M9 and "blown highlights"

[larryk34]

Advertisement (gone after registration)

I've read about everything that's been written so far about the new M9, but I haven't seen any comments about the new sensor's capacity to handle hightlights. I've never had one problem with my M8 and I'm delighted with what I've read so far about the improvements in the M9. But there's one problem shortcoming to all three digital cameras I have, M8, and Panasonic G1 and Z1. And that's the consistent problem with blown highlights. My understanding is that all digital sensors fall short when it comes to highlights versus shadows. Would the new M9 owners please comment on this for the rest of us? Is this a technological problem with sensors that has no solution in the near future?

[Advertisement]

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

[erl]

My first take on your observations is that possibly you are over exposing on all your cameras, by whatever method you are using to determine exposure. Run some tests to compare your results when varying, by reducing, your exposure. That could be your easiest and fastest solution.

[smb]

Have you read Steve Huff's review of the M9? In it he discusses what I believe you may be experiencing.

[dfarkas]

I've read about everything that's been written so far about the new M9, but I haven't seen any comments about the new sensor's capacity to handle hightlights. I've never had one problem with my M8 and I'm delighted with what I've read so far about the improvements in the M9. But there's one problem shortcoming to all three digital cameras I have, M8, and Panasonic G1 and Z1. And that's the consistent problem with blown highlights. My understanding is that all digital sensors fall short when it comes to highlights versus shadows. Would the new M9 owners please comment on this for the rest of us? Is this a technological problem with sensors that has no solution in the near future?

 

As I mentioned this in my review, my feeling is that the M9 is about one stop better in holding the highlights than the M8.

 

David

[lars_bergquist]

There is a radical difference between exposing negative film, and using digital. Negative film has exposure latitude, which means that the film can hold highlight detail at a luminance far beyond that of a normal diffuse highlight, which is about 2.5x the "mid exposure" or "Zone V" in the parlance of the zonists. With some burning-in, we could also put some of that extended highlight detail on the paper.

 

Digital is more like transparency film. There, what you exposed in the camera was what you looked at on the screen. There was no intermediate stage where tonality could be adjusted -- meaning, no latitude. Blow the highlight on exposure, and it cannot be saved. When a pixel well is overflowing with electrons, no more can be held, and no more detail can be recorded. No post production can put in details where none have been recorded.

 

With many others, I have concluded that the M8 as set by the factory exposes about one half f-stop too liberally. Therefore I have the exposure compensation constantly engaged for an 'under-exposure' of 1/3. (This is why I have asked for a permanent means to 'bias the exposure' without the infernal blinking light in the finder, but it sems that the people who just happily blaze away at whatever exposure the camera is handing them, cannot understand the thinking -- clearly because they do not know anything about exposure.) The supposed superiority of the M9 may come about because its exposures are factory-biased a bit lower. The sensor itself has in all probability the same dynamic range as before.

 

That's OK, but I still want the biasing feature, separate from and distinct from 'exposure compensation', which I don't use in any case. Most knowledgeable photogs knew that few films could be exposed at the ISO value printed on the box. Instad we set an Exposure Index (EI) that suited us. Pros did usually expose Kodachrome 64 at an EI of 80, to preserve highlights better, and get better general colour saturation. And the printers too liked that (a printer, in those days, was a fellow with ink on his fingers.) Now this is simply a firmware matter. Why not do it? Because Mama Leica always knows what is best for you?

 

The old man from the Press Room

[Leicaiste]

(This is why I have asked for a permanent means to 'bias the exposure' without the infernal blinking light in the finder, but it sems that the people who just happily blaze away at whatever exposure the camera is handing them, cannot understand the thinking -- clearly because they do not know anything about exposure.)

 

Lars,

 

I think you know that it was offered in the past by Leica on the R9 for multiple field metering (see attachment).

 

Lucien

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!

[erl]

Advertisement (gone after registration)

Lars, I read what you say and basically agree with you. But, like to old film box EI that you and I disregarded and used our own value, you can do that exactly the same way with the M8/9. Create, just once, a user profile to work as your 'standard' preferred ISO setting, apply it and forget it. How is that different from dialling in a different film speed from the 'coded' film cannister?

[pgk]

Shooting any camera using a centre weighted metering system which cannot take scene contrast into consideration will inevitably lead to some degree of inaccurate exposure if relied on without due consideration (that is not to say that the automated multi-pattern metering sytems are often substantially better). I find my M8 to be surprisingly accurate on its metering in many situations but I always check the histogram if in doubt and adjust accordingly (I assume the M9 will behave similarly). No disrespect, but 'blown highlights' are IMHO due to inattention by the photographer. Like with transparency film, precise exposure is an essential part of digital technique for optimal file production. We now have instant feedback via the histogram facility - a real time exposure review system - and this actually requires an adjustment in the mindset of 'traditional' film based photographers rather than modifications to a pretty effective metering system. A test shot and histogram review is better than pedantic metering IMHO!

 

I gave a talk on raw conversion last night and I always start by discussing exposure, and 'appropriate' exposure for optimal raw conversion require highlights to be on the very edge of the histogram. I always advocate test shots to establish precise exposure and the use of manual control if possible. I'd actually say that precise exposure is more efficiently established with my M8 than with my dSLRs much of the time because the interface is far simpler.

[Leicaiste]

Lars, I read what you say and basically agree with you. But, like to old film box EI that you and I disregarded and used our own value, you can do that exactly the same way with the M8/9. Create, just once, a user profile to work as your 'standard' preferred ISO setting, apply it and forget it. How is that different from dialling in a different film speed from the 'coded' film cannister?

 

One of the problem with the User Profiles on the M9 is that it memorize all the settings (save the language). And when you want to change one setting one and for all, like dialing -1/3, you have to do it in each of the four profiles.

 

I would prefer one main profile (in place of the snapshot profile) and 4 sub profiles.

 

Lucien

[ptarmigan]

I've read about everything that's been written so far about the new M9, but I haven't seen any comments about the new sensor's capacity to handle hightlights. I've never had one problem with my M8 and I'm delighted with what I've read so far about the improvements in the M9. But there's one problem shortcoming to all three digital cameras I have, M8, and Panasonic G1 and Z1. And that's the consistent problem with blown highlights. My understanding is that all digital sensors fall short when it comes to highlights versus shadows. Would the new M9 owners please comment on this for the rest of us? Is this a technological problem with sensors that has no solution in the near future?

 

Ultimately it's all about dynamic range, understanding exposure and knowing how your camera sensor is going to behave. Every sensor has a limit whether its 5 stops or 14 stops.

 

If you've ever shot slide film, you'd know all about narrow dynamic range and having to expose correctly. Of course different films behaved differently, slide film being the most exposure critical whereas negative film have maybe up to 2 stops exposure latitude.

 

I tend to treat digital exposure just like slide film exposure. Highlight control is often important and I generally find that the M8 sensor copes pretty well with the recovery of around 1 to 1.5 stops of highlights. As always though, if the DR of the scene exceeds that of the camera sensor then you either use a graduated ND filter, use HDR or you decide 'do I want to retain highlights or shadow detail'.

 

You then expose accordingly. The histogram in the camera is a reasonable guide and with practice you can learn to read it quite well. The ideal exposure has no lost shadow and no blown highlight - unless of course you want either or both!

 

Because we can 'fiddle' with exposure and recovery sliders doesn't mean you can be sloppy with exposure with digital. The more you do of both, the more effect there is on colour rendition, saturation, shadow noise etc. etc. So it's always best to expose accurately 'in camera' and that will solve most of your exposure issues.

 

One final point, smaller sensors typically have less DR that the larger ones and this is often why small format P&S cameras struggle with blown highlights or blocked shadows more than the M8 and others.

[jonoslack]

Excellent Ian

I've found that in bright contrasty light I tend to set -2/3 on the M9, and then in ordinary daylight -1/3 and in low light 0 EV compensation.

 

I don't like blown highlights, but rarely mind blocked shadows (each to his own).

 

I'm not sure that I agree with David - I feel that the M9 deals with highlights pretty much the same as the M8.

 

finally IMHO it's the photographer who blows out highlights, not the camera! (there is always a choice)>

[Guest guy_mancuso]

Need to realize the metering has changed in the M9 due to a bigger sensor as well and the metering pattern has changed slightly. But I did a review of the M9 as well along with a MF Phase P30+ which is pretty much the gold standard and the M9 faired fairly well. Yes it did have less DR but that is pretty much expected but any blown highlight was easily dealt with in C1 with highlight recovery but from memory slightly better than the M8. I did not deal with this specifically in my review but have many comparisons with uploaded raw files of both systems together and the M9 separately in conditions I would not call easy on any sensor.

If your interested it's here M9 Another Leica Journey (Report) - The GetDPI Workshop Forums

[Leicaiste]

I (re)discover that my M9 is blowing the highlights only with one lens: my 35/1,4 Asph.

 

I forgot about it because on the M8 I was almost never using that lens but was using the 28/2 Asph instead.

 

http://www.l-camera-forum.com/leica-forum/leica-m9-forum/100480-m8-versus-m9-regarding-frames-lenses.html

 

Looking back at pictures made with the M6 and that lens, I remembered that it was already there. As soon the light is coming from 180° in front of the camera, even if it's a grey sky, the picture looks washed.

 

It' is time to send it back to Leica.

 

Lucien

[robsteve]

With many others, I have concluded that the M8 as set by the factory exposes about one half f-stop too liberally. Therefore I have the exposure compensation constantly engaged for an 'under-exposure' of 1/3. (This is why I have asked for a permanent means to 'bias the exposure' without the infernal blinking light in the finder,

 

Lars:

 

I had Kindermann do this for me on my R9 and DMR combination when the R9 was in for service. I had them globally dial down the metering 1/3 stop. It was something the technician could do when setting up the camera via the service software. I wonder if the same could be done with the M8/M9?

 

Once this was done, the DMR exposed almost perfectly in just in auto and matrix metering.

 

Robert

[ptarmigan]

Excellent Ian

I've found that in bright contrasty light I tend to set -2/3 on the M9, and then in ordinary daylight -1/3 and in low light 0 EV compensation.

 

I don't like blown highlights, but rarely mind blocked shadows (each to his own).

 

I'm not sure that I agree with David - I feel that the M9 deals with highlights pretty much the same as the M8.

 

finally IMHO it's the photographer who blows out highlights, not the camera! (there is always a choice)>

 

Of course you are right about the photographer and we should not be surprised if the M8 and M9 are similar in their 'native' DR as I believe the latter just has more of the former in terms of real-estate. The real trick for me is understanding where you can 'push'. Like in the olden days with film, we knew how film behaved if we pushed/pulled in exposure/development. We should do the same with digital.

 

So when the DR of the scene is wider than that of the sensor I'm starting to understand how much I can open up shadow detail before I start to suffer with noise, how much highlight recovery I have and which raw converter is best for which conditions, high ISO etc. So I know where I need to expose 'in-camera' for highlights or shadows and by how much.

 

Strangely enough (or not! ) my D2x behaves differently to the M8 in most of these areas.

 

Digital can make you lazy if you are not careful. I honestly try to do as little PP as possible doing most 'in-camera' whenever I can.

 

I do recall some eegit on the Phase One stand at a show telling me I was mad using a dSLR and that if I switched to a MF back, I'd have so much DR that I could even adjust a non high key image into a high-key image using C1. He then showed me how to do it. I just replied, "can't I just sort out the lighting and do it properly in the first place - it's not that hard in a studio after all".He just shrugged and wandered off incredulous at my response. He told me I wa living in the dark ages! I think not.

 

I don't want to be spending hours in front of a omputer screen when I could be out shooting pictures

[hammam]

All digital cameras have issues regarding DR and highlights. Maybe, but some are worse (or better) in that respect than others. The Panasonic G1, for instance, that the OP mentions, is one of the worst offenders I've met, no matter what metering mode you're in, at least for high contrast scenes (except if you spot meter on the highlight of the scene, with the predicted very under-exposed image.) I just sold mine for that very reason. And haven't Fuji been praised for their very good DR in the Finepix S3?

[ptarmigan]

All digital cameras have issues regarding DR and highlights. Maybe, but some are worse (or better) in that respect than others. The Panasonic G1, for instance, that the OP mentions, is one of the worst offenders I've met, no matter what metering mode you're in, at least for high contrast scenes (except if you spot meter on the highlight of the scene, with the predicted very under-exposed image.) I just sold mine for that very reason. And haven't Fuji been praised for their very good DR in the Finepix S3?

 

Yes but you still need to know your equipment and how the sensor reacts. Just because you have a wider DR that shouldn't make you lazy. Some camera sensors will be better (or worse) than others, after all the technology can vary considerably. I'm talking RAW here but we should also acknowledge that for JPEG its even more crucial to know how the sensor behaves and how the in-camera processing affects the final outcome. You have very little headroom for manipulating JPEG in PP especially exposure/recovery and WB.

[ahl]

I agree with the previous correspondents. I came to Digital from an SLR with spot metering and slide film so was always very conscious of blown highlights and had my meter set at -1/2 stop.

I usually set one of my profiles on the M8 at - 1 stop but the M9 seems to require less adjustment: probably -1/3 rd to -1/2 stops. Could this be as a result of the different metering pattern on the shutter curtain?

We now have the luxury of a histogram and also blown highlights warnings so whatever the contrast of the particular scene we are recording we can always compensate for the dynamic range shortcomings.

We should not worry excessively about the characteristics of the sensor but just meter for the highlights and sort out the shadows in post-production.

I personally prefer highlights with some detail and darker shadows especially when printing.

Is the M9 much better than the M8? Who cares? Let's just enjoy whatever cameras we can afford: they are both wonderful tools capable of outstanding results and both a delight to use. Alain

[Advertisement]

Hallo,

Look here: Leica M

[lars_bergquist]

Lars, I read what you say and basically agree with you. But, like to old film box EI that you and I disregarded and used our own value, you can do that exactly the same way with the M8/9. Create, just once, a user profile to work as your 'standard' preferred ISO setting, apply it and forget it. How is that different from dialling in a different film speed from the 'coded' film cannister?

Why, that is EXACTLY what I have done! BUT that does not prevent that infernal finder light from permanently blinking "hey you idiot, you have forgotten to remove your exposure compensation!" And while I never use exposure compensation as such (if I must guess at half the exposure, why not guess at the rest?), imagine what mess this creates for people who actually do.

 

And therefore I want the two functions, which are entirely different in theory and in use, also to be separated in the menu system. It is that simple.

 

The old man from the Age of Kodachrome 64 at EI 80

[lars_bergquist]

Shooting any camera using a centre weighted metering system which cannot take scene contrast into consideration will inevitably lead to some degree of inaccurate exposure if relied on without due consideration (that is not to say that the automated multi-pattern metering sytems are often substantially better). I find my M8 to be surprisingly accurate on its metering in many situations but I always check the histogram if in doubt and adjust accordingly (I assume the M9 will behave similarly). No disrespect, but 'blown highlights' are IMHO due to inattention by the photographer.

I too find the M8 metering system amazingly good. But as you point out, this does not absolve us from using our brains. The contrast ranges of most scenes we encounter are vastly larger than anything that can be recorded on film, paper or silicon. Some intelligent adjusting is called for, and there is no gizmo that can do that for us -- because electronics can meter light levels, but cannot understand what the subject is!

 

Like with transparency film, precise exposure is an essential part of digital technique for optimal file production.

Back to dear old Kodachrome, in other words ... I do think I have shot more 'chromes than negative frames during the last half-century. That is in fact what lies behind my request for 'exposure biasing' or rather, 'meter biasing'.

 

I gave a talk on raw conversion last night and I always start by discussing exposure, and 'appropriate' exposure for optimal raw conversion require highlights to be on the very edge of the histogram.

Well, yes -- if we are speaking of diffuse highlights. Specular highlights, being actually mirror images of light sources, and the light sources themselves, like Old Sol, show luminances that are vastly greater than that of any 'ordinary' diffuse highlight. Like in the old days, we have to permit these to burn out. Any attempt to put these 'on the edge of the histogram' would reduce the rest of the picture to Stygian darkness.

 

And only YOU can decide what is a light source. A pleasant sunlit summer landscape is normally not a light source -- but it becomes one, when it is outside a window and you are trying to photograph something in the room!

 

The old man from the Age of Kodachrome 64 at EI 80