philip_pj wrote:
I shelved plans to acquire Leica's very good 35/2 Summicron (~$1000), to be used for groups and close work, when I saw the quality of images here at FM and at getdpi, some of which just take your breath away. Now, because quite a lot people don't either see it themselves or lean heavily on the measure sites' views, a small quote from DxO via the Sony rumour site:

"DxOmark published the RX1 Zeiss lens performance test. And they write: “ Its optical performance is outstanding, and particularly noteworthy for its consistent sharpness and homogenous imaging across the frame. With excellent image quality at maximum aperture becoming outstanding at f/2.8 and on, the Zeiss Sonnar T* 2/35 is likely to become a classic, against which all others are judged. Of course the lens can only be obtained with the purchase of the RX1, a camera that is not entirely without its own quirks and shortcomings. In spite of this, if you have the money to invest, then close to perfect imagery is assured.”

And for all these people that do think the RX1 is too expensive read this: “With a high overall DxOMark score of 33, the Carl Zeiss Sonnar T* is a superb performer optically. That score puts it comfortably ahead of the $1,850 manual focus Carl Zeiss Distagon T* 1.4/35 (35mm f/1.4) on a Nikon D3X, which achieved a DxOMark score of 30.”



Are DXO lens tests gaining credit? They used to give very strange results.

Anyway, we need to look beyond the total score and to the individual measurements.

If we take the short cut and only look at the 5 sub-categories on the front page, it is clear that it gains score mostly from high sharpness, and some from low vignetting. Transmission is not measured. Do we know how they got around that? Did they give it 100% score for transmission? If we look at the weight of the bar graphs, it looks like that is the case.

And the "sharpness" is really system resolution. Put any lens on a higher resolution sensor, and the score jumps up. Or the other way. The Canon lens suffers in the comparison because it is used on a 21 Mp sensor not 24 MP. If you really want resolution you should start with the highest resolving sensor possible and then select lenses.

Further, in one (for me) important metric, chromatic aberrations, the Canon 35 F2 IS is twice as good.

So, the total score doesn't really say much, and even worse, it may actually be cheating in this case (transmission).

Nikon lenses on 16 MP D4 body yield better results than any Canon lenses on 21 MP body.
By the way, the aberrations and distortions can be corrected by software, but you can't do anything with the resolution of lens.

tulaev wrote:
Nikon lenses on 16 MP D4 body yield better results than any Canon lenses on 21 MP body.


Not for resolution when you select the same lens, like the Z* Distagon 1.4/35.

Edit:

http://www.dxomark.com/index.php/Lenses/Compare-Camera-Lenses/Compare-lenses/%28lens1%29/391/%28brand%29/Zeiss/%28camera1%29/767/%28lens2%29/390/%28brand2%29/Zeiss/%28camera2%29/0

And there is something really strange going on there. In that example, the sub-scores are identical except +/- 0.1 on transmission and vignetting, that should actually balance. Only Resolution is significantly higher on the Canon, yet the total score is 2 points higher with the Nikon. I would call the scores total junk after seeing this.



By the way, the aberrations and distortions can be corrected by software, but you can't do anything with the resolution of lens.

The measured resolution is at a certain MTF. I haven't checked what they use here, but often resolution is given at 50% MTF. The thing is, that all lenses resolve detail far beyond this, just with a lower MT. So you can really bring up the "resolution" simply by sharpening the image. But there is no substitute for megapixels.

And those corrections resample the image, resulting in a slight loss of resolution.

Edit 2:
Just to be sure, I am not critizing the lens. But to use the DXO total score metric to compare it against other lenses on other cameras is just meaningless.

tulaev wrote:
Zoom?! Unrealistic.

Why? Think of the RX100 and simply super-size it.

Happy shooting,
Yakim.

http://www.imaging-resource.com/PRODS/sony-rx1/sony-rx1A.HTM

"it's clear that the Sony RX1's lens is the sharpest, most uniform full-frame 35mm optic we've ever seen. Period, full stop."

tulaev wrote:
http://www.imaging-resource.com/PRODS/sony-rx1/sony-rx1A.HTM

"it's clear that the Sony RX1's lens is the sharpest, most uniform full-frame 35mm optic we've ever seen. Period, full stop."

Mine certainly isn't.

Yakim Peled wrote:
Why? Think of the RX100 and simply super-size it.

RX1 lens quality is good, but not enough for an expensive FF compact camera. A high quality zoom lens will be too large.

Makten wrote:
Mine certainly isn't.
What is wrong with your lens?

tulaev wrote:
What is wrong with your lens?

Nothing wrong. It's just not the best 35 mm lens I've tried. I prefer both the ZF 35/2 and 35/1.4, except for their size of course.

tulaev wrote:
Yakim Peled wrote:
Why? Think of the RX100 and simply super-size it.

RX1 lens quality is good, but not enough for an expensive FF compact camera. A high quality zoom lens will be too large.

The word too is subjective. I'm sure some will like the added functionality of a zoom lens + a quality EVF despite the added bulk.

Happy shooting,
Yakim.

Makten wrote:
tulaev wrote:
What is wrong with your lens?

Nothing wrong. It's just not the best 35 mm lens I've tried. I prefer both the ZF 35/2 and 35/1.4, except for their size of course.

Is it the sharpness or the uniformity you question?

I prefer RX1 lens over my zf35/2 and I like it better than the zf.2 35/1.4 which has FC and is not as sharp due to undercorrected SA design.
I would not call it the best 35 as the 35 lux I think is sharper but the lux has FC and sometimes busier bokeh(latest version).

Yakim Peled wrote:
I'm sure some will like the added functionality of a zoom lens + a quality EVF despite the added bulk.

I think that the number of such people will be too small to justify the project. The best thing in RX1 is combination of quality and small size.

carstenw wrote:
Is it the sharpness or the uniformity you question?

It's the fact that you never know when it's gonna be sharp or not. Performance seems to be very random across the focusing distance range. Seemingly similar scenes can be tack sharp from corner to corner wide open one time, and just dull the next time, even stopped down. I also prefer the local contrast and bokeh of the ZF:s.

But hey, since there are no alternatives, I don't have anything serious to complain about.

I am kinda sorry I posted the DxO link, but some people do relate to their output. I call them 'measure sites' because they make plenty of assumptions, and are obviously not photographic inclined by their very nature, and so it devolves into another sharper/sharp/less sharp or good/better/best pissing contest.

What is important to some is that the camera/lens is so good it rates in the same general category as the 'best' and 'sharpest' available, and these results run contrary to the reports of several posters in this long thread.

And this only adds to its appeal based on less measure-sensitive perception characteristics, which please many very experienced photographers who spend time in the field.

Martin, maybe your RX1 copy is affected by moods or the weather, whereas the ZF lenses have more stgble personalities. You never know...

tulaev wrote:
Yakim Peled wrote:
I'm sure some will like the added functionality of a zoom lens + a quality EVF despite the added bulk.

I think that the number of such people will be too small to justify the project. The best thing in RX1 is combination of quality and small size.



The Nex-5 was about size, too, but many requested a built-in Evf, which came with the Nex-7 and proved very popular. The size difference between the two isn't that great.

Yakim and anyone else: is there in existence even a short zoom range, say 30-60mm or so, that is just a little larger than the 35/2 Sonnar, and is fast, say f2-f2.8?

I agree you risk losing the special mix that makes the RX1 such a great image machine. 'Just another zoom' compact...

A built-in EVF was a naughty omission, one that cost plenty of sales.

Makten wrote:
tulaev wrote:
What is wrong with your lens?

Nothing wrong. It's just not the best 35 mm lens I've tried. I prefer both the ZF 35/2 and 35/1.4, except for their size of course.

I agree. At least when comparing it to the 35/1.4 ZE/ZF. Not sure about the 35/2, because the bokeh of the sonnar is nicer than that of the distagon IMHO.
My 1.4 seemed to be even more detailed stopped down and because of the field curvature the foreground always is tacksharp when focusing at infinity, stopped down. That's why I really didn't see it as a negative most of the time. With the sonnar you need to stop down more to get the same thing.
I also still prefer the bokeh of the 1.4 ZE/ZF.

philip_pj wrote:
I am kinda sorry I posted the DxO link, but some people do relate to their output.

Why sorry? I think it is better to find out what they are up to than ignoring them. What I reacted to mostly was their own interpretation and statements, and the lack of connection to the relevance of the parameters included in the score, and the conditions for the measurements and result extraction.

Regarding sharpness across the frame, if you look at the actuance field maps, stopped down to f/2.8 to f/5.6, the Sigma 35 1.4 HSM is the top performer. Only at f/2 is the RX1 Sonnar a little stonger in the corners, but weaker in the center.

The impression I got from the RX1 lens is that it draws beatifully, and it is sufficiently sharp for landscape use, if you want to carry a small camera. But I would not substitute my D800E and C/Y Vario-Sonnar 3.4/35-70 for landscape because the RX1 is a better lens at 35 mm.

philip_pj wrote:
Yakim and anyone else: is there in existence even a short zoom range, say 30-60mm or so, that is just a little larger than the 35/2 Sonnar, and is fast, say f2-f2.8?

I agree you risk losing the special mix that makes the RX1 such a great image machine. 'Just another zoom' compact...

A built-in EVF was a naughty omission, one that cost plenty of sales.

Even a slow zoom with range to 50mm or more would be sticking out substantially and ruin the "compact camera" concept. (Compared to a MILC that can be compact but with a long lens). It would have to be a collapsible design to make any sense, and is that possible without compromising image quality?