Sean Reid s Review of the M9

[glenerrolrd]

I am hoping someone can set me straight because I don t get it? I always though the issue with noise in a digital rendering was more than just "digital grain". Noise seems to be used interchangeably for high ISO performance (not necessarily by SR but in most discussions).

To my eye .....there is absolutely no difference between the M8 and the M9 when it comes to high ISO performance. IMHO it still sucks. At 640 you have a loss in dynamic range, color saturation and resolution. Regardless if some photographers find higher ISO s acceptable .....I could see absolutely no improvement in high ISO performance. And I could easily see that the Canon 5DII was better in this specific test.

If I apply this to night photography( a joy with the M s ) ....I frequently lose images to lack of dynamic range. I can t expose for the shadows without blowing the highlights . It seems to defy logic to advise ..don t underexpose really anything at higher ISO s . I find about 1/2 to 3/4 stop latitude in bringing back shadows . The image quality suffers.

I understand the point about what you see in a print and this is a good point made frequently when comparing canon 5DII to the D3/D700 . But I believe thats only part of the discussion.

The other issue that jumped out at me in the review as the discussion of IR contamination . The M8 was unacceptable to most people without filters and the M9 is better.....but the contamination is still there? Compare the blacks from the 5DII to the M9 ....am I the only one that can see the difference?

In the big scheme of things ..the M9 ..is whatever Leica can build and I will certainly be trading my M8 s for M9 s...but in these two critical areas it seems that a lot of smoke is being blown about improvements.

(a qualifier..this is not a slam on SR review..it seems to show everything you need to know).

Is this stuff relevant or do I need another "education" ? :ROTFL:

 

[Guy Mancuso]

Roger part of the issue why it may not be in any of the reviews for that matter is the updates for LR and C1 profiles are missing when these folks actually did there testing and those algorithms for noise are just not being applied when they did there tests. The basic problem is the Raw software just cannot read the M9 properly. Now that C1 is out with a profile , now is the time to actually test that and run it in a profiled software. All the information from the M9 is wrapped in the DNG. Basically all the settings and in camera raw processing of vignetting, cyan drift and noise control is wrapped up in the DNG so any software with the M9 profile or update can read that, right now that is not being read by the software.

So these test on noise actually need to be reprocessed in C1 for the moment at least to update those tests.

We need to remember all camera data is wrapped in the DNG and until it can get properly unlocked by the software we won't see those benefits the M9 has like noise for example.

Geez i butchered the spelling here . Not enough espresso

 

[Guy Mancuso]

Roger hope that made sense but the reviews need to be updated now since there is probably for one new firmware which may have some real changes to the files themselves, since these folks most likely used some older firmware. But we honestly don't know if any color changes and such where actually changed. So that maybe doubtful but more important the raw software can actually start to read what's behind the DNG wrapper. So on noise I would not put to much stock yet on that until it is redone with the correct profiles or updates to LR. It's unfortunate that these folks had one hand tied behind there back without the correct raw processing profiles and or updates to LR.

 

[Riccis]

Roger - I am not one to care about technical comparisons but in my eyes I do see a better high ISO performance in the M9 over the M8/M8.2 (see the thread with samples I posted at the LUF http://www.l-camera-forum.com/leica-forum/leica-m9-forum/98714-debunking-online-myths.html).

Cheers,

 

[JSK Rangefinder]

Riccis these are very good examples

the only thing I see as far as noise is...
on the first image (upper right center) some Chroma Noise
but I can't say for sure based on this web sample
Riccis can confirm it, maybe I'm wrong

other than that I love it.

 

[JSK Rangefinder]

Guy don't tell me you don't have one M9 by now to play around?

You seriously have to get your hands on one a.s.a.p.

 

[Guy Mancuso]

Well actually I am hoping to try one out real soon here and give a users report. It does look nice and this M9 does have me missing the M system.

 

[glenerrolrd]

Roger - I am not one to care about technical comparisons but in my eyes I do see a better high ISO performance in the M9 over the M8/M8.2 (see the thread with samples I posted at the LUF http://www.l-camera-forum.com/leica-forum/leica-m9-forum/98714-debunking-online-myths.html).

Cheers,

Riccis

Thanks for the real world test. If noise = grain then your images prove the point that the higher MPs translate into a smoother look. And I acknowledge that a lot of people look at noise as essentially digital grain.

However it is also obvious that the images are struggling with limited dynamic range and desaturated colors. The colors you control and this could be your choice..they are quite attractive renderings for a wedding . But if say a face was caught in shadow you would struggle to bring it back.

You don t see these issues with a D3/D700 until you get to 1600 . I am not expecting D3 sensor performance (but for my use I would trade for it). Rather I am pointing out or suggesting that discussions of high ISO performance go beyond grain .

This in no way takes away from the improvements your images show or your conclusion that high ISO performance has been improved?

 

[JSK Rangefinder]

Well actually I am hoping to try one out real soon here and give a users report. It does look nice and this M9 does have me missing the M system.

I thought by now, Leica would send you one

they told me I'm gonna get my M9s in two weeks

is there a member that can let you use one over the weekend?

 

[Guy Mancuso]

Not here in town but trying to get one from Leica for a spin.

 

[nostatic]

I understand the point about what you see in a print and this is a good point made frequently when comparing canon 5DII to the D3/D700 . But I believe thats only part of the discussion.

What point are you referring to? That the extra resolution of the 5D2 is lost unless you print large or crop heavily?

 

[Guy Mancuso]

Well Leica just called me . I WILL have a M9 and several lenses here Thursday for a long week end and actually put it to use on a golf tournament job on Friday. So there you go plus I will do some testing as well. Guy

 

[Riccis]

What I see in some of those images is just the JPG compression (if I remember correctly I saved them anywhere between 6 and 8 to keep a decent file size).

I don't know what the D3/D700 look like but, IMHO, the M9 does an excellent job throughout the ISO range and there is no excuse (at least for me and my style of working) to not be able to come home with great shots regardless of lighting conditions.

Cheers,

 

[glenerrolrd]

Roger hope that made sense but the reviews need to be updated now since there is probably for one new firmware which may have some real changes to the files themselves, since these folks most likely used some older firmware. But we honestly don't know if any color changes and such where actually changed. So that maybe doubtful but more important the raw software can actually start to read what's behind the DNG wrapper. So on noise I would not put to much stock yet on that until it is redone with the correct profiles or updates to LR. It's unfortunate that these folks had one hand tied behind there back without the correct raw processing profiles and or updates to LR.

Guy

I follow you 100% but I expect that the raw conversion issues could actually be a limiting factor. Leica and C1 worked hard to optimize the DNG conversions with the M8 and they seem to have established a very high standard.....getting the most out of the files. Adobe isn t likely to have the same focus on optimization and Phase may not be all that motivated. You can t optimize a digital system by throwing the file over the wall to adobe to process.

I was just reacting to the very limited push back I have seen from Sean s review. We can only go with what he presented but the performance ...again as shown.....has limited improvement over the M8 files. So if his review is to taken seriously .....doesn t anybody else see the issues. There has been a lot of flap over the independence of the reviewers (as selected by leica) and I have no doubt that Sean has shown us some of the best tests out there.......but the conclusions don t seem to reconcile with what I am seeing. Again...from the tests...

(1) improved ISO has not been demonstrated beyond the obvious benefit to reduced noise in prints.

(2)IR sensitivity is better but without filters doesn t match the 5DII or the M8 using filters. Conclusion..don t have to use filters but IR contamination will slightly affect image quality.

(3)his example in B+W appears to have an unidentified or discussed "band" across 1/3 of the image.

:wtf: good are his reviews if the examples are not reflected in the conclusions reached ?

Again I think all the right stuff has been said by Sean and we all choose to see what we have already concluded. ( I am as guilty as anyone). Your response as I understand it is " you can really tell until the raw conversion software has the right profiles and the new firmware is tested" ......(good points). Riccis point is that in his practical application he sees an improvement and that he doesn t pay much attention to the tests?

Now if you tell me that my read of the report doesn t match yours....thats a totally different perspective.

 

[JSK Rangefinder]

Well Leica just called me . I WILL have a M9 and several lenses here Thursday for a long week end and actually put it to use on a golf tournament job on Friday. So there you go plus I will do some testing as well. Guy

huh Leica delivers

finally :thumbs:

 

[Riccis]

Riccis point is that in his practical application he sees an improvement and that he doesn t pay much attention to the tests?

This is correct... I am not one to pay attention to the technical stuff and as long as the tools I use are reliable don't really care about anything else... After all, I still shoot Holgas

My only goal when making images is trying to faithfully record a moment in time even if the results are not technically perfect. I will take a grainy, dark image that moves or invites the viewer anyday before a techically perfect but boring one.

Cheers,

 

[Vivek]

Good stuff, Riccis.

I also enjoyed your Noct shots on the other site.

 

[Guy Mancuso]

Roger I just got a call back from Leica . They are sending me a M9 24 lux, 35 cron, 90 summarit and thought of you on this one a 135 . They wanted me to make sure I did the firmware update, something I have been saying all along and I mentioned my C1 was up to date for the M9. All good on there end and I promise I will do a noise test like i did with my MF gear that i have shown on the forum. I don't have a M8 but we can at least see what is going on and i can shoot my P30+ next to it to compare noise because we know it's standard if we want which is the same 6.8 micron. Not sure it would benefit us but maybe interesting.

 

[glenerrolrd]

What point are you referring to? That the extra resolution of the 5D2 is lost unless you print large or crop heavily?

Sean explains that the M8 files maybe negatively impacted by upresing them for printing or conversely that the lower magnification required to print the larger M9 file may reduce the apparent noise. Its still there just not a prominent . This was a common discussion when comparing the noise of a 5DII file with the d3/D700. The argument is that you have to use similar magnifications to have an apples to apples comparison.

So an 18mp sensor with the same exact ISO performance will look a lot better than a 10MP sensor when compared at the same output size. My read of the tests would be that he proved the noise would be less but that the DR ,color and resolution look the same.

 

[nei1]

Thanks for the explanation Guy,you see I dont spend that much time on these forums but having checked it all out a little I see what you mean..handbags at dawn!
Glad youre getting an M9 ,be nice to read a working photographers views.
Dont suppose youd try putting it under the tap for me to see how it copes with adverse weather?