There is one difference between us though. I have posted my results for peer examination. No strict comparison using raw and showing the M8 outresolving a 5D or 1DsII has been posted. I am NOT suggesting you guys are wrong at all. I know you know a lot more about this subject than I do, but I don't understand why nobody has spent the ten mintues it takes to actually compare the same shot from the two cameras, with each at peak performance. The two lenses I used matched quite well, but as has been discussed, there are other choices and focal lengths which match well enough that distance can be used to match the scene exactly. It's getting to the point where I am about to try my Leica dealer's patience with a third visit. I am no expert in post processing though.
I think everyone is making a lot of sense, and I am grateful to everyone here for humouring my interest in the question of resolution.
Pondria, I agree with your suggestion about careful use of terminology. But shouldn't 'resolution' mean the camera's absolute maximum resolution for a particular image/angle of view rather than pixel pitch. That seems to me to be a peculiar interpretation of the term. Surely we should expect a 5D to outresolve a 20D.
brainiac wrote:
Pondria, I agree with your suggestion about careful use of terminology. But shouldn't 'resolution' mean the camera's absolute maximum resolution for a particular image/angle of view rather than pixel pitch. That seems to me to be a peculiar interpretation of the term. Surely we should expect a 5D to outresolve a 20D.
Right. I don't mean that Resolution=Pitch. I meant that 20D/30D has the highest resolution due to the Pixel Pitch.
With the same lens, from the same distance, we measure the minimum distance of the two black dots ( or pair of parallel lines ) that the camera can resolve. I would be very surprised if 5D beats 20D. Nature Bird folks already have done this
Mike, you are right 5 % more or less is not a big deal. though brainiacs 2nd comparision showed more difference in between them and it is difficult to see for me that on a 30x40 print the m8 would show more detail.
***
I might be the only one left here who reads Puts, though I found his 3rd text on the m8 can not be completly ignored. He tried to compare the 5D and the M8 with the same pixel area, means he used a wider lens on the 5D. While the M8 shots were very sharp, the 5D missed that sharpness in the beginning. Users downloaded these images and sharpend both, the m8 was quickly oversharpend, the 5D catched up to about the same sharpness and detail of the Leica and thus the 5D would have a benefit from the missing part of the crop
that's another version of the term that I think is misleading, and the reason why I used such an odd method for doing my test. The 20D outresolves the 5D with the same lens from the same distance, but it will not do so if you take the same picture with the same angle of view. I think most people's understanding of how well a camera/lens resolves is 'with the same _effective_ focal length, and the same picture bounds, measure the minimum distance between two just distinguishable dots'. That interpretation, and only that one, tells you how sharp your pictures are going to be.
And by the way I've never met a pro who uses only one system... :)
We should meet. I guess I am a pro, because I get paid to do it and have done for 20 years. My Rollei 6008 has languished in a cupboard for 10 years. My Mamiya 6's are sold. I really like using one system. It breeds familiarity, not contempt.
After have tested myself the M8 vs 1Ds, for pure curiosity as I already chose the M8 as my "semper-mecum" camera, I'm sure that the 1DsII (not to mention the upcoming 1DsIII) has a clear edge in maximum resolution, once used with excellent lenses.
But this has nothing to do with the wonderful look of Leica images, nor with the features of the M-system.
It's over fifty years that much bigger cameras yield better resolution than Leica-M cameras.
Everyone agrees that a Mamiya 7 with one of its great lenses did, does and always will outresolve any Leica M with the best of its great lenses...
But Leica images are great nonetheless, because technically there's more than pure resolution in a quality shot and aspects like colors and bokeh do play a big role in drawing a pleasing image.
Not to mention the portability of the system, which sometimes makes for capturing moments you otherwise couldn't.
For me, having a digital Leica-M that compares favorably with the 1Ds and 5D is thrilling.
Ok, now give me a Mamiya 7D 39Mp and I will eat my words...
ClubShooter wrote:
When you combine it with an AA filter it's quite noticeable. While 10MP couldn't show a difference in high-contrast areas and test targets, in actual real-world photography where the bulk of image information is low and medium contrast the combination of crisp optics and AA-free image formation makes a pretty big difference. AA filters start attenuating low-contrast detail well before 13 or 16MP or whatever, and on the 1Ds2 only high-contrast detail was resolved. I understand the AA filter on the 5D is improved, but it's not a brick wall OLPF, because such things don't exist outside matlab.
Digital photography is a pretty mature technology and a lot of design decisions are based on lessons learned in addition to theoretical engineering principles. So I'm just wondering, why are most DSLR's equipped with AA filters. Is this just bad judgement on the part of their designers? A slavish reliance on what is considered to be sound digital signal processing practice?
Do Canon, Nikon, et al test their products with various degrees of AA filtration and make decisions based on empirical results? If leaving off the AA filter would make the Pentax K10D or Rebel XTi produce images rivaling a Leica, why didn't the designers do it?
brainiac wrote:
that's another version of the term that I think is misleading, and the reason why I used such an odd method for doing my test. The 20D outresolves the 5D with the same lens from the same distance, but it will not do so if you take the same picture with the same angle of view.
Right. Because 5D will contain more "Details" although outResolved by 20D.
Marco wrote:
...
It's over fifty years that much bigger cameras yield better resolution than Leica-M cameras.
Everyone agrees that a Mamiya 7 with one of its great lenses did, does and always will outresolve any Leica M with the best of its great lenses...
...
I doubt. ( Forgive me. I don't have idrect experience with either of them )
Again the terminology issue.
The Lager format will capture more "Details" but M-lens probably out-resolve Mamiya in terms of Resolution (lppm).
The small Zeiss lens went into SONY 505V P&S can resolve 100 lppm. ( But it does only within the tiny image circle. Thus the total details contained within the small area was little short of 3MP worth )
Pondria wrote:
Right. Because 5D will contain more "Details" although outResolved by 20D.
This type of confusion is precisely why I refer to 'it' as "net image detail" -- the net quality after taking everything into account from capture through output...
But then I know my 'it' won't satisfy others because it is an entirely subjective definition -- and the factors that comprise it cannot be easily compared empirically. That's why I make prints to show other photographers so we can discuss the actual result and I can confirm or deny my own impressions... Clearly not a perfect system, and very difficult to "show" print differences on the web, but it works for me.
Jack Flesher wrote:
Carsten, to be more clear here is my main M8 issue:
There are only TWO possible fixes for the IR bleed issue on the M8: One is to replace the IR cut filter over the sensor with one that actually cuts ALL IR. Two is to filter every lens with that same filter and leave it on the lens 24/7/365. BOTH of those solutions will require new profiles and updated firmware and that is a very minor problem.
With that said, *MY MAIN GRIPE* is that Leica chose the "filter the lens" option...
<...>
I totally agree with everything you have written. Where I disagree is how we express it, respectively. I also really do not want filters on my lenses or no filters and purple funk all over. However, I think that calling Leica "arrogant" for making this choice is wrong, and unsubstantiated. I believe that we have sufficient proof to say that given that they do not want to waste the incredible M lenses, they actually had no choice.
I will bet you anything that half of the Leica M development team are "naked lens" fanatics! These are the people who design these lenses to the incredible limits we see. Why would they want filters there? Simply because a hot-mirror filter on the sensor would 1) introduce colour shifts for wide angle lenses, and 2) require coded lenses to be able to get the famous Leica colour and quality back. A thicker IR absorbing filter on the sensor would introduce ghosts and maybe other image quality problems, as seen on the R-D1. The only choice left is filters on the lens, which is why they chose it. Yes, we still require corrections, but it is optional, and you can choose not to use a filter. In the case of non-Leica lenses, or even Leica lenses, you can choose not to code them. There is a lot of choice this way, none the other way. For B&W fanatics, there is almost no issue here. No filter, period, and incredible-looking images.
And so I have chosen to accept that Leica has done what was possible, that they are unhappy with the choices available, and that they have made the best one they could.
FWIW, if you were repeatedly expressing your unhappiness that we have to use IR filters, I would have to join you in that. I just don't think that calling Leica arrogant is appropriate, especially repeatedly.
Pondria wrote:
Right. I don't mean that Resolution=Pitch. I meant that 20D/30D has the highest resolution due to the Pixel Pitch.
Haha, that is just classic, Pondria :) Read that again. You are funny.
I think resolution w.r.t. digital cameras is more commonly taken to mean the number of pixels distributed over the image than the number of pixels per unit area. The latter is really mostly of theoretical interest, whereas the former is of primary interest, as long as it is accompanied by high quality.
Given that the sensor in the DMR and M8 is almost the same, how come Leica didn't put an IR filter in the cover glass for the M8 as they did for the DMR.
Best regards,
Tim
This is incorrect. The M8 sensor cover glass is an IR filter; it is just that it is a weak IR filter because it is only 0.5mm thick.
Tom
Well, I went and pestered my dealer again today, asking this time for a *reason* that my M8 was taking longer to come back than other units shipped to Leica later than mine. Apparently there was a defective piece in it which took a while to come back into stock. I wonder what that was?
Anyway, I got my 50 Lux Asph back. It still feels a little stiff to focus, especially with the little focus-nub. Does anyone have one which is butter-smooth?
Mike Hatam wrote:
I also saw Jack work up some enlarged prints, and was stunned by how well the M8 files held up to very large print sizes. They held up better than the 5D when print sizes got very large. Again, not a tightly controlled test, but good enough for me to make that conclusion for myself.
Mike
Excellent point, one has to compare prints and not jpegs on the internet. A lot of what one sees and reads on the internet is pure BS.
Tom
Hi Guy, no, you did not make this goof! BTW, I am still waiting for my M8. It had not reached my Alaska dealer b4 leaving for the holidays - now on the east coast! Seems your fixed M8 is to be back today!
Merry Christmas!
Tom
Guy Mancuso wrote:
Jack well said kind of refers back to my earlier post about the logical thinking part adding up all the sums of the parts kind of gives you the whole in the end.
