This post was first published in my old Blogger blog (which is now offline). The articles I find noteworthy will appear here in the new blog over time, dated back to their original date.
I recently bought a red filter to try it with digital black & whites. Why? First, digital black & white has the stigma of being “just another post process manipulation” and second, post processing a color image into black & white can involve hundreds of little set-screws to tweak and fiddle with (ok, I haven’t counted, and it’s probably not hundreds, but still, a lot;-). This can be time consuming and bothersome – heavy manipulations can lead to banding or pixelisation – and specialized black & white plugins like Silver Efex add overhead to the workflow, and I generally like to keep it as simple as possible.
So after a friend who often uses film mentioned that he likes to combine a red filter and a polarizer for “that dramatic look” of his black & whites, and watching the interview/about video with Michael Kenna (whose work I much adore) where it can be seen that he uses a color filter (red or orange), I thought I might as well try that with digital black & white, in an effort to make the process simpler and give up some control in the process. Not necessarily to save time, but maybe to add some sort of veracity to my black & white images (even if it might not be visible in the final image to the viewer).
In the following paragraphs I will describe my experiences so that you may decide if you want to try that approach. Let me tell you my conclusion right away in case you do not want to read the entire article now: it is probably not worth the effort in the end. Those who do want to know the reasons now may continue reading. 🙂
The first thing worth noting when using a red filter is that the exposure times are getting way longer: depending on the situation, something between 2 and 3 stops. Combined with a polarizer (which costs another 2-3 stops) we’re talking about exposure times being 4 to 6 stops (!) longer than without filters. Using higher ISO and/or a stabilized lens, or better a tripod is necessary, most of the time, even in daylight situations (in the shade for example, or when it’s overcast).
The photos in this article were made with a Fuji S5pro – the camera is far more forgiving, highlight-exposure wise, making it a nice choice for using it with a red filter, since the red channel seems to be particularly problematic for most digital cameras (if you want to make a photo that contains a lot of vivid reds, it might be a good idea to dial in a -1 exposure compensation to avoid blowing out the red channel, for example).
I had set the camera to black & white so that the embedded preview JPEGs (which are the only thing you’ll ever see on the camera’s display, btw.) would give me some sort of a first representation of what the final image might look like. It turned out however that the red filter throws the JPEG engine off pretty much – it rendered a rather dull-looking image that totally lacked contrast and “pop”. After a first evaluation and hints by a friend I thought that the white balance might be causing this (I had set it to auto), but even after I manually set it to ~5300K (daylight-like), the results didn’t change.
But thankfully, the imported raw data didn’t show any of these problems – I used center-weighed metering (and overexposed one stop to operate the Fuji in it’s “sweet spot” and to exploit it’s dynamic range). Despite the red filter, I was able to use a totally normal metering, and the raw data was not dull and flat.
You don’t have to use Lightroom’s “Black & White” develop option (and mixer) since the recorded data is monochrome already – it’s just red instead of black & white (sic!). All that is necessary to convert to black & white is to completely desaturate the image by dragging the Saturation slider to -100. That’s a great simplification over using the black & white mixer (where Lightroom’s Auto mix hardly if ever provides an interesting starting point, anyway – the presets that Adobe provides, especially the “filtered black & white” ones, are much more useful for converting normal color images).
In reality, Lightroom can still differentiate a little bit between red, orange and yellow in the data captured through the red filter – so using the normal black & white approach with the mixer gives a little bit more control and fine-tuning (and is indeed the method I used for the two images above). The latitude for color-channel adjustments is quite small however, the majority of the post processing concentrates on the levels (via the Tone Curve and everything else that influences it, like Highlights/Shadows, Whites/Blacks, Contrast) – that’s the part which makes using the color filter quite attractive. All that fiddling with the black & white mixer, the white balance, maybe even the camera profiles, adjusting color noise control when things get blocky – gone, not necessary.
However, using a color filter essentially only means throwing away (the color) information, and you can do exactly the same in post processing. Here’s a gallery with two images for a direct comparison – you can click on the thumbnail to open the images larger in the slideshow module, and switch back and forth between them.
Can you tell which photo was made with the red filter, and which is the digital conversion? The most obvious difference when making the two photos was the exposure time: the version made with the red filter is a 5 second exposure (at ISO200, f/8) and the normal one is a 1 second exposure (same ISO and aperture, of course). This is the ~2-stop disadvantage of the red filter that I already mentioned above.
Upon careful inspection, you’ll probably notice a lack of detail and definition in the “red filter” photo, even in the web version. It has some sort of a “soft glow” on the detail level. The second photo was digitally converted by simple comparison with the first one. The “soft glow” is missing, but with a little bit of negative clarity applied in Lightroom (or ACR), it’s easy to add some of that in Lightroom. Adjust the sharpness, reduce the detail extraction, and the result should mostly be the same.
Here’s a closer look at this softness (click on the thumbnails to open them in the lightbox). Sharpening for both of these images is minimal (Lightroom defaults).
This may be a desirable feat for a not-so-digital look of the images, but then again… it’s easy to take some of the clarity and crispness of the normal version away by reducing clarity and sharpening – but it’s pretty damn hard to add it to the red filter version, where it wasn’t there in the first place.
Verdict: I think that the shortcomings of using a red filter (longer exposure times, less post processing latitude, loss of detail) outweigh the benefits (faster post processing, more “natural” black & white). It was fun and worth trying, nevertheless! The most interesting realization came with the look into the viewfinder: devoid of color, I noticed that it was much easier for me to pay attention to details that would influence the composition.
The same goes for switching the camera to black & white mode: it is so much easier to evaluate an image composition-wise, if the preview in the display is black and white. Unfortunately, that’s where two different interests collide: it doesn’t make sense to use UniWB for the best exposure control with a linear tone curve and then render the preview as black & white, rendering the color histograms (that are crucial for using UniWB) useless… oh well, can’t have it all. 🙂