The Maths of a Fine B&W Print

[paulcop]

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

 

[reub2000]

dpreview has the dynamic range curves for various cameras.

But you know what would be really fun? Comparing some velvia to a LCD projector.

 

[Ray Heath]

an interesting concept Paul, but I wouldn't put it in mathematical terms, i would guess that just like most analogue photographers most of them would not really understand the maths, but they should all appreciate a visual comparison

we analogue photographers constantly say our method is better, time to put up, or keep quiet

 

[boyooso]

I would say there is an 'intagible' nature which is lost which numbers can't explain. It is not necessarily the Dmin & Dmax is lost, it is also most inbetween.

Good Luck and sorry,

Corey

 

[Jed Freudenthal]

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

The transfer of any image is described with the Modulation Transfer Function (MTF) . This applies to analogue and digital photography, but also to many other optical systems. The MTF describes the quality of the transferred image. It is done with a mathematical function, but is part of physics. When you want to know the difference in the quality between specified analogue and digital photographic processes, all you have to do is to compare the MTF's of both systems. The result is often that the details ( high spatial frequencies) are lost in digital photography.
The use of MTF is common in photography, but photographers don't use it very much. This is strange because by using MTF, the process is better under control. e.g. the selection of lens and film is optimized.

Jed

 

[Roger Hicks]

Dear Paul,

Like others, I think you're on the wrong track. There's an old saying in writing: 'Show, don't tell'.

With the right inks you can get at least as good a Dmax from ink-jet as you can from halide, so that's not what it's about. For that matter the Dmax of a platinum print is comminly bloody awful. So?

MTF is a red herring. 'Sparkle' is a very high MTF at comparatively low frequencies (Zeiss and Ilford have researched this independently) and while MTF at high frequencies is important in some kinds of picture, it's totally irrelevant in others. Ignore MTF.

What does this leave? Nothing that can be demonstrated mathematically. So show them pictures and (above all) sell them YOUR enthusiasm. By all means bring in objective criteria such as quality/cost, durability, low-pass filters, etc. but look at it this way: how would you explain, mathematically, that alkyds are better than oils, or guitars better than violins? (Or, of course, vice versa in either case). You can't.

Cheers,

R. (www.rogerandfrances.com -- where there are both digital and silver halide galleries, but mostly halide...)

 

[jim appleyard]

Most of all, tell them that traditional darkroom prints are a whole lot more FUN to make!

 

[Donald Miller]

Actually I wouldn't mention the dmax difference at all, if you want to impress the studets with traditional silver prints. The newer digital printers are capable of dmax as high as 2.5...something totally unheard of in traditional silver prints.

The highest dmax that I have personally measured with a silver print is 2.30...that was with a print developed in amidol and toned in selenium.

 

[Willie Jan]

Actually I wouldn't mention the dmax difference at all, if you want to impress the studets with traditional silver prints. The newer digital printers are capable of dmax as high as 2.5...something totally unheard of in traditional silver prints.

The highest dmax that I have personally measured with a silver print is 2.30...that was with a print developed in amidol and toned in selenium.

Can you see the difference in a 2.2 and a 2.5 dmax print?
The are advertising but do not now what they talk about.

this printer prints 1200 dots an inch! Applause because we humans are not able to see the difference between a 300 and 600 dots an inch difference...

Until now at exhibitions of foto communities i only see crap. Blue tinted black&white photos, printed at 50 dpi so that we can save more images onto the card.... My pleasure in going to these kinds of exhibitions is over. Only museums are the place where they try to show quality these days. Besides that there will be a couple of photographers who can make nice prints, but i haven't seen them around until now....

 

[Donald Miller]

Can you see the difference in a 2.2 and a 2.5 dmax print?
The are advertising but do not now what they talk about.

this printer prints 1200 dots an inch! Applause because we humans are not able to see the difference between a 300 and 600 dots an inch difference...

Until now at exhibitions of foto communities i only see crap. Blue tinted black&white photos, printed at 50 dpi so that we can save more images onto the card.... My pleasure in going to these kinds of exhibitions is over. Only museums are the place where they try to show quality these days. Besides that there will be a couple of photographers who can make nice prints, but i haven't seen them around until now....

Yes, I can see the difference, if the prints are side by side. Taking each print by itself makes it more difficult. A 2.2 dmax is a convincing black. The 2.5 is more so. But dmax is only one part of what makes a fine print, as you well know. The tonal distribution and resolution are equally as important, to my way of thinking.

I have seen some very fine non traditional prints. The multiple blacks have totally changed the playing field.

 

[David A. Goldfarb]

I don't know that there is a useful mathematical way to describe the visual difference between an image made with an emulsion floated on a surface and ink sprayed on a surface. They just look different.

Make some comparison prints (or collaborate with someone who can make high quality digital prints to make comparison prints). A good silver print will usually show smoother gradation in the highlights than a good inkjet print. This isn't a Dmax issue.

 

[markbb]

To add further confusion, one can get digital images onto traditional halide paper using 'printers' like the Durst Lambda. I'm with Roger on this one - no one method produced better images, that's down to the artist themselves.

 

[Alex Bishop-Thorpe]

I've heard many discussion of Dmax and Dmin on these forums, but so far I haven't come across it in how I work. From a technical standpoint the numbers matter, but most of the art of photography is far from technical.
I agree, show some prints and say what's possible and unique with analogue.

 

[Jed Freudenthal]

Dear Paul,


MTF is a red herring. 'Sparkle' is a very high MTF at comparatively low frequencies (Zeiss and Ilford have researched this independently) and while MTF at high an frequencies is important in some kinds of picture, it's totally irrelevant in others. Ignore MTF.

The studies on photographic image quality were done by a German, a Swedish and a US group(Kodak research labs). The German panel included even 17.000 persons. All studies were in agreement, and were based on MTF. I am not aware of a Zeiss and an Ilford study. Although Zeiss, like the other industries use the MTF extensively (All lenses are sold with their MTF values). And Zeiss is producing instruments to measure MTF's. Recently I discussed this topic with Ilford people, and they are not aware of any study on image quality of their own. (They told me, they use an MTF related parameter)
I myself use the MTF approach to 'guide' the quality of the photographic print, an advice I got from Kodak (US) research lab people, and I use it with much succes.
But it is up to everyone what to use. The question here was to compare different image qualities. And then the MTF is the road to go without any question. The MTF is made to answer exactly this question. It works and it works in many image transfer areas like printing etc( Google MTF and you will see)

Jed

 

[Roger Hicks]

And then the MTF is the road to go without any question. The MTF is made to answer exactly this question.

Sorry, Jed, no. I know you are a great fan of MTF but is really is a complete red herring in this case, viz., trying to explain a 'look' in ways that his audience is going to understand. I do not deny its usefulness -- but it's useless here.

As a matter of interest, who did you talk to at Ilford?

Cheers,

R.

 

[Alan Johnson]

Some data has been published in Amateur Photographer magazine, but not on the internet.For example,G.W.Crawley Feb 10 07 reported MTF curves for the Canon 5D and D60. An MTF curve for 100 T-max is on Kodak's website.It seems reasonable to compare these if it is wished to compare resolution and the comparison is favorable to the film.
The dynamic range of DSLR's is usually about 7 stops and again I believe the comparison is favorable to film .
IMO apart from Dmax the data is generally favorable to the film/silver gelatin route but collecting the data is a bit like hard work at present.

 

[Photo Engineer]

MTF is, as Roger said, a red herring.

A print or a negative is the sum total of MTF, RMSG and the H&D curve of each material involved among just a few of the properties I could mention, and that is difficult to quantify. There are integrals that purport to put this together, but a picture is worth a thousand words.

To include some others there are the tone of the image, spectral sensitivity of the film and paper, light source for exposure and printing, goniophotometric response of the paper and surface (this includes to an extent goniophotometric response), and so many others that it is useless to try and state them all let alone study and understand them.

We had a team of people at EK studying each of the subjects named above and more, and then who would meet and review their results. The sum choice of the best and worst would then be shown to hundreds of people in blind tests to determine which was which. The results would be used to refine the math tests and come up with a better model. But, often the panels would confound the 'experts' by chosing something that was expected to be worse.

Truth and beauty, as I have often said, are in the eyes of the beholder.

PE

 

[Michel Hardy-Vallée]

If there was one thing I would use to pit film/digi side by side it's not the highest resolution, or the biggest dick.

Rather, it would be the faults. How does digital degrade? How does film degrade? What I like about film is that even in the situations where it gives "inferior" results (less sharpness, less detail, smaller dick), I like it because it degrades gracefully. Compare silver grain with CCD noise: which "defect" do you like the most? Is there anyone on this planet who likes JPEG artefacts?

On the other hand, you can sell a print made with 400 ISO Minox film developped in Rodinal 1+25. And the golfball grains are the equivalent of lumps in mashed potatoes: an intimation to reality, to the texture of things. Are the people doing Calotypes doing it because they can extract a bigger MTF out of it?

Let's stop measuring our dicks, please!

 

[Roger Hicks]

Cher Michel,

A magnificent analogy!

Amitiés,

Biggus Dickus atque Monty Pythonius.

 

[Jed Freudenthal]

Sorry, Jed, no. I know you are a great fan of MTF but is really is a complete red herring in this case, viz., trying to explain a 'look' in ways that his audience is going to understand. I do not deny its usefulness -- but it's useless here.

As a matter of interest, who did you talk to at Ilford?

Cheers,

R.

I am not explaining a look. But the question is to explain a difference in look.The difference between analogue and digital. And, as you admit, the road is via MTF. Another question is whether the audience will understand this.But, because a mathemathical procedure was asked for, I give the proper answer. I really know that many people do not understand MTF.
However, in Europe (The Netherlands and Belgium) Apuggers are eager to learn about these things. And I am being asked to give a workshop on the subject in Belgium ( and probably in France) this year. And I will illustrate the workshop on MTF with photographs, to link a more abstract entity with the reality of the photographic image.
In that way I hope to bring the MTF description to life within the APUG. And we all know from photographic industry; knowledge on this subject is extremely valuable for all photographers (just to know how to buy the right lens and camera etc.). And in that way, I address analogue as well as digital photographers. The fact that most photographers do not know anything about MTF at the moment, is no excuse. The APUG is a good site for learning. And to cite a Kodak (ex)employee ( I think it is Dick Dickerson, but it is long ago and I am not sure): to control image quality is to control MTF. And that is what it is. Of course, it is the control of image quality only. Every photographer has to make its own composition, light distribution etc. But, when we are talking on the technical quality of image quality, it can be controlled through the use of MTF. And that is why I would like to introduce other photographers via a workshop, using actual photographs into this way of thinking.
You learned in grade school to work with numbers. At a certain moment you may have learned to work with negative numbers. At another moment, you may have learned to use complex numbers. Learning is moving into another space of thinking. Photographers should just learn to think in a different 'space'. The space of MTF. After a while, you will notice, there is nothing special about it. MTFsounds difficult, but it isn't. Just pass a threshold.

Jed

 

[Photo Engineer]

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

 

[Alan Johnson]

Of course there is no "theory of everything" for silver gelatin.
But it is interesting to hear in what areas there is some theory relevant to electronic imaging ,acknowledging the limitations of that theory.

 

[David A. Goldfarb]

A Steichen or Alvin Langdon Coburn or Irving Penn print could and very likely would have the same curve as what most people would consider a bad print. Many are low resolution, low contrast, low Dmax, high Dmin, and yet they made prints with all these "negative" qualities that are beautiful. Where do artistic intention, history and context fit into the calculation?

 

[Michel Hardy-Vallée]

Cher Roger,

Coming from a gifted writer like you, the compliment is most heartwarming!

 

[Michel Hardy-Vallée]

On a more serious note, I think we should perhaps make a difference between the theory of reproduction and artistic photography.

Artistic photography depends on principles of tone reproduction, image-formation in optical systems, resolution of details, but only in the same way as sculpture relies on gravity, fresco painting on chemistery, and dance on kinetic physics.

In other words, you cannot (and should not!) ignore the realities of gravity in sculpture, you cannot avoid the principle of tone reproduction in photography, but you do not optimize their parameters in the same way that someone copying continuous-tone documents does. The technical aspect of photography is a field of parameters, of which many local optimizations are possible according to the requirements of the problem to solve.

I understand the need to anchor a qualitative impression in a quantitative factor, and that's why concepts like Dmin/max, MTF curves and so on are useful to answer specific questions, but it is an error to consider that they are either necessary or sufficient to explain why one would prefer a silver gelatin print to an inkjet one. Gravity does not explain why people fall in love, as Mr Einstein said.

You can explain certain differences with quantitative factors, but you have to limit reasonably the scope of their application. As others have said, a higher Dmax does not necessarily contribute to a better image quality. But you can describe how the tone of a silver image is related to the way metallic silver is formed, and how it reflects certain lightwaves more than others, thereby creating a reddish or blueish or greenish tone. You can describe how CCD captors create noise, which is non-existent in analogue photo. You can describe how the demosaicing+interpolating of pixels have an impact on subjective color perception, fringing effects, and so on. In other words, you need very specific questions if you want to answer them with the specific tools of science.