How to Cure the Digital B&W Nasties

The biggest change in film > digital from a visual, artistic, results standpoint* is the wholesale changeover from B&W to color. When I got into photography in 1980, almost all serious photography was B&W. Color was considered "literal" and "decorative" and widely denigrated as being the province of commercial photography on the one hand and amateur snapshooters with their machine-made 31/2 x 5-inch prints on the other. There were a few serious color photographers, but they were token. When serious photographers began committing to color in the '80s, it took some courage.

That's been turned on its head. Carl Weese opined way back in the 1990s that "digital is the coming of age of color photography," and the years since then have borne out the truth of his words. Cameras shoot color now. People shoot color now. Color is ubiquitous. B&W has become as quaint as an old-fashioned wooden boat.

(...Also as beautiful, but if you don't feel that, you don't, and if you do, you already know what I mean. This is a 1941 Chris-Craft barrel back, photographer unknown.)

And here's one reason why: most digital B&W is nasty. I mean NASTY. I'm going to give you the simplest recipe to cure that.

There are two reasons why almost all digital B&W is like drinking rotten pond scum:

1. The first is that B&W film is biased toward richness and detail in the highlights and lacks information in the shadows (hence the old adage "expose for the shadows and develop for the highlights," because if you didn't get the shadows right in camera you weren't going to be able to get them by prestidigitation later), and this happens to match how our eyes see. Digital does the opposite: it has a low amount of detail in the highlights and has vast reserves of information in the shadows. Ever taken a raw file with a camera high in DR and tried to dig the shadows in a good raw converter? It's amazing. This does not match how our eyes see. Bad highlight detail isn't just a problem with digital B&W: it plagues all digital. We're getting used to digital highlights now, thanks to the habituation of constant bombardment, but that's like learning to love an ugly dog. It still doesn't make it pretty.

2. The second reason why almost all digital B&W is slathered with pasty poodle excrement is that digital depresses the middle values. Compared to a scale of theoretically ideal tones, Zone VII is rendered as Zone VI if not Zone V, Zone V is rendered as Zone IV if not Zone III. It looks terrible. Rodinal used to have this effect on film. Mildly, not as bad as a lot of digital. Okay, okay, granted, it's a "look," and a few photographers have done it on purpose and can make it look aesthetically coherent. Emphasis on "few." Especially if their subjects are dark and airless. But most digital B&W is that way by default and it looks awful.

The combination is a one-two punch of harsh grating ickyness. Vacant highlights as stupid as a lobotomized lobster, and all the grays crushed toward black like the Devil's own fog of depression. Shoot me now!

It's no wonder most people prefer digital color.

(May I just point out here in passing that most scanned film also looks terrible? It's because the film's characteristic curve was designed to be counteracted by the paper curve. Virtually everyone scans film and reproduces the film tones exactly—literally—which is why scanned negatives also look harsh and tonally wrong, and nothing like the same negative printed on an appropriate photo paper. Here's the best way to scan B&W film: 1. Take the negative. 2. Make an 8x10 print on a suitable photo paper. 3. Scan that.)

So here are three simple steps if you want to make oh-so-nice black-and-white in digital:

1. NOT MANDATORY: Use a Fuji X-Trans sensor. They just yield better B&W. I won't go in to the reasons why here.
2. Underexpose everything by 2/3rds of a stop for headroom. You'll want to be be able to place your highlights "just so" in post, and this will give you more information to work with. If you are using a high-DR camera (look at DxOMark Sensor and click "landscape" for a ranked list of these, if you don't use a Fuji. They don't rank Fuji). Shadows? You don't hafta worry about no steenkeeng shadows. Digital will let you dig out dark-value information to your heart's content in B&W.
3. Correct the tonality by lifting the midtones.

Here's how to lift the midtones: in whatever editing program or you use for post processing or conversion, find the curves function. Grab the middle and move it up, to taste. Viz.:

That's the simplest formula. However you do it, a) get enough highlight detail and b) restore the midtones to their proper values and relieve them of their horrible depression, and you'll be way ahead of 94% of the digital B&W out there.

DISCLAIMER: I know there are dedicated practitioners out there who do very nice digital B&W and I don't mean to "damn them by generalization" with my playful rhetorical hyperbole about pond scum and poodle poop. I'm just slagging the typical desultory, rote B&W conversions you run across here, there and all over everywhere. Nothing personal, fellow B&W mavens. I don't mean you.

Most digital B&W is never going to quite match film B&W, for the simple reason that it is no longer the era of the classic wooden boat. But that doesn't mean dedicated workers can't make perfectly nice, non-atrocious B&W files and prints in this, the Era of Color.

Mike

P.S. Two hot equipment tips: the original Sony A7, now available used for $700–$1,000, is a great camera for B&W because of its very high DR. And the new Fuji 23mm ƒ/2 lens is well matched to B&W because it's biased toward high lens contrast like a classic Leica or Kodak Commercial Ektar lens.

*The changes to the industry and in convenience might be larger but are already well parsed.

Original contents copyright 2017 by Michael C. Johnston and/or the bylined author. All Rights Reserved. Links in this post may be to our affiliates; sales through affiliate links may benefit this site.

It's a gray area
Give Mike a “Like” or Buy yourself something nice

(To see all the comments, click on the "Comments" link below.)
Featured Comments from:

Bill Wheeler: "Thanks for the informative article, Mike. As a rule, I prefer black-and-white photography, so I have been working on improving my digital technique. Thanks again for the tips. By the way, I love hyperbole."

greg brophy: "Ha, that's funny. I once asked Carl to help me with a B&W photo and he basically lifted the midtones."

Andre Y: "Interesting thoughts. A while back, wasn't there going to be a post with annotated examples of good B&W photos? I might have missed it though...

"Two random thoughts:

1. I just started reading Keith A. Smith's Structure of the Visual Book, and he rails against color photography as it was practiced when the book was written (maybe 20 years before digital became popular), by saying, among other things, that people think it's literal, when that's the worse thing color does.
2. And underexposing to save highlights is an interesting topic, which on the face of it seems to contradict ETTR practices (it doesn't), but there are a few cameras which do that automatically now, including your Fujis with their DR modes, and the current Hasselblad and Fuji medium format digital cameras.

"What they do is that above a certain ISO (1600 is common), they no longer brighten the images in the camera, i.e., apply digital gain. Instead the files are exposed for metering at ISO 1600, even if you're using IS0 6400, and as a consequence they get increased dynamic range at high ISOs. You can see the measurements on photonstophotos.net, for example, of how the dynamic range curve flattens out above ISO1600 for the two mirrorless medium format cameras instead of decreasing like say on a Nikon D810.

"The RAW converter on your computer then automatically applies the correct brightening when the RAW conversion is done so that the photos appear to have been exposed at the intended ISO. The noise in the raised photo is basically identical to a shot that was metered and exposed at the intended ISO because of the way the sensor works. It's too complicated to explain here, but look up on DPreview, for example, ISO-invariant or ISO-less sensors. Basically for those types of sensors, there is no difference between exposing at ISO 6400, or using ISO 1600 and raising the brightness two stops afterwards: the noise is identical.

"There are cameras without these kinds of sensors, and some sensors have regions of ISO-invariance separated by sensor gain bumps at some ISOs. The Nikon D5 is sort of an example of the first kind, and some of the Sonys have a dual-gain structure that are an example of the second kind. So do your research first, and experiment!

"What you gain is an increase in dynamic range at high ISOs, which is especially useful for situations where you might like to use high ISOs in the first place, like night shooting or theater stage shooting where the brightness range of the scene can be huge due to individual lights being very bright against a dark background, and it's easy to blow out your highlights.

"The great thing about this practice is that it can be applied, somewhat inconveniently, to any camera that has an ISO-invariant sensor, like the Nikon D810 and D500 that I use. Of course, my camera previews will be dark (unlike the Fuji and Hassie images), and I need to manually raise the exposure slider in Lightroom, but it's more than compensated by the fact that I now have at least three more stops of dynamic range in my camera at high ISO than before (my ISO is set to 200 and I expose as if I was shooting at 1600), and that helps quite a bit with post-processing. I was also going to get a D4S, and I don't think that's necessary anymore, at least for high ISO dynamic range reasons when I use this trick."

Andrew McFaul: "Any experience or comments regarding the use/benefits of using Foveon sensors (either Merrill or Quattro) for B&W photography?"

Mike replies: My experiments with the Merrill were extremely lovely but I found the camera I tried (not a current one) and its workflow pretty difficult to use.

Al C.: "Agree: most digital B&Ws are ugly, unless their 'digital hardness' serves an artistic purpose. But Merrill Foveon B&Ws are gorgeous. The transparency and smooth tonality cannot be achieved by any amount of post of a Bayer file. Have not used the Monochrom, but people I trust say the Merrill is almost as good."

mike: "I'm going to have nightmares about that dog. And maybe the Omega too!"

Stan B.: "Sorry, Sternfield's American Prospects was published in '87 (could've sworn sooner), Meyerowitz's Cape Light in '78, and Shore's Uncommon Places in '82. Soon photojournalists like Paul Graham got into the color mix with smaller formats in essays like Beyond Caring [the link is to Jeff Ladd's book on the book —Ed.] and the incredible Troubled Land, and the rest as they say is...."

Ken Ford: "Very timely post for me, Mike—I have a new X-Pro2 and 23mm ƒ/2 and have been starting to create a B&W workflow for them. Thank you! "

Douglas Chadwick: "I would advise one modification from your advice, 'the best way to scan B&W film: 1. Take the negative. 2. Make an 8x10 print on a suitable photo paper. 3. Scan that.' Change step 3 to 'match that.' And match it from a scan of the negative, not the print. I spend a lot of time matching digital files to darkroom prints and have learned a lot from doing that."

Mike replies: That's a great idea. Wish I'd thought of it.

Jeff: "On the other hand, I've seen a lot of mediocre silver prints, and a fair amount of gorgeous inkjet prints. While tools, techniques and materials help (e.g., large sensors, all B&W inksets, fine papers. etc), the most important tools remain between the ears. Nothing substitutes for a good eye and good judgment, film or digital."

Oskar Ojala (partial comment): "I'm not 100% sure what exactly is referred to in this article; examples would help a lot to clarify!"

Mike replies: It's a problem. Probably the biggest problem, historically, with this website, as a format. I could find hundreds of examples of what I consider "bad" B&W, easily, but how could I present them? If another person's  picture is used as a subject of a critique, it's legal under Fair Use (probably—Fair Use is never 100% certain), but I'd have to be taking someone's work and holding it up as a negative example. As a normally courteous person I have trouble with that idea. I certainly wouldn't like it if it were done to me, so how can I do it to someone else? What if I were to use as an example of "bad" tonality something someone did very deliberately, because they like it that way?

Over the years I would have loved it if I could have put up pictures I find on the Web to discuss—many pictures of many kinds, to discuss in all kinds of ways. But I just don't think it's polite to do that to the people the pictures belong to.

It's a big difference between the Web and a classroom. In the privacy of a classroom you can use specific pictures as examples, easily. But I have a problem with doing the same thing here.

Comments

[Chuck Albertson]

Or, buy a Monochrom and use an incident meter to nail the highlight exposure (just like shooting slide film).

[Dennis]

Can't wait to try it. One side benefit of the -2/3: I have a 'thing' about converting photos to b&w if I didn't intend to shoot them in b&w in the first place - it feels like I'm cheating (I'll typically see an image in Lightroom that makes me wonder if it might look good in b&w and sometimes it does and sometimes it doesn't, so it makes me feel like a hack who's just playing around). I'd have to be deliberate about it to expose differently in the first place, so I'd alleviate that (silly) concern.

[Sal Santamaura]

Now two months into a hybrid workflow, I'm entertained to see your advice. I bought an Epson V850 scanner and P600 printer.

The curve you suggest is almost identical to what has worked best with my 8x10 320TXP negatives. Of course, some of that relates to the TRI-X being "all toe," but I suspect it's also partly due to the digital printing process' linearity compared with darkroom paper.

As an aside, if anyone decides to follow this path, note that TRI-X sheet film's unique retouching base side coating makes dry scanning on a flatbed extremely simple. No Newton's rings.

I've been trying lots of inkjet papers for my 'quaint' B&W images. So far, Hahnemuhle Photo Rag Ultra Smooth pleases me most. I settled on simple gray scale files, the downloadable Hahnemuhle profile and color management turned off. No ABW necessary for a neutral result. The resulting prints are an excellent match for Bainbridge Alpharag Pure White board. Best of all, they're matte, and can be easily viewed without reflections. I've been frustrated for quite a few years by the ever-increasing glossiness of darkroom papers, making even "air-dried F" surfaces far too mirror-like. This advantage of digital printing is a bonus beyond the greater tonal control it affords.

You haven't gotten me to give up the view cameras yet, Mike, but incremental steps could lead there eventually. Perhaps a Fuji GFX 100S, with its rumored 100 megapixels, will be close enough to the 115 megapixels an 8x10 negative on the Epson scanner provides to inspire complete conversion. :-)

[Robert Newcomb]

Color is still "literal" and "decorative" even in this day when its shot alot.
Not any better then years ago, just easier to do then on film.
Isn't that what we're all about now, quick, easy and so so.

[Bill Poole]

Thanks, Mike. I'll be interested in reading the comments on this one. I convert a lot of my Fuji digital files to black and white, and yes, I often underexpose and then boost the mid-tones using the Lightroom sliders. I also use the color sliders to create digital filters for the files, for example, to darken a sky.

But I also shoot and scan medium format black-and-white film, adjusting those files in Lightroom also. And I've got to say that even ink-jet prints from these images are far more satisfying (to me) than even the best prints from digital files.

As a lover of traditional black-and-white, I feel fortunate to live in San Francisco, where there is a thriving analog photography culture, especially among young shooters, and where film, processing, and public darkrooms are readily available.

[Larry Gebhardt]

Mike, sorry to say I think your advice to underexpose by 2/3 stop is widely off-base. In order to capture as much of the shadow detail as possible [That's not the aim! What I'm saying is that you don't need to worry about shadow detail--it will take care of itself. What you need is to capture as much of the highlight detail as possible. --Mike] what you want is to expose as much as possible without clipping any channel. This is know known as exposing to the right (push your exposure to the right side of the histogram). Under exposing one stop throws away half of the precision of every "zone". Not a huge loss in the highlights with a 14 bit raw image, but at the lowest zones you will going from say 8 available tones to 4 per color.

This is as close as you will get to exposing for the shadows with digital, while still needing to avoid blowing out the highlights (like positive film). Unlike transparencies we can adjust the exposure in post and fix the overexposed look.

My advice for digital black and white is:

1) Shoot raw

2) Expose to the Right. Set exposure so one channel is pushed to the right without clipping. Hard to see with the cruddy jpeg based histograms in cameras, so bracket if possible and pick the correct exposure later. Once you learn your camera's metering behavior in various lighting you can set exposure compensation pretty intuitively. Depending on the lighting this may require positive or negative compensation from the camera's meter.

3) Bring the exposure down in post processing until the exposure looks right, for me that's usually a bit more than the amount I had to tell the camera to "overexpose".

4) Convert to black and white. Edit to taste, and remember you can adjust the colors behind the conversion like we did with filters.

[David Babsky]

I like "burned out" highlights. I like white skies which have no detail. They look like, and remind me of, the old photos of Whatsisname ..er, you know, the Frenchman, Eugene Atget, who used blue-sensitive film back in the days before Verichrome Pan B&W was sensitive to other colours than blue.

I like the rapid fade to white of distant objects in B&W pictures, rather than the "faithful" tonality which we're used to in Ansel's 10x8" smooth-toned "performance" enlargements.

Burned-out skies are nostalgic for me. I'm not after perfection; I'm after olde worlde charm.

[What you mean is panchromatic film, not Verichrome Pan. And actually white skies in albumen prints were masked; blue sensitive emulsions led to failure of the reciprocity law, and mottling. But as for your taste in sky rendition, it's okay by me if you do what you want! Whatever you like is the way you should have it. --Mike]

[Ernest Zarate]

Excellent, as usual, Mike. I am glad to see you regularly bring up this topic on TOP.

I remember reading two essential books when I was a youngster: David Vestal's The Craft of Photography and Fred Picker's Zone VI Workshop. My primary take away from both of those books (which are here, somewhere...) was that making a good print was a matter of time and patience.

While your commentary divorces the "convenience" of digital from the main topic, I'd posit that ease of use is one of the main reasons B&W has suffered so. Digital, at its foundation and by its very nature, is quick. One and done. It is the polar opposite of what we took as required in the darkroom: time consuming. Few go through the process (and, yes, it is a process) of making a top quality image.

As I recall, one of those two authors stated that an important darkroom tool was to have a few high quality B&W prints available to view as one was printing. These prints would serve to calibrate the photographer's eyes as to what a "good print" would look like, to give them a sounding board by which to judge their own efforts. A corallary to that is that perhaps many current photographers, who have limited, or no, darkroom experience simply don't know what a good B&W print looks like.

Maybe your next print sale might include a few sample B&W prints, at a reduced price that digital photographers could pin to their walls next to their computers and printers so they could judge their efforts next to some solid, well made samples.

[Martin Fritter]

Thanks. Will try. Could you elaborate on lenses? Seems key. A7 does do nice b&w, plus wide range of lenses available via adapters.

[John Krumm]

Great post, very useful. I think it's more or less what I do with all my cameras. I've played with lifting the bottom of the curve a bit and then pulling it down slightly up the curve (the classic S) but I'm not sure if I ever do much better than with a straight midtone boost.

[Larry Gebhardt]

[That's not the aim! What I'm saying is that you don't need to worry about shadow detail--it will take care of itself. What you need is to capture as much of the highlight detail as possible. --Mike]

Exposing to the right does that (assuming raw format) across the board shadows to highlights, even if the image looks over exposed when you first open it. By underexposing one stop you are throwing away half the tonality at every zone and cutting the sensor's effective dynamic range. There's nothing to be gain by underexposing, besides an increase in shutter speed.

It's easy to test. If you want to get really geeky open the images in Raw Digger and see your true exposure relative to the sensor's capabilities.

If you intentionally bury your shadows when printing it's probably not too important to worry about. But neither is buying a sensor with high dynamic range.

[adamct]

Mike,

I think you and Larry aren't really disagreeing, just coming at this from different perspectives (practical vs. technical/theoretical). If (and ONLY if) you don't clip your highlights by not overexposing any channels, then digital actually has most of it's information in the highlights, and the "least" information in the shadows.

BUT, as a practical matter, today's digital cameras frequently overexpose some portions of the frame. When that happens, it's not that you lose some detail, it's that you lose all detail (in theory, there may be some detail if you only clip one color channel, but then you have color problems). It isn't gradual. Exposing to the right yields greater and greater information, right up until the point where you go too far and have zero information.

Moreover, your point is well taken. Just because digital cameras have the "least" amount of information in the shadows, that is only a relative statement, not an absolute statement. On an absolute basis, there is usually plenty of information that can be recovered from the shadows to produce pleasing prints.

As a result, I agree with your advice from a practical perspective. I think the metering systems on all modern, digital, "large" sensor (micro 4/3rds or larger, I don't have enough experience with 1" sensors to speak to them) cameras should be revised to avoid overexposure in all but the most extreme situations. Whether setting exposure compensation to -2/3 is enough to do that or not, I don't know.

The primary reason people don't do this is because it makes pictures appear dark and unnatural when viewed on the camera's LCD screen. But there is not reason that manufacturers couldn't apply a tone curve when displaying images to make them look better, while retaining all of the original data in the raw file.

Best,
Adam

[Stan B.]

Digital B&W made in the shade can look phenomenal- in contrasty/direct sunlight... Yipes!

I will say however, that I am quite happy with my scanned B&W neg results- it allows me more overall tonal control (except in Totally blasted highlights), as well as in the smallest of details (shoot with a W/A- that's what you get)! You don't waste a ton of paper, or the money for it, and you don't inhale noxious fumes.

And nitpicking here, but when Eggleston burst unto the scene in '76, color (particularly large format color as propagated by Meyerowitz & Sternfield) had already begun to dominate the photo art scene by the end of the '70s. It was a quick and furious takeover...

[Omer]

You really think William Albert Allard, Joel Meyerowitz, John Pfahl, Larry Burrows, William Eggleston, Helen Levitt, and many others were not taken seriously?

I don't remember ever reading anything that dismissed the work of these photographers, and my reading goes back to the '70s. Perhaps you should consider the crowd you were involved with.

[Peter Wright]

Interesting post. I work mainly in B&W in both film and digital. For film I use a hybrid approach, and sometimes find myself thinking that the difficulty in obtaining a good end result must be something I'm doing wrong. Getting a good commercial scan appears to be virtually impossible (and I've tried many labs, each has their own variation of the 'poodle poop'). So I tried my own scanning using flat bed scanners and dedicated film scanners. These have their own problems, not least of which is the supporting software.

So I finally decided to 'scan' using a camera, neg holder, and flash to light the negative. I now have this down to a fine art, and find that the OMD-EM1 mk2 with it's high pixel density and ability to do high resolution tricks with static objects by shaking the sensor, is ideal. I invert the files in batch mode in photoshop and then use the curves tool on an individual file by file basis.

With the films I know, I tend to lean to over-exposure (rate Tri-X at 200) and slight over-development. With digital it's the other way round. I use a Monochrom and tend to start with an underexposure of 2/3rds stop as you suggest. This allows the texture of the highlights to be preserved. If I simply go by the histogram I may not entirely wreck the highlights, but the texture will suffer as the highest pixels of the highlights can be clipped but not seen.
Usually, a quick trip to the curves tools is necessary before all is done.

I was beginning to think it was just me that was finding getting good B&W so difficult. You made my day!

[Sylvain G.]

Mike, very interesting post.

I'm about to start "scanning" negatives using the Hi-Res mod of the E-M5 mkII. The idea is getting enough resolution with a single shot without the ultra painful hassle of my epson flatbed. There are been many very good results on this approach, not without other issues to deal with, of course.

But this raises a question about your curves tips.

You detail how to correct digital files.
You mention scanning doesn't quite work how it should as it's supposed to be printed. I guess "paper curves" could also be applied, to taste.

Then what happens if you "scan with camera" applying its curve on a negative curve supposed to match the curve of a paper" ??

I'd of course try to get the most neutral curve possible from the olympus, but what about the rest...
Any help appreciated :-) and I haven't even spoken of the orange mask issue for color negs...

Nice anyway, Mike !

[Dogman]

I'm a muddler. Reading the instructions almost always either frustrates me or leaves me with the proverbial thousand yard stare. I had just as soon wing it and see what works.

Soon after I started using Fuji cameras, I rekindled my interest in B&W. After all, B&W photography had been my priority beginning in the 1970's and had only shifted to color with my use of digital cameras around 10 years ago. So I muddled along blindly for a time until I found methods that worked for me. As noted in Mike's suggestions, my muddling led to the eventual conclusion that I should routinely underexpose a bit while shooting and then pull out the shadow detail in Lightroom when developing the Raw files. Imagine that!

It's nice to have my conclusions confirmed today.

[Steve Jones]

My cure was to switch from digital to film nearly a decade ago. And since I still struggle to love inkjet prints, I'll soon be going into the darkroom. But whether using digital files in the past or scanned negatives, curves has been my primary software tool for contrast and tonal control. Definitely recommend it if it's included in your software.

[scott kirkpatrick]

That ChrisCraft is so beautiful that I haven't looked at anything else in the article yet.

[David Saxe]

I am not so sure I agree. I have been shooting B&W for 40 years—first on film and since 2006 digitally. Its true there is a difference but I don't think either one is better than the other—they are only different. What I see these days is that people just simply convert from RAW to grayscale and make no further adjustments because they do not know how. They just accept what they get. In my work, I make adjustments in Lightroom, open in Photoshop and use Nik SiverEfex (soon to disappear) and then finish up with some dodging and burning and levels adjustments in Photoshop. My the RAW file is only the beginning and a good black and white image is at least 20-30 minutes away. Once I print the image, some levels adjustments are still necessary. Most important, an image is never complete until it is printed.

[psu]

I mostly enjoy digital B&W because I've come to peace with the conclusion that basically what you do is shoot it just like digital color (which you shoot basically like color slide film, but with a longer DR), but then afterwards you make the thing look like a B&W picture in Lightroom or Photoshop or whatever. This is convenient because in the film days I could never do that trick where you shoot one camera with color and one with B&W because I'd always shoot one or the other wrong, esp. if the color was slide film.

I always find myself curious about exactly what I'm not seeing in my various B&W shots that would make you think they were poodle pucks (I even have a lot of scanned B&W film that as far as I can tell invokes the spirit if not the actual look of the prints I made from the negatives back in the day). But then I remember the rule: all that matters is that it looks good to me. 😃

[Ken Tanaka]

"There are two reasons why almost all digital B&W is like drinking rotten pond scum:"

Hilarious!

[Bill Mitchell]

One of the BIG advantages of mirrorless cameras is the ability to show the B&W image on the EVF.

[Chris]

The paradoxical side of this "everything digital is colour"thing is that of course the sensor is entirely monochrome, and has to have colour added by extra colour filters and digital jiggery pokery! So taking black and white on digital (the Leica Monochrom aside) means reconstructing the mono that was there before it was destroyed!

I do hate a lot of digital black and white that I see, possibly for different reasons. I mean the sort of image that looks like the black and white version of HDR. Super slick and glossy.

I use film; I like the "intentionality" of film black and white, though I will sometimes take two cameras so I do have a choice. I've a full hybrid workflow; I haven't noticed the issue you describe on scanning; it's certainly not the same as wet printing, but it seems possible to get nice images with quite long tones (as far as I know what that means!).

Maybe the worst thing about black and white on digital is this view that it is something you do when an image hasn't quite worked out right. No, no, no! It's what you do when you see an image that is just going to work supremely well in black and white; at that point, the colour is irrelevant, and only getting in your way!

[David Cope]

In a digital camera I'll take tonality over resolution anytime. I would like to see manufacturers use 16-bit ADCs with ultra low noise amplifiers with plenty of headroom in some models rather than just high end digital backs.

And can we PLEASE have more dedicated monochrome cameras. Ideally what I want is some boffin to invent is a voltage controlled (similar to liquid crystal) layer which can be "pixel" addressable and placed in front of the mono sensor to act as a programmable CFA - transparent for mono shooting, or a bayer (or X-Trans) pattern for colour. It could act as a grad ND as well I suppose if both colour and density were programmable.

Or I could just keep on shooting film :-)

But the big question Mike - are you going to spill the beans as to be why Fuji X-Trans is good for B&W in part II of this article?

[Dave Millier]

I'm not really sure I agree with your comments, criticisms or cures in this piece (actually, I'm pretty sure I strongly disagree). And the post was a bit too Ken Rockwell-ish in style for my taste, although I'm sure that was at least partly tongue in cheek for effect.

But rather than get all DPReview and argue pointlessly, perhaps I can suggest there is an opportunity for a follow up post or two in which you could illustrate/demonstrate with examples, the things that you believe afflict digital B&W? It would be educational for those looking to improve B&W renditions and perhaps the collective wisdom might even lead to some new solutions to the faults you see.

[William]

Those who choose to use in-camera JPEGs probably will benefit from your advice. However in-camera exposure and JPEG rendering parameters could be counter productive.

I agree with Larry Gebhardt.

The analog sensor response is essentially flat. A raw file is essentially flat. What happens after that is up to you. There are no fundamental reasons a properly exposed digital photograph can't be rendered in B&W to "match how our eyes see". Unrealistic shadow regions and depressed midtone renderings are just two of many options. In my view these are mostly bad habits. The high signal-to-noise ratios of shadow regions permit them to be selectively pushed without technical consequences. This doesn't mean they have to be pushed. In fact realistic shadow region rendering is an important aesthetic tool.

Intentional underexposure has significant disadvantages in digital photography.

o The analog signal-to-noise ratio is lower than necessary. In contemporary technologies photon (a.k.a quantum or shot) noise is the main contributor.
o The highest possible dynamic range can not be achieved (DR depends directly on SNR).

By the way, whenever ISO is set higher than it must be to freeze motion or achieve adequate DOF, the sensor is needlessly under exposed.

Purposeful overexposure maximizes the analog SNR and DR. The only proviso is to avoid overexposure of all thee channels for objects that are vital to the photograph's purpose. This means reflections, interior lamps, bright street lights, headlights, etc. should be overexposed. Also, gross overexposure of unimportant highlights is not purposeful.

This may sound like the expose-to-the-right cliche. But it's not. Instead, the mantra would be maximize the signal levels when the shutter is open.

My cameras are psuedo-ISO invariant. So, I use base ISO, choose a purposeful shutter time and aperture and then optimize global brightness in post production rendering. Often I auto-bracket three aperture by -/+ 1/3 stops. I keep the one with optimum overexposure and delete the others.

Shadow region and mid-tone rendering can be optimized In post-production. Silver Efex Pro is just one platform that offers many convenient yet effective options.

[Timo Virojärvi]

The same unfortunately applies to much of current color photography. Harsh highlights with dark midtones are so common in Instagram for example that people are beginning to like it.

Years ago I had a discussion with a person who said she can tell apart film and digital black and white pictures and like you write there are telltale signs. My argument was that while it is the default, it doesn't have to be so. My first move in Lightroom more often than not is Contrast -100, Highlights -100 followed by a curve. It doesn't work for every picture and sometimes Contrast -100 alone is good, sometimes Hightlights -100 and sometimes -50/-50, but at least you see the highlight tones the behind-the-scenes curve had compressed. From the resulting flat picture where the highlights are visible I can get long highlight range and raise the midtones like you showed. Examples here https://www.instagram.com/p/BSbz1vnhNIg and here
https://www.instagram.com/p/BDyJw09uJih/

[Lothar Adler]

Again a really wonderful article, which is very necessary in these colorful days … .
I myself came to the finding too, to slightly underexpose my RAW exposures to surely save the highlights and then apply modest HDR in the RAW converter, so that the shadows come up again and they do have a lot of “footroom” under the carpet ... . This approach often leads to a much more classical B&W look than to expose correctly.

On the other hand, I’m also very much into camera scanning my B&W negatives, often combined with stitching (which leads to impressive results) and I did run a lot of tests. All in all I’m very convinced that this hybrid route contributes to better B&W tonalities than the average conversion of a digital color file, because the inherent curve shape of the B&W film, preferably the S-shaped type, is transmitted and preserved in this workflow, despite the linear reproduction of the sensor.
But I have to admit that the second curve of a photographic paper isn’t there, but at least the characteristic curve of the film come into play.

[hugh crawford]

An interesting exercise in lightroom or the raw converter or the raw filter in photoshop is to convert to B&W by lowering the saturation rather than clicking the B&W button. Then adjust the white balance ( yellow blue slider) and the cast ( green magenta slider )

The bulk of what the B&W film emulators do to differentiate between emulsions like tri-x and tri-x professional ( the iso 320 kind ) is right there.

[Jackson Bart]

I always assume if a digital photo from a quality camera doesn't look as good as you think some other camera or film looks, it's because you haven't figured out the right settings in Lightroom or you don't have the right plugin.

Of course, with tiny-sensor cameras like an iPhone, you get a lot of blown highlights.

[Michael Perini]

Articles like this really need Examples (of both good and bad) to back up statements.
Without more examples they are not really very helpful.
(I do like the one B&W picture included a lot)
Expose to the right, came from Thomas Knoll as a way to get the most out of Digital Data. It does not mean over expose, or blowing highlights.
True there are a lot of 'less than ideal' B&W conversions out there because lots of folks rely on presets. There is also a lot of beautiful B&W work out there done by folks who take the time to understand the available tools. In addition to curves the careful use of channel mixing goes a long way toward creating pleasing B&W tonality .
I understand you were being purposely hyperbolic, but in this instance I think it got in the way of your message.
But that's just my opinion.........

[Darlene]

I was taught to expose 4x5 transparency film at ±1/2 stop brackets, not just for exposure, but for additional file copies as well. When I started shooting digital, I noticed its similarities to the transparency exposure latitude. I have been using your 2/3 recipe for a while now. I always shoot my Sigma Merrills at ±2/3, but with the Fujis, I find ±1 stop works well. I like having additional file copies. ☺

[Leon Droby]

Mike, curious what you think about Fuji's Acros simulation?

[Glenn Allenspach]

A couple of years ago, give or take, I came to the conclusion that the Old Saw "expose for the shadows, develop for the highlights" was great advice for film. Just not so much for digital. In fact, flip it around and you're better off, for color or BW.

[Matt]

The way I elevate midtones is by playing around with white balance and color channels.

By increasing color temperature or by increasing the orange channel, skin becomes lighter. Skin tone in digital b&w is often too dark. Using this method, white skin becomes light grey again instead of muted gray.

Another trick I am willing to share, is my 'secret S curve' :) Instead of randomly picking points in the curve, I divide the tone curve into logical parts: value 32 are deep shadows, 64 are mid shadows, 128 are the very midtones, 192 are highlights. The values I use are the following.

INPUT -> OUTPUT
32 -> 16
64 -> 48
128 -> 160
192 -> 224

This creates a strong S curve that does a few things: it lifts the midtones as Mike describes, but it also adds punch by darkening the shadows and increasing global contrast.

Usually, this needs to be combined with a toned down exposure, which can be as little as -1/3 stop or as much as -1 stop.

[Oskar Ojala]

I've read what Ansel wrote, done some analog BW, done a fair amount of digital photography and at times I'm not 100% sure what exactly is referred to in this article; examples would help a lot to clarify!

While there are certainly differences in sensors, modern full frame sensors have excellent response and are low noise to the point that I feel post processing is the key to getting the look of a picture right rather than the exposure itself (provided that the automatic exposure didn't do something weird). What's so special about Trans-X? I never figured that out and I consider myself technical.

[Len Metcalf]

I set exposure using the histogram in the electronic viewfinder or in live view. I always check my rgb histogram afterwards when exposure is critical. This is the huge advantage of electronic viewfinders, let alone the o e of seeing in bw when your framing up.

The other thing is treating digital like transparency film. Personally I love to work in soft light with a limited contrast range. Mist, rain, in the shadows etc.

Digital / film. Two different mediums in my humble opinion. Different workflows and methods. Treating them the same is the error.

I have lots more to say on this issue, but that will have to wait until another time. Thanks Mike.

[Tsukasa]

Now I really want to know why the X-Trans does better B&W. Will you go into that in a later post?

Very relevant for me because I'm planning to buy a mirrorless system, having sold my Pentax DSLR outfit, and leaning toward Fuji but also waffling between Sony and maaaybe Olympus. And B&W is all I do*.

*poorly, but still.

[Earl Dunbar]

I agree with:

1. Use an XTrans sensor. It's not perfect for b&w JPGs, but it's close enough for jazz and government work. To really get a better b&w final image, I work with the (colour) RAW file and

2. Lift the mid-tones. Actually, most often I make an S-curve, even if the "S" is quite subtle.

3. With film I tend to zone meter to capture adequate shadow detail and develop for the highlights. This leads me to have no problem with ...

Rodinal. Rodinal is my go-to developer. I absolutely love how it treats mid-tones. And at 1:50 (sometimes 1:100) it is a compensating developer and I rarely have problems with blowing highlights. Sometimes I mix Rodinal and Xtol and I get the "best of both worlds". It especially works magic with TMax 400 (TMY-2) and (the much lamented) APX100.

I find that scanning b&w negatives (Epson V750 and Epson Scan) works quite well. I admit a "real" darkroom/optical print is the optimum, but I currently have no darkroom. I've been having Peter Thomas at https://editionsprinting.com make my digital/inkjet prints, and I've been more than satisfied. After printing one of my shots made with Rollei RPX 100 (my new favourite medium speed film) on 6x9, he remarked "Maybe I'll get out my RB67 again." That gave me more joy than the print itself.

I have several rolls of expired Agfapan 100 (pre-APX) in 120 and I'm chomping at the bit to shoot it. Anyone jonesing for shooting old/"expired" sensors (other than the Olympus E-1)? Didn't think so.

[Jim Natale]

A point that a few other comments have made a nod toward is that the capabilities of contemporary software makes it possible to achieve very fine results with B&W. Specific brightness levels can be manipulated via the TAT adjustments in Lightroom, the control points in Silver Effects Pro II, the Light EQ adjustment in ACDSee Ultimate 10, etc. Tone mapping, which when overdone in color can bring one to (and beyond) the point of nausea, works wonderfully if used with subtle restraint with a B&W file. As have many here, I began with Tri-X and a wet darkroom. Digital darkroom tools make it so much easier to achieve various looks. To wit: http://www.jimnatalephoto.com/gallery.html.

[Adrian Malloch]

Hi Mike, an interesting post on a subject which had long vexed me. I spent years in the darkroom. I felt that I intuitively understood what a black and white photo should look like, yet struggled long and hard to get that look with digital. The breakthrough came about 10 years ago when photographing a snow-clad mountain against a bright but stormy sky. From then on, every (raw) photo was exposed for the highlights regardless of the brightness range or contrast. For my tastes, black and white seems to drop into place just right with those files. Here's a link to 6 photos from a series I shot last week. No significant processing heroics apart from a tidy-up crop, maybe a little edge burn and a mid-tone lift: https://adrianmalloch.photoshelter.com/proofing/G0000P7eSlcT2dWo Please feel free to publicly criticise, praise, condem, post the images or anything you feel like about them. It would be really helpful for me, and perhaps for T.O.P. readers too.

[Steven Willard]

I'm fining it interesting that all of the comparisons so far have been digital vs gelatin silver as if those are the only black and white choices that exist (or have ever existed), leaving out all the other processes that we could be comparing digital to. My point is why do we insist on comparing digital to anything? Or if we are, why not compare it to daguerreotypes, or tin types or salt prints or albumen prints or calotypes, or even Carl's lovely palladium/platinum prints? They are all black and white processes, each with it's own particular look, and I haven't listed anywhere near all of them.

Why do we insist that in order for a digital print to be "good" it should look as much as possible as a gelatin print?

Just wondering.

[Jonathan]

As a side note i have found the Panasonic Leica 25mm f1.4 lens to be great at rendering B & W images, i usually use it with a an Olympus EM1 set to monotone with a few tweaks in camera and use the JPGS for black and white. The Panny Leica to my eyes seems to bring out tonality a little better than some of my other lenses. Also having rarely shot with B & W Film and only digitally its great reading articles like these, look forward to playing around a little in post now :) another aside, the other Olympus i own the EM5 mkII seems to render smoother gradients/transitions when used much the same way as the EM1 which the dxo figures seems to back up

[Eric Peterson]

Great dialogue. I still shoot film, occasionally still make a print in my 19th century darkroom. But a digital print has become the new default. For an older negative, I always reference the darkroom print, and for new work on film I just feel lost without a traditional contact sheet.

[John Camp]

Really good post. Thank you.

[Chris]

Regarding the need for "bad" examples, I'd suggest seeking out individuals that would be amenable to the idea of offering up their work for the exercise. I, for one, volunteer. The knowledge gained would far outweigh any embarrassment.

[Herman Krieger]

Call me quaint, but I'll stick with my black and white film camera.

[Moose]

"But Merrill Foveon B&Ws are gorgeous. The transparency and smooth tonality cannot be achieved by any amount of post of a Bayer file."

There's an interesting question. The HD mode of the top Oly µ4/3 bodies moves the sensor so that each pixel location is sampled by a sensel of each color. In effect the same as Foveon, although achieved quite differently.

So there is no demosaicing of a Bayer array, or any array. In color, shooting HD, then downsampling to the normal size results in smoother, more accurate color and quite a bit more fine detail.

I wonder what that would do for B&W.

[Gary Nylander]

When it comes to shooting black and white I still prefer film. I have not cared all that much for black and white images from digital files, although I must say that there are photographers out that who's work I greatly admire who make black and white images from their digital cameras.

I have been scanning my film since 2004, (mostly 4 x 5) directly from the negative, rather than making a print and scanning that. I have simplified my work flow, my darkroom is downsized, although I have enough space to make contact prints. However I prefer to scan my negatives directly. There is an option that I have discovered with my scanning software,and that is to scan my negatives using their "raw" setting, I avoid these so called presets of various black and white film types. I then use various curves in different parts of the image to achieve my "vision" in Photoshop, I don't bother with layers, thus creating something that I feel is every bit as good as I something I could have made in the darkroom. Film (and print paper) tends to have a nature "S" curve while digital has a straight line with no curve.

At least thats what works for me, each photographer will have to discover what best works for them. BTW the all the work on my website are scans directly from negatives.

[sebastel]

another volunteer for B&W critique here.
in my flickr stream, you'll find plenty. a lot of that is scanned from (rodinal developed) film, but you also find black-and-whityfied digital colour ...

[Ernest Zarate]

Ahem...

"As I recall, one of those two authors stated that an important darkroom tool was to have a few high quality B&W prints available to view as one was printing. These prints would serve to calibrate the photographer's eyes as to what a "good print" would look like, to give them a sounding board by which to judge their own efforts. A corallary to that is that perhaps many current photographers, who have limited, or no, darkroom experience simply don't know what a good B&W print looks like.

"Maybe your next print sale might include a few sample B&W prints, at a reduced price that digital photographers could pin to their walls next to their computers and printers so they could judge their efforts next to some solid, well made samples."

[Nigli]

Ditto what Chris said. It's only going to help me get better.

[MikeK]

Mike, in response to your response to Oskar Ojala's partial comment – what if some of us readers submitted a B&W picture and you could critique it freely? We could also supply the Raw file, so you could make your own interpretation. I've taken to liking B&W a lot more in the last few years, but, being essentially entirely self-taught and untrained, I often wonder what a more visually literate person would make of them (I would fully anticipate you may find them rather unpalatable).

It's probably an idea full of its own problems (not least the hundreds of pictures you'd likely receive), but it would really be interesting to see what you make of them and it would be great learning experience for a great many of us to see a real world example.

And I'd promise not to take offence.

[Oskar Ojala]

Mike, I understand your conundrum. While legally you would do nothing wrong, it would place the spotlight on someone for doing things badly.

Would it then be possible for you to take a few photos, convert them to BW in ways that demonstrate common errors and also show the way you feel things should be done? In theory this could be easy, as subject matter isn't important per se as long as it contains sufficient amounts of tonality in critical parts of the brightness spectrum.

[Bob Johnston]

In days of yore when we used film most black and white was just as bad as it is today -'muddy' midtones, blocked shadows and/or empty highlights just as we see now. You had to learn not to underexpose and overdevelop, which is what a lot of people did. You had to be able to test for the true film speed and tailor the development to suit and then learn how to print. In the same way it is necessary to learn how to treat a digital image correctly,and many people don't know how. With the exception of Fuji and probably Leica most black and white in-camera conversions are also pretty dire.

[Anthony Shaughnessy]

Surely you must have some B+W pictures of your own that you could use as examples? Can't you try deliberately doing a bad B+W conversion to show us what you mean?
Anthony

[Kenneth Tanaka]

Mike: Now that I have some time I offer the following comments to this article in the spirit of constructive debate.

1. Is your opinion of digital b&w imagery really based on prints or on screen images found on social media sites? That is, how often do you really see digital b&w prints?

2. I second Steven Willard's excellent earlier comment to the effect: what version of "black and white" are you using as reference? Of course silver gelatin is the most commonly-used medium in the past half-century. But I know you know that there are many other styles of monochrome printing, each with its own "look".

3. The tonal curve adjustment you recommended essentially reduces contrast. Is excessive contrast the "pond scum" that really bothers you? Understandable. Contrast is arguably a primary visual tool of emphasis in b&w imagery. Contrast is how you yell in b&w. Excessive or careless contrast is like the CAPS LOCK of b&w.

4. And finally, I wonder if the true bogie here isn't actually the digital medium but rather the general public aesthetic? Careless excessive contrast and excessive color saturation -- photography's CAPS LOCKS -- are hallmarks of undeveloped eyes. Never in history have we been able to see so much amateur work, which might lead to the impression that the digital medium is to blame. But I greatly disagree. In fact, today's digital medium offers us the potential for an order of magnitude more subtlety in both b&w and color than the analog/chemical era ever could. But you won't find much subtlety in social media sites.

-----
Addendum: One of the delights of my recent life has been my involvement with an international exhibition project of Japanese post-war photography titled "Provoke: Between Performance and Protest". (The catalog, a wonderful simulation of deep gravure by Steidl, recently won the prestigious 2017 Kraszna-Krausz Foundation award for Best Photography Book.). So much of the work in this show represents CAPS LOCK intentionally applied toward messaging. I offer this a an example of chemical-era b&w photography that used contrast and abrasion as visual tools.

Sorry for the lengthy comment. I finally had some time to jot my thoughts on this good subject.

[Edward Fritzen]

I'd be more than willing to provide examples of bad black and white conversions. I'm sure I must have many such examples and I might learn something in the process.

[Dori]

"As a normally courteous person I have trouble with that idea."

There is a really easy fix, solicit photos from users willing to be subjected to it. I know I wouldn't mind tagging my flickr photos allowing for TOP critiques.

[Andrea B.]

I have a set of Labsphere diffuse reflective standards for use with my reflected UV photography. The 99% white standard is used for white balance (of false colors) and for setting a white point, if needed. The 2% reflective black standard can be used to set a black point, if needed.

I'm wondering if these targets would be useful for setting the tones in B&W photos? The assumption would be that reflectivity corresponds reasonably well to brightness in an HSB model? No/yes? If so then the 25%, 50% and 75% standards could be used to set the midtones on a curve. I'm def in agreement with your observation that midtones get smunched in digital conversion because I never get 25/50/75 brightness when photographing those grey standards in visible light.

[Andrea B.]

P.S. Mike, you can make your own bad example of a B&W photo to use as a teaching aide. We will all know you did this on purpose.

[Jackson Bart]

Kenneth Tanaka wrote: "Careless excessive contrast and excessive color saturation -- photography's CAPS LOCKS -- are hallmarks of undeveloped eyes. Never in history have we been able to see so much amateur work, which might lead to the impression that the digital medium is to blame."

The jacked-up look garners more likes on photography sharing sites like Flickr. And people want likes, they want their photos to reach the most-popular page and get more views, so Flickr and other sites like that TRAIN people to jack up their photos.

[John Ironside]

Thank you, Mike, thought-provoking.
As an alternative to the kind offers of samples made by other readers, I'm sure you could set up and take some of your own which were effective examples of what you don't like, and perhaps contrast them with better versions of the same image.

[Geoff Smith]

Will second (or third or, uh, "nth") the original A7 Mk I for B&W. The term "poor man's Leica" gets thrown around a lot but this seems to come as close as anything (modulo, you know, having an actual _rangefinder_). Suspiciously similar IQ to the CMOS sensor digital Ms. I had a 50mm Leica R Summicron that I struggled to focus manually on OVF DSLRs and this little guy makes it easy... and fun. And the shallow flange E-mount provides a host for basically anything on a helical ever made. So my Pentax M42, Canon FD, Russian ebay finds, etc. all have a home now. Not my only camera but probably the one I'm using the most atm.

[Brian Stewart]

A few comments coming to this a bit late:
(i) My impression and limited experience is that ETTR was a very useful too in the early digital years, but less so now given the advances in sensor technology. Yes, Larry Gebhardt is right - that the exact same theory still applies - but the practical benefits are much less in my view. Which I think is one of your points, Mike.
(ii) ETTR is tricky with Fuji X anyway, since it lacks an RGB histogram, even an imperfect one based implicitly on JPG. I don't understand why Fuji haven't implemented RGB histograms - this omission is my one big outstanding gripe. I wonder if your -2/3 recommendation is based partly on this issue.
(iii) I think Kenneth Tanaka has nailed it on the head - photos with 'pop' are popular, and that means excessive contrast and, for colour work, colour saturation. Digital images and prints that seek to portray subtlety and sensitivity will tend to be regarded poorly when viewed alongside those with pop. There's also a current tendency towards images with a lot of pure black in them, i.e. with little shadow detail. It's a style and personal preference I guess, but I can't see the point.
(iv) I wonder how much of the poor online B&W that you see comes from uncalibrated and unprofiled screens. A lot of screens are set too bright (and too cool) for proper soft-proofing, and the same digital image will look a fair bit darker on a properly adjusted screen. This is an issue I often confront when printing for other photographers, and I have to resort to either to viewing the image on the photographer's own monitor to see what they want, or calibrating their monitor for them and teaching them how to soft-proof. Screen calibration hardware is not all that expensive nowadays and seems more widespread, but nowhere near widespread enough IMHO and experience.

[Jerry Walsh]

I just started converting color to B&W using Photoshop Elements. After reading these comments I have a LOT to learn. Any generally accepted books out there that may help me with my growth? Thanks.