The Internet seems to offer very little actually useful information on the M typ 246’s response to color contrast filters. One would observe that this is a product of authors who have little understanding of color theory and the use of filters with traditional film materials. Let’s supplant all of these blind-leading-the-bind pages with a totally new, semi-ignorant page on the subject.
What do we know about filters? Or not?
The first thing to confront is that we don’t know as much about filters as we think we do. Part of the problem is expectations formed by reading filter pamphlets. These contain such questionably useful statements as:
Especially useful for clear contrast between blue sky with clouds and foreground. Provides a natural tonal rendition. Often used for subjects at intermediate distances. (Hoya Yellow K2)
…has a very pronounced effect and darkens violet and blue very strongly, green quite strongly, and even yellowish green a little. Landscape and architectural photos show an increased, almost “graphic” contrast, while a cloudy sky may already appear dramatic. Because the skin tones rendered by the filter color are noticeably lightened in comparison with plant greens, this filter is often used in nude photography outdoors to increase the contrast between the lighter body and the darker landscape. (B+W 040)
The only statement that is almost universally true in filter literature is that the 022/K2/#12 yellow provides a natural tone rendition. This is due to the fact that it makes blue darker than red.
But as far as contrast goes, the reality is that the effect of a filter is profoundly influenced by the color of the light hitting the scene – as well as the color of the objects in the scene. If the predominant light color is complementary to the filter color, then the scene will simply be muddy – because the filter acts as a neutral-density filter. Shoot a red filter in the mountains, and sometimes you don’t cut through the haze. Sometimes you create one.
Applying “standard” filter factors usually has a similarly disastrous effect. To make filters work, you absolutely have to start with a scene featuring colors that look different in real life. Depending on what you are doing, the actual correction may be far less than the factor implies. With TTL metering, you may have to decrease exposure to get what you want.
Depending on the vintage and construction of a lens, the use of filters can also cause focus shifts. Green filters can cause front focus with some lenses; red filters can cause back focus. These are accentuated on digital, where the imaging surface is much thinner than on conventional black-and-white films. This may be the focus (tehehe) of a future installment.
By experience, we can at least understand the following with the M typ 246: filters can have a very exaggerated effect, starting with yellow filters. The following from two months of testing and feeling this out:
- A yellow-green filter (060, X0, Green 11) has small effects on exposure but puts more texture in people’s skin. Same as with film.
- A green filter (061, X1, Green 13) has a noticeable effect on foliage. Same as with film.
- The standard yellow (022, K2, Yellow 12) is still useful for correcting scenes to make color relationships work as expected. It actually makes a very good standard filter because it does not alter exposure much at all. It also has a noticeable effect on blue skies with the new Monochrom, much more so than on film.
- The yellow-orange (023, G, Yellow 15) acts more like an orange (040, Orange 22). The picture at the top of this article was taken with a Hoya G, though you would think it was something stronger.
- The deeper oranges (040, Orange 22) act like reds, and they are the practical limit of what you can use. Stop at 041. In daylight scenes, particularly in seaside settings with much blue light, these start to show diminishing marginal returns because they can affect most things in a scene.
- Red is almost impossible to use.
This is somehow not surprising in light of the fact that Leica only offers filters in light green, yellow, and yellow-orange.
Testing M typ 246 with filters, by the numbers
Do these anecdotal observations have any real basis? The one way to find out is to move away from infinitely variable real-world scenes to see what the sensor does with white light. So here is a quick set of tests:
- Sunlit and bright overcast conditions.
- 90%+ reflectance target.
- 90mm lens (M-Hexanon), defocused (lens at infinity; target closer than the lens’ close-focusing limit).
- Manual exposure keyed so that white clips (the obvious limitation is that setting white to 255 might diminish slightly the differences between no filter and some filter).
- Values are expressed as levels as measured in the center of the frame. This is to avoid having the results influenced by vignetting.
- Blue line is bright overcast; red is direct sunlight. You won’t be using these filters in other conditions.
And the results with a whole pile of 46mm B+W filters is…
Holy exposure cliff, Batman! Your stops are at 256, 128, 64, 32, 16, 8, 4, 2, and 1.
How do we translate this?
It is instructive to check out the B+W response charts that explain what we would expect to see (assuming, of course, a perfect sensing medium that could see all the way from UV to IR). Each filter does not make a perfect cut but rather has its own curve. Reds and yellows ramp up pretty quickly, and greens and blues are something of a free-for-all.
What does Leica say about the M typ 246’s response?
Below is Leica’s own test data for the M typ 246 (I asked, and they sent it to me). The values are fairly consistent with the testing above using filters, taking into account:
- This is chart still computed arithmetically.
- Leica was no doubt able to run this more systematically, with a light source that could crank out much more narrow wavelengths than a filter would admit.
- We don’t know at what stage in the imaging process these values are measured.
Regardless, Leica’s values are generally consistent with what you see mounting filters.
What real-world quirks should we expect?
There are a couple of strange things one might expect, and they were borne out by this test.
- The 092 is labeled by B+W as 20-40x – on the M typ 246, it is more than 256x, meaning that the IR rejection of the camera is incredibly good.
- The 081 (blue) has an unexpected spike in overall brightness. This cyan filter (used for B+W contrast reduction) also causes the camera’s metering system to overexpose (I was able to confirm this in testing).
- The 060 (yellow-green) reads a little lower than Leica’s data on both Monochroms might suggest. This may be explained by peak camera response falling between the two filters (060 and 022) that simply wouldn’t show up on this chart.
Is it really different from the Monochrom Typ 230?
Now, in terms of the “rendering” between the M typ 246 and its predecessor, the Typ 230, we are told (anecdotally) that it is “different.” Is it? So let’s take the plot with real-world filters and convert it to log values (which for some reason fascinate the publishers of exposure data).
Note: every logarithmic chart has its own scale determined by the base; here, I used 255 (because that is how Photoshop reads out). Leica may have used a number based on raw output from the sensor. Kodak has a standardized negative density that informs its chart. The one thing that can be done with these charts is to compare shapes to see if the response goes up or down and where on the spectrum.
And then compare it to the M typ 230 chart for the original Monochrom (below), which is expressed the same way. You can see that these are not very different in shape – except that the Typ 246 seems to be a tiny bit more friendly to oranges and reds. It would actually be nice to get the raw data to plot these together, but alas…
This in turn can be compared to film. Note that neither Monochrom camera has the little blip in the red range that TX and TMY do. But that has a lot to do with the dark science of film design. One thing these charts do reveal is why you actually need to make sure that no UV light hits TMY – although the film is a little less sensitive to blue (and therefore needs less of a correction), it has quite a bit of UV sensitivity (consider a B+W 415 strong UV filter).
All numerical silliness aside, the point here is that the responses of the Monochroms are fairly similar – and that on both, you should go easy on the strength of contrast filters – instead working to make sure you have a sufficiently varied scene to make their use meaningful.