« Song Five (OT) | Main | Song Six (OT) »

Monday, 29 December 2014


Feed You can follow this conversation by subscribing to the comment feed for this post.

When I do post processing, my perception of good contrast, brightness/darkness, etc. changes (creeps) as I stare at images for a while. How do you recalibrate your perception of what "looks right" periodically during a session?

[I notice that too—sometimes I "go after" certain effects such that I get misled and get off track. For instance, I'll be loving the richness of dark blacks and I'll way overdo it, making the image much too dark.

The best way to cope with this is to set the image aside and come back to it later. That's a common workflow tip in darkroom printing--you print it once and then come back to it and do it again a few days later, or make a good workprint during one session and then make the final print the next morning.

If you can't do that, then keeping a few "reference prints" handy could help--these would just be excellent full-range prints that have held up for you over time, which you can go look at to recalibrate your eye after looking at the image you're working on for too long. --Mike]

This was very useful and motivating for me, though I still prefer doing my local adjustments in Lightroom. But I should try Photoshop more. I'm 16 prints into my OC/OL/OY project, and so far so good. Nothing for the Smithsonian yet, but it keeps me moving on two fronts: in the chair, and behind the lens.

+1 on Mike's suggestion to study the printing vision and technique of photographers you like. Being introduced to Roy DeCarava's work exploded my understanding of how to print some of my bodies of work. It'll be different photographers for other people (and for other bodies of my own work).

I've found, especially for people who never printed in the BW darkroom, that reading Tim Rudman's "Photographer's Master Printing Course" can be extremely educational. If you're printing in the digital darkroom, the specific techniques aren't going to be that helpful, but throughout there's great inspiration on how to take the "SOOC" image to somewhere that expresses something for you.

How about some props for the green filter in B&W. This has been a favorite with portrait shooters for generations because of the nice skin/lip tones. The green is also very nice for landscapes because it lightens the green leaves and helps the camera see into the shadows of the foliage and also makes nice tones in the sky. For outdoor portraits it just makes everything better.

Thank you!

This is probably your best style of article - I've actually never done B&W conversion since I went to digital, and now I know why... There are a lot of options now, and it takes serious effort to do it right. Very worthwhile - in what types of images?

An obvious companion article is selecting and shooting images for B&W. I usually shoot for impressive color, with Tri-X I looked for structure and shadows. No doubt you could explain this better...

Personally, I'm having too much fun with color, but I love to know this stuff - it will give me something to do when I can't get out to get new images.

Great post. I think sometimes we all underestimate the interest others might have in things that we do that we take for granted - but came from decades of experience. In saying you are not sure how much help you could be, you are guilty!:)

As primarily a view camera fanatic, I can't help but comment on your "image geometry" thoughts! For those of us who either tilt their camera backs, or use shift lenses on smaller formats, completely vertical verticals look perfectly correct, and convergence looks wrong. A case of "beauty lies in the eye of the beholder!"

A propos your comment Jeff, do you find yourself with different degrees of discernment twixt morning and evening? I think I'm more discerning in the morning and the effect seems to be independent of Chardonnay, food, seasonal variation etc: basically I'm a harsher critic in the a.m. and yet I often do a great deal of editing and post-processing in the evenings and, worse, post to Flickr etc. I then have to remove the photos the following morning.

Black line borders -- Do you go back to the tradition of filing out a 35mm neg holder until the space is just a bit larger than the exposed frame, then printing the rough black edge, sometimes with camera-induced shape marks at the edge of the exposed frame? This was a testament to "do not crop" purism.


Something easy to overlook, that I never expected before I took one, is that some pictures should be cropped differently for color vs B&W. These would tend to be scenes with larger bright areas towards the edges. In color they may look good while in B&W those areas can be just bright and annoying.

Excellent post.

I understand what you're saying about image geometry, but non-vertical verticals and non-horizontal horizontals drive me crazy because I spend so much time looking at paintings, in which the verticals are almost always vertical, and the horizontals are almost always horizontal.

Because painters can't really give you the realism of photography (even in the most realistic paintings) I believe there's a tendency to emphasize those things that can provide the sense of reality, and that means especially the geometry.

Even in impressionistic paintings (by good impressionists) where lines are not sharply defined, if you lay a ruler along the "fuzzy" edges of buildings, you'll find that they're usually geometrically correct. Photographers don't have to worry so much about lens effects, because, as you say, people expect them, and the other aspects of photographs are realistic enough that you'll get your expected shot of realism.

The images that I find really disturbing are paintings and other hand-made graphic images which are produced by projecting photographic images, and using the projection directly, without correction, so you get lens effects in paintings.

Note to Jeff Hartge -- I'm a professional writer (novels) and when I'm writing, I lay down a first impression of a scene, and then I continue writing forward, only to come back later and work through the first scene again, adding, in Mike's terms, sharpening and local contrast, etc. There's almost a direct analogy between the two forms, at least the way I do them. It usually takes three or four passes, with some time between each pass, before it's satisfactory. You need the time between passes so you can see the photo/writing passage with a refreshed eye.

Have to disagree on correcting the converging parallels. If you were shooting this with the 8x10 for architects you would correct in camera. Why not do it here?

One approach that works for me is to start out by setting the saturation to zero and shifting the color temperature and tint sliders in Lightroom.

It's a quick way of seeing what the possibilities for tonal separation are in about two seconds.

For example shifting color temperature towards blue and tint towards green is sort of like the old tri-x professional that Kodak claimed was optimized for male portraits.

A comment not specifically related to this, but to the more general question of mono or color. I am vexed by this question, but I think I have found my own personal answer.
To me it comes down to "mood color". Even when printing black and white, I will tone for mood and often when printing color, the saturation and vibrancy of the color will be reduced to introduce a feeling, not an over powering beacon. I have never responded well to "pure" black and white for my own printing (although others work I do?).

If I correct verticals I will nearly always under correct by just a small amount. Making the verticals parallel result in the building looking top heavy.

I'm also a bit of a one for ensuring that horizons are spot on, to within a tenth of a degree. I can usually tell by eye if it's out by more than a degree, but checking at 100% with the grid overlay on (Photoshop) lets me get it right.

Not every photo is corrected for these two things, but if it is, it will be right. Each to their own, of course. In your example above I would probably not correct the verticals at all!

Mike, three years ago I found this tutorial for a B&W workflow in The Luminous Landscape and it really help me improved the quality of my images. Now I do basic correction in LR, then I search for the look I like in Silver Efex and go to PS to do the final but critical adjustments following George DeWolfe advice:


The perspective correction is a nice tip. I haven't really thought much about it before, but it makes sense. Amateur work can often be a bit stiff and lifeless (my own work included), perhaps because the photographer concentrates too hard on perfecting the composition and not breaking any rules. Many of the greatest photographs on the other hand have the looseness of a casual glance yet still somehow manages to communicate the scene perfectly.

Although those proceedings are rather complicated, the results speak for themselves. Nice job indeed - but wouldn't it be simpler to wipe the dust off your film camera and expose some Tri-X? ;)

[I can't shoot Tri-X (or P3200) at ISO 2000 and get this kind of quality, so, no. --Mike]


Thank's for picking this topic up. It's probably worth a series of posts, isn't it?

I am usually reluctant to manipulate the image in a way it would be difficult or impossible to do in a darkroom alone. Doing such it makes me feel better but I would never insist this the only right way. Just my way.

That said I also find that digital B&W images look more "juicy" if they have much higher contrast (globally) and when there is a good amount of grain visible. The grain is giving "a structure" to objects / planes which have no or subtle texture. I personally find it even more relevant for small web images, where the texture is usually not visible (prints are different - there you can usually enjoy subtle textures).
Yes, I know adding grain noise and manipulating with high contrast will affect the sharpness, details etc, but somehow it still makes them look closer to 35 mm negative prints, which I always liked.

A very good example (even exaggerated) of what I mean is Jacob Aue Sobol work (which I admire in a way): http://www.auesobol.dk/work/index.php?Sabine?image=2#photo


Useful insights, thanks. A couple of questions:

Camera manufacturers have their own "signature" color outputs. Are some easier to convert to B&W than others? Is there any one that yields better results than others or is it just a matter of having to work longer to get what you want?


Really nice post, Mike, with great insights into black and white conversion. Silver Efex Pro is still my favorite B&W conversion plug-in.

Regarding the comments about converting black and white for print compared to conversion for Web viewing. As most images viewed on the Web are JPEGs, it makes sense that an image that has the desired tonal range as a TIFF file in ProPhoto RGB would be too contrasty for viewing as a JPEG on the Web. The JPEG algorithm applies an S-shaped tone curve to images, which results in increasting the contrast, so your tip to increase the brightness is a really good one.

Regarding Fuji X-Trans files in particular: It appears you're using Photoshop (as opposed to Lightroom) for editing. After two full years of examining Fuji X-trans RAW file conversion, my current RAW conversion app favorites are 1) Capture One 8.1. This Phase One app does the best overall RAW conversion for any camera I use (Canon, Olympus, Fuji, Sony), but it is also the best I've used for Fuji X-trans files in particular. It's sharpening is the best I've found until the most recent release of Iridient Developer, but still does the best job with color, correcting keystoning and now supports up to 16 layers in adjustment layers. It's color performance, and ability to pull up shadows and pull down highlights w/o squashing the contrast curve is quite remarkable. In general, I find RAW files converted with C1 8.1 superior to any other editing app. Also C1's ability render skin tones from full-color Fuji X-Trans files is a wonder to behold. 2) Iridient Developer (version 3, beta 4) The latest version is really very good and as it works as a plug-in for Lightroom or Photoshop, makes it easy to achieve nearly perfect sharpening of X-Trans images, which Adobe apps still cannot achieve. I think Capture One does a superior and more accurate job of rendering color, but Iridient Developer is hot on it's tail. If folks are interested in the sharpening settings I've found most natural-looking for Capture One 8.1 and Iridient Developer, I'd be happy to post them here.

There's essentially no difference between fully corrected converging verticals and a wider shot with a straight and level camera, cropped to the same view. Is the first somehow wrong and the second right?

So Thomas Struth and anyone else who uses a technical camera is an amateur! I beg to differ ;-)......photography does not need recepies Mike, you should know better!

Greets, Ed.

I find these sort of posts quite interesting and useful. Although I have been doing black and white digital for about a decade now, I still could be better at it. (I hated darkrooms and never really learned decent printing there.) It's always interesting to see how someone who knows black and white well approaches it.

Only comment I have Michael is to try Iridient instead of ACR, try the demo and bring the same image up on both and you will be amazed. I use nothing else every day in my professional work. GB.

I find that printing is absolutely the slowest part of my skill set to evolve. This is in part due to the fact that relatively few photographers make any prints at all, and there seem to be no good books on printing. Perhaps you should consider holding a 2-3 day 'printing workshop'. I would sign up now.

A good write-up.

Regarding Jeff's comment, I keep a few good files handy on the computer so I can compare - files I know print well, or I know look good on the web. The longer I work the darker the photos get, unless I check myself. That was true in the darkroom and is still true for me today at the computer.

(I try to look at my web sites on as many screens as I can to see how the photos stand up on different monitors and settings. I aim for something that looks at least good on the majority of screens, not something that looks perfect to the tiny minority using calibrated monitors.)

As to "creeping up," I think we're using the same idea but describe it differently. One of my darkroom mentors used to say you can never get it right creeping up on it - you have to go past it and come back. Which sounds like what you're doing. It seems to work for me, especially with slider adjustments where I can go back and forth and narrow in on the exact point.

I've really been enjoying Peter Turnley's B&Ws that he has been posting on Facebook. It would be great if you could get his recipe as well...or is he just using the Leica Monochrome?

Best article I have read in a long while, thanks! And a peaceful new year to you all.

I would like to take a step or two back and ask you to consider the remark from various sources that digital images look digital. I have seen this regarding both color and black and black and white but mostly I think, black and white. The only explanation I can remember is that with digital the shadows are blocked up. I think maybe it is also too easy move the clarity and contrast sliders too far to the right. I was especially interested in your image of Butters without true blacks vs with blacks and higher contrast.

I think this is a topic that could use a knowledgeable article or several plus a mass of comments. I just don't understand and I really, really want to.

This was a interesting and worthwhile post, and I'm glad Mike Potter asked the question.
I use Elements 12, and after using Photoshop for years for just sizing and enhancing an image prior to using it as basis for a line illustration, I really know little about using software to make an image,- gray scale or color, look better.
My question is can Elements be used in the same fashion as Lightroom? I assume Nik Silver is a plug in; can it be used with Elements as well?
I would like to know more about black and white conversion. So anytime you want to divulge more words of wisdom, I'm listening!

Lots of good advice here and it's always good to get a reminder every now and again of methods you may have got lazy about or let drift to the back of your mind. However I don't like the use of the term 'amateur' here. You seem to use it in a derogatory sense. It only means unpaid and has no reflection on ability. Perhaps novice might be better. With apologies for pedantry as always.

Great post. Sad to say that if it is true that "genius is knowing where to stop" then I am sometimes dumber than paste.
I keep a reference image next my computer but it is not a good print. It is an example of my first venture into the world of Photoshop vandalism.
It is a nice enough image of Roughlock Falls in the Black Hills and the negative is probably twenty five years old. I have a straight silver gel print of it and that's a perfectly pleasant if not amazing picture. By the time I got done "improving" it the water looked like it had some kind of rash, the falls looked like three vertical fog banks and all the little ambiguities that give an image resonance had been tracked down and excised.
That was ten years ago and I hope I have developed some taste in the interim but I still keep that picture around to remind me of just how fast you can go off the rails even with the best of intentions.

For years I used faux film rebates to "hold" the absolute white areas bleeding off the print (usually an unfiltered sky). Then I grew out of pretending that my digitals and scans were somehow darkroom prints. A few seconds with the ACR vignette control introduces just enough tone to hold a delicate edge.

Also I never understood the need for BW plug-ins once you have control of the conversion using the colors and curves. I mean what else do you need?

And finally, toning or colorizing a BW image seems rather pointless when your website is on a bone-yellow background! We're already getting our dose of sepia radiation! Where is that decade-late modernization so that's so sorely needed?

;-p Happy New Year Mike!!!

Regarding the offer in the post from Stephen Scharf, I for one would be interested in knowing the settings he uses in Capture One 8 with the Fuji X-Trans files. I currently have version 7 but (so far) I'm so disappointed with the results I'm getting from my Fuji X100S images that I'm seriously thinking of selling it! My favourite default camera is still my old Leica Digilux 2, for colour or (especially) the out-of-camera B&W.

I would add two caveats to all of this. First, you have to understand that not all images work well in monochrome (I prefer this term to black and white, which I find to be misleading). Images that have great color contrast can often have very poor tone and/or luminance contrast. You won't know this unless you carefully examine luminance values across the image. If you try to work with an image with poor luminance contrast, you'll end up with absolutely tons of noise that result from pushing saturation values to their extremes as you try to introduce contrast into the conversion.

Second, and just a real sticking point for me, is to make sure you're not just creating a grey image. I thoroughly detest the washed-out look that many photographers produce as they attempt to get an "art" image. If the scene allows, try to get a true white and a true black; you can add drama and force yourself to separate tones more, giving the image more accutance and vibrance.

Having a border is something that I haven't done before, but I like the look. It seems that SEP reduces the image size to make room for true border rather than expanding the canvas size. That doesn't sound like a great idea to me. Mr Nik resizes my image using who knows what algorithm???

Just adding my two cents worth to the comments about correcting verticals. Mike: ".. if you made the verticals exactly vertical you'd lose the sense of the building looming above you and of looking up at it (plus you'd emphasize perspective distortion). It's just not a good subjective impression of the way we see." Mike is not alone in this view of course - many photographers will tell you something like this, or that verticals should be slightly converging.

Could it be that this preference for converging verticals is a product of photographers expecting photographs to look this way because, well, so many photographs look this way? It is certainly not true that my eyes and brain, when looking up at a tall building, translate the vertical lines into converging lines; I see vertical lines. I can have the sense of a building looming over me without perceiving the verticals as not-vertical. I suspect that everybody else's eyes/brain do the same - try it, am I right?

So the only time I see converging verticals is in photographs taken by cameras pointed up at the building. And the reason there is so many of these is of course that the vast majority of cameras have lenses fixed in relation to the image plane - from a low viewpoint you can't keep the verticals vertical without losing the top of the structure, as you can with lenses that shift. So, maybe, the aberration caused by tilting the camera up becomes the preferred norm because that's how buildings look - when you are looking through a viewfinder.

By the way, all correction of verticals in post will alter the ratio of height to width in the image, slight corrections are barely noticeable, severe correction of converging verticals will result in short, stout buildings that are a long way from the designer's intentions. I'm not sure that this is accurately described as perspective distortion.

One of my initial steps before taking it to the Silver Efex Pro is to "Flaten" the image in Lightroom before taking it over. (Sepia 19 @20% was a GREAT tip)

Happy New Year.

Another personal tip for the Silver Efex Pro... turn the darn "Film Types" off (it's ON/Checked by default it seems). It adds way too much noise/grain

Dear Mike (and others),

Sorry for the slow comment to this column; been celebrating the New Year.

(“For THREE days, Ctein?” “Yup… My life sucks.”)


On this business of local contrast enhancement, I think you may be missing out on a trick, Mike. Indeed, the most effective way to use it is selectively. Some photographs do benefit from overall local contrast enhancement, and, as I wrote some years ago most all photographs destined for printing benefit from a subtle enhancement: http://theonlinephotographer.typepad.com/the_online_photographer/2007/08/how-to-improve-.html .

But as often as not, you will want to apply it selectively, and in different strengths and forms to different parts of the photograph. The most controllable way to do that is with masked layers. Duplicate the image layer in your photograph (or create a new merged one from the existing layers if you've made a bunch of manipulations already, using the intuitive [ahem] keystroke combination splat-alt-shift-e). Apply your contrast enhancement tool to that layer. Add a mask to that layer and invert it to black. Paint in white where you want the enhancement to occur. It makes it very easy to control the degree to which it is applied and you can change your mind later if you want to apply more or less or add it to or remove it from another part of the image.

Create as many such layers as you need for the different degrees and styles of enhancement you want.

This will work with any of the enhancement tools from the simplest brain-dead ones like the aforementioned unsharp masking, to nearly uncontrollable powerhouses like ContrastMaster. In fact, the more powerful and flexible tool is, the more useful this approach is.

pax \ Ctein
[ Please excuse any word-salad. MacSpeech in training! ]
-- Ctein's Online Gallery http://ctein.com
-- Digital Restorations http://photo-repair.com


SEP doesn't resize the image to add borders. It just overlays the border on the image. No resizing.

It does this beacause it's really just a TIFF (or JPEG) editor. You give it a image and it just keeps the same size canvas.


You can modify the given film types then save them into your own preset so they don't add grain/noise if you prefer that style (I do). I have presets for the film types I like "TriX (no grain)".

This is part of a bigger tip of make a preset (or presets) of your favorite look ("yellow filter; Tri X; no grain; some midtone structure") and use that as a starting point. It's just like choosing a film and a paper to print on. Then fiddle with the controls to move around your favorite look.

Then add local adjustments.

In addition to applying effects to the whole image you can use Nik’s U Point® Technology (gotta love that name for their "control points"). The control points can apply an effect locally in a circle. You can control the position and size of the circle, the amount of the effect and the feathering at the edge of the circle. I'd use one or two control points on the girls in Mikes image to add structure to her without adding structure to the blank wall. No need to go to Photoshop if you don't wish to.

For the borders try using negative numbers as parameters e.g. you can get white and black borders. Not obvious :-)

Finally Google is finally getting more info on their web site (though it's still not enough). They can be helpful with understanding some of the less obvious features of SEP.

For example,


For more see:


The comments to this entry are closed.



Blog powered by Typepad
Member since 06/2007