The Maths of a Fine B&W Print

Heads in a freezer

A
Heads in a freezer

  • 2
  • 0
  • 485
Route 45 (Abandoned)

A
Route 45 (Abandoned)

  • 1
  • 0
  • 596
Sonatas XII-48 (Life)

A
Sonatas XII-48 (Life)

  • 2
  • 3
  • 906
Waldsterben

D
Waldsterben

  • 2
  • 0
  • 1K
Microbus

H
Microbus

  • 3
  • 1
  • 2K

Recent Classifieds

Forum statistics

Threads
199,674
Messages
2,795,270
Members
99,999
Latest member
inukadk
Recent bookmarks
0

Michel Hardy-Vallée

Membership Council
Subscriber
Joined
Apr 2, 2005
Messages
4,794
Location
Montréal, QC
Format
Multi Format
Word, I must be in good writing shape today, thank you PE! That darn MA thesis is heating my Wernicke area.
 
Joined
Nov 22, 2004
Messages
226
Location
Bilthoven, T
Format
4x5 Format
Jed;

I would have to disagree to some extent in what you have said. The proof is often only in the images, not MTF, RMSG, curve shape etc.

As one manager often said to us when we quoted math figures "We sell photos, not curves".

PE

The interesting result found in the three studies ( in fact the Kodak research labs made more studies as well) is the almost 100% correlation between MTF and the SUBJECTIVE image quality judgement by a panel. After all, this might be not so surprising because the MTF of the human eye is a desicive factor. However, I must note, we are talking on the technical image quality. A technical perfect photo is not necessarily a perfect photo. But it is a prerequisite for a good photo, that it is technically perfect. And therefore experience with MTF is a valuable tool in your photographic bag.
In digital photography, an US firm has developed software to modify the MTF ( and therefore the appearance) of digital prints.
In analog photography we use the MTF to select the lens, the camera/ negative size and film as a first step to our final goal: the print. And after this selection ( you are in the ball park) there is so much left before your print is made. And once the print is made, and you are not satified with the result, MTF is a great help to make the proper modifications in your next photographs. I consider the MTF as a helpful tool before making photographs ( planning stage). Planning the selection of lens and camera, film and developer , size of the print etc. Based on this principle, I make another selection in case of landscapes than in the case of portraits. And, I differentiate in Europe between landscapes in different countries, because of the difference in light.
It is now 60 years ago that MTF was created at the Zeiss factory. However the old 17 th century (Dutch) masters were using (high definition) already. A painter like Rembrandt was using a magnifying glass and a brush with only a few hairs only. A painter like Salvador Dali was using even a microscope in the 20 th century. When we translate this in MTF language: we are using high spatial frequencies or high definition. The insight and the language have changed; but the idea is centuries old. Rembrandt was using the technique; but was Rembrandt just selling technique?
MTF is a tool to achieve technique, but do not expect that it will solve everything. As a matter of fact, the German study is mentioning the result of image quality, noticed by the panel. The effect of grain is an example. I agree, that it is unlikely to find a 'math formula' that will solve everything. Nevertheless, the laws of physics are pushed as far as possible. I know that Einstein said: 'religion starts where my knowledge stops'; therefore the boundary is always moving.
Many photographers use a lightmeter, because it is a useful tool. Even when many photographers know that a misreading is certainly possible. Every tool has its limitations. You have to be aware of these limitations; and we are back to Einstein.

Jed
 
Last edited by a moderator:

Photo Engineer

Subscriber
Joined
Apr 19, 2005
Messages
29,018
Location
Rochester, NY
Format
Multi Format
Jed;

I am aware of the studies and the actual people involved in conducting the studies. The key word though was that these studies were conducted by varying the MTF but keeping everything else constant. If that was so, then the MTF was of primary importance. However, if the image characteristics changed such as contrast or tone, often the person would pick another print.

In some cases, the subjects were told to 'select the sharpest print' not 'select the best print' nor were they given, in all cases, the optimum print or sometimes even a beautiful print.

I was a subject in one of those tests for color, and I was one of the people who measured the prints in another test. I know how they were conducted and how they were tested.

MTF is, as Roger said, a red herring. It is only ONE factor among many.

Also, I have seen digital prints where the MTF is exaggerated, and in large images they look exceedingly poor. In fact, they get worse as magnification increases. I have done this exercise to prove to my own satisfaction that this is not the be all and end all for digital.

PE
 

gainer

Subscriber
Joined
Sep 20, 2002
Messages
3,699
Maybe I should go back and read the beginning of this thread. I am a bit puzzled about the term MTF. I know it as a Fourier Transform from my ancient days as a so-called "Expert on non-linear modelling of the human operator" which was not my description but appeared on my position description while working at NASA Langley Research Center. Are we talking about a transfer between scene and picture, or between camera and picture, or simply talking about frequency content? Does the shape of the filtering curve correlate with anything? We used to think we could boost the highs and lows of an audio amplifier and get high fidelity. After all, a human can only hear 15 to maybe 20,000 hz. The trouble was that such things as aliasing of unresolved highs could fold back into the lows and cause audible distortions. Does the same sort of thing happen with photographs? I need to be educated.
 

Joe VanCleave

Member
Joined
Jan 20, 2004
Messages
677
Location
Albuquerque,
Format
Pinhole
Target The 'Target Audience'

Hi all,

I am giving a talk on B&W prints to people who are digital photographers who know nothing about traditional B&W process. So I need to know some of the maths involved to enable them to understand how a B&W print can have so much detail in the shadow and highlight areas. If anyone knows the maths involved here of the difference between the DMAX of B&W compared to digital I would appreciate it greatly. Many Thanks Paul

Paul, since your target audience seems to be people fundamentally unfamiliar with the techniques of silver gelatin photography, and your stated goal is their understanding of the qualitative tonal aspects of B&W prints, that should in itself guide the creation of your presentation.

And, as stated earlier, don't focus on just one measurable artifact, like DMAX - which is a bad choice, anyways, if your goal is to espouse the unique qualities to a B&W print.

Perhaps you should focus your talk on the basic techniques that photographers use to capture such a rich tonal scale; a dip of the toe into the ocean of information about the various 'Systems' used to capture a given scene's bightness range onto the film and paper media (Zone & BTZS are two such popular systems). Again, you don't need to 'baffle them with BS' and spout forth sensitometry data; just impress them with the wide range of tools and techniques available to photographers working in silver gelatin media.

You could also talk to them in general terms about various film development techniques and their effects on the distribution of silver grains in the resulting fixed emulsion, and how this can affect subtle but important changes to the resulting prints. Impress upon them that our 'optical image sensor' is 3-dimensional, and can be altered at the microscopic level by customizing our 'work-flow' (i.e. the film development process).

You could also talk a bit about the merits of contact printing verses projection enlargement, and the differences between cold light and condensor optics on the resulting print, and the necessary changes required in one's film developing regimen if one were to change enlarger optics or printing technique.

And then there are the wonderful effects of staining developers on controlling print contrast.

Etc, etc, etc.

In summary, impress upon them that the fundamental differences between the worlds of electronic and chemical photography is that electronic photography works much more in the area of the manipulation of abstract 'data', whereas chemical photography operates entirely in the realm of physical processes; as such, B&W silver gelatin photographers are by nature much more attuned to the subtle physical nuances of the materials at their disposal, and therefore more sensitive to disruptions and unexpected changes in their quality, working methods and availability.
 
Last edited by a moderator:

gainer

Subscriber
Joined
Sep 20, 2002
Messages
3,699
Another point. The frequency content of a work of art by a painter, say, is set by the brush strokes of the painter which were governed by the painter's eyes. When one of us views directly one of those works of art, we do not see artifacts like fold back. We see what the painter saw. It is comparable to hearing a symphony orchestra in concert, which certainly produces frequencies beyond human hearing, but which do not cause distortions, even when one is in the middle of the orchestra. But when the orchestra is recorded, the playback can certainly be distorted by the MTF of the recording and playback mechanisms.
 
Joined
Nov 22, 2004
Messages
226
Location
Bilthoven, T
Format
4x5 Format
Jed;

I am aware of the studies and the actual people involved in conducting the studies. The key word though was that these studies were conducted by varying the MTF but keeping everything else constant. If that was so, then the MTF was of primary importance. However, if the image characteristics changed such as contrast or tone, often the person would pick another print.

In some cases, the subjects were told to 'select the sharpest print' not 'select the best print' nor were they given, in all cases, the optimum print or sometimes even a beautiful print.

In the German and Swedish studies, the objective was to find if there is a relation between MTF and subjective image quality. And there is, with a high correlation coefficient. One should, of course, not change the tone and contrast. They did that in the German investigation, but that was part of another study ( one should never change more than one parameter). The German study consists of several parts: selection on sharpest prints and selection on image quality,and further: influence of contrast and tone, preference for brillance and rejection of adjacency effects.
The Germans found a strong correlation between MTF and subjective image quality, using only one spatial frequency; the Kodak group is using a range of spatial frequencies. But, the results of both groups are fully in agreement.
I tested the results of those research groups myself, using a photoclub as panel. I found that one can predict what a panel will choose as the print with the best subjective image quality. But, the most important point is that photographers can select their lenses, negative size, film etc in a 'non-random' way. That is why that kind of information is given by the manufacturers. And the manufacturers can ( and will) use the information in the design of their lenses etc. And I have experienced that one can use the information in a practical way.
The main problem with MTF I notice with photographers, is a 'fear for graphs'. But, how to explain the general acceptance in other areas than photography?
An US firm has recently (last year) developed software to control the MTF in digital printing. The software, based on the Kodak research results, came recently out. May be you didn't see the software before. The reason, they made the software is that there was an overkill ( too high MTF) in previous software ( imitating adjacency effects to increase 'acutance').

Jed
 

Photo Engineer

Subscriber
Joined
Apr 19, 2005
Messages
29,018
Location
Rochester, NY
Format
Multi Format
Jed;

In one study, prints of a given MTF at a moderate contrast were given to subjects to compare with prints of lower MTF but higher contrast. The results were mixed, but seemed to prefer lower MTF and higher contrast.

This led to the inescapable conclusion that you could not pin an optimum print to MTF alone. And, btw, the high MTF print at high contrast was rejected.

PE
 
Joined
Nov 22, 2004
Messages
226
Location
Bilthoven, T
Format
4x5 Format
Maybe I should go back and read the beginning of this thread. I am a bit puzzled about the term MTF.

The MTF is very simple the way contrast is transferred. Suppose, you photograph a sinusoidal pattern (modulation) with a varying period. Then, the modulation (sinus) will be compressed , and more the higher the frequency ( shorter period). If the compression of the pattern is 50%, the MTF=50%. Therefore, the MTF describes the quality of image transfer, and therefore the quality of lens and film.

The MTF can be measured in a different ways. This involves mathematics of all kind, but is irrelevant for ther photographer. Some people are talking on Fourrier transforms. That is just confusing. I explained the MTF to photographers, and it took them only a few minutes to undertsand what it is all about. It is 'didactics'.

Jed
 

David A. Goldfarb

Moderator
Moderator
Joined
Sep 7, 2002
Messages
19,974
Location
Honolulu, HI
Format
Large Format
Though such studies are interesting from the perspective of understanding cognition, I don't find them terribly motivating as a guide for making art. Perhaps commercial photographers should worry about what kinds of technical qualities of images appeal to the broadest audience, but when artists start worrying about such things, the results are comic.

Here are my favorite satirical paintings based on statistical surveys of popular taste--

http://www.diacenter.org/km/

I also highly recommend the CD.
 

bjorke

Member
Joined
Aug 17, 2003
Messages
2,263
Location
SF sometimes
Format
Multi Format
I would be VERRRRRY wary of using Math attempting to explain anything to anyone unless you genuinely understand it yourself! A two-minute overview is unlikely to do it justice, and has hazards:
  • Someone in the audience just might understand it as well or better than you, and may call you on the least mistake, or
  • you will hopelessly torpedo your own argument if you try to make assertions based on information that your audience doesn't believe you really understand (whether they themselves do or not)
  • people who don't grok math will just get bored
  • people may think you're deliberately talking over their heads, which isn't just boring (see above) but comes off as hand-waving. To convince people of anything, you have to make sure they believe there's something in it for them!
All that being said, the primary errors in digital imaging, compared to film, have to do with quantization -- that is, the conversion of a varying analog signal (brightness) into discrete digital values. These errors can occur at several points in digital or hybrid image chains -- at the digital camera, in the film scanner, in the input stage (i.e., from camera or scanner to your PC), and in manipulations such as the Photoshop curves layer.

The number of different gray values available on a film stock (except in extreme cases like, say, 100x enlargement or other deliberately high-grain circumstances) is very large. Effectively, compared to most computer images, it's infinite.

Computer images have a fixed number of gray values, based on the bit depth. 256 values for 8-bit images, a lot more for 16-bit images (though for most 16-bit images, the input is usually LESS than 16 bits - most good digicams and scanners have less precision than that, they just pad-out their values with zeroes because 16 bits is a standard computer format).

The quantization criteria -- that is, the choices of which continuous value is going to get mapped to which fixed number -- are made early in all digital processes (at the begining -- before that, the image isn't digital!).

What often happens in shadows for digital images is that the dark values get the smallest number of choices. Between "8" and "16" is a one-stop range, but only gets eight distinct values (while the one-stop range between "128" and "256" gets 128 values -- hence the well-known digital dictum "expose to the right")

Sure, you can use a Curves or Levels layer to boost those dark grays -- but you can't undo the quantization. They will just look chunky and crappy, because even if you expand the original 8-to-16 range to 1-to-258, there are still only eight unique values, and you'll get a posterized result. The effect is a kind of aliasing. Most gamers know aliasing as "jaggies" along the edges of 3D computer objects. That's spatial aliasing, because the true 3D shape is being broken-down into discreet pixels. The same thing happens to the extra "dimension" of color for each pixel.

Enlarging directly from the negative to photographic paper doesn't have these same problems*. An example, I believe, are Ralph Gibson's famously "over-developed" negatives, which he prints in a way that doesn't look over-developed, looks high-contrast, but doesn't show obvious banding. It's my belief that he's managed to become skillful enough that he can extract the VERY narrow range of his over-exposed negs into a full-range print. He has gone on record as saying he can't do it with digital, and I have tried -- it's really hard for a scanner to get at that compressed range of densities (scanners and digicams are usually tuned to "standard" ranges), but photographic paper can. (Lest people accuse Gibson of Luddism, he is a big advocate of digital -- he speaks excitedly of wanting to scan his prints and sometimes is even eagerly scanning them wet before they've dried).

KB
Imaging Editor
GPU Gems

* yes, it has a different set of problems, like repeatability and stinky fingers. What do you want? Perfect AND easy?
 
Joined
Nov 22, 2004
Messages
226
Location
Bilthoven, T
Format
4x5 Format
Jed;

In one study, prints of a given MTF at a moderate contrast were given to subjects to compare with prints of lower MTF but higher contrast. The results were mixed, but seemed to prefer lower MTF and higher contrast.

This led to the inescapable conclusion that you could not pin an optimum print to MTF alone. And, btw, the high MTF print at high contrast was rejected.

PE
It is strange that they made such a study at Kodak. I cannot believe they changed more than one parameter. In particular, because they followed the example set by the Germans. Anyway the results of such a study has never been published, and has no influence on the outcome of the published results.

Jed
 

Roger Hicks

Member
Joined
May 17, 2006
Messages
4,895
Location
Northern Aqu
Format
35mm RF
Dear Jed,

I am prepared to accept that PE knows a lot more about this than I do, not least because it squares with everything else I have ever read or heard or observed. His reputation and background helps.

I am not prepared to give a great deal of credence to your views, because they are one-dimensional and not outstandingly well informed. Also -- forgive me -- I am not familiar with your reputation as an authority.

Why don't you, just for once, accept that there are other people who know a LOT more about this than you do? I do not make any great claims to greater knowledge than you. Indeed, I gave away my last book on Fourrier transformations to someone who would understand it better than I. But equally, I have yet to see a great deal of evidence that you do, in fact, know or understand more than I. You may, or may not; quite honestly, it doesn't matter a whole hell of a lot, because even if you do, you have yet to demonstrate it convincingly.

This is, in Michel's memorable analysis, a dick-measuring contest. My own dick is big enough for my purposes, even if it isn't as big as yours. Now is the time to stop comparisons.

In Britain, there is Healey's Law of Holes. It is this: when you are in a hole, stop digging.

Stop digging.

Cheers,

R.
 
Joined
Nov 22, 2004
Messages
226
Location
Bilthoven, T
Format
4x5 Format
Though such studies are interesting from the perspective of understanding cognition, I don't find them terribly motivating as a guide for making art. Perhaps commercial photographers should worry about what kinds of technical qualities of images appeal to the broadest audience, but when artists start worrying about such things, the results are comic.

Here are my favorite satirical paintings based on statistical surveys of popular taste--

http://www.diacenter.org/km/

I also highly recommend the CD.

You may like it or not, but the two dimensional arts were always being considered to be a science. I quote the english painter Constable from one of his Hampstead lectures for the academy in 1816: 'Painting is a science and should be pursued as an inquiry into the laws of nature. Why then, may not landscape painting be considered as a branch of natural philosophy, of which pictures are just the experiment'. I tell you: Constable was a painter and scientist as well. And many 2 dimensional artists feel that way.
And this is not just a phrase but is mentioned by E.H. Gombrich on the first page of the first chapter in his famous book ' Art and Illusion , a study in the psychology of pictorial representation'. And Gombrich adds: we call natural philosophy nowadays physics.
And, my personal statement is: when painting is a science, photography is a science for sure. There is so much physics and chemistry involved in the technique. But that is not all; like in painting, much thinking is involved how to translate light in feelings. Feelings that should radiate from your photographic print. And to understand that I recommend reading the first chapter in the book of Gombrich, I mentioned above.

Jed
 

Michel Hardy-Vallée

Membership Council
Subscriber
Joined
Apr 2, 2005
Messages
4,794
Location
Montréal, QC
Format
Multi Format
You may like it or not, but the two dimensional arts were always being considered to be a science. I quote the english painter Constable from one of his Hampstead lectures for the academy in 1816: 'Painting is a science and should be pursued as an inquiry into the laws of nature. Why then, may not landscape painting be considered as a branch of natural philosophy, of which pictures are just the experiment'. I tell you: Constable was a painter and scientist as well.

That is a knock-out argument.

You may like it or not, but plumbing was always considered to be a branch of accounting. I quote my cousin who is an accountant: "plumbing is in fact a branch of accounting because you cannot run a successful plumbing business without proper accounting." I tell you: my cousin is a plumber AND an accountant as well.
 

Photo Engineer

Subscriber
Joined
Apr 19, 2005
Messages
29,018
Location
Rochester, NY
Format
Multi Format
Jed;

I am trying to convince you that it is impossible to 'quantize' the Mona Lisa. Art defies description.

My comments come from people at Kodak such as Zwick, Kriss, DeMarsh, Bartleson and Breneman, among others. I worked closely with all of them and worked for Zwick on many occasions, sometimes for several years. Jim Bartleson was a close friend until his untimely death.

I am not in their league by any means, but they knew what they were talking about. It was Daan Zwick who said to me "Ron, we sell pictures, not curves". He and I put many pictures up for measurement in a MacBeth cabinet and used a spot meter to measure curves while the actual MTF was measured by others. I participated in many of his blind studies as a judge and tallying results in other tests.

So, go ahead with your beliefs, but human opinion may belie them, sorry to say. Roger was right before. He remains correct.

I'm sorry, but not everything done at Kodak was published. Over 15 years of my work remains unpublished to this day. This is in spite of the fact that I worked with every emulsion maker at Kodak, world wide. Live with the fact that things exist that you know nothing about, I live with the fact that my work is largely unpublished.

It does not matter. What you say is not wrong. It is just not the entire picture.

PE
 

Photo Engineer

Subscriber
Joined
Apr 19, 2005
Messages
29,018
Location
Rochester, NY
Format
Multi Format
Jed;

Your argument about Constable is compelling until one asks oneself, "If he understood painting so well as to reduce it to a science, how come he is not as well known as some of the great painters".

This too is telling.

In spite of his apparent erudition about art vis-a-vis science, he was not one of the 'greats'.

PE
 

David A. Goldfarb

Moderator
Moderator
Joined
Sep 7, 2002
Messages
19,974
Location
Honolulu, HI
Format
Large Format
Constable was a product of the age of Enlightenment and early Romanticism, so he placed exactly this sort of value on science as described in the quoted passage. Science included ideas like mesmerism and the belief that the sensation of beauty resulted from corpuscles in objects that exerted pressures on the brain to produce a state of relaxation of the solids (Edmund Burke believed that one).

Of course science is relevant to painting and to photography, but Constable's assertion is more of a historical and cultural artifact than evidence for any claim about the relation of painting to science. (I would consider Constable a great painter, by the way).

Similarly, Gombrich was writing in an age of psychologism. Contemporary academic psychologists wouldn't take him too seriously (they don't even teach Freud in modern psych departments typically), though Gombrich is certainly important to understanding intellectual history.
 
Last edited by a moderator:

Roger Hicks

Member
Joined
May 17, 2006
Messages
4,895
Location
Northern Aqu
Format
35mm RF
Jed;

Your argument about Constable is compelling until one asks oneself, "If he understood painting so well as to reduce it to a science, how come he is not as well known as some of the great painters".

This too is telling.

In spite of his apparent erudition about art vis-a-vis science, he was not one of the 'greats'.

PE
Dear PE,

Not only that: why is he commonly cited as a great painter, and not as a scientist in the ranks of his (approximate) contemporaries such as Lavoisier, Herschel and the like?

Cheers,

R.
 

Ed Sukach

Member
Joined
Nov 27, 2002
Messages
4,517
Location
Ipswich, Mas
Format
Medium Format
Ah!! An expert on "Modulation Transfer Function"!!!

I had NO idea that the MTF was so closely and indelibly related to the image, and that it would be so useful in the actual execution of the photograph.

Here are three .jpgs of MTFs, and an image taken with a lens represented by one of them. Can you tell me which MTF applies to the image, and which lens I SHOULD have used?

Oh, by the way.... Dali - Salvador Dali used a microscope" ... for ...?

His "Dream Evoked By the Flight of a Bee Around a Pomegranite" was an interpretation of one of his dreams. Was he using a microscope in his dream, or when he painted the canvas?
 

Attachments

  • EC#55.jpg
    EC#55.jpg
    96.2 KB · Views: 126
Last edited by a moderator:
Joined
Nov 22, 2004
Messages
226
Location
Bilthoven, T
Format
4x5 Format
Similarly, Gombrich was writing in an age of psychologism. Contemporary academic psychologists wouldn't take him too seriously (they don't even teach Freud in modern psych departments typically), though Gombrich is certainly important to understanding intellectual history.

The science is now within the branch of psychophysics, a seperate part of physics . Within universities civil research is done and military research in specialized institutes. That is where the research on ( the interpretation of) images is done.[ In the last few years, fysiology has also an important position in this area]. There are periodicals specialized on psychophysics etc.

Jed
 

Tom Stanworth

Member
Joined
Sep 4, 2003
Messages
2,021
Format
Multi Format
I agree, the wrong tree is being barked up.

Analogue and D$#@al are different beasts which can arriive at a similar end point. Just like sharks are fish and dolphins mammals. Both do the same thing but are very, very different.

For me, as I go along, the more the 'analogue approach' diverges from that of digital. Its about the way you approach image making knowing you have to get it fairly right in the camera and how this affects the relationship between the photographer and subject and the photographer and his/her equipment. Its about a print which has been thru your hands, from your neg, which you developed.

If you see a difference in the end state either way, that is fine to point out using examples. The images will either speak for themselves or they wont. But for me the process, which leads up to and directly affects the image, is of as much importance, just as the ownership aspect does after it is on the wall, sold or bought. We all know what a traditional black and white reportage/street shot represents and this is reinforced by the knowledge of what it is. For me it is a slice of something directly or loosely tethered in reality. with di#$#al I am left with an uneasy feeling, not knowing for sure what I am really looking at........(with so many di$#@% photographers being accustomed to telling such outrageous, hideous lies about their images (and lying to themselves), this matters to me, a LOT).
 
Photrio.com contains affiliate links to products. We may receive a commission for purchases made through these links.
To read our full affiliate disclosure statement please click Here.

PHOTRIO PARTNERS EQUALLY FUNDING OUR COMMUNITY:



Ilford ADOX Freestyle Photographic Stearman Press Weldon Color Lab Blue Moon Camera & Machine
Top Bottom