M9 Colors at Night — Best Way to Shoot High ISO?

[nggalai]

Advertisement (gone after registration)

Addendum: More extreme desaturation example, with really awful mixed lighting – mind, not a good image, but hey, it was late and I was weirded out.

 

 

 

50mm, f/2, 1/60s, ISO 500, pushed by 1.2EV in Capture One, Saturation -25. Added a shadow area crop. Again, grain added for aesthetic reasons:

 

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!

[Advertisement]

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

[zlatkob]

Here are two comparative shots at approx 100% crop of 640ISO pushed two stops and 2500ISO not pushed

 

The ISO 640 shots are terribly blotchy in the big dark rectangle that makes up the side of the piano. The ISO 640 shots also have much more noise in the white cloth on the left side and on entire upper right of the photo, including the wall and the framed pictures. In your comparison, ISO 2500 not pushed looks much better.

[erl]

I think I am returning to my earlier held belief that in high contrast scenarios pushing one stop @ 640 is OK but two stops is pushing it.

 

Noise resides in underexposed areas. Extending underexposure extends the habitat of noise IMO. For 'street' work, Mitch's favoured genre, I think it has real application, but I doubt its great value in 'normal' lighting. The example scenario I used above has a range of both types of lighting in it in the one shot. The noise abounds in the shadows if I lighten them. 100% crops are savage, however, at normal en;largement, all ISO's, even 2500 become acceptable, depending on your level of acceptance. I tend to accept more noise than many I suspect, in appropriate shots, eg. street. Not in commercial work, unless 'art' effects are part of the brief. As I type this I have just finished a commercial shoot of a hairdressing salon, all shot at 160ISO, on a tripod with the M9.

 

the blowups will be gorgeous, no noise except for the applause.

[Guest malland]

...The example scenario I used above has a range of both types of lighting in it in the one shot. The noise abounds in the shadows if I lighten them...

Erl, it's not clear to me whether you're pushing the exposure of lightening the shadows. If you're doing the latter, it's best to try just increasing the exposure and doing only minor lightening of the shadows.

 

The technique of "shoot at ISO 640 + push in post" as providing a better signal-to-noise ratio than raising the ISO in-camera has been well proven by Jim Kasson, douglas3f and others. The question, it seems to me, is whether the right, or best, post-processing techniques, including Noise Reduction are being used.

 

EDIT: Another issue occurs to me: in your test you're simply pushing two stops to match the ISO 2500 shot. What I'm doing is pushing the minimum necessary to get the look I want. That includes picking the lightening that occurs from increasing the Clarity slider as well, which reduces the need for some of increase in Exposure. I find that I often push back on the Exposure slider after increasing Clarity.

 

The following thought occurs to me: in making a comparison between the two methods the "scientific" way would, at first be to increase Exposure by 2 to match the ISO 2500 of the other picture. But that is not the reality what would be done in either case. It seems to me that the best way to make a fair comparison is to process both pictures to their "optimum look, including Noise Reduction." For the ISO 2500 picture that might mean to push back on Exposure a bit; for the ISO 640+push picture that might be to push Exposure substantially less than 2 stops. My experience that, in such a comparison the ISO 640+push picture should generally look better. I would be very much interested in seeing the results that you would get with a few comparisons done in this way.

 

Finally, I should also state that the ultimate test is not which picture has less noise, but which picture is better aesthetically.

 

—Mitch/Bangkok

Bangkok Obvious [WIP]

Eggleston said that he was "at war with the obvious"...

[erl]

Mitch, I will attempt to address your suggestions ASAP, but currently I have some pressing diversions demanding my time. I will come back to this, maybe in 24+ hours.

[baci]

Interesting discussion - and very helpful. To add to erl's comments, last weekend my little boy had his third birthday party (held inside because of unappealing weather) and I tried disregarding ISO and simply shooting for ideal depth of field and sufficient shutter to freeze movement. The lighting in the room was low and flat, and I have to say that the results were less than stellar. It was a first attempt at a (for me) novel process, and practice makes perfect, but the results I got were consistent with erl's inside shots, and a long ways from the sort of result which seem to be achieved in situations with large amounts of contrast between illuminated areas and naturally occurring darkness. Flat lighting does not seem to work to the advantage of this process. For now. For me.

 

YMMV!

 

Michael

[Guest malland]

Advertisement (gone after registration)

I don't know what the answer is, but I would reflect on the fact that flat, dull daylight won't make for good pictures either — or flat, fluorescent lighting that is bright enough not really to require high-ISO will also result in ugly pictures. Isn't that what you're talking about, rather than real, low-light photography?

 

—Mitch/Bangkok

Bangkok Obvious [WIP]

Eggleston said that he was "at war with the obvious"...

[baci]

Hi Mitch

 

Not really - I've shot lots in this space and have in the past respected the holy trinity of ISO / Aperture / Shutter speed. I'm using a Zeiss Biogon 35 at f2 most of the time because of the need to get enough light to the sensor. I can get nicely textured well balanced images, but with limited depth of field and lots of blurred subjects due to 1/30 and lower shutter speeds. The thought of the kind of freedom of capture provided by disregarding ISO is attractive indeed in this setting

 

I know what you're saying, photography is first and foremost (literally) about light, but the quality of the light is OK, there's just not that much of it. Using an aperture of f8 and a shutter speed of 125 in those conditions gave me too much noise across the board. In post I set the white point and black point then used NR to take out the obvious noise but the loss of detail was severe. No problems, I'll keep exploring. Next step is to be less extreme with my settings (f4, 1/60) and to fix the color temp in camera.

 

Great fun this exploring. will fire off some night shots soon to see how those frames respond.

[Guest malland]

Baci, if you're trying to shoot at f/8.0 and 1/125 sec it seems to me that you're simply not getting enough light to the sensor. Also, it occurs to me that I've been shooting at fairly large apertures because, when it's dark, you simply don't need that much DOF because light falls off so rapidly that you can't see that far. Perhaps you should indeed try the same scene at f/4 and 1/60 sec. I'll be bery interested in knowing how it goes.

 

—Mitch/Bangkok

Bangkok Obvious [WIP]

Eggleston said that he was "at war with the obvious"...

[jip]

Lets say you want to shoot 1/60th and you need ISO 1250, but you want to shoot at 640 can't you just set the EV correction to -1 and shoot at ISO 640, then push +1 in post?

[baci]

Baci, if you're trying to shoot at f/8.0 and 1/125 sec it seems to me that you're simply not getting enough light to the sensor. Also, it occurs to me that I've been shooting at fairly large apertures because, when it's dark, you simply don't need that much DOF because light falls off so rapidly that you can't see that far. Perhaps you should indeed try the same scene at f/4 and 1/60 sec. I'll be bery interested in knowing how it goes.

 

—Mitch/Bangkok

Bangkok Obvious [WIP]

Eggleston said that he was "at war with the obvious"...

 

In he interests of science I'll give it a go tomorrow morning and report back!!

[dwbell]

Couldn't it be that "night" shots don't have many values in the II III IV area, but rather represent either lit or crushed black?

 

I wouldn't call Erl's examples flat lighting, but they do contain a good cross section of illumination which night shots may not.

 

Simple to test the theory by just shooting a step wedge or even a Gretag target.

 

I can well believe that off camera post processing to increase the gain will get better (since the M9's introduction), especially compared to Leica's in camera processing power. So I do believe it should 'better' - how much better is unclear to me in non-night situations.

[erl]

Just quickly 'looking back in' here for a moment.

I am coming to the conclusion that pushing 1 stop ie. -1 comp or manually doing it is very manageable. After that, noise can become an issue. It will depend on the light quality. By that I mean contrast ratio, not intensity or lack thereof.

 

As a side issue, maybe of interest:

Yesterday I shot for two hours in a salon solely lit by fluoescnt light, designed to be flat. I shot at 160ISO (no pushing) and produced some excellent contrast in the files. Colour temp was set at 'Daylight' (my default setting) in the M9 and I corrected from the tungsten illumination very quickly in post. Colour is perfect. It surprised me.

 

I will return through the weekend to further address the O.T.

[Guest malland]

Lets say you want to shoot 1/60th and you need ISO 1250, but you want to shoot at 640 can't you just set the EV correction to -1 and shoot at ISO 640, then push +1 in post?

You could but, in my view, this negates the underlying idea of the technique we're discussing. I'm starting with the assumption that the lighting in these scenes (at least in the photos that I've posted) is so difficult that exposure measurement is problematic, and that you can't go over with an incident meter to take a reading (street photography). That means that you won't know in advance that 1/6oth at ISO 1250 is what you need. In this situation, you expose at ISO 640 at, say. f/2.8 and I/60th. Then in post you push the Exposure Slider in LR the minimum that you need to get the look that you want — this may turn out to be less than 1/60th and ISO 1250.

 

Again, it helps to think of the camera not having an ISO after ISO 640: all you have is an electronic gain control,which after ISO 640, gives better result if it's pushed up in LR rather than in-camara. In other words, I feel that it's helpful not to think in film terms at all.

 

—Mitch/Pak Nam Pran

Paris au rhythme de Basquiat and Other Poems [download link for book project]

[Guest malland]

The attached file was shot in a shopping mall lit by fluorescent lights, with an Elmarit-21 ASPH lens, at ISO 640 pushed 0.35 stops, f/4.0 at 1/1/180 sec. What it shows is how ugly this lighting can be when it is the only lighting in a indoor situation; not at all like my earlier night shots with fluorescent lighting on a dark night. (The attached file is a reduced PDF file, which compromises it's quality as well).

 

—Mitch/Bangkok

Bangkok Obvious [WIP]

Eggleston said that he was "at war with the obvious"...

L1000849.pdf

[JimKasson]

Mitch, one of the subtleties of "ISO-less" exposure is that not all raw processors push the same way. It's actually more confusing than that: Lightroom Process Version 2010 works very differently than Lr PV 2012. The PV 2010 Exposure control works like the ISO dial on the camera; it applies gain equally to all channels. In that regard, it works like the Exposure adjustment layer in Ps.

 

The Lr PV 2012 Exposure control attempts to reproduce the soft clipped highlight behavior of film. Take a look at pages 26 through 30 of this presentation:

http://www.kasson.com/ll/RawJMK.pdf

 

You can see see film and digital camera tone curves, and also Lr PV 2012 and Ps (equivalent to Lr PV 2010) tone curves.

 

If you're trying to reproduce the saturation behavior of the M9 when the ISO is cranked up and just using ISO-less means to get lower noise, you should use PV 2010. If you want soft clipping of the highlights, you should use PV 2012, but you may experience a bit of highlight color shift.

 

Other raw processors use other algorithms. YMMV.

 

Jim

[Guest malland]

Jim, thanks for posting here. Your article that I linked in post #1 of this thread has been very helpful in a practical sense. I don't know whether you saw the example night photos that I've posted earlier in this thread. I used the Lr PV 2012 process version and was happy with the results, maybe because the scenes were all of high contrast.

 

I'm not sure in what type of light and contrast situations "soft clipping of the highlights" is preferable to the "saturation behavior of the M9," and vice versa. Or is the main issue the possible highlight color shift that you refer to? Can you elucidate a bit?

 

—Mitch/Pak Nam Pran

Bangkok Hysteria (download link for book project)

[JimKasson]

I'm not sure in what type of light and contrast situations "soft clipping of the highlights" is preferable to the "saturation behavior of the M9," and vice versa. Or is the main issue the possible highlight color shift that you refer to? Can you elucidate a bit?

 

Mitch, I like the PV 2012 Exposure algorithm a lot. I have found the color shifts, although a theoretical issue, to be no problem at all in practice. Hard clipping causes its own color shifts, too. I don't like the hard clipping inherent in digital cameras, and always preferred the more forgiving behavior of film cameras, and consider the soft clipping of PV 2012 to be a beneficial sade effect of pushing in post.

 

The only reason I bring up PV 2010 is that some of the people posting in this thread seem to be seeing differences in tonality between cranking up the ISO and pushing in post, and if you want to keep things apples to apples, you'll at least test with PV 2010.

 

Jim

[Advertisement]

Hallo,

Look here: Leica M

[Guest malland]

Jim, thanks. Very interesting. I hope that we can call on you to answer questions that come up when more people try the technique being discussed here.

 

BTW, have you done or seen any tests on how the M-Monochrom would act in terms of the point at which pushing in post becomes better than increasing the ISO in-camera. If it's a question the configuration of the sensor photo sites, I would assume it would think it should have the same ISO 640 breakpoint as the M9...

 

—Mitch/Pak Nam Pran

Paris au rhythme de Basquiat and Other Poems [download link for book project]

[JimKasson]

Jim, thanks. Very interesting. I hope that we can call on you to answer questions that come up when more people try the technique being discussed here.

 

I'd be pleased to help if I can.

 

BTW, have you done or seen any tests on how the M-Monochrom would act in terms of the point at which pushing in post becomes better than increasing the ISO in-camera. If it's a question the configuration of the sensor photo sites, I would assume it would think it should have the same ISO 640 breakpoint as the M9...]

 

If it's the same sensor and the only change is the removal of the Color Filter Array, then the ISO where the noise starts to fall off relative to pushing in post should be twice as high, since the removal of the CFA makes the sensor twice as efficient at converting photons to electrons. Therefore, I'd expect that the point where you should stop increasing the ISO is 1250.

 

If it's a different sensor, I don't know the answer, since I've never tested -- or even seen -- an M Monochrom.

 

Jim