Densitometry, exposing for the shadows, and effective speed. And those highlights?

[trypdal]

*Long post warning* (apologies in advance)

I have a perhaps boring question for the more technically minded in this forum if I may.

A bit of background first. I am what people nowadays call 'a hybrid film shooter'. Essentially someone who has nothing against digital photography, but dislikes expensive/complicated DSLRs, loves film, film cameras, grain etc, and wants to achieve a scan as an end product, as opposed to a darkroom print. I like film cameras, I like the beauty of scanned black and white film, I like the process that gets me there. I do not have, nor I'm interested in setting up, a darkroom, and I do my film development using a dark bag, which suits me fine.

From the above it follows that I share with many film photographers who discuss the medium and the art online an interest for a) exposure/getting great exposures, b) development/processing my film well, BUT NOT c) darkroom printing, as my negatives need to be optimised for one purpose only: the best scan I can get.

Now to (slowly) get to the point of my post: whereas I have found that most advice on negative exposure and negative development applies equally, regardless of end product (whether a scan, a print, a slide projection etc), I'm not quite so sure this holds for one critical technical variable - what the zone system aficionados call 'effective film speed' or 'effective sensitivity'.

A rapid google forum search of any film stock currently in production will return endless threads on how 'it should be exposed so that it prints on contrast 2 paper'. So most advice on film regards the optimisation of its characteristic curve to fit the signal on the negative to an extremely non-linear, technically inferior (relative to the negative), but creatively interesting medium, such as the photographic paper. Cue endless posts on home-made densitometric tests, discussions on true film speed, metering for the shadows then placing any interesting shadows in Zone III and so on and so forth.

The key thing here is that, almost inevitably, the outcome of these discussions is that 'real' film speed is a fraction of the one declared by the manufacturer on the box (with a few exceptions - some films seem to be 'truer' to box speed). It is not uncommon to read that 400ISO film 'should be exposed at 160' and that 100ISO film 'is a 50EI or 25EI at most'.

Now to my question, which is very simple really. The theory behind the practice of finding the true, much lower, film speed is steeped in densitometry. I believe the end result sought is that of getting enough light in as to retain detail in the shadows above base+veil. One outcome of this is that the negative will be 'easier to print on normal contrast paper'. I have seen some prints, or scans of prints, obtained using this method, and they seem to focus on readable detail even in deep shadow areas (for example, undergrowth in shade - a rock in shade - etc).

But what does the above do to the highlights?

We know that exposure mostly affects the shadow density, whereas we fine tune highlights and contrast via development. However, we know most films have a characteristic curve which is linear, if at all, only in a middle segment, becoming non linear in its extremities (the 'toe' and the 'shoulder').

I have noticed that many prints or scans of prints shared online, and produced according to densitometric 'true film' speed principles seem to feature weirdly compressed, mangled up highlights. I have been wondering why that might be the case. Could it be that using a densitometrically-correct film speed that lifts the shadows 'above the toe' (pushing them into the linear section) might lead to pushing the highlights way beyond the linear section of the curve, into the shoulder - even if we reduce development considerably?

So my question is if by trying to save the shadows to make the negative easy to print, we might (for contrasty scenes) end up pushing the rest of the histogram so far up that we're now getting suboptimal mapping of the highlights (which is not a massive concern, I understand, for wet lab printing, but might results in not exploiting the film's tonal range to the fullest).

So essentially is a different mindset in order when shooting film that will be scanned? A scanner sensor, unlike an enlarger head, is an almost completely linear device (excluding non linearities determined by thermal noise I'd assume). So a scanner has no problem in digging into 'thinner' shadows that would make wet printing difficult. As a consequence, if my reasoning is correct, placing the histogram lower on the curve (ei exposing at box speed or higher speed than what results from densitometric tests) might mean we can map those highlights before the shoulder kicks in (depending of course on film used) thus getting a truer representation (read: more linear mapping) of the negative wrt what could have been achieved via a print.

Any thoughs or comments/corrections appreciated.

 

[StephenM]

I have a negative that I thought was completely unexposed and clear film. I subsequently found that against a white background there were traces of something. A scan revealed a poor but recognisable image. So you're correct that a scanner can recover detail that conventional printing can't.

The slope of the characteristic curve depends on the development, or, put another way, you're going to reach the higher densities sooner in the highlights with more development. The late Barry Thornton said that almost everyone overdeveloped and should cut development by 15%. And if you read the instructions carefully, you'll usually find that development times are determined according to the required contrast for the enlarger, giving different times for diffusers and condensers.

Giving less development and more exposure reduces the contrast of the negative; and most people seem to find that scanners are easier to use with a low rather than high contrast negative.

My own practice is to expose FP4 Plus at 80 not 125 ISO to ensure shadow detail is lifted. Your development time will depend on how you develop (continuous or intermittent etc.) so I won't give mine as not being useful - but in Rodinal 1:25 or 1:50 with my agitation technique I find that those in the massive development chart a good starting point (higher dilution = lower contrast). These scan well on my Epson flatbed.

Note that conventional papers are what the curves are tailored for, and these can't easily cope with densities above a certain value, hence curves are only plotted up to a certain density. If you look at the curves for Pan F compared to FP4, the latter shows a shoulder at the top end whereas the former is still a linear response. Read into that what you will...

Edit to add. I am not overly concerned with technicalities, don't test scientifically and am satisfied with what works for me. So long as the negatives print easily to give the results I want, I'm happy. The "cut the film speed half a stop or so" works for me with my developer, agitation and equipment.

 

[sirch]

I wonder if the thing with the highlights is that for some subjects, e.g. landscapes, they typically matter less? Trying to lift shadows tends to emphasize the grain but bringing down a highlight doesn't, also as long as there is some detail in a sky then who is to say what the clouds looked like at the time of shooting?

 

[Terrywoodenpic]

in days of yore, every one understood that the film speed on the box only applied if the standard Recommended development was used and they they were processed to the recommended gamma. If you did so the film speed was indeed accurate.

However almost every one came up with their own Exposure index that was correct for their own developer choice and processing parameters and type of enlarger and paper to be used. As I used Kodak D76 Full strength with large format, in 3 gallon tanks, and kept topped up with replenisher. I used the Boxed ISO. with out any problems.

For medium format I also used D76, but this time diluted for one shot use. and ended up with a slightly lower personal Exposure index. To allow me to make full tone prints on Grade 2 bromide paper. ( or multigrade)

Medium speed film of around 400 ISO has an enormous brightness range. allowing both the sholder and toe of the characteristic curve to cover an equally large tonal range.
It was normal to expose so that the toe contained detail in the tone immediately above that you wished to print pure black.. In practice this required taking a shadow reading using a Weston meter using the "U" setting as the index. This would ensure that all tones up to the "O" index would contain usable detail in the resulting image. the exposure was then read from the Arrow position in the usual way. Every thing from your "chosen" deepest shadow would be Printable.

However you could equally work in reverse and use an incident light meter and "Peg" the highlight exposure ( as was normal for Transparency film) .
This works very well especially on slower high contrast films. An incident meter pointed from the subject position towards the camera position will peg the highlight so as to never burn them out ( spectral light excluded) as the tonal range on such films is more restricted The position of the toe is come what may, but will be as good as is possible in the given situation. Negatives produced from incident light readings are exceptionally easy to print as the print Exposure times fall in to a surprisingly constant band.

Very slow films developed in a highly dilute Beutler type compensating developer, like Neofin Blue. can provide exceedingly sharp images with an extremely wide tonal ranges.
However this is today almost a forgotten technique.

This was recovered from a small album print Made in about 1956 using Adox R17 film developed in Neofin Blue. though the film was only 17°Din the tones were exceptional right from the deepest shadows to the highlights.

 

[StephenM]

If you look at Way Beyond Monochrome you'll find a section on digital printing, where the authors say that if printing digitally, it's easier to maintain separation in shadows and highlights, and in fact digital prints exceed conventional printing in that respect. This is from memory, so a disclaimer in case I've misunderstood or misquoted them.

And I should perhaps add that in my case, although I still maintain a darkroom with enlargers, I haven't made a conventional print in a dozen years, finding that my digital black and whites are better than I could achieve in the darkroom.