Expose for the shadows, develop for the ... scanner?

[Acticus]

Fair enough. It should be noted, that this can be done (in my circumstances) only with 35mm B/W negatives. I get 100% of information out of them in terms of resolution and dynamic range, no need for a lecture here (trying to be preemptive), my B&W 35mm scanning is superb. Plus, I have the comfort of doing data management professionally (RAID, checksums, four backups in four separate locations).

I am nowhere close to achieving this with medium format or color, so perhaps I should have added some disclaimers.

Are your backups EMP proof?

 

[Bormental]

Yes! I said four separate locations. So anything short of a nuclear war I'm fairly certain my digitized negatives will be the "last man standing".

 

[Acticus]

Yes! I said four separate locations. So anything short of a nuclear war I'm fairly certain my digitized negatives will be the "last man standing".

You don't need a nuclear was, just one EMP device detonated 300 miles over Kansas. That would fry all unprotected electronics in the entire continental United States. Your negatives would survive though.

 

[MattKing]

You have to be a gluten for punishment.

You can now get punishment gluten free!
It is called scanning!
(Ain't auto-correct grand?)

 

[MattKing]

How does micro-contrast work? This is probably unknown to me. I mean, isn't film "linear" in sense that the tones "map" no matter what kind of macro contrast you have?

Micro-contrast is the secret ingredient in photographic joy!
It is the variable that you control that most affects how the important mid-tone and lower highlight details appear.
You adjust contrast to change mushy, blah images into interesting and crisp images.
The tone of the shadows - of some importance.
The brightness of the highlights - of more importance.
The detail, tones and rendition of the mid-tones - of utmost importance.
In the following image, controlling the mid-tone micro-contrast makes the birch bark what it is - the main strength of the image:

 

[Bormental]

Yes, I meant macro-contrast. How does micro-contrast work? This is probably unknown to me. I mean, isn't film "linear" in sense that the tones "map" no matter what kind of macro contrast you have?

Micro-contrast was popularized by mass adoption of digital cameras around 2010. It never existed in engineering literature where these matters have always been expressed via MTF

It's a property of any image-capturing device, i.e. how well it records a signal change at a given spatial frequency. Perceptually MTF can be interpreted as how "sharp" an image is from a certain distance (or zoom level). When viewed from afar, that's simply "contrast" and when you get closer (nobody defined how close) that's "micro-contrast", and when you get very, very close - it suddenly becomes "sharpness" in the pixel-peeping circles.

At Matt points out, people often bring it up describing textured surfaces in the midtones of an image.

 

[hsandler]

I don’t see any discussion of grain here. In my limited experience in which black and white negative film was overdeveloped to a high contrast and density, the scan, which already overemphasizes apparent grain relative to optical printing, looked terribly grainy. On the other hand thin negatives sometimes scan surprisingly well. So although I tend to shoot and develop for box speed and normal recommended density, pulling a stop to reduce grain seems to work well for scanning. Even though it uses less than all the available scanner bits, I have never had any banding effects saving Epson flatbed scans to 16 bit files with subsequent moving tones around with curves.

 

[Alan Edward Klein]

You can now get punishment gluten free!
It is called scanning!
(Ain't auto-correct grand?)

It wasn't auto-corrected. I think I spelled it that way.

 

[grat]

Plus, I have the comfort of doing data management professionally (RAID, checksums, four append-only backups in four separate locations).

I would never throw away negatives, but I respect your data storage. How often do you verify the checksums? Most people assume RAID is good enough, not realizing that RAID by itself has no bitrot detection mechanisms. I use a snapshot based RAID called 'snapraid' which includes a parity "scrub" (validates parity and data consistency), and that process makes it through my entire array about once a month.

 

[MattKing]

Micro-contrast was popularized by mass adoption of digital cameras around 2010. It never existed in engineering literature where these matters have always been expressed via MTF

And more generally, expressed as "great midtones and highlights!".
I'm a late adopter of the term. I've been talking about mid-tone rendition in darkroom prints and slides since long before 2010.
When I adjust contrast settings in the darkroom or while post processing, that is what I am focusing on.

 

[Bormental]

I would never throw away negatives, but I respect your data storage. How often do you verify the checksums? Most people assume RAID is good enough, not realizing that RAID by itself has no bitrot detection mechanisms. I use a snapshot based RAID called 'snapraid' which includes a parity "scrub" (validates parity and data consistency), and that process makes it through my entire array about once a month.

Bi-weekly. I have two sets of alternating nightly jobs: one is uploading backups to S3 and Glacier and another verifies integrity of two local copies. The most common reason for corruption is not hardware failures, but software glitches and user errors. I only use RAID for convenience, rebuilding the array is just easier than restoring from a backup. It's all based on Synology, which I highly recommend. Snapraid looks cool, I will check it out, thank you.

 

[MattKing]

I'm old school. From 40 - 45 years ago, arrayed for artistic effect:

There are a lot more that are newer and in PrintFile pages.
A relatively small percentage have been digitized - mostly the colour ones that I'd like to have inexpensive prints made from, or which I decide to share here.
I don't have much desire to increase that percentage.

 

[radiant]

Micro-contrast is the secret ingredient in photographic joy!

I'm glad there is still some joy I haven't really looked into

The detail, tones and rendition of the mid-tones - of utmost importance.
In the following image, controlling the mid-tone micro-contrast makes the birch bark what it is - the main strength of the image:

Could micro-contrast be also described by how "sharply" tones separate? Then blurred image is opposite to good micro-contrast?

But how one controls the micro-contrast in film & darkroom? On computer photography it is probably done with wallet?

 

[Koen Van Crombrugghe]

I don’t see any discussion of grain here. In my limited experience in which black and white negative film was overdeveloped to a high contrast and density, the scan, which already overemphasizes apparent grain relative to optical printing, looked terribly grainy. On the other hand thin negatives sometimes scan surprisingly well. So although I tend to shoot and develop for box speed and normal recommended density, pulling a stop to reduce grain seems to work well for scanning. Even though it uses less than all the available scanner bits, I have never had any banding effects saving Epson flatbed scans to 16 bit files with subsequent moving tones around with curves.

That's a very interesing point which deserves the necessary attention.
Even though grain is often a matter of taste I fully agree that it would not make a lot of sense to try to get the maximum amount of levels only to have massive grain.

Would this be an important factor in a fine grain film such as T-Max?
I suppose this calls for some more experiments

 

[Adrian Bacon]

Is my reasoning correct that this means that an image with a density range of 1.8 only has about 6 bits of tonality range for mapping these 10 stops of exposure?

roughly, yes, however, it's the top 6 bits of the ADC where most of the discrete tone values are. Increasing your developing time will give you more discrete tone values, but it very quickly is diminishing returns, at the expense of more grain.

For a 14 bit ADC, the top 6 bits make up 16128 discrete tone values of the the 16384 total available for 14 bits. Increasing your development time so that your density range is 2.1 will only add 128 discrete tone values at the very top of your highlights. Adding another 0.3 of density will only add 64 discrete tone values, etc, very quickly turning into chunky looking highlights because you don't have enough discrete tone values per stop of exposure on the film to avoid posterization artifacts and other bad things.

 

[Koen Van Crombrugghe]

roughly, yes, however, it's the top 6 bits of the ADC where most of the discrete tone values are. Increasing your developing time will give you more discrete tone values, but it very quickly is diminishing returns, at the expense of more grain.

For a 14 bit ADC, the top 6 bits make up 16128 discrete tone values of the the 16384 total available for 14 bits. Increasing your development time so that your density range is 2.1 will only add 128 discrete tone values at the very top of your highlights. Adding another 0.3 of density will only add 64 discrete tone values, etc, very quickly turning into chunky looking highlights because you don't have enough discrete tone values per stop of exposure on the film to avoid posterization artifacts and other bad things.

Thanks, this makes a lot of sense!
I think this is the ultimate response to my question.

 

[MattKing]

But how one controls the micro-contrast in film & darkroom?

For film - the lighting conditions, film, lens contrast, developer and how long you develop the film.
For prints, the light source, the enlarger, the paper, the paper grade, the developer and the developing time.
In other words, you use the same controls for contrast as you have probably used before. But you evaluate the results based on how those mid-tones and lower highlights render, not on how wide the dMax to dMin range in the negative or print is.
In most cases, if you choose over-all contrast based on those criteria, darkroom prints will turn out very well, with shadows and highlights falling at or close to the desired density.
If you are scanning from a well exposed and well developed negative, the scan will give similar results - that of course is what I have been saying from the start in this thread.
When you get to the point where you are really fine-tuning your final results, this approach goes hand in hand with split contrast printing, because it frequently leads to making different (micro) contrast choices for different parts of the image.
Similarly in the digital world, one can fine tune your results by using layers to make different (micro) contrast choices for different parts of the image.