Forgot Password
Pentax Camera Forums Home
 

Reply
Show Printable Version Search this Thread
07-28-2015, 08:35 PM   #226
Pentaxian
philbaum's Avatar

Join Date: Oct 2007
Location: Port Townsend, Washington State, USA
Posts: 3,659
QuoteOriginally posted by mohb Quote
I sometimes wonder how those forever calling for increased dynamic range and high ISO would have coped with slide film where 1/2 a stop either way was crucial and ISO 100 was fast
I fail to see what your point is. Are you suggesting we all return to film cameras? I have no problem with higher ISO cameras, the higher the better - don't get all this nostalgia business.


Last edited by philbaum; 07-28-2015 at 09:08 PM.
07-28-2015, 10:24 PM   #227
Veteran Member
Ikarus's Avatar

Join Date: Apr 2010
Location: California
Photos: Albums
Posts: 462
QuoteOriginally posted by Ratcheteer Quote
that article needs updating cmos technology has improved quite a bit in 14 years
Feel free to point us to a more recent scientific article of similar quality demonstrating that with modern technologies, pixel size has become irrelevant for dynamic range. I'd be very surprised to see that. The laws of physics pertaining to scale usually aren't subject to revision. 4-cylinder turbo engines now make power figures that needed 12-cylinder engines in the sixties, but all this progress doesn't change the fact that all else being equal, a large displacement engine still makes more power than a small displacement one. Of course you can argue that nobody really needs a 700hp 12-cylinder monster in the garage for taking it shopping on the weekend, and I'd agree with that. In photographic terms, however, dynamic range is still in short supply. Otherwise, HDR wouldn't be so popular.

Last edited by Ikarus; 07-28-2015 at 10:30 PM.
07-28-2015, 11:15 PM   #228
Pentaxian




Join Date: May 2009
Location: Ponoka Alberta Canada
Photos: Gallery
Posts: 541
QuoteOriginally posted by Ikarus Quote
Feel free to point us to a more recent scientific article of similar quality demonstrating that with modern technologies, pixel size has become irrelevant for dynamic range. I'd be very surprised to see that. The laws of physics pertaining to scale usually aren't subject to revision. 4-cylinder turbo engines now make power figures that needed 12-cylinder engines in the sixties, but all this progress doesn't change the fact that all else being equal, a large displacement engine still makes more power than a small displacement one. Of course you can argue that nobody really needs a 700hp 12-cylinder monster in the garage for taking it shopping on the weekend, and I'd agree with that. In photographic terms, however, dynamic range is still in short supply. Otherwise, HDR wouldn't be so popular.

ERIC R. FOSSUM and G. AGRANOVG
(From memory)

Published a paper on Sub 2μm Pixels for CMOS (in about the 1.75 μm -1.4 μm)
With pixel of this size if the SNR of that pixel are scaled proportionally to the pixel area no DR is lost.
The conclusion from the paper from what I gathered was that at this time they are able to scale SNR right down to at least 1.7 μm. It should be within the realm for manufactures to easily scale SNR in sensors in the range of 2 μm and I am sure they are doing this as we speak

Iíll look for papers Maybe Falk knows of this paper and can point me in the right direction
07-29-2015, 12:09 AM   #229
Pentaxian




Join Date: Feb 2010
Photos: Albums
Posts: 1,436
QuoteOriginally posted by philbaum Quote
I fail to see what your point is. Are you suggesting we all return to film cameras? I have no problem with higher ISO cameras, the higher the better - don't get all this nostalgia business.
I was just wondering if some of the 'higher ISO more DR' fanatics would have persevered with photography in the film era.
Perhaps some of my generation appreciate the abilities of modern cameras without feeling a need to shout for more because we remember 'the good old days' when anything less than sunshine made taking photos a challenge

07-29-2015, 12:19 AM   #230
Veteran Member
Ikarus's Avatar

Join Date: Apr 2010
Location: California
Photos: Albums
Posts: 462
QuoteOriginally posted by Ian Stuart Forsyth Quote
ERIC R. FOSSUM and G. AGRANOVG
(From memory)

Published a paper on Sub 2μm Pixels for CMOS (in about the 1.75 μm -1.4 μm)
With pixel of this size if the SNR of that pixel are scaled proportionally to the pixel area no DR is lost.
The conclusion from the paper from what I gathered was that at this time they are able to scale SNR right down to at least 1.7 μm. It should be within the realm for manufactures to easily scale SNR in sensors in the range of 2 μm and I am sure they are doing this as we speak

I’ll look for papers Maybe Falk knows of this paper and can point me in the right direction
There's no denying that a lot of progress has been made for increasing the SNR, but the question still stands - based on what effect should the benefit of larger pixels for dynamic range have vanished? Doing a cursory search on Dr. Fossum's work, I found this article which in turn references this web page specifically for SNR and dynamic range of SLRs, where you find this clear and unambiguous statement:

QuoteQuote:
with the larger pixels potentially collecting more photons, those larger pixels can also have a higher dynamic range
I still don't see a smoking gun here and I really don't see how it could be any different, given that larger pixels just can't help it - they are bound to gather more photons.
07-29-2015, 01:09 AM   #231
Pentaxian
Fogel70's Avatar

Join Date: Sep 2009
Location: Stockholm, Sweden
Photos: Albums
Posts: 1,741
QuoteOriginally posted by mohb Quote
I was just wondering if some of the 'higher ISO more DR' fanatics would have persevered with photography in the film era.
Perhaps some of my generation appreciate the abilities of modern cameras without feeling a need to shout for more because we remember 'the good old days' when anything less than sunshine made taking photos a challenge
Back in the days, the obsession was about fast lenses instead of better high ISO performance.

While we today think that a f/2.8 zoom lens is a fast lens perfectly usable in low light.
Most low light shooter back in the day wanted to use FI 50/1.2 or 135/1.8 lenses, as f/2.8 was considered too slow for low light.

---------- Post added 29-07-15 at 10:43 ----------

QuoteOriginally posted by Ikarus Quote
I still don't see a smoking gun here and I really don't see how it could be any different, given that larger pixels just can't help it - they are bound to gather more photons.
The same physics that apply to pixel size do also apply to sensor size. So when you compare two images you compare the total amount of photos each sensor captures, not how many photons each pixel capture.
The amount of photos each pixel capture is just a small sample of the whole image, and the sample size of each pixel will depend on the total amount of pixels on the sensor.

Last edited by Fogel70; 07-29-2015 at 01:43 AM.
07-29-2015, 02:36 AM   #232
Site Supporter




Join Date: Mar 2009
Location: Gladys, Virginia
Photos: Gallery
Posts: 16,005
QuoteOriginally posted by Ikarus Quote
Feel free to point us to a more recent scientific article of similar quality demonstrating that with modern technologies, pixel size has become irrelevant for dynamic range. I'd be very surprised to see that. The laws of physics pertaining to scale usually aren't subject to revision. 4-cylinder turbo engines now make power figures that needed 12-cylinder engines in the sixties, but all this progress doesn't change the fact that all else being equal, a large displacement engine still makes more power than a small displacement one. Of course you can argue that nobody really needs a 700hp 12-cylinder monster in the garage for taking it shopping on the weekend, and I'd agree with that. In photographic terms, however, dynamic range is still in short supply. Otherwise, HDR wouldn't be so popular.
My understanding is that dynamic range of a sensor is related to the total number of photons that sensor can capture, not how much each of the individual pixels can capture. This is why the D810 (36 megapixel sensor) out performs both the D610 and the Sony A7s when it comes to dynamic range when the results are normalized.

I agree that more dynamic range would be useful, but simply increasing pixel size and decreasing number of pixels doesn't necessarily do much.
07-29-2015, 03:00 AM   #233
Pentaxian
falconeye's Avatar

Join Date: Jan 2008
Location: Munich, Alps, Germany
Photos: Gallery
Posts: 6,863
QuoteOriginally posted by Ikarus Quote
I would still like to see you respond to the article I was referring to
QuoteOriginally posted by Ikarus Quote
The laws of physics pertaining to scale usually aren't subject to revision.
QuoteOriginally posted by Ian Stuart Forsyth Quote
I’ll look for papers Maybe Falk knows of this paper and can point me in the right direction
QuoteOriginally posted by Ikarus Quote
I still don't see a smoking gun here and I really don't see how it could be any different, given that larger pixels just can't help it - they are bound to gather more photons.
This entire confusion is probably due to my inability to explain properly.

First, if you read my post #215 again, I explained that even the ideal camera has a finite dynamic range which is a function DR(ISO,f) of both (equivalent) base ISO and spatial frequency f. DR(ISO,f) is both strictly monotonically decreasing with arguments ISO and f. And because it is physics (optics), there is no CMOS or sensor tech involved. No need to even read those papers.

Therefore and of course, a larger pixel (which corresponds to a smaller spatial frequency) has a larger DR. I need not to read the paper because it says the same. If you look at the dynamic range of a pixel, you look at DR(ISO,f=sensorheight/pixelpitch). However, we photographers only care about DR of the image (i.e., at some predetermined f), not DR of a single pixel.

If anybody wonders how sensor tech could influence these findings: not much, provided read noise isn't much larger than shot noise and quantum efficiency isn't much smaller than 1. Both being true now for the best sensors out there.

Therefore, any understanding of dynmic range starts with understanding its property for the ideal camera (*).

This is what I advocated and it is where equivalence (the attempt to really understand it) does actually help.

Therefore, I won't point to further literature of SNR properties of actual pixels on sensors. It would be misleading. Understanding is always more important than finding a reference for a claim.

__
(*) An ideal camera is a camera registering angle and energy of each photon passing its aperture and shutter, with the finite precision allowed by the quantum nature of photons including Heisenberg's uncertainty principle for the complementary variables of angle/aperture-size and energy/shutter-duration. Aperture and shutter may be virtual, i.e., they are just boundary constraints for photon wavefunctions in space and time, resp. The angle/aperture-size uncertainty is known as lens diffraction to photographers. The energy/shutter-duration uncertainty means that at ultra-short exposure times, color becomes uncertain. This is a non-issue for photographers (so far). The quantum nature of photons means images must be noisy with a finite exposure time.

An ideal camera does never clip, it has a native base ISO of 0. In order to define an ideal camera with a finite base ISO, I define an ideal camera with equivalent base ISO 100 to clip after 1.5E12 (1.5 trillion) registered photons. That's roughly the ISO standard as applied by DxO (for a 24x36mm sensor).

Therefore, DR(ISO=100,f=1/0.5) = 0.75 trillion:1 = 237 dB = 39.5 EV. And DR(ISO=100,f=2309) = 17.5 EV. For the ideal camera, using DxO's 8MP value for f when assuming a 3:2 aspect ratio. An ideal camera has no limit for f except for its diffraction limits. f=1/0.5=2 corresponds to an image half white and half black.

ERRATUM (added July 30):
The ISO 100 clipping level of 1.5 trillion registered photons applies to the current level of quantum efficiency (which is about 50%) and color discrimination filter absorption (which is about another 50%). Therefore, a truely ideal camera would have 2 more stops DR at ISO 100 which is close to 20 EV at ISO 100. Nevertheless, the disclaimer changes nothing about the remainder of the argument. As noted, QE doesn't affect the maximum possible DR.


Last edited by falconeye; 07-30-2015 at 04:18 AM.
07-29-2015, 08:19 AM   #234
Veteran Member
Ikarus's Avatar

Join Date: Apr 2010
Location: California
Photos: Albums
Posts: 462
QuoteOriginally posted by Rondec Quote
My understanding is that dynamic range of a sensor is related to the total number of photons that sensor can capture, not how much each of the individual pixels can capture.
That still doesn't make any sense to me. The sensor is made up of those pixels, so surely, it inherits their dynamic range. Two sensors of the same size but with different pixel DR necessarily have the same difference in DR.

QuoteOriginally posted by Rondec Quote
I agree that more dynamic range would be useful, but simply increasing pixel size and decreasing number of pixels doesn't necessarily do much.
It does for dynamic range at the expense of resolution. That's what led Sony to release the A7s with its 12mp full-frame sensor and 15.3bits dynamic range. If they could have hit that number with a 36mp sensor, they would have.
07-29-2015, 08:46 AM   #235
Site Supporter




Join Date: Mar 2009
Location: Gladys, Virginia
Photos: Gallery
Posts: 16,005
QuoteOriginally posted by Ikarus Quote
That still doesn't make any sense to me. The sensor is made up of those pixels, so surely, it inherits their dynamic range. Two sensors of the same size but with different pixel DR necessarily have the same difference in DR.



It does for dynamic range at the expense of resolution. That's what led Sony to release the A7s with its 12mp full-frame sensor and 15.3bits dynamic range. If they could have hit that number with a 36mp sensor, they would have.
I think the 12 megapixel sensor had more to do with video (and maintaining performance at high iso) than dynamic range. Actually the A7s has less dynamic range than a K5 II up to iso 400.
07-29-2015, 09:09 AM   #236
Veteran Member
Ikarus's Avatar

Join Date: Apr 2010
Location: California
Photos: Albums
Posts: 462
QuoteOriginally posted by Rondec Quote
I think the 12 megapixel sensor had more to do with video
Sure, but that's just because the resolution/DR tradeoff for video is different. Why record 36mp and throw away much of the data, when you can get close to the DR of a professional movie camera at 12mp?
07-29-2015, 09:16 AM - 1 Like   #237
Pentaxian
falconeye's Avatar

Join Date: Jan 2008
Location: Munich, Alps, Germany
Photos: Gallery
Posts: 6,863
QuoteOriginally posted by Ikarus Quote
The sensor is made up of those pixels, so surely, it inherits their dynamic range.
Please, make an effort to understand.

The dynamic range of a sensor is NOT the dynamic range of the pixels it is made from.

Please, try to understand that dynamic range is a function of spatial frequency. If you have a problem with this sentence, then please ask questions.

Maybe, imagine an image made from black and white dots only (as produced by laser printers). Such images can have great tonality when printed at high enough resolution and viewed from a distance. Still, the tonality of a single dot is little.

Yor are making the same mistake when assuming that dynamic range is a property at the pixel level. It is not. An image has dynamic range even if not made from pixels.

---------- Post added 29-07-15 at 17:24 ----------

QuoteOriginally posted by Ikarus Quote
That's what led Sony to release the A7s with its 12mp full-frame sensor and 15.3bits dynamic range. If they could have hit that number with a 36mp sensor, they would have.
Sony's 36MP A7r has 1 stop better DR than the A7s (at base ISO). And this number isn't 15.3 bits, it is 13.2 EV.

This 15.3 bit figure has little to do with DR. It is a detail of the A7s raw file format:
QuoteQuote:
The sensorís dynamic range has also been further extended to 15.3 stops sensor RAW output. Technical note: as the bit depth remains unchanged and is presumed to be 14-bit ARW, this enhancement implies a modified raw gamma curve.
source: Sony confirm silent mode in A7S
Ironically, a higher MP camera has less problems with the RAW file bit depth to achieve high DR. In the case of the A7s, it didn't help anyway as its read noise at base ISO is much larger than quantization noise.

As the 15.3 bit file feature came as a firmware update, I verified with DxO staff (via a thread in their forum) that DxO did measure 13.2 EV with that firmware indeed.

Last edited by falconeye; 07-29-2015 at 09:49 AM.
07-29-2015, 09:25 AM   #238
Pentaxian
Class A's Avatar

Join Date: Aug 2008
Location: Wellington, New Zealand
Posts: 9,176
QuoteOriginally posted by Ikarus Quote
I'm referring to those who are with the article "How Small Should Pixel Size Be?", Proc. SPIE 3965, Sensors and Camera Systems for Scientific, Industrial, and Digital Photography Applications, 451 (May 15, 2000); doi:10.1117/12.385463" by T. Chen et al., available online here:
I had a very brief look at the article and it seems they arrive at their result because they do not consider microlenses and hence obtain atrocious effective fill-factors for small pixels.

This is not realistic and with modern real sensors you will find that smaller pixels do not lead to noisier images.

BTW, the A7s does not owe it's high-ISO performance to a low pixel count but due to distinctively not being an "ISO-less" camera, a design that indeed compromises its low ISO performance to a small degree.

The low pixel count is indeed meant to support fast read-out times and to support 4K without the need for supersampling.
07-29-2015, 10:26 AM   #239
Veteran Member
Ikarus's Avatar

Join Date: Apr 2010
Location: California
Photos: Albums
Posts: 462
QuoteOriginally posted by falconeye Quote
Maybe, imagine an image made from black and white dots only (as produced by laser printers). Such images can have great tonality when printed at high enough resolution and viewed from a distance. Still, the tonality of a single dot is little.
First of all, it's good to see we agree that there is trade-off between DR and resolution, since stepping back is equivalent to lowering the resolution by virtue of the eye integrating over an area of pixels, which results in a DR that exceeds the one of the individual pixel. You can, of course, display a greyscale image based on 1-bit pixels if the viewer is just far enough away, but that does not change the fact that the DR at the original resolution is still just 1 bit. It also does not change the fact that the DR at the larger viewing distance scales with the DR of the individual pixel.

Furthermore, if you want to build a capture device based on 1-bit pixels you will find that it won't come out this way. Instead of the nice dithering that conspires to a greytone at a downsampled resolution, you end up with whole areas clipped to black or white.
07-29-2015, 10:50 AM   #240
Senior Moderator
Loyal Site Supporter
Parallax's Avatar

Join Date: Apr 2008
Location: South Dakota
Photos: Gallery
Posts: 15,591
QuoteOriginally posted by falconeye Quote
Please, make an effort to understand.

The dynamic range of a sensor is NOT the dynamic range of the pixels it is made from.

Please, try to understand that dynamic range is a function of spatial frequency. If you have a problem with this sentence, then please ask questions.

Maybe, imagine an image made from black and white dots only (as produced by laser printers). Such images can have great tonality when printed at high enough resolution and viewed from a distance. Still, the tonality of a single dot is little.

Yor are making the same mistake when assuming that dynamic range is a property at the pixel level. It is not. An image has dynamic range even if not made from pixels.
That's a great explanation.
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!
24x36mm, 600mm, ability, aps-c, camera, cameras, depth, depth of field?...can, equivalence jihadis, ff, frame, full-frame, image, inch, lens, lines, lot, lots, magnification, pentax, people, picture, pictures, post, sensor, sensors, stop, sunset
Thread Tools Search this Thread
Search this Thread:

Advanced Search


Similar Threads
Thread Thread Starter Forum Replies Last Post
Depth-of-field bracketing -- is there a way to do this? baj Pentax K-5 16 07-04-2015 02:17 PM
Depth of field and focus trouble with manual lens Craig Barber Troubleshooting and Beginner Help 10 07-09-2014 07:18 PM
diffraction vs depth of field for Q hoffo Pentax Q 6 02-07-2014 09:38 PM
K10D and depth of field rimofheaven Troubleshooting and Beginner Help 7 01-04-2010 09:21 PM
portraits and depth of field raz Pentax SLR Lens Discussion 31 06-01-2007 08:35 PM



All times are GMT -7. The time now is 07:07 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