I would love to be able to find some WA RF M-mount lenses for the A7r. But if I can't find them there are the questions of the 28mm and the 21m focal lengths for SLR lenses. Since my normal operation would be to carry both a Leica M to NEX adapter and a Leica R to NEX adapter for M-mount and my Leica R lenses I am leaning toward the Leica R 28mm f2.8 Elmarit V2 and I wouldn't need another adapter. In the case of the 28mm focal length I am also considering Contax 28mm f2.8 Distagon though the Leica is considerably heavier, much more expensive and sharper than the Contax. I really am trying to avoid the Contax or the Zeiss 21mm f2.8 Distagon lenses due to their weight and their 82mm filter thread size.
Personally, I am not willing to go with the Contax 28 f/2.8. I had that lens and liked the high micro contrast, but my experience with it was that it had quite a bit of CA that I am not willing to put up with. My copy was also only sharp across the frame at f/8 or narrower, which I think is less than ideal with a high megapixel sensor. I think the Leica R 28 Elmarit v2 is a good choice. For me it looks at this point it will be between that lens, any new FE mount lens Zeiss makes, and the Pentax 31 f/1.8 limited. Currently, I would probably go with the Pentax, but I am hoping Zeiss makes something that makes me change my mind. At 21mm I will probably go with my ZE 21 Distagon, but again I am less concerned about size than many people here.
coffeeshakes wrote:
Thanks for running the test! A quick question:
What is the 24mm SE asph? There is an Elmarit ASPH and an Elmar ASPH, not sure what the SE is? I have an elmar I am hopeful for, so just looking for some more detail
Its the Elmar ASPH (f/3.8).
SE stands for super elmar, used by the 21/3.4 and 18/3.8. The 24/3.8 is only a elmar indeed. But I find it super as well
naturephoto1 wrote:
It is a possibility. When I had the original 19mm R lens (in about 1985?) I did find it hard to control the angle of view and replaced it with the R 21mm f4 Super Angulon which I still have. But I doubt the 21mm Super Angulon will be sharp enough into the edges and the corners. I could probably learn to work with such a wide view and it is a super performing optic. But it is as heavy as the the 21mm Distagons and heavier than the Contax version and I am somewhat concerned about protecting the front element. Also, I am not sure of the lens thread size for using ND or Grad filters.
Rich ...Show more →
The Elmarit-R 19mm f2.8 V2 doesn't have a filter thread, it has a built-in turret with 4 filters: NDx1 (neutral density), yellow-green YG, orange OR, and blue (conversion filter KB12). So probably not so great for landscape.
Some questions to the pixel peepers:
- Have you seen the MTF curves for the Biogon G 28 ?.-At f/2.8 corners are far, very far, from perfect and vigneting is roughly 2 stops.
- Do you expect then that the corners will be sharp at f/2.8 looking at a 100% crop of a 36Mp. sensor in a a monitor ?. ( you realize that this equals to a.picture of 133x200cmtr. )
I'd like to see a comparison with a Leica M240 upsized to 36mp -provided that somebody made a M mount conversion of the Biogon G 28 or, failing that, to compare side by side the ZM21/ZM28 on the A7r and the M240.
We may suppose than the sensors of the A7(r) will have some issues with RF wides but it's not clear yet which part of the 'corner trouble' is due to these issues or to the lens itself.
In case anyone is interested, I went back to the same vantage point from which I did the a7 with rangefinder lenses test, and shot the scene with the M9 and the following lenses:
This wasn't really something I planned ahead of time, rather, I was in the area with that gear and the weather happened to be similar to that during the a7 test, so thought I'd give it a try.
Bottom line, other than the CV12 and 90 Summarit, which look pretty similar on both cameras, there is noticeably better edge sharpness/detail in the M9 images with the other lenses. While the test isn't exactly apples to apples (different day/time/weather/light; slightly different framing; M9 files processed from RAW vs. the a7's in-camera Jpegs), one can still get a pretty good sense of the differences between these two cameras with these lenses.
rscheffler wrote:
In case anyone is interested, I went back to the same vantage point from which I did the a7 with rangefinder lenses test, and shot the scene with the M9 and the following lenses:
This wasn't really something I planned ahead of time, rather, I was in the area with that gear and the weather happened to be similar to that during the a7 test, so thought I'd give it a try.
Bottom line, other than the CV12 and 90 Summarit, which look pretty similar on both cameras, there is noticeably better edge sharpness/detail in the M9 images with the other lenses. While the test isn't exactly apples to apples (different day/time/weather/light; slightly different framing; M9 files processed from RAW vs. the a7's in-camera Jpegs), one can still get a pretty good sense of the differences between these two cameras with these lenses....Show more →
You don't think that using native lenses might involve automatic in-camera correction?
This is corroborated by the similar effect of the CV 12 (no correction available in either camera) and the Summarit 90 (no correction needed in either camera).
Thank you for the sobering post Ron. Back to reality, the A7/R is not the holy grail of RF lenses, I'm afraid. We keep on hoping and wishing for a new test that would show that RF lenses below 50 work well, but I have the feeling we will be disappointed. At this time, the best platform for shooting RF lenses remains surprisingly the M9, not even the M 240 with its non offset microlenses.
edwardkaraa wrote:
Thank you for the sobering post Ron. Back to reality, the A7/R is not the holy grail of RF lenses, I'm afraid. We keep on hoping and wishing for a new test that would show that RF lenses below 50 work well, but I have the feeling we will be disappointed. At this time, the best platform for shooting RF lenses remains surprisingly the M9, not even the M 240 with its non offset microlenses.
After trying the A7r and M240 yesterday I became even more happy with the M9p.
rscheffler wrote:
In case anyone is interested, I went back to the same vantage point from which I did the a7 with rangefinder lenses test, and shot the scene with the M9 and the following lenses:
This wasn't really something I planned ahead of time, rather, I was in the area with that gear and the weather happened to be similar to that during the a7 test, so thought I'd give it a try.
Bottom line, other than the CV12 and 90 Summarit, which look pretty similar on both cameras, there is noticeably better edge sharpness/detail in the M9 images with the other lenses. While the test isn't exactly apples to apples (different day/time/weather/light; slightly different framing; M9 files processed from RAW vs. the a7's in-camera Jpegs), one can still get a pretty good sense of the differences between these two cameras with these lenses. ...Show more → Albi86 wrote:
You don't think that using native lenses might involve automatic in-camera correction?
This is corroborated by the similar effect of the CV 12 (no correction available in either camera) and the Summarit 90 (no correction needed in either camera).
The M9 certainly applies correction, but this is for color shift and vignetting. AFAIK, Leica has never stated there is any kind of sharpening enhancement (as seems to be the case with others, such as the Fuji X cameras). Also, forgot to mention that the CV12 has been hand coded as the 21/2.8 non-ASPH, which minimized the degree of color shift it would otherwise exhibit on the M9.
edwardkaraa wrote:
Thank you for the sobering post Ron. Back to reality, the A7/R is not the holy grail of RF lenses, I'm afraid. We keep on hoping and wishing for a new test that would show that RF lenses below 50 work well, but I have the feeling we will be disappointed. At this time, the best platform for shooting RF lenses remains surprisingly the M9, not even the M 240 with its non offset microlenses.
I think your statement needs to be revised:
"At this time, the best platform for shooting certain wide RF lenses [...]"
ulrikft2 wrote:
I think your statement needs to be revised:
"At this time, the best platform for shooting certain wide RF lenses [...]"
Not really. From all what I have seen so far, nothing beats the M9 in the corners in terms of lack of smearing and minimal color shift. Certain wides work well on the A7/R like the CV 12 and 35/1.2 but not as good as on the M9. I am sure the culprit is the AA filter, which even on the A7R that is advertised as AAless, the AA filter is still there, just rotated 90 degrees.
I am confident that when Sony has the balls to remove the AA filter, we will get stellar results, probably much better than the M9. Until then, the A7R is not an option for RF wides.
edwardkaraa wrote:
Not really. From all what I have seen so far, nothing beats the M9 in the corners in terms of lack of smearing and minimal color shift. Certain wides work well on the A7/R like the CV 12 and 35/1.2 but not as good as on the M9. I am sure the culprit is the AA filter, which even on the A7R that is advertised as AAless, the AA filter is still there, just rotated 90 degrees.
I am confident that when Sony has the balls to remove the AA filter, we will get stellar results, probably much better than the M9. Until then, the A7R is not an option for RF wides. ...Show more →
Edward,
What is your source that the AA filter is still there but rotated 90 degrees in the Sony A7r?
the A7R that is advertised as AAless, the AA filter is still there, just rotated 90 degrees.
Aaargh ... I've been (skeptically) wondering (with hope otherwise) all along if they were using the approach similar to the D800E instead of a true AA-less sensor. Do we have confirmation of this?
_julian_ wrote:
My opinion is that the best bang for the buck for M lenses will be the A7 since the larger pixel wells are a more relaxed design less affected by light falloff and color shift. The low-pass filter/IR hot mirror can then be physically removed leaving only the thin IR absorption plate which will dramatically reduce corner smearing following the same principle that Leica use for their digital Ms.
I believe from posts earlier in this thread that the IR absorption layer is between the two layers of the OLPF, and that the second layer of the OLPF doubles as the sensor coverglass. Removing it leaves a very delicate sensor surface underneath.
You would have to then source a replacement IR filter, but if you did this then yes you would reduce smearing with symmetrical lenses, but at the expense of performance with lenses designed for digital. The refractive properties of the filter stack cause a degree of field curvature that lens designers now take into account (attempting to induce the opposite curvature in the lens itself).
edwardkaraa wrote:
Not really. From all what I have seen so far, nothing beats the M9 in the corners in terms of lack of smearing and minimal color shift. Certain wides work well on the A7/R like the CV 12 and 35/1.2 but not as good as on the M9. I am sure the culprit is the AA filter, which even on the A7R that is advertised as AAless, the AA filter is still there, just rotated 90 degrees.
The only reason to go the rotated-90-degrees route is to make it cheaper to offer two varieties of the same sensor (D800 and D800E), given the A7r sensor is only offered sans AA filter it is much more likely they used an optical flat of the same thickness and RI as opposed to two AA filter crystals. However, the overall effect on smearing is the same.
edwardkaraa wrote:
I am confident that when Sony has the balls to remove the AA filter, we will get stellar results, probably much better than the M9. Until then, the A7R is not an option for RF wides.
Unfortunately I don't think this is going to happen due to the need to offer different models with and without AA filter, and have autofocus (with phase detect via one of the SLT adaptors) and lens performance be the same (particularly lenses designed with the effect of the filter stack in mind).
It's much more likely that Sony will keep the filter stack height and just design retrofocal wide angle lenses.
Toothwalker wrote:
I think that a lens "with the filter stack considered in the design", is nothing else than a lens with an exit pupil farther away from the sensor. A smaller maximum aperture also helps a bit. If there are other measures, I don't know what they could be. It would also not be a strong selling point for a lens if it were matched to one particular sensor.
It's not a case of lenses being matched to one particular sensor, as much as it is being matched to one filter stack thickness.
Brian Caldwell mentions, on separate occasions, accounting for the effect of the filter stack in the design of both the Costal Optics 60mm UV/VIS/IR and the Metabones speed booster.
Given the filter stack induces field curvature the effect could be neutralised by designing a lens with the opposite curvature.
From what understand, it's the microlenses that help/hurt with noise, vignetting and color shift, but the sensor toppings like the IR, AA etc, that influence smearing. The oblong microlenses of the m240 might actually be doing a bit better job than the M9's microlenses, since the camera actually has more colorshift than the M9 (better tuned microlenses equals better reception of light rays but worse color shift at the edges,) but Leica must have changed the whole IR and cover glass setup for the worse. TheSuede had a post about microlens design recently talked about this.
Of course, there also must be differences in the pixel designs of the CCD vs CMOS, so that could be part of it. edwardkaraa wrote:
Thank you for the sobering post Ron. Back to reality, the A7/R is not the holy grail of RF lenses, I'm afraid. We keep on hoping and wishing for a new test that would show that RF lenses below 50 work well, but I have the feeling we will be disappointed. At this time, the best platform for shooting RF lenses remains surprisingly the M9, not even the M 240 with its non offset microlenses.
RustyBug wrote:
Aaargh ... I've been (skeptically) wondering (with hope otherwise) all along if they were using the approach similar to the D800E instead of a true AA-less sensor. Do we have confirmation of this?
I'm curious as well as to how we know this. Is it just speculation based on the D800E using the same basic sensor and/ or ease of manufacture for Sony OR has it actually been confirmed as fact? Source?
What is your source that the AA filter is still there but rotated 90 degrees in the Sony A7r?
Rich
This is what my camera dealer in Bangkok told me, and he has close connections with Sony. Also I remember having read it in one of the interviews with Sony guys but it would be difficult for me to relocate the source with all these posts and threads about the A7R.
Agreed, Douglas. Maybe the oblong microlenses are better for vignetting but they do induce more color shift. I have no indication that the M is worse than the M9 in terms of smearing, as both seem not to smear at all.
douglasf13 wrote:
From what understand, it's the microlenses that help/hurt with noise, vignetting and color shift, but the sensor toppings like the IR, AA etc, that influence smearing. The oblong microlenses of the m240 might actually be doing a bit better job than the M9's microlenses, since the camera actually has more colorshift than the M9 (better tuned microlenses equals better reception of light rays but worse color shift at the edges,) but Leica must have changed the whole IR and cover glass setup for the worse. TheSuede had a post about microlens design recently talked about this.
Of course, there also must be differences in the pixel designs of the CCD vs CMOS, so that could be part of it.
