Getting the best B&W rendering from digital.

[250swb]

If anybody remembers the Leica M9 it was greeted with such enthusiasm in part for the way the CCD sensor rolled the mid-tones into whites in much the same way as film, and this carried over into B&W conversion. But being digital the rule was still expose for the highlights. Unfortunately the CCD sensor was soon dropped and the next generation, the M240 and the monochrome M246, were never as good 'out of the box' at converting to B&W because of the way the sensor handled the whites. Indeed it goes further than the sensor itself for any camera, photographers have been brain washed into fearing white, the whole 'don't blow the highlights' mantra has shifted acceptable highlights to Zone 9.

Of course a Zone 10 isn't needed in every picture, but I'm sure the fear of white is why so many B&W conversions fail. And it goes further still, the Leica M10M, best B&W camera in the world blah, blah, blah, has a range greater than the human eye, so if you photograph something that you know is white, it still doesn't come out white in the photo, the camera can see another tone. Hence the images frequently come out looking dull and flat, but people are still scared to countermand the camera and introduce the white or near white that they saw in real life. So nail your white/highlight tone and work back from there.

And obviously it works in a similar way for blacks. Black in film photography was never feared, there was acres of it because of films poor sensitivity to shadows, but digital rules say not to ever lose shadow detail, so be prepared if necessary to throw that mantra out the window if you want to emulate film.

This is why Silver Efex and it's Zone tool is so useful, it not only tells you where your whites and blacks are, but how much you have. You have to adjust your approach between emulating a 35mm film or a large format film, but I think to emulate film you basically need to throw out some of the digital repertoire and go rogue.

 

[MGrayson]

High DR sensors created "lifted shadows syndrome". Not every pixel needs to be visible. This is an M9 conversion, if a somewhat extreme one, (Silver FX for the first cut). Perhaps this just reinforces the Digital is not Film point. (But oh! how I miss the look of an 8x10 contact print...)

L1004007 (2) by Matthew Grayson, on Flickr

Matt

 

[Jorgen Udvang]

I wonder if bayer sensors with larger pixels render BW more closely to film ?
There is a thread on this site about Medium Format fat pixel backs, so the 22Mpixel Blad backs etc.
They seemed to still command a good price.

Cameras like the Sony A7s (I, II, III) and the Nikon DF at 12Mpixel might be worth examining their images to see if there is anything to see.

https://www.flickr.com/search/?q=nikon df portraits bw

Just a thought.

In my view: yes and no.

I find that larger pixel sensors offer some of the same qualities as film, but then add a kind of brilliance (particularly in colour) that I don't see in film. I suspect though that it's not only a question about pixel size, but also about technology or lack thereof. I find that the less perfect older sensors often offer an artistic dimension to portraits that I have problems recreating with the "latest greatest". It's as if something has been lost during the hunt for perfection.

The D2Xs has some of this, the D700 and D3 even more so. The Olympus E-1 is quite good also, as is the Panasonic L1. The real bargain however is the Nikon D200 (and little brother D80). Search for "D200 portrait" on Flickr, or go to the group "Nikon D200 Extreme Machine" on Facebook. It's a great camera with lots of filmatic soul, and they can be bought used from $100. It's the only camera one needs really, plus a few good lenses.

 

[Godfrey]

Digital imager sensors are imperfect, have discontiguous spectral response, and (sometimes) have limited dynamic range, for sure. So do all B&W films.

The basis of "good B&W photographs", therefore, is not the technology of the capture medium but the skill of a photographer in understanding how tonal values of a multi-spectral scene will translate into grayscale tonalities. More acutance/resolution helps when detail is paramount to the intent of the photograph. Better tonal sampling is critical when subtle interactions in the subject need to be expressed in the final rendering. Two things for sure can be agreed upon, I think: it's always easier to reduce the tonal scale after the capture than to expand it, and it's impossible to increase the resolution of a scene after a capture.

Film captures and processing have a very different kind of versatility when it comes to rendering and adjustment when compared with digital capture and digital image processing. Film capture and processing is a chemical game, with the chemistry of the medium and the processing solutions, along with the physical process of the development workflow, determining the baseline for how far a rendering can be manipulated to achieve a desired end. There are pretty tight constraints on how far things can go due to the nature of the chemical processes.

Digital capture and processing is a game of mathematics as to what can be achieved, within the constraints of whatever the set of numbers the equipment is designed to generate. Digital image processing tools try to mimic notions of film image rendering in many cases, drawing on the familiarity of photographers with the chemical processes of film notions, but these control notions are really completely arbitrary and just a matter of UI design. How you beat the numbers up to achieve a given end is what is most important, aside from capturing the subject with a range of numbers that allow maximum flexibility for representing it.

The trick to making B&W that is pleasing and satisfying with digital capture, therefore, relies upon letting go of film notions when capturing exposures and understanding the limitations/characteristics of whatever digital capture device you have at hand in order to get the most data. And then, understanding enough about the image processing tools beyond the notions of film processing simulation and whatever its UI might be to punch the image's numbers about in a manner that produces the desired effect.

I work with both film and digital capture cameras. I can tell you beyond any speck of doubt that what I do when working with film is completely different in evaluating exposure for a given scene, what expectations I have for the final rendering work is radically different, and how to get there is utterly different. It's a learning process. I long ago cast off any notion that I was going to try to get "Tri-X look" or "Panatomic-X in Rodinal" look. I look to what kinds of photograph I want to make and push my digital cameras to make it in almost completely different ways from how I previsualize what my film cameras can and will do.

Sewer Grate with Graffiti - Santa Clara 2020
Panasonic GX9 + Voigtländer Color-Skopar 28mm f/3.5​

Whether I'm successful enough at it for other people only they can say. I'm pretty satisfied for my own purposes that I get what I want, now, after 35 years of banging this stuff around.

G

 

[Jack]

I find that digital sensors of almost any kind do a better --better here meaning more film-like-- job with B&W when the scene is low contrast overall; when there are only 6 or 7 stops total stops of range to capture. Which when I think about it, starts to make more sense.

 

[Wall-H]

I stumble across this old thread while i was searching for any ideas on how to replicate film look for my digital images.
I'm way out of my depth here. I think (probably) some people who posted here are taking pictures long before i was born and i'm seriously doing photography for very few years.
I read about Ansel Adams zone system 2 days ago in book. So apologies for all next in advance.
But i'm picking up this fellowship of B&W photography ( lame way to put my decision in heroic way. I do it because i just like it.)

When we are talking about any conversion (in here, editing or retouching technique to replicate film), before you start, you have to take a scene that would be exactly same for all the way from background + subject + light on subject + composition + lens + size of sensor + settings for light. In my opinion, before we attempt to edit digital picture and take it to film like look, we have to feed same picture from "background + subject + light on subject + composition + lens"
This discussion is missing very much key point of Lens ? Lens are very important factor about image rendering ?

Again i don't have much experience on lenses (Probably ~50 Nikon F, Nikon Z, Nikon old lens, Canon and their mounts, Sony, Hasselblad and now phase one. I also rented few Leica) . This is my personal opinion and I could be very wrong in this very very stereotyping, Except of few exceptions ( major leica lenses, Nikon 58 F/1.4, few voigtlander ? ) all digital lens (and some even older lenses) were more in the "race for sharpness" than rendering.

Why i'm typing this post: I want to try this (simulating film on digital files.) and i could use help from all the experience and wisdom here.
Why I think I'll even have shot at this: Because i've solid test plan in mind. I'm going to put "background + subject + light on subject + composition + lens + size of sensor + settings for light." exact same and then compare B&W film and Digital conversion for that specific setting. That is the best shot in achieving or even evaluating this on possible terms.
Would this help in generic and landscape work: Most likely no but i think there should be enough in studio work ?
What is exactly that i'm looking for: Input on my thought process. Pointers if my calculations are on the point. nudge to right direction if they are not. audience for blind taste.
What one would get out if this: Apple to apples comparison of film and digital black and white in terms of pictures. This is 100% guaranteed at some point in time. If possible, steps to replicate film look on digital files.

Anyone interested ?

 

[pegelli]

It seems you are assuming there exists just one "film look" which is probably a gross oversimplification. First there's different kind of films which all render differently, There's Ilford (where the FP5/HP5 is again different from their "Delta" series), Agfa, Kodak, Fuji Acros (all available at different speeds) and then all these different films can be developed in different ways, different developers, different dilutions, stirred or still, pushing iso or box speed, etc. etc. Every combination results in a different overall rendering. And even size matters.

So the first thing you need to do is determine which film look you want to emulate digitally, best way is to get an analog camera and start experimenting with different films and development methods. And on that camera use a lens that you can (via an adapter) also mount on your digital camera so that once you have determined what you're after a different lens doesn't throw you off base.

My own approach is much simpler, if I want a "film look" I shoot film and fully accept digital is different, not better or worse, just different.

 

[Wall-H]

Great point. I've tested Kodak Tax 100 , Ilford Fp4 plus 125, Cinestill 200-800. I'm having hard time to tell the difference between Ilford Fp4 plus 125 & Cinestill 200-800.
I'm clear with preferred speed. (As much as possible < 200). I've fuji GSW690III for landscape work but i'm looking for a portrait medium format camera not at insane prices.
I'd to give up my 503CW for the Phase one purchase.
I'll put this question on separate thread.
I agree with your point about going on separate looks. but please continue suggesting flaw you see in the plan.

 

[Paul Spinnler]

As Pegelli says, there is no B&W film rendering given the stocks. But it goes far beyond that with B&W.

With film it is a print rendering, so film is shot in a way, developed in a way and printed in a way. With B&W the last two parts are highly complex and require a lot of experience. In the big laboratories in Paris, London, it is not uncommon to distinguish on the print lab side between colour and B&W. I am talking about full analogue process, the digital print process is another layer of complexity because you need to also see how you scan and as soon as you break out of the analogue domain you both limit and expand your transformation options.

It is complex.

To master film emulation you need to master film and printing first.

This takes years.

Mastering printing takes years as you need to learn how the print reacts to your input on shadows and light parts of the negative during the printing process. Within printing you further distinguish between high key and low key which, if mastered, can create results from a negative difficult to repliacate digitally because the tone curve is mapped totally non-linearly through the development and printing process.

Start with the film side and in a few years you will have the eye and can try to emulate. You need a lot of time and resources do to film emulation right and B&W film requires a deep practical experience of printing. It is not about just retouching it with a Wacom. You need to gain a feeling how toanlity reacts to over- and under exposure THROUGH a print which both are non-linear steps. That's why a master printer can create dramatic skies difficult to replicate just in one image editing app.

Looking at Illfords instagram, you will see some amazing B&W landscapes. Truly masterful stuff.

Best is to learn it at an art school and print, print, print. Even better, start working for a print master. You won't earn a lot and will have a steep and long learning curve, but in the end you'll master an arcane art form. Almost like learning how to craft violins in Italy or learning how to create hats in Paris or London from a hatmaker.

Working in C1 won't cut it as you need to understand how the analogue medium reacts to changes in intensity of light which is non-linear.

I havent even yet touched development! LOL.

I think 90% just think of film today as shooting, developing, scanning it with some retouch. But this is not the traditional way; negative printing is a technology that is rarely used because of the costs and lack of access to printers, chemicals or just general lack of means to do it the fully analogue way.

Go into a huge print lab in a large European capital with enlargers and master printers and you will see that it is precisely not just scanning and C1, but a whole world beyond that.