M9 iso 1000

[Bartie]

Advertisement (gone after registration)

This image was taken on a recent trip to NY....iso setting of 1000 ,the file was put through Nik Software Dfine 2....I think it puts to rest the moans and groans about poor high iso performance of the M9.I hardly ever go higher than iso 1000 even on my slr cameras,so I`m more than happy with this result.....wot do you guys think?

 

Andy

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!

[cheewai_m6]

i think that is fine. i can't say i've got an m9, but when i use the m6 and summicron, at f2, i think for all situations that i've personally needed, iso1000 would be enough. for me anyway, i only ever print to 6x4, i'm not a pro. noise reduction reduces the sharpness, but at 6x4, you can't even see any effects. i know this from my d90 and using noise reduction from lightroom.

[ndjambrose]

Not sure I'd call that low light. It looks pretty well lit to me. I'd be quite happy shooting 400 speed film under the illuminations in NY, especially round Times Square :-)

 

But you raise an interesting question, and I'm always interested in seeing quality low light results because I'm a frequent low light shooter, and performance in that area would make a real difference to me. It would be great to put the myth about low light performance to bed once and for all.

 

But to do that I think shots need to contain a range of shadow and highlight tones, be close enough to the subject to be able to view detail (especially in typical noise areas), and be in conditions with average light value of EV 4 or less. EV 6+ (which is typical of most illuminated city scenes) isn't the best test case.

 

An experiment is usually more convincing if it's designed to demonstrate clear success or failure conditions. In other words, pick the difficult subjects where performance window is narrower and see how it goes.

[rosuna]

It is impossible to get an idea from an image so small...

[scc]

The reds are a bit off there, real magenta cast to the traffic light for example.

[Michael Tyler]

In my opinion, the reputation for poor high-iso performance in the M's is in relation to today's digital slrs. If you're comparing to a film M then yes, the noise/grain performance is quite good. But compared to a Nikon or Canon full-frame dslr, it's lagging by several generations. ISO 1000 should be almost noise-free on a D700 or 5DMKII, better than 100 speed film. As a professional shooter (whose main subjects are for delivery to clients, not for personal enjoyment) and who has shot extensively with an M8, I can tell you that the poor ISO performance is counterbalanced by the joy of the rangefinder mechanism and the insatiably beautiful rendering of the images when light is ample. It's a quandary I still wrestle with in my head.

 

I will say this though. If Leica could equal the noise performance of those other cameras, there would be no difficulty justifying the $7K price of the M9.

[Rolo]

Advertisement (gone after registration)

It would be great to put the myth about low light performance to bed once and for all.

 

But to do that I think shots need to contain a range of shadow and highlight tones, be close enough to the subject to be able to view detail (especially in typical noise areas), and be in conditions with average light value of EV 4 or less. EV 6+ (which is typical of most illuminated city scenes) isn't the best test case.

 

An experiment is usually more convincing if it's designed to demonstrate clear success or failure conditions. In other words, pick the difficult subjects where performance window is narrower and see how it goes.

 

Neil, perhaps you have an example from digital or film that illustrates an image that would be more convincing ?

[Jerry_R]

You can only answer youself - when you would have two shots 160 and 1000. And compare them and see how many details you loose. And probably with bigger resolution.

 

As for WWW - it is fine, but then you do not need M9 for that.

 

When I first used DFINE - I was also amazed. Pictures looked great. Until I compared them to native ISO. But fro WEB, full HD projector - it is more than enough I would say.

[scott kirkpatrick]

I'm using AUTO ISO with the max ISO at 1000 for casual shooting with the M9. Results on well lit shots that I don't plan to greatly enlarge are just fine at 1000 (as the first example shows). But I'll lock in 160 or 320 (actual, not actually 200 and 400 as was the case with the M8) when details matter the most, and work around slow shutter speeds. Similarly, I'll lock in ISO 640 or 1000 when shutter speed matters more than grain.

 

No big deal. Doesn't everyone do this?

 

scott

[Jerry_R]

I do similarly.

 

I got two friends, tripod and external flash. Using them - my photography gained new quality. They both bring more benefits than having useful ISO 25 600.

[Bartie]

Not sure I'd call that low light. It looks pretty well lit to me. I'd be quite happy shooting 400 speed film under the illuminations in NY, especially round Times Square :-)

 

I didn`t mention low light,it`s obviously a well illuminated scene ,I`m just happy with the performance at the iso setting.

 

 

Andy

[innerimager]

Andy- In fact, using an excellent noise reduction program like define or noise ninja does yield excellent results. I remember when the canon 5D came out and achieved a new level of high iso, low noise. I was using the D2H at the time. In camera, the D2H did not match the canon, but applying noise ninja yielded results every bit as good in all ways, and took seconds. I do particularly like define, using it's selective masks, allowing application to shadows ( or wherever needed), only.

[scc]

I agree NR programs are help a lot, but the output often has a 'plastic' quality, a bit unrealistic looking.

Often I prefer the unadulterated image than what comes out of NoiseNinja.

[innerimager]

I agree NR programs are help a lot, but the output often has a 'plastic' quality, a bit unrealistic looking.

Often I prefer the unadulterated image than what comes out of NoiseNinja.

 

Yes, agreed, you have to be very careful, and using a selective approach which I think can be done in Ninja but definitely in Define helps a lot.

[ndjambrose]

Neil, perhaps you have an example from digital or film that illustrates an image that would be more convincing ?

 

Alas, I am not blessed with an M9. But if you lend me yours I'll be happy to shoot one.

 

Actually, any test scene will do. A fairly typical example might be a low light portrait - say a figure sitting close to a light source in a dark room. There would be a range of well lit highlight tones on the face, some midrange tones on the torso, and shadow tones in increasing distance from the light source. And then we'd have an understanding of how the M9 performs at higher ISOs, and the effects of chroma and luminance in areas of different tonal value.