High ISO - M10 vs Q

[Guest Nowhereman]

Advertisement (gone after registration)

M10 vs. α7R III low light comparisons at same ISO here: https://www.smugmug.com/gallery/n-Jfdr66/

M10 looks fine to me in low light. See α7R III + 55 Otus vs. M10 + 50 Noctilux at ISO 6400, for example. Love the M10.

 

Yes, I've looked at your comparison but my experience is different. But first, someone pointed me to this DXO graph (go to the Measurements tab). It shows that at the ISO1600 setting the measured ISO of the Leica Q is 1575 and that of the M10 is 787; at the ISO3200 setting the measured ISO of the Q is 3037 and that of the M10 is 1566; and at ISO6400 the measured ISO of the Q is 5999 and that of the M10 is 3274. I don't think that DXO gets ISO measurements wrong — and am not ecstatic about Leica inflating the ISO of the M10.

 

Now, I've just looked at some of my street photography shot in BKK with the M10 and the Summaron-M 1:5.6/28 (so I know that I used an aperture of f/5.6): I see that I shot a lot of pictures at ISO 3200 and 6400 in a type of daylight in which I would have used no more than ISO1250 with the M9 and MM. I was surprised to be using ISO3200 and 6400 for those images, but now I see why. 

 

In addition, as stated earlier, I felt that I had to underexpose much more with the M10 than with the M9 and MM in order not to blow highlights excessively — come to think of it, that is what drove me to using the ISO 3200 and 6400 settings in daylight:  increasing the shutter speed by as much as 1 stop and 2 was necessary because the minimum aperture of the Summaron is f/5.6, which meant that, if I didn't go to the ISO 3200 and 6400 settings, I would have had to go to shutter speeds that would be too slow to avoid motion blur of the subject. 

_______________

Alone in Bangkok essay on BURN Magazine

Nowhereman Instagram

[adan]

Whoa! You'll have to walk me through why you are using higher ISO settings to avoid blowing highlights and "underexpose more." Normally, using a higher ISO reduces DR, and thus increases the chance of blown highlights.

 

I mean, when I want to avoid blown highlights (which is all the time) I dial in a permanent EC of -0.7 - I.E. the camera is set to, for example, ISO 200, but I am feeding it less light (higher shutter speed)  than a straight exposure calls for - effectively exposing as though it was set and metering at 320 (but it is not - it is still at ISO 200 on the dial). But that is different from cranking the actual camera ISO setting up to 320.

 

As to DxO's results - they point out they are using the Saturation-based method of measuring ISO, one of the five methods available. Which measures ISO according to where highlights blow. Given the propensity of the M10 to blow highlights, it rather makes sense that they get "special" values for the M10. When I tested my M10 using Standard Output technique (more or less, metering and exposing for 18% gray), the M10 ISO came out right on target (correct gray value, as exposed with both the M10 "classic" meter, and a Sekonic hand-held).

[Advertisement]

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

[Mark II]

As to DxO's results - they point out they are using the Saturation-based method of measuring ISO, one of the five methods available. Which measures ISO according to where highlights blow.

 

 

Exactly. And the exact number that DXO produce for the ISO rating is very much less important than their use of a consistent methodology. Comparisons where data is plotted against the manufacturer's notion of ISO would otherwise be misleading because of the manufacturers' scope to interpret ISO standards rather differently.

 

The one thing that DXO do not show is the effect of pattern noise on the effective dynamic range, not least because it is a highly subjective thing to characterise. Personally, I do not see significantly different pattern noise between the M10 and M262, although someone else would probably have a very different perspective on this.

[Guest Nowhereman]

...Given the propensity of the M10 to blow highlights...

 

Andy - To me this is the key point: the M10 has a huge — I should say "humongous" — propensity to blow highlights compared to the M9 and MM that I used to have, whether I'm shooting in the harsh high-contrast tropical light of Bangkok or that of Paris when the sun is out. I expose manually; and find that often I have to underexpose much more than a –0.7 exposure compensation would accomplish. Indeed, I feel that the major, and unfortunate,  issue with the M10 is the extent to which its necessary to underexpose not to blow highlights. 

 

Now, below are two examples of shooting with the Summaron-M 1:5.6 28 mm lens, whose minimum aperture is, of course, f/5.6. I was shooting with this lens because I like the way it diffuses the light around fluorescent lights (exaggerated by burning-in around the lights in the first picture): the sky was heavily overcast, but with a high luminescence; below the large, dark awnings of the market, covering most of the sky, it was dark enough that with the M9 I would have shot at ISO1250, and would have lifted the shadows in post, if necessary. Because of the greater flexibility of the M10 for raising the shadows, here I shot at ISO3200 and 6400 — same shoot, same market.

 

You are right there are two different issues here: the excessive propensity of the M10 to blow highlights, but perhaps also the overrated ISO of the M10. It seems to me that the OP has a point here.

 

 

M10 | Summaron-M 1:5.6 28mm | ISO6400 | f/5.6 | 1/180 sec

Bangkok

 

 

3200M10 | Summaron-M 1:5.6 28mm | ISO3200 | f/5.6 | 1/60

Bangkok

_______________

Alone in Bangkok essay on BURN Magazine

Nowhereman Instagram

[Simon]

I don't understand this discussion except to say that I like that b&w shot above.  The blown sky and glow makes the shot.

[Guest Nowhereman]

Thanks. Actually, the sky and the fluorescent lights are not blown: the brightest parts still 99.4%. There was no detail in the sky — just evenly overcast with great luminosity, like a fog.

[250swb]

Advertisement (gone after registration)

I don't understand this discussion except to say that I like that b&w shot above.  The blown sky and glow makes the shot.

 

I think exactly the same, make it more so, feather and spread those highlights even more! There comes a time when a photograph is either about passion or technical details, and it's a loosing game when sad tech freaks care about technical details beyond the pleasure to the eye.

[pgh]

I don't understand this discussion except to say that I like that b&w shot above.  The blown sky and glow makes the shot.

That's nice and all but what's hard to understand about Leica says their sensor is one thing, and in practice many people find it inferior to their claims, and that thing being a point of contention, especially when people put down 7k USD for it? 

[adan]

Nowhereman - when you say you meter manually, what exactly do you mean? M10 "classic" meter or 3rd-party hand-held? Reflected or incident? Metering for shadows or highlights or an "average gray" part of the subject?

 

In playing with your posted images above, I note there is a large amount of shadow detail that can be pulled out of those. That is, compared to how you want them to look and have showed them here, they probably are unnecessarily over-exposed (and thus unnecessarily pushing the highlight limits).

 

Because the converse to the M10's humongous tendency to blow highlights if not carefully metered, is its humongous capacity to recover shadow detail.  Are you aware of the "humongous" contrast curve the M10 applies to its .DNG files by default, that can be opened up 2-3 stops via the "Shadows" slider in LR or Camera Raw, without significant color distortion or excess noise (above the normal amount for the chosen ISO)? Given the "extra" shadow detail I was able to find in your two shots above (even as compressed jpgs), I think you could have used the same exposure time and aperture at ISO 800-1600, and gotten the same final pictures in the end - minus the blown highlights.

 

In other words, you can't use the M10 as though it was an MM or M9 (or M6, for that matter). But you can get pictures without blown highlights if you use the camera as an M10, and process the images accordingly.

 

Denver has quite contrasty lighting. Even though I don't have the stalls of SE Asia here (or at least not in my normal range of subject matter), here is a sample that comes as close as I can get to your situations and contrast range (recently, anyway). Very high contrast lighting, people under an awning, backlit.

 

M10, 28mm Elmarit v.2, ISO 200, EC -0.67, "Auto" exposure with classic meter (not EVF), 1/350, f/stop unknown (older firmware) but ~f/5.6-6.8, probably, .DNG.

 

First picture - straight from the camera. No matter how dark it looks here (or on the camera screen), this is a "correct" exposure for the M10 in this situation, since it just barely avoids clipping the highlights except in the sun reflection off the trash basket, left.

 

Second and third pictures show the "real" picture that is in there, with shadows slider pulled all the way to the right, and other tweaks (exposure, highlights sliders to compensate for the postprocessing exposure adjustment). BTW - if you don't like my taste for contrast or saturation, that's another story. I can make 'em duller or punchier as desired.

 

To get back to the original topic - the exposure I used to get a good picture out of the M10 (shadow detail, 18% grays, etc.) in this case means I got an effective ISO of 320/400 with the camera set to 200 (Sunny 16/backlit, shutter speed 1/ISO, f/5.6). Which is why I take DxOMark's tests with a grain of salt. They didn't necessarily understand the M10's admittedly weird "sensitometry".

 

Welcome, dear visitor! As registered member you'd see an image here…

Simply register for free here – We are always happy to welcome new members!

 

 

 

[Guest Nowhereman]

Andy - Thanks for taking the trouble to answer so fully, appreciate it. The point I'm making is that the M10 does indeed have, as you so well put it, a  weird 'sensitometry'". I'm aware of the huge headroom of the M10 in lifting shadows — and that is what I do most of the time: underexpose substantially, and lift shadows 1 or 2 or even 3 stops. The shots above were made when I had recently acquired the camera and was much taken by how much I liked the look of the images shot at ISO3200 and 6400; also, I was using a new lens, the Summaron and conscious of the limitation of not being able to open up to a larger aperture than f/5.6. I use the classic M10 meter and meter for the highlights, unless I know that the highlights won't be a problem and meter for the mid-tones. 

 

The trouble is that the weirdness of the sensitometry often seems unpredictable, so that I sometimes make a few test shots before deciding on the parameters of the ISO and exposure, when I don't simply meter for the highlights. Can't say that I like this about the M10. In one way, it's somewhat like using the M9 for low light photography, by shooting at ISP 640 and pushing in post — although the M10 lets you lift the shadows more.

 

BTW, I've never been called a "sad tech freak" before, and dislike the personal attacks of 250swb, which make this place more unpleasant than it needs to be.

_______________

Alone in Bangkok essay on BURN Magazine

Nowhereman Instgram

[Chaemono]

I agree. Please, stop the name-calling, guys. Back to topic. This issue has been beaten to death in other threads but the Energizer Bunnies can’t help it, they have to keep going and going.

Edited September 12, 2018 by Chaemono

[erl]

If I may interject here and adjust the thinking about 'personal attacks'.

My reading of 250swb's comments is very generalized and definitely does not name or personalize anything that I can see.

 

I see two lines of (maybe) opposing thoughts about how to use the M10. I have an M10 and I suspect I have a foot in both camps. ie. 1. technical 'perfection' and 2. to hell with it all, just shoot the image if it's worthy of my attention.

 

Generally my best efforts derive from the latter, because I tend to be lazy about precision and maybe life's experience has trained my brain to get 'close' to accurate anyway. For those who want/need to analyze the finer points of highlight repro of the M10, I say go for it. When you write about it I can maybe crystallize something that was in my head anyway, without doing the hard yards. I did say I was lazy!

 

For those, which includes me quite often, just go shoot the M10 more or less as you instintively do, according to your own taste. I bet everyone is a winner.

 

Personally, I find the DR of the M10, regardless of how I expose the sensor, to be marvelous. Is it the equal or better of other cameras? I have no idea.

[adan]

I do wish Leica would - rethink - exactly what they are doing with the in-camera processing of M10 images (especially .DNGs). (LEICA - ARE YOU LISTENING?)

 

What I called the "weird sensitometry."

 

My theory all along has been that the extreme contrast mapping of the M10 was to avoid the reaction the M240's higher-DR CMOS sensor got from some people (including me) regarding color saturation and contrast - less punch than the M9's CCD colors. It could, of course, also be to suppress shadow noise.

 

And correspondingly set the metering (at least the classic off-the shutter metering) too high, thus biasing "normal" exposures towards overexposure and blown highlights, to compensate for their own "darker" contrast mapping choice.

 

(Or, of course, maybe reknowned Leica artist Ralph Gibson was a beta-tester and consultant on M10 imagery - he loves punchy contrast and black-black shadows and white-white highlights. ) - http://www.ralphgibson.com/gallery.html

 

But I think it is a bit overcooked, and not necessarily the right shape of a curve. Some firmware maven early on (sandymc?) reported that he found many things in the M10 firmware and EXIF, at least in the beginning, that were "sloppy and rushed," and this may be another such.

 

Cut back on the severe contrast-mapping, and simply tune the meter to provide less exposure (higher shutter speeds) - which (back on topic) would also get the apparent ISO back in line, at least as far as highlight clipping goes.

 

That is to say, to darken the gamma of the image, simply reduce the exposure (metering at ISO 200 gives a shutter speed of 1/1000th in lighting X, instead of the current 1/500), rather than dragging down the brightness and shadows in the process of getting the picture from the sensor to the SD card.

 

There's not much they can do with the ISO as marked (short of shipping everyone new ISO knobs) - but I think the real issue is with the metering and tone mapping. Both addressable in firmware.

 

Of course, they'd have to warn everyone in the fiirmware notes that "everything you knew about exposing with the M10 is now wrong - re-learn it!"

[Guest Nowhereman]

Thanks, Andy. What you're proposing strikes me as a better solution that leaving it as it is. What surprises me, though, is that more M10 shooters have not reacted to this — although, because the sensitometry is so weird what people have been reacting to may be unclear because, they may have reacted to various different aspects, so that a clear voice has not yet emerged.

Edited September 12, 2018 by Nowhereman

[Chaemono]

I don’t mind if they correct it by firmware but then they should give me a sticker with the new ISO markings for the knob.

[Guest Nowhereman]

...Seems rather strange to me why none of the beta testers and reviewers picked up on this. Instead we got misleading plaudits about the M10's high ISO capability, e.g. "Generally speaking a 2 stop advantage seems to be about right, getting slightly less as one goes into extremely high ISO."

 

You can say that again!

[jaapv]

I'm rather surprised to see this thread a year after the camera hit the shops. Surely such severe issues would have been spotted as soon as photographers started using it.

[Chaemono]

You guys are going to make me use the same lens on the α7R III and the M10. It’ll be the APO 50 Summicron-M at ISO 1600, 3200, 6400, and 10000 which I’ve done before. The M10 will come out looking just fine in terms of noise vs. the Sony. I’ll see when I have time because I just got another ‘toy’ that I wanted to try out.

[pgh]

I'm rather surprised to see this thread a year after the camera hit the shops. Surely such severe issues would have been spotted as soon as photographers started using it.

Active threads have discussed sensor/imaging pipeline quirks for about a year now. We all experience the quirks differently and so the topics have varied, but generally any issues that have been brought up with regards to the sensor were, for the most part, met with dismissive remarks. I rarely shoot at high ISO, so it never bothered me much, however the lack of highlight retention when exposing in a way that is consistent with other digital cameras - which is tied in with this as previous posts suggest - have been remarked on consistently. That said, issues regarding the ISO have been posted about too, if not fleshed out so much because, again, of the tendency to dismiss the possibility of inherent flaws (or, um, differences) in the imaging pipeline.

 

It doesn't mean you can't make good pictures with the M10, but the way the sensor works is generally weird, and in many situations subpar - compared to other cameras out there. 

 

But to your point, these issues were spotted as soon as photographers started using it - even if they didn't know exactly how to say what was wrong. 

[Advertisement]

Hallo,

Look here: Leica M

[pgh]

You guys are going to make me use the same lens on the α7R III and the M10. It’ll be the APO 50 Summicron-M at ISO 1600, 3200, 6400, and 10000 which I’ve done before. The M10 will come out looking just fine in terms of noise vs. the Sony. I’ll see when I have time because I just got another ‘toy’ that I wanted to try out.

You do all of this at your own leisure! 

Looking at Raws, raw conversions, and prints from my own camera kits (leica, sony, and old nikon raw files) is enough for me, and, I suspect enough for most other users who are simply trying to get the most out of the thing they own. I adjust to the tools I hold in my own hands. The definition of "fine" varies from one user to the next. Tests etc can be helpful, especially when in the market, but once I own the thing, I go with the results I see.