2015
Photos: Albums
Posts: 3,527
QuoteOriginally posted by Digitalis Quote
The Best is sometimes a very different thing from what is Optimal. The problem is with that kind of pixel pitch the photosites would be smaller than any of the component frequencies of visible light. Off the top of my head I can't think if any reason someone would want to photograph subjects illuminated by monochomatic 14nm light (extreme vacuum UV, bordering on X-rays) - you wouldn't capture anything due to the absorption of such high frequencies by the air itself. Larger photosites are superior when it comes to low light imaging, larger sensors have more area for the photosites to be distributed. It is possible for small photosites to be merged into a larger one - achieving greater charge capacities than they would alone, however this approach reduces overall resolution, it makes the chip design considerably more complex, increases electrical cross talk, and makes spurious effects like noise and dark current more problematic and the manufacture of such chips is inherently more failure prone.
I was referring to the sentence I quoted. Commercially mass produced sensors are not limited by the photo-lithography process. At 14nm, we don't need electron beam litho, photolithogaphy is still usable. We use focused ion beam litho , but only for engineering debug, because it would be far too slow for production. I know that larger photosites are better for low light photography. Using multiple smaller photosites and averaging their results does not decrease overall resolution, provide better overall noise by getting same averaged thermal noise but with the advantage of having the read noise averaged, involves more circuitry as you mentioned. The reason why D610/D750 are cheaper than D810 is not only a marketing reason, it is also that for the same fab defect density (D0), the yield of D810 sensor is lower than the yield of the 24Mp sensor. Lot to lot yield variation is likely higher for the 36Mp sensor than for the 24Mp sensor.
10-16-2015, 08:42 AM   #7
Pentaxian
Digitalis's Avatar

Join Date: Mar 2009
Location: Adelaide.
Posts: 8,795
QuoteOriginally posted by biz-engineer Quote
Commercially mass produced sensors are not limited by the photo-lithography process
No. they aren't, however there are limitations on how small things get before the sensors cease to be efficient enough for visible light photography.
10-16-2015, 10:38 AM   #8
Pentaxian




Join Date: Feb 2015
Photos: Albums
Posts: 3,527
QuoteOriginally posted by Digitalis Quote
No. they aren't, however there are limitations on how small things get before the sensors cease to be efficient enough for visible light photography.
Yes, you are right; optically down scaling hit some limits. I have to do a bit of research regarding the drop of efficiency at smaller pixel sizes.

Reply

Bookmarks
  • Submit Thread to Facebook Facebook
  • Submit Thread to Twitter Twitter
  • Submit Thread to Digg Digg
Tags - Make this thread easier to find by adding keywords to it!
645d, 645z, camera, canon 5ds, d810 and pentax, medium format, mm, phase, phase one, pixels, sensor, sensor pixel
Thread Tools Search this Thread
Search this Thread:

Advanced Search


Similar Threads
Thread Thread Starter Forum Replies Last Post
Pentax 645 lens resolving power BrunoL Pentax Medium Format 13 04-03-2015 01:38 PM
Help me understand the 645Z sensor's resolving power 6BQ5 Pentax Medium Format 21 04-26-2014 09:14 AM
Variable Neutral Density Filter to over power the sun mictadlo Photographic Technique 9 12-09-2013 04:25 AM
Nature DA 40 Ltd shows resolving power lectrolink Post Your Photos! 3 12-25-2011 12:07 PM
K20D: Relationships among Pixel Density, Sensor, Lens starstuff Pentax DSLR Discussion 12 08-27-2009 03:23 AM



All times are GMT -7. The time now is 06:13 AM. | See also: NikonForums.com, CanonForums.com part of our network of photo forums!
  • Red (Default)
  • Green
  • Gray
  • Dark
  • Dark Yellow
  • Dark Blue
  • Old Red
  • Old Green
  • Old Gray
  • Dial-Up Style
Hello! It's great to see you back on the forum! Have you considered joining the community?
register
Creating a FREE ACCOUNT takes under a minute, removes ads, and lets you post! [Dismiss]
Top