Monochromatic Rendering - Comparing the Leica M10, Leica M240, Leica M246, Leica M9, Leica MM, Leica SL, and the Leica M6 + Agfa APX 400
Déjà vu. Yes, I've already done a blog post comparing monochromatic rendering between the Leica M246 and the Leica M240. In fact, I've done two blog posts, which I have just removed from this blog. However, I've never been satisfied with my work on those two blog entries. On my first attempt, I forgot to bring my M240, which clearly was a rookie mistake... not to mention it made the comparison pointless. But thankfully on my second attempt, I remembered to bring it. However, I was never quite satisfied with my second attempt... given how incomplete it was. You see, I didn't include the Leica M Monochrom or the M9 at the time. I had left them both in Hong Kong, when I was conducting the comparison in New York.
Well, if at first you don't succeed, try, try again. Third times the charm, so I'm told. So that's why I’m conducting this comparison yet again - and this time with a more complete list of cameras, including the following:
Leica M Monochrom
Leica M6 + Agfa APX 400
I decided not to compare the raw unedited images, because each of the seven cameras experienced ±½ stop variance, despite using the same exact exposure setting from a hand held meter. To some, this may appear to be an odd decision, given the assumption that raw images would naturally offer a more accurate comparison. However, the variance in exposure made the selected sample images difficult to compare. With this being the case, I made a judgment call, and decided to compare what each camera's image file was capable of rendering after post processing.
In order to insure consistency in processing, I decided on using a black and white film reference to provide a standard to follow - specifically using Anna's skin tone and facial accents as the tonal value to match in post processing. That said, it should be noted that I did not purposely select Agfa APX 400 for any particular reason. I ran out of Tri-X, which would have been a more common film to use for this comparison. In addition, I had also brought along a roll of Ilford Delta 3200, which I had intended to push a stop. However, none of the images on that roll were usable. I suspect the roll expired. So even on this third attempt, I goofed.
The lens I used for this comparison were the following:
Leica 28mm f/1.4 Summilux-M ASPH
Leica 28mm f/2 Summicron-M ASPH Version II
Leica 28mm f/2 Summicron-M ASPH Version I
Leica 28mm f/2.8 Elmarit-M ASPH Version II
Leica 28mm f/2.8 Elmarit-M ASPH Version I
Lecia 50mm f/2 Summicron-M Version I "Dual Range"
The images shot on the first four image set were all shot with one of the 28mm lenses, in no particular pairing. For the interest of time, I didn't want to be overly burdened by sharing one lens between seven cameras. Although, for the last image set, I was forced to do that. However, since I was shooting at table level, sharing one lens between seven cameras didn't prove too much an imposition.
Last, I didn't have much time to do this investigation. But, I still wanted to squeeze out as much from it as possible. So naturally, I chose to shoot at a reasonably incorrect exposure and/or under unideal light conditions to test dynamic range of the resulting image files. Given the scope of the comparison, I wasn't able to do more than five image sets.
Set 1: Normal Light Conditions - shot at 28mm, ISO 400, f/4, -1 Stop
It is never easy to evaluate rendering of the same documentation between so many different cameras. It's just so tedious. They all look more or less the same - except for the sample taken from the Leica M6, which being film capture looks like film capture. Having said that, I was mindful not to cause sampling error by knowing which image was taken by which camera - opting instead for a blind side-by-side evaluation. Under that approach, the one image that caught my attention on the first image set was the sample taken from the Leica M246.
It's difficult to explain what it is I saw in the rendering from the Leica M246. To my eyes, there appeared to be more fullness in tonal gradation which created a greater impression of a three dimensional quality. I suppose if I were to articulate myself better, I would say the tonal range of the shadows and highlights is noticeably more complete, ranging from true black all the way to true white. By comparison, the desaturated images from the other cameras appeared less certain in gradation, with tonality being more gray by comparison and falling short of rendering true blacks and whites in shadows and highlights respectively, unless impacted by incorrect exposure.
In terms of tonal range, the sample taken from the Leica M246 appeared most capable to emulate what's typically rendered in black and white film, based on the sample taken from the Leica M6 + Agfa APX 400. Also worth noting is the comparatively clinical rendering from both the Leica M10 and Leica SL. To my eyes, the rendering appeared tonally flat with its grayscale missing true black and white in the shadows and highlights respectively.
NOTE: For some reason, all the images shot on the Leica M Monochrom on this image set appeared to have either missed focus or shot at a thinner depth of field. How that was possible, I don't know. All images on this set were shot at the same aperture setting, and with the camera on a tripod at the same distance from the subject - verified with a tape measure.
Set 2: Low Light Conditions with Strong Back Light - shot at 28mm, ISO 400, f/4, -2 Stop
On the second image set, I admit I couldn't find any discernible differences between the six digitally captured images on a blind side-by-side comparison. Because of that, I had no choice and was forced to uncover the source of each sample image. Invariably, this led to bias and sampling error in the resulting opinion of my evaluation. So not surprisingly, much of the evaluation in this second image set went to corroborating the opinion from the first image set.
The differences are very slight. But the Leica M246 and the Leica M Monochrom appear to have greater tonal range than the Leica M240 and Leica M9 respectively. The demonstration of tonal range is evident in the ceiling moulding and panels (which I admit is somewhat overreaching in observation). In my opinion, there seems to be marginally more details in the monochrome files than in the desaturated files. Even so, I still find the Agfa APX 400 image files tonally more complete, albeit obscured in presentation by the presence of film grain reducing overall resolved details.
As for the Leica M10 and the Leica SL, they both appeared more contemporary in rendering, with noticeably clearer definition of resolved details. However, the gradation of tonal range was noticeably flatter in the shadows and highlights - or at least that was how it appeared to someone in my position who was tasked to split hair in providing an opinion. To be honest, I really didn't find any material difference in rendering between the six digital capture.
Set 3: Direct Sunlight - shot at 28mm, ISO 400, f/4, +1 Stop
The obvious take-away from the third image set is just how poorly the Leica M246 and Leica M Monochrom renders overexposed highlights. Even though this image set was only intentionally overexposed by a single stop in direct sunlight, the highlights from the image samples taken by the Leica M246 and the Leica M Monochrom were blown and unrecoverable in post processing. Of course, this comes as no surprise. It is already well known that both the Leica M246 and Leica M Monochrom are designed to perform better under lower light conditions, and not especially well in direct sunlight - unless underexposed a stop or two.
That being said, I do like the image files from both the Leica M Monochrom and the Leica M9 on the first three image sets. The tonal range, even with the desaturated images of the Leica M9 are surprisingly complete. In fact, it could be argued the Leica M9's desaturated images are better than the Leica M Monochrom, under optimal light conditions, given how much more forgiving the dynamic range of its image files are.
Again, the images from the Leica M10 and Leica SL appear to be somewhat clinical and noticeably flatter in the shadows and highlights, when compared to the digital rendering from the other cameras. Also worth noting was how direct sunlight from the window drowned out details in facial accents. What this does is reduce visual definition, which then makes the sample image appear less impactful in presentation.
Having said that, the sample image from the Leica M6 + Agfa APX 400 had the best tonal rendering of the seven cameras.
Set 4: In Darkness - shot at 28mm, ISO 6400, f/8, -2 Stops
On the fourth and fifth image sets, I did not have a black and white film reference to provide a standard to follow. As I said earlier in this write-up, I suspect the roll of Ilford Delta 3200 had expired outside of proper storage. Even so, low light conditions isn't exactly where film naturally shines in performance. With that being the case, I can forgive myself this oversight in not having a black and white film reference for the last two sets.
Having said that, low light isn't particularly ideal for the Leica M9 (which tops out at ISO 2500) and the Leica M Monochrom, which suffers from excessive noise beyond ISO 1600. And at ISO 6400 pushed two stops, the Leica M Monochromes suffers from horizontal banding in addition to just noise. As for the Leica M240, it too exhibits some signs of noise at ISO 6400 pushed two stops. However, I don't believe it impacts overall resolved details significantly - at least not to my eyes.
By comparison, the Leica M10 and Leica SL were both able to perform under the low light conditions of this image set. However, the rendering from both the Leica M10 and Leica SL is notably flatter than the Leica M246. In my opinion, this demonstration of tonal range is evident on Anna's face. The image file from the Leica M246 clearly appears more three dimensional in rendering than the desaturated images files from the Leica M10 and Leica SL.
On the final image set, I didn't edit the sample images. Other than the image from the Leica M9, they all look more or less the same. But I suppose if I had to split hairs, the Leica M Monochrome and the Leica M246 appear to have more tonal range.
Set 5: Higher ISO Exposed Properly - Shot at 50mm and minimum focusing distance, ISO 6400, f/2
In conclusion, this comparison was really done in good fun for the sake of providing a reference for public record. It's not as if I have some special innate ability to eyeball marginally observable traits between native monochromatic and desaturated images. That we can leave to the diagnostic testing at DxO. Still, I suppose the real lesson learned from this entire exercise is how capable color sensors have become, and how unnecessary monochrome sensors are for most practical use under normal shooting conditions. Of course, if low light performance is material to your needs, then the Leica M246 is demonstrably the most capable in monochromatic rendering. With that said, the Leica M10 and Leica SL isn't too far behind.
However, being most capable in monochromatic rendering doesn't make the Leica M246 best in this comparison. In my opinion, the Leica M6 + Agfa APX 400 was the best. Admittedly, film is less capable in low light, and even less convenient to process and digitize. But, from the perspective of rendering, it still looks the best and continues to be the standard by which digital photography aspires to emulate. It is after all why I included it in this comparison.
No digital images were cropped. Film images have been cropped as close to full negative as possible. Any editing performed on Lightroom have already been disclosed in the caption of each image. All image sets shot at the same exposure settings.