M10 color rendition compared to M9-M240-SL cameras

[Guest]

Advertisement (gone after registration)

Having the M9-P and the MM, the color rendition of the M10 is what will determine whether I eventually buy the new camera, particularly as I am these days shooting color with an M3 and Portra 400 film (glorious!).

 

One of the reasons that I didn't buy the M240 was its color rendition compared to that of the M9. My view was reinforced reading posts by two photographers with excellent color vision and perhaps the same taste in color as I have: Marc Williams ("fotografz") and Charles Peterson. I don't want to refight old battles, but they both felt essentially that, if you fixed the colors you wanted (say, skin tones), other colors would go out of whack. From what I've read, very impressionistically, is that the SL colors in this sense are "worse" than those of the M240.

 

From this point of view, can we get a reading of how the color rendition of the M10 compares to those of the M9, M240 and SL?

_______________

Alone in Bangkok essay on BURN Magazine

[eujin]

I just got to play with the M10 at my local dealer. My impression is that the color rendition is going to be different from the M240...big caveat: this is based on looking at shots on the rear screen. I will look at the files when I get home. But based on the rear screen, colors look cooler than what I'm used to seeing on my M240-P — how much this is due to the rear screen's calibration, we will see. 

[Advertisement]

The Leica Q3 is available for pre-order at B&H Photo Video and Adorama

[John Black]

I'm expect a color look / rendition akin to the Leica SL.  I never bonded with the M9 color rendition, so was happy to move past it.  That said, the M-240 wasn't nirvana either.  I've been using the SL for awhile, and I think I like it's color, but takes me a long, long time to decide on color - I have to go through the seasons.  The SL files are definitely brighter than the M-240.  The M-240 files were dark - the mid-tones were repressed (IMO).  So I'm looking forward to seeing what the M-10's color turns out to be.  My solution for the M-240 color was the M-246 - and that totally solved it

[rscheffler]

Mitch, I think a lot of us are waiting to see how this evolves. According to Jono's M10 write up, he feels M10 skin tone rendition has improved over the M240, but only touched on this very briefly.

 

I just had a look at DPReview's sample images and pulled the DNGs from some of the people photos. First impression is the M10 files feel extremely similar to the M240. Whatever differences there are, are subtle. Artificial light situations can still result in overly orange 'pumpkin-like' skin tones. I'm liking skin tone results better when saturation for orange and yellow is pulled back a fair amount, which is the same thing I do with the M240... The low ISO files do push nicely and I don't see any signs of banding. I pushed one ISO 100 image 3.3 stops and I wouldn't guess this degree of push based on (lack of) image noise.

[Guest]

Thanks, Ron. I didn't think we would have any conclusions on this so early, but thought it's worthwhile for people to start thinking about this. BTW, I only included the SL because someone wrote that the M10 sensor is the same one as that of the SL, but others have contested this.

[Printmaker]

This was my first thought. The first few color samples I saw looked too contrasty with blown highlights and a few oversaturated colors. Then I came across Jonathan Slack's review and breathed a sigh of relief. The skin tones looked pleasant and the color was rich without being oversaturated.

 

I have a M10 on order and was planning on selling my M9 when it comes back with a new sensor. But I may shoot both for a few days before putting my M9 up for sale.

[otto.f]

Advertisement (gone after registration)

My impression of the photo's that Steve Huff took for his review, which are mainly shot at night or in indoor artificial light, is that the yellow is a good deal less conspicuous. If I wouldn't have known the M240, I think I wouldn't have noticed it coming from my M9

Edited January 19, 2017 by otto.f

[adan]

1. Download some of the raw (.DNG) test shots from dpreview or Leica's own M10 downloads selections. We should be able to manipulate all the image parameters enough to get rid of the excessive contrast choices of some shooters and see what the real colors, practical ISO limits (where banding kicks in) etc., are.

 

(I'll do it myself but I'm out of town until Friday)

 

2. Frankly, I found some tests shots by a forum member with a 262 to already be an improvement in color over the base 240s I tried. Which just goes to show the embedded profile Leica installs can and likely has been adjusted by Leica over the years, as well as whatever the sensor itself outputs.

 

Some of the available sample images - as processed by Leica or the dpreview shooters - do look harsh, contrasty and/or with worrying red shadow casts. But until I can work with them myself, I don't take that as a serious test of what the camera can do.

 

You know the old saying: "If you want something done right......."

[Chaemono]

Having the M9-P and the MM, the color rendition of the M10 is what will determine whether I eventually buy the new camera, particularly as I am these days shooting color with an M3 and Portra 400 film (glorious!).

 

One of the reasons that I didn't buy the M240 was its color rendition compared to that of the M9. My view was reinforced reading posts by two photographers with excellent color vision and perhaps the same taste in color as I have: Marc Williams ("fotografz") and Charles Peterson. I don't want to refight old battles, but they both felt essentially that, if you fixed the colors you wanted (say, skin tones), other colors would go out of whack. From what I've read, very impressionistically, is that the SL colors in this sense are "worse" than those of the M240.

 

From this point of view, can we get a reading of how the color rendition of the M10 compares to those of the M9, M240 and SL?

_______________

Alone in Bangkok essay on BURN Magazine

 

I own an M9, M240 and SL. Agree with all your points except for "From what I've read, very impressionistically, is that the SL colors in this sense are "worse" than those of the M240." This is not correct based on my experience. It's much easier to maintain naturally looking skin tones and get punchy colors from the SL files than what is possible with the M240 if one uses the right color profile. In LR I either use the embedded one or the Huelight Leica Camera Profiles, not the Adobe default one. Also, see the last comment by vanhulsenbeek here: http://www.l-camera-forum.com/topic/262001-alternate-lightroom-profiles/?do=findComment&comment=3160910

Edited January 19, 2017 by Chaemono

[rscheffler]

I believe it's Erwin Puts who stated the sensor is the same as in the SL... As already appears in this thread, you're going to get a different opinion from each photographer about the M10's color. That said, I stand by my assessment that it's much closer to the M240's file 'behavior' in post than the M9's.

[UliWer]

I believe it's Erwin Puts who stated the sensor is the same as in the SL...

Sean Reid and Jono Slack wrote that it is a different sensor. Andreas, the Admin, asked the Leica people yesterday at the event and they also told him it was not the SL-Sensor or based on the SL-sensor. Though knowbody seems to know who else makes it.

[eujin]

I took a quick look at the shots I fired off with the M10 in my local store last night. From what I've seen, yellow is less prominent in the M10 files. The overall color rendition SOOC appears to be cooler than the M240. This was my initial thought when looking at the rear screen; in LR, the files look a little more neutral (conclusion is that the rear screen looks to have more blue). There's more dynamic range with the M10 files, which I think is why the DNGs look flatter. ISO 12500 looks very usable as long as the exposure is somewhat correct. Otherwise, at that ISO there's definitely noise in the shadows but no banding that I could see. 

[thighslapper]

c'mon guys ..... we have been down this road a gazillion times before.

 

If you use DNG it is almost entirely down to the WB algorithm, WB setting (particularly AUTO) and the profile used by the RAW processor you use. 

 

With JPG you are stuck with what Leica thinks things should look like. 

 

The M240 profile and WB was crap initially and corrected later by Leica and Adobe. The SL colours also changed when the LR profile appeared from Adobe (greens initially were garish).

 

The posted images to date look fine to me ...... and the SL and Q images are also fine, and I guarantee in a blind testing with images taken under identical conditions and settings you would not find a difference worth arguing about.

I did this years ago with the M9/M240 .... and although the images were 'different' for other reasons, the colour palette once WB was corrected was so similar I couldn't tell which was which. 

Edited January 19, 2017 by thighslapper

[otto.f]

Common guy, I can still tell the difference between M9 and M240. There is a test running somewhere on the internet for this, I had 70% correct. And that, given the fact that the files of the M240 were postprocessed to a great extent, needing a fair amount of time, to get them neutral enough to compare with M9.

Which is not to say that the images out of the M240 cannot be beautiful with the right knowhow and effort, sometimes more so than from the M9 because in the end you see that they have 24Mp and images can show richer therefore.

Edited January 20, 2017 by otto.f

[adan]

 

If you use DNG it is almost entirely down to the WB algorithm, WB setting (particularly AUTO) and the profile used by the RAW processor you use.

 

Ennnh - I would beg to differ. There are many issues that a global white balance can't really correct for, once the image is baked into even a .DNG

 

The obvious examples would be - photograph a yellow sheet of paper next to a white sheet of paper under very yellow light, and then try to find a white balance that will remove the yellow cast in the white paper without erasing the (correct) yellow of the yellow paper.

 

Or the classic M8 IR purple neutrals - photograph a guy in a gray suit (which the M8 turns purple) next to a guy in an actual pale magenta tux - and you can't get rid of the sensor-induced cast in the neutral gray without killing the natural purple of the purple tux. No WB adjustment exists that won't affect both.

 

One of my "acid test" colors - for sensors or film - is medium to dark gray-green. Evergreens (important here in Colorado), or architectural dark greens like lamp-posts. The M240 sensor simply added so much red (instead of magenta) to those (white-balanced for a gray card) that they became browns or grays. I could "white-balance" to get the green correct with a massive correction - but the whole picture had to turn green as a result.

 

Additionally, if I photographed someone in a red nylon hoodie, for example, the red cloth would go solid tonally undifferentiated red. No shading for highlights or shadow areas in the cloth folds. Unless I underexposed 3 stops, which was not good for the other colors and tones in the picture. It simply had a hot-hot-hot red channel (so much so that the red channel was the source of the banding at as low as ISO 3200). The red channel was so over-amplified that it distorted the relationships between colors.

 

(The firmware and embedded profile were whatever was current June 2016)

_____________

 

I did download some of the M10 examples from both Leica and dpreview. The M10 sensor handles those greens much better, while keeping other colors correct. I can't say it is M9 colors, without a direct scene-to-scene comparison (and with my cyanny Mandler lenses). Feels a bit like E100G vs K25, though. With a mid-tone saturation and/or contrast bump, it seems reasonably close.

 

As asides not directly related to color rendition:

 

Especially the Leica samples have the "Clarity" ramped way way up, adding a ton of contrast - apparently this is an "in" look, which is fine, but they don't represent what the sensor's tonality can be. The added contrast also hyper-saturated some colors.

 

The M10 at ISO 10000 (picture of 2 guys in red and white light indoors, in dpreview's samples) - converted to grayscale, and pushed 1.4 stops (original was underexposed) so actually about ISO 28000 - has softer and less obvious noise than my CCD MM at ISO 8000. (!) I turned off the noise reduction the original photographer had applied in Adobe.

 

EDIT - so just imagine what an M10M without bayer filters could do!

Edited January 20, 2017 by adan

[helged]

EDIT - so just imagine what an M10M without bayer filters could do!

Interesting, indeed... Does this also imply that M10 rivals the territory of (the existing) M246?

[bigben]

given the M10 has come out, I think Im going to buy the M9  - to add to my MP240/Q. Im guessing the M10 has improvements over the MP240 - the low light capabilities sounds exciting (although can you really see to manually focus if light is that low...) and if they can get rid of the horrible banding!!...

Edited January 20, 2017 by pop

[markforce]

Interesting, indeed... Does this also imply that M10 rivals the territory of (the existing) M246?

Erwin has something to say about that in his write-up.

[Guest]

...As already appears in this thread, you're going to get a different opinion from each photographer about the M10's color. That said, I stand by my assessment that it's much closer to the M240's file 'behavior' in post than the M9's.

...One of my "acid test" colors - for sensors or film - is medium to dark gray-green...The M240 sensor simply added so much red (instead of magenta) to those (white-balanced for a gray card) that they became browns or grays. I could "white-balance" to get the green correct with a massive correction - but the whole picture had to turn green as a result...I did download some of the M10 examples from both Leica and dpreview. The M10 sensor handles those greens much better, while keeping other colors correct. I can't say it is M9 colors, without a direct scene-to-scene comparison (and with my cyanny Mandler lenses). Feels a bit like E100G vs K25, though. With a mid-tone saturation and/or contrast bump, it seems reasonably close...

 

Thanks, gentlemen that's useful at this stage. Of course, it will be interesting to do more testing once there are better M10 DNG files to work with.

 

BTW, I never would claim that the M9 has the perfect color rendition for other cameras to emulate: in certain light it produce horrible, screaming electric blues...

 

Another factor is color vision. For example my daughter who, on the basis of color test should have had much better color vision than I did — and apparently women in the mid-twenties on the average have the best — played a computer color game in which we had to mix colors to match a target color: unaccountably, I did much better than she; perhaps it was because I was doing oil painting at the time.

 

Therefore, while I agree with "adan" that it's best to test the camera yourself, it's also good to read the conclusions of people whose color vision one trusts — and I hope that both of you will chip in here when we get better files to test.

[Advertisement]

Hallo,

Look here: Leica M

[thighslapper]

Erwin has something to say about that in his write-up.

 yes .... and he also basically declines to comment much on colour as he states that whole issue is very subjective .... and as Adan points out colour is dependent on the lens as well ... and they DO vary .... particularly the older M lenses. 

 

I personally have a slight but noticeable difference in colour balance between both eyes ...... on renders slightly 'bluer' than the other .... almost certainly down to incipient age related lens changes.... and it is 'exposure' dependent as well to an extent. 

 

Folk who have cataract surgery and new lenses are usually astonished at how blue the world appears .... after years of looking through muddy brown cataracts .....

 

oh .... and anyone posting comparison shots needs to ensure EVERY parameter is identical and include a grey card in the photo so WB can be rendered identically. 

 

.... and even then you have the variability of the display screen and then viewers retina to contend with. 

 

Almost every camera seems to have some colour fidelity issues at certain points in the spectrum and exposure which pop up under the 'right' conditions. I am not sure that the technology is completely correctable for all situations ... just most. 

Edited January 20, 2017 by thighslapper