Pages: [1]   Go Down

Author Topic: Ink Jet paper fStop dynamic range..  (Read 6724 times)

AndreG

  • Jr. Member
  • **
  • Offline Offline
  • Posts: 82
    • photohippique.com
Ink Jet paper fStop dynamic range..
« on: June 19, 2010, 03:52:07 pm »

Hi,

Would you know the dynamic range in fStop of Inkjet paper?

Please excuse me if the topic has already been discussed. We read a lot about the gamut of the inks but rarely about the papers...

Anyone..
Logged

Geoff Wittig

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 1023
Ink Jet paper fStop dynamic range..
« Reply #1 on: June 19, 2010, 07:09:45 pm »

Quote from: pratic
Hi,

Would you know the dynamic range in fStop of Inkjet paper?

Please excuse me if the topic has already been discussed. We read a lot about the gamut of the inks but rarely about the papers...

Anyone..
Good question!
Below is a good general discussion of the central problem of photography: mapping the huge dynamic range of the real world down to the much smaller dynamic range of film or a digital sensor, then mapping that image down to the much smaller yet dynamic range of printing paper. The number that sticks in my head somewhere is a contrast range for cotton rag inkjet paper of about 80:1, but this paper quotes dynamic ranges from 275:1 for an ideal inkjet paper (obviously using photo black ink and lots of optical brightening agents) to 100:1 for platinum prints and as low as 40:1 for newsprint.
Unless my math is totally off, a dynamic range of 128:1 corresponds to 7 stops, and this is the absolute best you'll ever get from any inkjet or darkroom printing paper. Something like 6 stops is probably a more realistic usable dynamic range.

http://wwwimages.adobe.com/www.adobe.com/p...renderprint.pdf
Logged

Bart_van_der_Wolf

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 8914
Ink Jet paper fStop dynamic range..
« Reply #2 on: June 19, 2010, 07:32:25 pm »

Quote from: pratic
Would you know the dynamic range in fStop of Inkjet paper?

Please excuse me if the topic has already been discussed. We read a lot about the gamut of the inks but rarely about the papers...

It is not quite clear why you want it expressed in F-stops. The reason may require a different answer. Also, gamut is more about saturation at a given brightness level than about maximum density/brightness level differences.

However, assuming a glossy paper, a net density above base can reach a value of 2.1 or a bit more, which would be similar to 7 'stops' of exposure difference. BTW, this has little to do with the dynamic range of a camera, because the signals that the camera records will be altered by Gamma adjustment and non-linear tonemapping. With HDR tonemapping it is possible to fit a huge capture range into a small output range of densities.

Cheers,
Bart
Logged
== If you do what you did, you'll get what you got. ==

AndreG

  • Jr. Member
  • **
  • Offline Offline
  • Posts: 82
    • photohippique.com
Ink Jet paper fStop dynamic range..
« Reply #3 on: June 19, 2010, 10:06:29 pm »

Thank you for your answers, I have some studying to do. I never thought that the dynamic range was so narrow.

Again thank you, I am very grateful.
Logged

neile

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 1093
    • http://www.danecreekfolios.com
Ink Jet paper fStop dynamic range..
« Reply #4 on: June 20, 2010, 12:16:43 am »

Quote from: pratic
Thank you for your answers, I have some studying to do. I never thought that the dynamic range was so narrow.

Again thank you, I am very grateful.

If you want a list of how the different papers compare in this regard, try my inkjet paper list. It lists the max density readings I get for each paper on my printer, as well as what my eye can see on a ramp of black and white patches. The really good papers with good profiles will give a dmax of around 2.4, and I can see ink laydown at 254 (almost completely white) and can distinguish down to 2 (almost completely black).

Neil
Logged
Neil Enns
Dane Creek Folio Covers. Limited edition Tuscan Sun and Citron covers are now in stock!

ErikKaffehr

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 11311
    • Echophoto
Ink Jet paper fStop dynamic range..
« Reply #5 on: June 20, 2010, 01:28:33 am »

Hi,

Just a few points.

1) Density Range in f-stops is (Dmax-Dmin)/log(2). Log(2) is very close to 0.3

2) Dynamic range in capture and on paper are different things

3) You may check this page: http://www.normankoren.com/digital_tonality.html
 it's a bit heavy on graphs but go to the bottom of the page for some good illustrations

4) In practice there is always some kind of tonality mapping. Normally we use an 'S-curve', thus adding toe and shoulder characteristics. This is done by film, in cameras shooting JPEG and also normally in RAW-processing.

Regarding DR in capture things are more complicated. There is a technical definition of DR and that is that Maximum Signal/Noise floor. On digital sensors both values are measured in electrons, on DSLRs the maximum signal (known as "Full Well Capacity") used to be around 50000 and the noise floor about 3-15 electrons. That would give a dynamic range of about 5000:1 that is about 12 stops, this assuming that we expose to the right, that is maximize exposure without blowing out highlights.

Now things are not simple as that! To begin with about one percent of the light passing trough a lens will be reflected on the glass surfaces and just bounce around in the lens. Also lens barrel and mechanical parts of the camera will also be illuminated and will reflect some light on the sensor. This will be seen as either veiling flare or ghosting. So even if we had a sensor of infinite DR it would be limited by camera and lens. How well camera and lens is made matters. Some are probably better than others. large white surfaces (like windows) cause much flare while small bright areas (specular highlights) cause little flare.

Regarding the "noise floor" the engineering definition of DR says SNR = 1, signal noise ratio is one. This means quite noisy. On the other hand the dark areas in print will be compressed due to the S-curve so the noise may be much less visible.

The most important factor in transferring the dynamic range of a scene to paper is the tonality mapping.

Finally, the perceived dynamic range of a print will depend much on how the print is illuminated. Bright light increases the perception of contrast, the brain works that way. A print that is boring on a dark wall will come to life if illuminated with a bright spotlight.

Best regards
Erik


Quote from: BartvanderWolf
It is not quite clear why you want it expressed in F-stops. The reason may require a different answer. Also, gamut is more about saturation at a given brightness level than about maximum density/brightness level differences.

However, assuming a glossy paper, a net density above base can reach a value of 2.1 or a bit more, which would be similar to 7 'stops' of exposure difference. BTW, this has little to do with the dynamic range of a camera, because the signals that the camera records will be altered by Gamma adjustment and non-linear tonemapping. With HDR tonemapping it is possible to fit a huge capture range into a small output range of densities.

Cheers,
Bart
« Last Edit: June 20, 2010, 01:29:21 am by ErikKaffehr »
Logged
Erik Kaffehr
 

Colorwave

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 1006
    • Colorwave Imaging
Ink Jet paper fStop dynamic range..
« Reply #6 on: June 20, 2010, 02:33:51 am »

Geoff-
The paper you referenced from Karl Lang is the best written explanation of the process of capturing and processing light that I have ever read.  I'm far from a imaging scientist, but I've read quite a few books on the subject, and this short paper explains it all in such clear and easy to understand terms, but with enough detail to be significant, that it trumps all others in succinct bang for the buck.  I will certainly save it and encourage all of my clients and colleagues to read it, if I sense any sort of inquiring mind and desire to understand the subject.  Unfortunately, I think many of them will fail to grasp the concepts from a lack of desire to truly understand them, not any shortcoming of Karl to explain it well.  I personally find it fascinating, and while there was very little in it that I hadn't already been exposed to, I found myself admiring the explanatory process of Karl so much that it was fun to read and am sure that it focused some of the details for me as a part of the process.*  Thanks for sharing the link.

*My photo metaphors were strictly subliminal on my part.
Logged
-Ron H.
[url=http://colorwaveimaging.com

Damir

  • Full Member
  • ***
  • Offline Offline
  • Posts: 237
Ink Jet paper fStop dynamic range..
« Reply #7 on: June 20, 2010, 05:37:42 am »

Quote from: ErikKaffehr
Regarding DR in capture things are more complicated. There is a technical definition of DR and that is that Maximum Signal/Noise floor. On digital sensors both values are measured in electrons, on DSLRs the maximum signal (known as "Full Well Capacity") used to be around 50000 and the noise floor about 3-15 electrons. That would give a dynamic range of about 5000:1 that is about 12 stops, this assuming that we expose to the right, that is maximize exposure without blowing out highlights.

Not true Poisson statistic law define noise as the square root of signal, you can find many references, most easy to follow is:

http://en.wikipedia.org/wiki/Shot_noise

so if Full Well Capacity is 50.000, noise floor of the signal is 224 electrons, that is only one kind of noise in digital systems, there is also noise of semiconductor compunds, dark current noise and few others which renders dynamic range of the sensor more to the 256:1 or 8 stops. This is real life useful range - if you shoot test patches you can get 12 stops in case where you have full saturation of sensor therefore you have only white in picture - but in real world you will have shadows in picture, dark areas with much less signal, so the noise will be quite obvious.

Conclusion:
- if you shoot only high key images, you can get with big photo site on sensor dynamic range of 10 stops
- in real life situation 7 - 8 stop is what you get from the very same sensor

Print on regular photo paper can have in ideal situation dynamic range of 100:1
This is known for the long time, the same question people asked in predigital era.
With optical brightener that number can be higher.
« Last Edit: June 20, 2010, 05:39:45 am by Damir »
Logged

ErikKaffehr

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 11311
    • Echophoto
Ink Jet paper fStop dynamic range..
« Reply #8 on: June 20, 2010, 08:42:52 am »

Hi,

I'm familiar with shot noise. The 224 electron you mention is the standard of deviation for a Poisson distributed signal of 50000 samples. If you have a dark area in the image which yields on average 100 electrons/pixel the standard of deviation would be 10 electrons/per pixel. In the first case the Signal Noise Ratio would be 224 and in the second 10.

You seem to assume that you would have 224 electrons of shot noise on each pixel, independent of exposure? That's clearly wrong.

You may check this: http://theory.uchicago.edu/~ejm/pix/20d/te...oise/index.html

and this: http://www.clarkvision.com/articles/evaluation-1d2/


Best regards
Erik


Quote from: Damir
Not true Poisson statistic law define noise as the square root of signal, you can find many references, most easy to follow is:

http://en.wikipedia.org/wiki/Shot_noise

so if Full Well Capacity is 50.000, noise floor of the signal is 224 electrons, that is only one kind of noise in digital systems, there is also noise of semiconductor compunds, dark current noise and few others which renders dynamic range of the sensor more to the 256:1 or 8 stops. This is real life useful range - if you shoot test patches you can get 12 stops in case where you have full saturation of sensor therefore you have only white in picture - but in real world you will have shadows in picture, dark areas with much less signal, so the noise will be quite obvious.

Conclusion:
- if you shoot only high key images, you can get with big photo site on sensor dynamic range of 10 stops
- in real life situation 7 - 8 stop is what you get from the very same sensor

Print on regular photo paper can have in ideal situation dynamic range of 100:1
This is known for the long time, the same question people asked in predigital era.
With optical brightener that number can be higher.
« Last Edit: June 20, 2010, 08:47:12 am by ErikKaffehr »
Logged
Erik Kaffehr
 

Geoff Wittig

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 1023
Ink Jet paper fStop dynamic range..
« Reply #9 on: June 20, 2010, 09:36:06 am »

Quote from: Colorwave
Geoff-
The paper you referenced from Karl Lang is the best written explanation of the process of capturing and processing light that I have ever read.

I was also very impressed by this reference. I stumbled across the paper through a simple google search looking for paper dynamic range, since the 80:1 number was all I could remember. My favorite go-to source for clear thinking on perception, dynamic range and tonal mapping is still Ctein's wonderful book Post Exposure. It's mostly directed at film capture, but the first few chapters are a model of clarity in reviewing the differences between human visual perception and how film and paper manage contrast. Ctein includes a very explicit discussion of how the vast dynamic range of the real world is gradually mapped into progressively smaller ranges, from capture to print, and how to optimize the translation.
Logged

Damir

  • Full Member
  • ***
  • Offline Offline
  • Posts: 237
Ink Jet paper fStop dynamic range..
« Reply #10 on: June 26, 2010, 06:48:56 pm »

Quote from: ErikKaffehr
Hi,

I'm familiar with shot noise. The 224 electron you mention is the standard of deviation for a Poisson distributed signal of 50000 samples. If you have a dark area in the image which yields on average 100 electrons/pixel the standard of deviation would be 10 electrons/per pixel. In the first case the Signal Noise Ratio would be 224 and in the second 10.

You seem to assume that you would have 224 electrons of shot noise on each pixel, independent of exposure? That's clearly wrong.

You may check this: http://theory.uchicago.edu/~ejm/pix/20d/te...oise/index.html

and this: http://www.clarkvision.com/articles/evaluation-1d2/


Best regards
Erik

I know that is what I am saying - you can have 12 stops on completely white patch - in reality you want in your picture more information that completely blown out whites!
Logged
Pages: [1]   Go Up