sebboh wrote:
why? as long as it doesn't effect the performance with wide angle lenses is the difference in per pixel sharpness that important to you on a 24mp sensor? i bet it still outresolves the m9.
You know, I used to believe that, until I started using the GXR. Believe me the difference is night and day. In my estimation, the M9 and the M will deliver the same resolution. Besides, there is such a great satisfaction looking at the crisp files at 100% that do not require any sharpening. Once you experience that, no camera with AA filter will be acceptable.
edwardkaraa wrote:
You know, I used to believe that, until I started using the GXR. Believe me the difference is night and day. In my estimation, the M9 and the M will deliver the same resolution. Besides, there is such a great satisfaction looking at the crisp files at 100% that do not require any sharpening. Once you experience that, no camera with AA filter will be acceptable.
I've owned a camera without an AA filter. The problem is that artifacts and aliasing can give an impression of false detail, and if you properly sharpen an AA filtered image, the resolution is comparable with less moire and artifacts.
douglasf13 wrote:
I've owned a camera without an AA filter. The problem is that artifacts and aliasing can give an impression of false detail, and if you properly sharpen an AA filtered image, the resolution is comparable with less moire and artifacts.
I have not seen artifacts with the GXR. Aliasing is somewhat more frequent but I don't find it too serious as only photographers notice it. Ok, I know that the GXR pixel density is equivalent to 30 mp, which is a somewhat safe territory where aliasing is not as problematic with lower density sensors. No matter what, whether real or false details, the files look fantastic and in my experience you need to downsize a AA filtered file that is 50% larger to obtain a comparable detail. Sharpening will only give sharpened detail, the look is not the same.
sector99 wrote:
It will come out soon enough when the M's proof images arrive for the Elmarit M 28 f/2.8 ASPH and the biggie末the Super Elmar 21 f/3.4 ASPH. Images from those will tell the whole truth about sensor performance on wides.
I would say the critical M tests are the Elmarit 28 version 1 and the biggie -- Super Angulon 21/3.4. If those results fly, we have a digital RF winner.
edwardkaraa wrote:
I have not seen artifacts with the GXR. Aliasing is somewhat more frequent but I don't find it too serious as only photographers notice it. Ok, I know that the GXR pixel density is equivalent to 30 mp, which is a somewhat safe territory where aliasing is not as problematic with lower density sensors. No matter what, whether real or false details, the files look fantastic and in my experience you need to downsize a AA filtered file that is 50% larger to obtain a comparable detail. Sharpening will only give sharpened detail, the look is not the same.
I'm not sure if you saw theSuedes window frames artifact example with the M9, and how to sharpen AA files to be better than AA-less files, but I don't have the link, and this is a topic for another discussion. Cheers!
douglasf13 wrote:
I'm not sure if you saw theSuedes window frames artifact example with the M9, and how to sharpen AA files to be better than AA-less files, but I don't have the link, and this is a topic for another discussion. Cheers!
Sure, I have seen them. I can't remember in which thread either. However, I think this is really a very subjective matter that divides many experts. And to be honest, that would be the first time I do not fully agree with the Suede's opinion. However, I fully understand why most users prefer to have an AA filter.
sector99 wrote:
The M9 isn't a stellar tool in color so the hope here is a change in sensor architecture will regain a modicum of critical respect for the M.
You don't use an M9 do you? Colour separation is excellent, however high sensitivity to reds can cause some interesting ruddiness which I correct by having colour profiles in different lighting conditions.
sector99 wrote:
It's fairly simple with the CMOSIS boys末they deliver a wider dynamic range with a thin but still present AA filter (According to patent information at CMOSIS).
AA filter does not affect dynamic range. You might have CFA (eg. Bayer/x-trans) and AA filter mixed up, but a CFA won't affect DR, it only impacts demosaicing, noise and colour separation.
sector99 wrote:
As for the flat "Pixel architecture"末I've taken Leica's wording at it's core meaning: Flat means flat末No micro lithographic wedding cake "Lenses".
The sensor topping WILL have micro-lenses. They are referring to flatter circuitry in front of the photosites, the depth of which usually impacts wide incident angles on CMOS sensors quite negatively.
sector99 wrote:
It will come out soon enough when the M's proof images arrive for the Elmarit M 28 f/2.8 ASPH and the biggie末the Super Elmar 21 f/3.4 ASPH. Images from those will tell the whole truth about sensor performance on wides.
I doubt these will be anywhere near the most demanding lenses. the 21/3.4 even with coding switched off has nowhere near the colour shifts of the voigtlander 15/4.5 and 12/5.6 on the M9. The older 21/3.4 Super-Angulon will be interesting since it is almost unusable on the M9.
The ultimate test will be a converted 16/8 Hologon
Tariq Gibran wrote:
The M-E is just like the M9 with the manual frame selector removed from what I have read.
The M really does look great. Wish it was coming out sooner actually.
Glad to see Leica breathing life into the M line while respecting the historical DNA as much as possible. Absent frame selector is nothing new: early DS M3 had none and I want one for that reason. Leica could also drop the RF (as with M1), or the entire optical VF (Ig, MD series). Such reductions could allow a modest price drop but, more importantly, encourage the radical variation: manual "advance" and no electronic displays. Variety is the spice of life, even for Leica, and I want the original form factor and a proper place to rest my thumb! Epson got the manual advance right, but no-one has dared to drop the LCD in this digital age. I would even remove the LCD frame counter. Now that's what I call an M to concentrate on the photographic essentials, my fellow chimpers.
I have one question guys, everyone is talking about the AA filter, but I can't find any mention of it anywhere. Could someone give us a link or something? Thanks.
edwardkaraa wrote:
You know, I used to believe that, until I started using the GXR. Believe me the difference is night and day. In my estimation, the M9 and the M will deliver the same resolution. Besides, there is such a great satisfaction looking at the crisp files at 100% that do not require any sharpening. Once you experience that, no camera with AA filter will be acceptable.
i've seen lots of sharp raw files from the gxr and found the difference between them and the NEX-5 at 100% to be much smaller than i expected. the m9 has a much better per pixel sharpness than the gxr. nevertheless, i don't care that much about ogling photos at 100%. the d3x still outresolves the m9 and i'm sure this new m10 will too whether or not it has an AA filter. even playing with the raw files of the m monochrome didn't give me any strong desire for an AA filter free camera. i wouldn't mind it, but i don't really understand making that a deal breaker, especially from someone who primarily shoots film.
sebboh: did you run the GXR rawfiles in RPP? I would have to say that in my opinion the difference between the M9 and the GXR with regards to per pixel sharpness was quite small, though using ACR took its toll with the GXR in that aspect. Loved the M9 output though and I would too say it is better. Colors are better from the GXR though, by quite a margin.
During a short period I had all three cameras actually, and the NEX5 would have sufficed had it not been for the wide angle problems, it ticked most of the other boxes when it comes to cost, size, resolution, speed and handling (much faster than either the M9 and the GXR) and the per pixel sharpness was in practicality on par with the GXR though the GXR require less work in post pro and very little sharpening, almost none.
I still have the GXR, it is the best compromise for me, sluggish shot to shot speed and buffer is putting me off at times though.
EDIT:
Just checked out the M-E announcement on DPR, isn't it more or less an M9 at virtually the same price?
Does not make sense...
If indeed the ME is purely a repackaged M9, in terms of internal components, then I am somewhat disappointed as this means it will suffer from the same M9 problems of sluggish buffer rate and random banding issues (which might be card related).
I agree with Joe these ME 'features' makes the new M even more attractive, if it is indeed $7K. The sensor performance is promising, and with the Maestro processor, should be a more responsive camera...
I find it interesting that the CMOSIS site indicates the sensor can do 5 fps, whereas the M spec list indicates 3 fps... would be nice to see even 4 fps.
Regarding high ISO performance: I've yet to use a camera where the max ISO is more than 'usable.' As such, I expect the M's ISO 6400 to be in line with the M9's ISO 2500, though perhaps a bit better. That said, decent ISO 1600-3200 would be great, as I find I rarely go above this even with my DSLRs.
kosmoskatten wrote:
sebboh: did you run the GXR rawfiles in RPP? I would have to say that in my opinion the difference between the M9 and the GXR with regards to per pixel sharpness was quite small, though using ACR took its toll with the GXR in that aspect. Loved the M9 output though and I would too say it is better. Colors are better from the GXR though, by quite a margin.
During a short period I had all three cameras actually, and the NEX5 would have sufficed had it not been for the wide angle problems, it ticked most of the other boxes when it comes to cost, size, resolution, speed and handling (much faster than either the M9 and the GXR) and the per pixel sharpness was in practicality on par with the GXR though the GXR require less work in post pro and very little sharpening, almost none.
I still have the GXR, it is the best compromise for me, sluggish shot to shot speed and buffer is putting me off at times though.
EDIT:
Just checked out the M-E announcement on DPR, isn't it more or less an M9 at virtually the same price?
Does not make sense......Show more →
i did run the gxr files in rpp and they looked better, but to me the m9 files actually looked worse in rpp than they did in lightroom (kinda had an oversharpened look). i do like the gxr for some of its controls and it's performance with rangefinder wides. i like the shape and shooting experience with the NEX-7 much better though.
the new m looks like a welcome improvement that should satisfy many of the complaints users had about the m9, though still not a camera for me. to many things i don't want compromising the things i do like about it. i'm just not looking for the m experience.
The M typ 240 looks interesting, LV and R lens compatibility is not too bad on a M camera huh? Neither the ME nor the M typ 240 is for me, though the latter is an interesting bridge camera.
I wonder who Leica has collaborated with (if any) with the EVF, smells Ricoh or Oly. Hopefully the peaking is well implemented.
Not that I want to sound negative on this new camera, which is deservedly well received, but CMOSIS claim 42 employees on their Website. That is an awfully small number, and it is not M sensor production that can keep them alive. So buying an M is taking a gamble on this minnow staying alive in a world of really big fish.
