cam
the need for speed
i totally agree with that. i sometimes wonder who makes those decisions? but then again, i think their press conferences suck as well, so what do i know?They need to hire an editor, then. Not a curator, but somebody with high-brow universal appeal (i.e. National Geographic) not high-brow five-or-six-figures print to match my studio humidor appeal (i.e. New York gallery of mohdern photougraphee) or technically-good-but-so-snapshottish-it-must-be-art (i.e. Foam Magazine).
Maybe I'm just cranky 'cause I got sick last night.
truly, though, i'm hoping there will be more images released soon.
and i'm sorry you got sick. feeling better today or is it that nast bug going around that hangs on forever?
Wow, the impressions/comments here are diametrically opposed to what I'm seeing on other sites. Interesting. Brutal, some would say!
I think people expedct miracles based on the fact that it is a Leica and it is expensive. However, it is not expensive because it is going to give one IQ above and beyond anything else in its class. It is expensive because it is a luxury good.
Spicy
Well-known
the only thing about those (i only looked at the jpgs) that was impressive was the london shot at 1/25 with a 135 (probably on a tripod). impressive detail if you cropped, although i haven't the slightest doubt that a d800 would absolutely monster it.
pretty pathetic for a 7k body, if you ask me. yes, i know they're OOC jpgs, but for $7000, you'd like to hope that at least they could develop it enough to look like something a bit better than a snapshot from a 2 year old canon powershot.
pretty pathetic for a 7k body, if you ask me. yes, i know they're OOC jpgs, but for $7000, you'd like to hope that at least they could develop it enough to look like something a bit better than a snapshot from a 2 year old canon powershot.
mani
Well-known
If you mean over on LUF? No-one dares stick their head above the parapet there any longer.
As far as I can see, the CMOSIS sensor isn't quite a match for the OMD/E-PL5. Especially when it comes to the tartan pattern noise in any areas that are underexposed. As for the skintone examples we've seen - they look fine if all your friends and relatives have faces made of rubber.
douglasf13
Well-known
Semilog is right in that it seems people confuse artifacts and artificial detail as "more detail," which makes AA-less cameras en vogue these days, even to some pros. AA filters are relatively expensive, so it isn't as if most camera manufacturers use them for the heck of it. Of course, as pixel size keeps shrinking, the need for AA filters diminishes, but we're not to that point, yet.
All of that being said, the real advantage of Leica removing the AA filter is in regards to astigmatism with angled light rays. Edge performance, especially with Leica's short registration distance, is improved by removing the AA layer, since it does cause astigmatism leading to softer corners. Leica has to sacrifice moire and false detail for better performance across the frame. Conveniently, this false detail has become known as more sharpness or "sparkle" to many shooters, so it became win-win for Leica.
I don't get moire all that often, but, as Semilog suggests, it is likely because of the nature that I often shoot my M9 (handheld, slight focus errors, lower light, etc.). Throw the M9 on a tripod at F5.6, and moire certainly shows up.
maitani
Well-known
I agree, partly,
I don't expect it to beat D3S or D4 in IQ, but at that price, for me it's imperative to match at least D700 or so output at higher iso, I'd be extremely happy if yes, extremely disappointed If not.
bests
maitani
Not speaking of LUF, no. I visit there maybe a once a year, if that.
1joel1
Well-known
Keeping my M8 for now.
furcafe
Mentor
I wasn't disagreeing w/his technical explanation, but rather the implication that a camera without the anti-aliasing filter is just "amateur-bling." Even if all the detail seen without an anti-aliasing filter is "false" or an artifact (& I'm not sure this is true, not being an engineer), if someone likes the look & it works for their photography, I see that as a positive, not negative. The best analogy I can think of is resolution v. grain in film photography. But then again, I don't view photography as equating to scientific imaging. I would also question whether or not Canon or Nikon or Sony or Leica chooses to sell, or not sell, cameras without an anti-aliasing filter has anything to do w/"professional pride" or engineering integrity; I think it has more do whether or not they can make a buck (or rather Yen or Euro) out of the deal.
SDK
Exposing since 1969.
The debate about the desirability of antialiasing filters in digital photography is moot, both in the sense of being endlessly debatable, and in the other sense of the word moot: not always being relevant in practice. To insist that there is only one correct solution to the problem of aliasing, the hardware solution, is dogmatic. There are excellent cameras commercially available that have AA filters (Canon DSLRs, Nikon D600, D800, D4) and cameras that have no AA filters or very weak ones (Nikon D800E, Leicas and Phase One equipped medium format bodies). I don’t think the variety of approaches to the problem we se are all market driven, because engineers do evaluate the upside and downside to hardware AA filters. Physical AA filters will lower the resolution of a camera system a bit, and this effect is always there, a built-in limit that cannot be undone in postproduction. The alternative is software AA filtering. Fixing moire in post processing has become easier for cameras that lack hardware AA filters, and can be applied when there is a problem, either to an affected part of an image or to the whole image if needed. Again, resolution will be decreased if software AA filtering is done. In practice, moire may or may not show up in images made with a camera without an AA filter. Things that prevent moire include oversampling, defocused images, lens aberrations, diffraction, motion blur and low contrast or low detail subjects.
Because I mainly do landscape photography and use f/8 or f/11, diffraction (and focusing errors, etc.) takes resolution well below the Nyquist frequency of the sensors of the D800E and M9P. For me moire usually only shows up at f/4 or f/5.6 with subjects like high contrast ice, with really good lenses like my 75mm Summicron, or 100mm/2 Makro Planar. A 35mm/2.8 C Biogon I used to have was the most prone to provoking color moire on my M9P, and I decided to give it to my nephew who is learning B&W film photography along with my old Hexar RF body.
I realize that there are many kinds of photography that do really benefit from hardware AA filtering, such as fashion and wedding photography and portraiture, because cloth is one subject matter that is very likely to provoke color moire, and busy commercial photographers don’t have the time for extra postproduction to correct extra image flaws.
PS. In our lab we use digital microscope cameras that use a 3 shot system for producing color. The cameras we have (SPOT RT3 Color) have monochrome sensors with 3 LCD filters for red green and blue light that rapidly turn on and off during 3 exposures to produce a full color image. This is great because it prevents color moire completely, but it’s only optimal for static specimens, as moving subjects wind up with funny colored edges, which is actually a lot more disturbing than moire. I suppose some day we may have sensors that solve the problem, probably be even more oversampling and the use of better detector material, like some modification of Eric Mazur’s lack silicon (that is too sensitive to infrared right now).
Because I mainly do landscape photography and use f/8 or f/11, diffraction (and focusing errors, etc.) takes resolution well below the Nyquist frequency of the sensors of the D800E and M9P. For me moire usually only shows up at f/4 or f/5.6 with subjects like high contrast ice, with really good lenses like my 75mm Summicron, or 100mm/2 Makro Planar. A 35mm/2.8 C Biogon I used to have was the most prone to provoking color moire on my M9P, and I decided to give it to my nephew who is learning B&W film photography along with my old Hexar RF body.
I realize that there are many kinds of photography that do really benefit from hardware AA filtering, such as fashion and wedding photography and portraiture, because cloth is one subject matter that is very likely to provoke color moire, and busy commercial photographers don’t have the time for extra postproduction to correct extra image flaws.
PS. In our lab we use digital microscope cameras that use a 3 shot system for producing color. The cameras we have (SPOT RT3 Color) have monochrome sensors with 3 LCD filters for red green and blue light that rapidly turn on and off during 3 exposures to produce a full color image. This is great because it prevents color moire completely, but it’s only optimal for static specimens, as moving subjects wind up with funny colored edges, which is actually a lot more disturbing than moire. I suppose some day we may have sensors that solve the problem, probably be even more oversampling and the use of better detector material, like some modification of Eric Mazur’s lack silicon (that is too sensitive to infrared right now).
icebear
Mentor
Maybe I'm just lucky that I haven't encountered the false detail or artificial detail yet
but then I'm also not running around shooting "street" by mounting my cameras on a tripod
.On a side note :
All shots being discussed here are labelled with "M240 pre-production model"
Oh, believe me, from a consumer perspective I agree. However, shouldn't people really be comparing the M-240 vs. the M9 regarding improvements in ISO? The reason I say so is that for many M users, a DSLR is not an alternative anyway.
ramosa
B&W
Keeping my M8 for now.
In contrast, I will be getting the M and then selling my M8 (or maybe keeping it as a backup). I have played some with the images in LR4 and think they work out fine with simple edits.
douglasf13
Well-known
While I would agree that semilog's wording was a little harsh, I understand his point, and he's technically correct. There are two advantages to not using an AA filter: the performance of angled light rays at the periphery of the sensor and cost. The D800E is a particularly unusual specimen, in that their AA-canceling filter (they don't actually remove the AA) isn't any cheaper, and the camera's registration distance probably doesn't warrant the improved edge performance. I'd imagine that's why semilog characterized as "amateur-bling," as there isn't much reason for it. Proper sharpening of an AA'd sensor will not have a resolution disadvantage when compared to an AA-less sensor, and it will show less artifacts. Heck, in the studio, MFDB shooters are usually battling moire by stopping down and using diffraction to their advantage. My MFDB used to drive me nuts in the studio, back when software didn't do as good of a job at correcting moire.
Don't get me wrong. I love my M9, and it's currently my favorite camera, but I think it's important for M9 users to understand why the camera doesn't have an AA filter. Again, cost and angled light rays. If Leica could figure out the astigmatism thing and still use an AA filter, I'd likely welcome it, unless they make a big jump in pixel count, and then the AA wouldn't be needed, anyways.
p.s. Leica also uses one of the cheapest IR sensor filters out there from Kyocera, FWIW
ramosa
B&W
For me, yes. I can't see myself ever going back to a DSLR, so my options are M8 (which I have), M9, or M. (The MM won't work for me, nor will a film RF.) The new M's imagery seems better than I expected. I wanted images to be as good or better than the M9 at base ISO--and I think they will be (see Chris' image of the woman and dog). And better imagery at higher ISO (which is a given IMO). Plus, I will enjoy the better processor, LCD, weather sealing, battery life, etc. (But don't need to the video or live view or focus peaking, etc.)
douglasf13
Well-known
semilog
curmudgeonly optimist
Many professional photographers (and doctors, lawyers, etc.) do all sorts of things that have no sound technical basis.
Again, note that the major camera companies that fabricate their own sensors do not sell models that lack AA filters.
semilog
curmudgeonly optimist
semilog
curmudgeonly optimist
For most applications in microscopy, it's pretty straightforward to match the numerical aperture of your objective to the camera, so that you're oversampling. Diffraction then becomes a de facto AA filter.
We're approaching that point with camera sensors in many cases as well, as pixel sizes descend into the 2µm range. At that point only very good lenses used at wide apertures with immaculate technique will be capable of showing aliasing artifacts. But we're not quite there yet.
The point about the M sensor not having an AA filter because of a design trade-off with astigmatism in the peripheral field (when legacy lenses with a short node-to-sensor distance must be employed) is a very good one. That is an authentic design trade-off, and a reasonable one for Leica's engineers to have made. The rationale for leaving the AA filter off of the D800E is weaker.