rscheffler Offline Upload & Sell: On
|
p.117 #9 · Sony RX1 FF Mirrorless (fixed lens) | |
Thanks Matt, I too really prefer the ZM lenses on the GXR. There's something about the higher contrast of these lenses that seems to work well with the camera compared to other lenses. For example I wasn't all that blown away with the 50 Lux ASPH or the CV50/1.5 on it, but love the look of the ZM50/2.
The Sonnar on the RX1 is really sweet. It might be enough of a reason alone to get the camera if that focal length is one of your favorites. Judging from comments in the RX1 images thread, it would seem at least a few have found the RX1 sufficient to replace their DSLRs.
philip_pj wrote:
The ZM 35/2 at f5.6 still has these swathes of colour engulfing whole branches in magenta CA in the extreme top right corner ('near4' I am looking at). Still there at f11, much reduced though. Frankly it's horrible wide open.
The RX1 at same aperture same place shows very little CA. Much more controlled bokeh as well, kind of soft and smooth, very natural. By f11 CA is almost as good as my CY 21/2.8, you have to look closely for it.
The 35C f2.8 does about as well at f5.6 with CA as the RX1, and is a better choice for landscape going on this result, quite a yellow/green cast however. By f11 it is exemplary, a very nice lens.
The CV lens is also well-behaved if a little dead, good CA control except the very extreme, don't go much on the fuzzy bokeh. It fades badly at f11, CA still clearly there, poor greens. Just some reactions.
Just had a quick look at the f2 images in 'far' (dark tree and manor house): that is the sensor at play (RX1), brighter highlights and snappy contrast, excellent shadow detail. The fine branch detail is phenomenal, typical vignetting. Better shaping of the building/seat. I would feel much more comfortable doing post on these files than the others in this set. It's a great result really from what I saw so far.
Thanks for your feedback Philip. I agree the RX1 files are very pliable. As I think I noted, I shot duplicate sequences with ND filtration because the sunny ISO 100 1/2000 f/2 images looked really blown out on the camera's display, but in LR it was no problem to pull them back and they looked absolutely 'normal' compared to the properly exposed images in the sequences. Same can't be said for the M9 files. For those I could recover around .5 to .7 stops but beyond that there was no point. I also think the Sonnar is really impressive for its out of focus qualities at stopped down apertures, not just wide open, though of course with it being a 35mm lens, once stopped down there is much less background separation.
As for colour casts to images, please don't read too much into that. I had to pick a baseline towards which to normalize the images and for some viewers it might be too much on the yellow/cyan side. This might be in part because I'm somewhat 'visually allergic' to strong magenta tint shifts, which I find the RX1 biases towards, as do my Canons, especially in environments with a lot of greenery.
The magenta CA in the Biogon 'near 4' scene is purple fringing. If you look closely, it's also evident, though to a much lesser degree, in the Sonnar files and all the others as well, The C Biogon seems to handle it the best once stopped down a ways. I think the infinity scene is a bit better from which to judge CA. All of the lenses are really good and the CV is the weakest.
zhangyue wrote:
Ron, My bad. I thought you gonna give a thumbnail look for all of them since you left a bar under each image
Anyway, I just download the infinity test and your VC35 is incredible well behaved. (Mine is pretty good, but definitely not these close to Biogon) You should hold it for sure after f5.6, I still feel Zeiss biogon and c-biogon are slightly better, have more contrast. I feel Biogon is slightly better than c-biogon at f8, though difference is extreme small.
One thing I notice before is Biogon has focus shift. given Leica has no LV, I am not sure either I or you reach its full potential or not. I planed to test it on NEX7 but sold it since then. Judge from f5.6 and f8 center, it seems definitely reaching infinity though.
Your test also confirm M9's IQ is still very high even compare with Latest sensor. hopefully, this can mute some sound of Leica basher The detail is just incredible. This kind of test remind me why is justified for high price Leica or Zeiss lens. They perform as should be most of the time. not like a lottery pick for Canikon pro lens.
I agree Sonnar is the strongest performer at f2 at infinity by quite a large margin. but after that, it seems doesn't reach Zeiss m twin performance for infinity landscape shot. But I want reinforce that is pure technically difference, not as much important for real world images. But at f2 OF rendering is the selling point of this lens, which should be more important for most people for real world shooting.
Same as you, I wish they can bring us a M version of this lens, I will be the first on the line. assuming it is in biogon price range.
Thanks Michael. Sorry for the confusion.
I agree I might not have given the Biogon a fair chance by not predetermining whether focus shift is an issue. You and Edward say it has it, so it's very likely. But by testing it in this manner, without focus compensation, it's how the lens would initially respond for a new owner. Maybe more can be eked out of it with some care, but I doubt it would be a significant improvement. I'm not that thrilled with it wide open, especially compared to the 28 Cron. In hindsight I'm glad I started with the C Biogon when I first got into the M system because it's much more of a no nonsense type lens, like the 50 Planar or 21 Biogon.
Yes, the M9 files hold their own. But I do wonder how much of that snap and sparkle is related to false details and sharpness.
rscheffler wrote:
Here's my RX1 vs. M9 & ZM35/2, ZM35/2.8 and CV35/1.2 comparison, including links to the high rez files at various distances: http://www.ronscheffler.com/techtalk/?page_id=188
I think the RX1 does pretty well at infinity, but it's not the best. It definitely, from my observations has some mid zone dip in resolution/sharpness:
Is this the end of the world? Of course not, just depends on what your expectations are from this camera and the type of work you use it for.
colonelpurple wrote:
I am still not convinced there in anything different at infinity.
1. I don't know how or if you are correcting distortion. Some distortion prevention algorithms blur detail
2. f4 is hardly the right f stop for landscape infinity, even at 35mm (or wherever the RX1 is in fact at). I can see things not close to the center veering off at this aperture
3. I have plenty of indistinct infinity at the pixel level with the D800E, ME and OM-D, especially at f4. It also depends where focus is,
I am not sure this pixel level gazing is that useful.
At the end of the day, if you dig too deeply you can also be drilling into sample variation, which produces confusion.
Its better to think of "zones" of lens/sensor competence. The RX1 is definitely in the top zone, where it is exactly is just masturbation!
BTW I like your blog, the comparisons are crisply written. And I prefer examples anytime to words 
best rgds
Thanks Colonel.
Regarding a couple of your points:
1) I believe I stated in my write up that LR was used at standard defaults and extra corrections such as CA and distortion were disabled. What I may not have mentioned was that I also disabled these corrections in the camera just in case they might have been baked into the RAW files. I'm not sure if the RX1 does that and wanted to avoid any possibility it might.
2) The effect can be seen at f/5.6, f/8, f/11. It never improves to a uniform result, though slightly diminishes as the lens is stopped down. I chose f/4 because the effect is somewhat more visible, as is the case when I wrote about this same problem with the Leica 21/1.4 at middle apertures. I agree it's pixel peeping to an extreme, but did jump out at me quite early on in my image evaluations, perhaps because I was used to seeing it with some of my Leica lenses.
While it would seem to be common knowledge that one should stop down more for infinity shots, it depends on the lens. Many peak in absolute sharpness already at f/4-5.6 and when wanting to preserve as much fine detail as possible, would be a logical aperture to use. But I guess it's also a matter of balancing across-frame performance against absolute sharpness. Some lenses are also surprisingly good wide open at infinity. For example the Leica 50/1.4 ASPH is very impressive, so it's not necessarily unheard of to shoot infinity scenes at or near wide open (will be interesting to see how the new Zeiss 55/1.4 fares).
I agree with you about sample variation. It was definitely on my mind but didn't comment on it in my write up and I only had access to one copy of each. The results could very well be different with another camera, which I found when doing my 21mm comparison when I had access to a second ZM21/2.8. Unfortunately I don't have the resources of Roger at lensrentals at my disposal!
|