In digital photography all images are captured in colour. I think every digital camera offers a black & white or sepia conversion feature in-camera. This post is intended to convince you never to use it. The images were kindly provided by Passerine.

Black and white photography is a vast and demanding subject. It’s not one I feel qualified to talk about authoritatively, but I can offer some ideas. They are mostly lifted from Leslie Alsheimer’s book, “B&W in Photoshop CS3” and from Michael Freeman’s seminal “The Complete Guide to Black & White Digital Photography”, both of which I heartily recommend. I’m hoping other, more skilful practitioners of black and white digital photography (hello, Peter et al.) can contribute.

Subject matter and approach

You don’t need a high-end camera, but it really helps if you shoot in RAW. JPGs lose detail when edited. The conversion to B&W should be done in post-processing, never in-camera, but I’ve found I have to be a B&W mindset to capture the image. Colour photography can feature two shades of green (or red, etc.) that complement each other. In B&W, they just disappear. It helps to expose for the shadows in B&W and for the highlights in colour – two very different approaches. In B&W, texture, shape, profiles, lighting and patterns emerge far more prominently when there is no colour to ‘distract’ you. Andri Hery said: “To see in colour is a delight for the eye, but to see in black and white is a delight for the soul”. Trite, but true.

Grappling with black and white

This is an example of what a compact can do in-camera as a black and white conversion.


I think a more effective approach is to shoot in colour and post-process. Unfortunately, this image is not available in colour, but Katarina was kind enough to provide a similar image in colour. I would like to use ACR or Lightroom to post-process this, but the interface may be unfamiliar to some of you. I’ve decided to use tools available in Photoshop CS5 (equivalents must be available in your image editor), but I strongly recommend using ACR or LR for black-and-white post-processing.

To begin with, the camera has produced a rather dark image.



Use the brightness / contrast (bad), levels (better) or curves adjustment layer (best) to add punch and then a hue / saturation AL to add colour. Don’t worry about how it looks in colour; it’s the B&W result we want. The advantage of using adjustment layers is that you can, ahem, adjust them later.



An easy approach when using the levels AL is align the tabs on the horizontal axis left-hand side (shadows) and right-hand side (highlights) under the main part of the histogram “bell curve”. Be careful with the levels AL. It is not a subtle tool. I think of it as trying to mend a watch with a hammer. In the above example, it’s introduced chroma noise and further softened the image.

Now it’s worth trying a black-and-white conversion. In Photoshop CS3 had a rather clunky mono converter, but the point is to adjust the colour channels to achieve an effect you find pleasing.



This is where my B&W skills are clearly lacking. The image is softer and a lot of noise has been introduced. I am going to use a noise reduction technique explained here, which involves using the despeckle filter on each colour channel. This further softens the image, but we might be able to recover the image later. If this has taught me one thing, it’s the usefulness of shooting in RAW.

Below is the final sharpened image using a rather drastic technique: first merge all visible layers. Then copy the single background layer and change its blend mode from “normal” to “overlay” or “hard light” (your choice). Then Filter > Other > High Pass at as high a pixel count as you feel comfortable with. When working on images I set it at 10-15. Here, I went up to 30.



This is the final image:

Compared to:


I am fully aware that this is a very subjective matter. In fact, I think I prefer the soft and rather gentle original. My point is that you should get to make these choices, not your camera. Post-processing is, after all, what allows us to add our own view.

Please feel free to improve on this thread!

There's something I don't understand. Surely, if I'm shooting in RAW I'm capturing all the information that the camera has. That being the case it wouldn't actually matter what setting the camera is on as you can always convert to a different setting out of camera later. Or are you suggesting that there is RAW and RAW?

Not entirely. RAW means only that noise reduction, colour saturation (Canon: "picture style") and sharpening algorithms have not been applied in-camera. Your exposure and aperture settings still apply. The advantage of editing RAW files is that the original remains untouched; all modifications are added in a "sidecar" file.

If picture style is not applied in camera, surely it doesn't matter which style I choose to shoot in as I can change the style to whatever I want without loss of quality later on. No?

@ Jon: I don't use DPP for post-processing. I use LR3 and the histogram does change significantly depending on which RAW conversion method I use. Not sure if I answered your question, so I'll assume that by 'loss of quality' you mean 'loss of detail' or clipping. The way to ensure maximum detail retention is to shoot in Adobe 98 colour space (because we haven't got anything better) and RAW and convert into 16-bit ProPhoto colour space in LR3/PS CS5. The files are humongous, but sometimes worth it.

Sorry for jumping in.

@Jarvo:
No, picture style settings do not affect RAW information. Even if you set style to B&W, the RAW file will still be in colour.

@ Rokas: no, please jump in!

It's my understanding that picture styles are combinations of saturation, contrast, sharpness and other settings. While they can be applied in post-processing (both DPP and LR have them), so the RAW file remains untouched, I think that they do affect the RAW image when set in-camera. I've just checked my Brian McLernon Field Guide to the 5DII and although he doesn't say so outright, it's inferred in the text.

I'm not a fan of picture styles. I prefer to leave my Canons on "standard" and apply changes in the post-processing.

I might be totally wrong here, but my understanding is that RAW is RAW, except setting a picture style adds some flags to the output file and DPP adjusts the file accordingly, while "non canon" raw converters ignore these flags.

To Stéfan and Rokas, thanks very much for your explanations. I'd love to tell you that I'm wiser, but in reality this stuff still makes my head spin ;-)

@ Jon: If you prefer to see the results immediately on your camera in black and white, that's no problem IF YOU SHOOT in RAW. The setting on your camera (to shoot in black and white) is independent of the RAW data. Each photo taken will be black and white on the displaying screen of your camera, but all the color information is still present in the RAW file. So, you can still create your own black-white variant. And while you are shooting a picture you will have a good idea of how your photo can look in black and white.

Thanks Sandra. That's what I thought all along actually.