Sony RX1 Ranked #4 on DXO Mark at 93

[barjohn]

Ad

This is quite remarkable performance and at the top of the heap with only very small differences between it and the very top D800E. High dynamic range (14.3EV), great SNR across the ISO range and color depth (>25 bits). The new M-240 will have quite a challenge producing a sensor that out performs the state of the art Sony sensors and the Zeiss lens is rated the equal of the Leica 35 by Lloyd Chambers and other reviewers very familiar with Leica..

[kdriceman]

It is, no doubt, a great camera, but at $3,000 I would expect a built in viewfinder (hybrid would be nice, but at least an EVF), the ability to change lenses with an ability to use an adapter for 3rd party lenses and phase-detect autofocus instead of contrast-detect. The NEX 6 now has all these features, lacking only the full frame sensor. My guess is that Sony is milking the top of the market for now and will introduce a more complete version of a full frame CSC before long.

[iedei]

Originally Posted by kdriceman

It is, no doubt, a great camera, but at $3,000 I would expect a built in viewfinder (hybrid would be nice, but at least an EVF), the ability to change lenses with an ability to use an adapter for 3rd party lenses and phase-detect autofocus instead of contrast-detect. The NEX 6 now has all these features, lacking only the full frame sensor. My guess is that Sony is milking the top of the market for now and will introduce a more complete version of a full frame CSC before long.

i kind of like the idea of a fixed lens, however i agree that i just can't get over the idea of the RX1 not having a viewfinder......i can't take it seriously without it! With the added accessory, it becomes much larger....taller than an M-E, in fact.

[Paul J]

There's just something about the colour and contrast and the way it swallows up the detail and micro contrast that makes pictures look a bit amateur IMO. Like a cheap dSLR albeit with a decent lens. Perhaps it's just the sample I've seen.

[thedigitalbean]

When the RX-1 was first announced I too was ready to dismiss it. The lack of a viewfinder made it difficult to take seriously. However, after several individuals whose opinions on equipment I trust completely highly recommended the camera I gave it a shot, with the intention that if it really didn't work out I would return it.

Upon first using the camera, it became clear with the goals of the design team were. They wanted to build a FF compact to be as absolutely small as possible. Its clear from handling the camera that adding an EVF would increase size, making the mount take interchangeable lenses would also increase the size. Of course, not everyone will be comfortable with the compromises made to get to a FF compact of this size. Then again, there are many people who think the compromises made to get to a compact FF rangefinder like the M9 are too much.

For myself, the 35 ASPH FLE is usually the lens stuck on my M9, its a lens I love and 35mm has always given me a FOV I've liked. After getting the RX-1, I think in many cases I'd rather use it than the M9+35 FLE. That said, optically speaking the RX-1 lens is simply superb, but its also no 35 FLE (no idea what the 35 'cron is line since I've never owned one).

[Tonki-M]

Originally Posted by thedigitalbean

but its also no 35 FLE (no idea what the 35 'cron is line since I've never owned one).

the lux FLE has the advantage of 1 stop, FLE helps close up performance and focus shift issue and faster transition from focus to blur. in real world use, at equal aperture, they are EXTREMELY close, you would really need to pixel peep to see any discernible difference. so basically, you pay extra for that 3D look @ F/1.4 (and the 1 stop ofc)

[Ecaton]

Originally Posted by Paul J

There's just something about the colour and contrast and the way it swallows up the detail and micro contrast that makes pictures look a bit amateur IMO. Like a cheap dSLR albeit with a decent lens. Perhaps it's just the sample I've seen.

You might refer to the Sonnar lens rendering. Have a look at M9 files shot with the ZM 50 Sonnar at f1.5 and f2 for comparison.

[barjohn]

There are three basic links in the production of the image, the lens, sensor and processing software. You can never get better than the weakest link. In the M9, the weakest link is the sensor. Leica obviously works hard in the software to try and overcome the weakness in the sensor (see DXO Mark evaluation) but there are limits to what can be done in processing and they do a very good job with what they have. The RX1 has one of the best sensors on the market today and an excellent lens. The software is very good but probably not as mature as the software on the M9. The images are superb as has been reported by many M9 owners and reviewers.

[jaapv]

Uh- no. The weakest link analogy is faulty. The final quality is a combined result of the factors involved. Improve one without the other and quality still goes up.

Please note that DxO measures the results of the files out of camera, not the sensor output.
All digital cooking by the firmware is included, making sensor comparisons based on their measurement rather doubtful, to put it mildly.
But the site has very pretty graphs.

[Peter D Lux 2]

Whatever. Fact remains that the RX1 combo yields really good IQ.

[jaapv]

Ad

Nobody is disputing that.

[Ecaton]

Originally Posted by jaapv

Uh- no. The weakest link analogy is faulty. The final quality is a combined result of the factors involved. Improve one without the other and quality still goes up.

Please note that DxO measures the results of the files out of camera, not the sensor output.
All digital cooking by the firmware is included, making sensor comparisons based on their measurement rather doubtful, to put it mildly.
But the site has very pretty graphs.

As a user what I get is the "result out of the camera". Even the RAW files get some cooking, don't they.

[jaapv]

True - but there is a difference between the very strong in-camera processing of a small-sensor CMos P&S and the minimally worked-on output from a CCD based MF camera (or Leica M for that matter). DxO does not take that into account with their rankings, which makes for some weird results.

[barjohn]

Originally Posted by jaapv

True - but there is a difference between the very strong in-camera processing of a small-sensor CMos P&S and the minimally worked-on output from a CCD based MF camera (or Leica M for that matter). DxO does not take that into account with their rankings, which makes for some weird results.

The M's work on moire reduction, color shift problems, and vignetting in processing the RAW files in camera based on lens identification. I don't think one can call that minimal processing. For all we know they may also do work on noise reduction.

[barjohn]

Originally Posted by jaapv

Uh- no. The weakest link analogy is faulty. The final quality is a combined result of the factors involved. Improve one without the other and quality still goes up.

Please note that DxO measures the results of the files out of camera, not the sensor output.
All digital cooking by the firmware is included, making sensor comparisons based on their measurement rather doubtful, to put it mildly.
But the site has very pretty graphs.

I have to disagree. If the lens produced a weak blurry picture all of the improvements in the world to the sensor and processing would not improve the image. If the sensor produces smeary images with color shifts and other strange artifacts, no matter how good the lens or how you improved the processing would you make for a better image. The final product can be no better than the weakest link in the chain.

[jaapv]

Originally Posted by barjohn

The M's work on moire reduction, color shift problems, and vignetting in processing the RAW files in camera based on lens identification. I don't think one can call that minimal processing. For all we know they may also do work on noise reduction.

The M8 does no noise reduction, the M9 has a minimal amount for noise that cannot be addressed in post at high iso and a shifted black point. Yes you are right Leica does apply some necessary manipulations to the raw data, but compared to some other cameras it is very minimal.

[jaapv]

Originally Posted by barjohn

I have to disagree. If the lens produced a weak blurry picture all of the improvements in the world to the sensor and processing would not improve the image. If the sensor produces smeary images with color shifts and other strange artifacts, no matter how good the lens or how you improved the processing would you make for a better image. The final product can be no better than the weakest link in the chain.

No that is not correct. A blurry lens will still deliver an improved image on a better sensor. Or, if you prefer to put it the other way around, an even worse image on an inferior sensor.

[lanetomlane]

I'd love to be able to comment on the image quality from the RX1 ..... I received mine yesterday, but it didn't work!
________________
Cheers, Tom

[barjohn]

Originally Posted by jaapv

The M8 does no noise reduction, the M9 has a minimal amount for noise that cannot be addressed in post at high iso and a shifted black point. Yes you are right Leica does apply some necessary manipulations to the raw data, but compared to some other cameras it is very minimal.

I guess it depends on your definition of minimal. Some cameras do noise reduction on RAWs and that is about it. One could call that minimal compared to the intense processing required to deal with color shifts in the M9.

An improved blurry image is still a blurry image.

[myshkine]

Originally Posted by kdriceman

It is, no doubt, a great camera, but at $3,000 I would expect a built in viewfinder (hybrid would be nice, but at least an EVF), the ability to change lenses with an ability to use an adapter for 3rd party lenses and phase-detect autofocus instead of contrast-detect. The NEX 6 now has all these features, lacking only the full frame sensor. My guess is that Sony is milking the top of the market for now and will introduce a more complete version of a full frame CSC before long.

I agree. Maybe the RX1 is just an experiment, to see how the market reacts and what technical results it achieves. It may well be followed by a "full size" version, one with EVF (or even a hybrid) and interchangeable lenses, on the footsteps of Fuji.