Crazy Comparison: A not fair Leica M9 RAW vs NEX-5n JPEG comparison!

After the last crazy comparison the other day many Leica owners complained that I should not and could not compare a Leica JPEG to a NEX-5n JPEG. Some e-mailed me in anger when all I did was produce a fair JPEG comparison that showed how each camera would pump out a JPEG with the same lens and aperture. It was indeed fair but all of us Leica M9 owners KNOW that the JPEG engine is not the greatest (though it is not really that bad).

I was asked by at least 20 of you to post a Leica M9 processed RAW file next to a NEX-5n out of camera JPEG. While this is not a fair comparison at all to the NEX-5n you can see the richness of the M9 file here. Both are full size and the 5n is direct out of camera with no tweaks at all. The M9 RAW file had the exposure taken down a notch in the RAW conversion.

So click on each image for the full size file. As soon as I get RAW support for the 5N, I will do a RAW comparison between both to satisfy those who have been asking .

On another note, I am traveling home today back to Phoenix AZ and will have a few cool posts up this week so check back later today and all week for more!

The M9P and 50 – click image for full size file

The NEX-5n out of camera JPEG with the 50 – NEX was on the VIVID setting

Now just to be clear – the Leica file was shot at 2.8 (exif says f/2 but it’s never accurate with the M9) and the NEX was shot at f/4 so this is not only NOT fair for the fact that the M9 was a RAW file, but the NEX was stopped down a bit. SO this is a flawed comparison all the way! BUT!! You can get an idea of how gorgeous the M9 RAW files come out. At least I think so. If the NEX RAW conversions are as good as the M9…look out.

BTW, I stood in the same spot so you could see the effect that the crop sensor has on the 50mm focal length .Enjoy!

[ad#Adsense Blog Sq Embed Image]

64 Comments

  1. “Garvin: The sony is a much brighter exposure. At least a stop and a half. And its because of this that I find the sony to be a lesser quality image, mainly due to the blow out around the trees in the upper left. ”

    The reason for the exposure difference is probably because of the metering. The M9P being full frame included more of the bright sky thus the exposure is slightly darker.

    “Michiel: What a difference, much bigger than I’d have expected. The Sony shoot looks horrible, not like from a decent aps-c senor camera. ”

    To tell you honestly, based on the two images above, I think the image shot by NEX is probably more accurate to the actual scene. If you have a preference towards vivid looking images, you can set it via menu >> brighness/color >> creative style

    I’m not trying to be bias here. I use both Sony and Nikon cameras (A900, A850, A700, Nex 5N, Nex 5, D3S). The Sony cameras can produce very good results. Do look at my gallery over here http://gallery.darrenf.com. Most of the images were created with Sony cameras. Some were with Nikon. See if you can spot the difference.

  2. Hello Eric

    I like your efforts.

    Personally, I don’t think the difference between your edit and and the M9 version is worth the price difference between the NEX and the M9. But, of course, there is no accounting for taste.

  3. Since the M9 has the unfair advantage of being processed from RAW, I thought I’d try to even the score a little by pushing the NEX-5N JPEG through Adobe Camera Raw, and see what I come up with.
    I think I managed a pretty good grade:

    [img=http://www3.picturepush.com/photo/a/6537936/640/6537936.jpg]

    View Full size: http://www3.picturepush.com/photo/a/6537936/img/6537936.jpg

    What do you guys think?

  4. I downloaded the two images and tried a more fair comparison.
    I opened the jpeg file from the Nex with Camera RAW for adjusting the exposition and to recover something from overexposed areas. Then, in CS5, I adjusted shadows, color balance and constrast.

    After that, I cropped the Nex JPEG image to the same resolution of the APS-C equivalent resolution of the M9’s FF sensor (that is, the M9’s sensor has an APS-C “embedded into it” with a resolution of 7,6 Mp). In ths way the resolving power of the lens is put into a more apple to apple comparison between the two bodies.

    The results are now much more comparable. There is a different focusing between the two shots, in effect the shot with the Nex seems to be focused nearer to the shooter. The details are almost on par between the two shots, with the Leica image having somewhat clearer output thanks to the absence of the anti aliasing filter. The image from the Sony seems to resolve better some yellow midtones (strange, the A/D conversion of the Nex 5n should be at 12 bits, while the Leica implements 14 bits per pixel conversion). The Sony seems to apply more sharpening, it would be very interesting the comparison of the two RAW images.

  5. And another thing I’ve noticed, Steve:
    Although both cameras were set to A mode (Aperture Priority) , the shutter speed on the NEX-5N was 1/160 sec, and 1/1500 sec on the M9.
    Since you took these photos hand-held, it explains why the NEX-5N photo is more blurry (about 10 times slower shutter speed).

    People should take that into account.

  6. Steve,
    I also see that you’ve chosen different metering modes.
    On the M9, it was set to “Center-weighted average”, while on the NEX-5N it was to “Pettern-metering”.

  7. Steve, some questions regarding thr NEX-5N:
    ————————————————————
    What was the HDR/DRO setting on the NEX-5N?
    Why Vivid? Portrait gives much better DR than Vivid! It’s more important to register as much detail into the file than to make the colors pop a little more. We can adjust the colors later in post, but we cannot make up detail that is not there in the first place.

    Did the MM indicate 0.0 when you took the shot?

    Why didn’t you move back so the photos would cover the same landscape? Would have made the comparison much easier.

  8. i prefer the leica m9p color,it’s so green,the same place,leica looks like in the summer days,the sony looks like in autumn days.

  9. Thanks for comparing these two. It is great fun to see the relative strengths. I have seen plenty of photographers take great pictures with Leica, but a good enough photographer could take great pictures with a folding pocket Brownie, so I never paid much attention to the camera.

    Only when you see crazy comparisons do you really understand what the fuss is about. I am happy with my Sony, but I can see why others might not be.

  10. Well, to be “fair” Steve has to walk back or forth to do the same frame on both shots because of the crop factor the number of pixels in nex5 arround the house is higher than M9.
    M9 wins here, but we need to wait for raws.

    • Why bother with these comparisons at all if you’re not going to spend an extra 5 minutes to ensure that focus point is consistent and that apertures are set appropriately to ensure equivalent depth of field for sensor size.

  11. Honestly, why bother with these comparisons if you don’t take the time to equalize settings. If you shot the NEX at f4 then you need to shoot the M9 at f8. If not, your comparison is worthless.

      • Because of the difference in sensor size and depth of field. You need to stop down to about f8 on full-frame to get similar depth of field as f4 on APS.

        • Nah. your math is flawed. You get a stop more DOF by using a 1.5x crop sensor as compared to FF, i.e. f 5.6 and not f8.

          • I am afraid that both of you are incorrect. Steve used the same lens for both shots. Therefore the depth of field should be identical between both cameras if the same aperture was used. Depth of field is a function of focal length, aperture, and lens to subject distance. The image size (i.e. format) determines the field (or angle) or view.

            It appears that Steve used different apertures between images simply because he wanted to.

            PaulB

  12. Steve – we get the point 🙂 But I am obviously not the only one who would like to see how much headroom is there for NEX in RAW. Hope you will bring that one too.

  13. I have to admit that I don’t really see anything special in that particular Leica shot probably as I haven’t invested fortune in Leica gear. Regarding sony nex, it would be interesting to see raw conversion when proper profiles in ACR and Lightroom are available. Based on my limited experience with Nex 3 (used friend’s camera on 2 day trips) Sony’s jpeg engine is nothing to rave about, I got much much better results when processing raw which, in the end is logical, the same goes for my 5dmkII as well (i shoot ony small jpeg as orientation when I review photos on the computer) and pretty much any big(er) sensor camera apart from offcourse Olympus pen series.

      • Zoran

        I would say no they were not. Not that Steve did not attempt to do so, But to me, the NEX appears to be focused forward of the barn while the M9 is focused on the barn.

        PaulB

    • Duhhhh – no-one noticed …… read my reply (in comment 10) – I’ve already said focus point differs …….

  14. The highlights are blown in the NEX shot, giving it that “digital” look. I think it would help if you under exposed by 2/3 EV.

  15. When the NEX-7 arrives can you do a comparison use the 35mm Summilux on it vs. the 50mm Summilux on the M9? With the 1.5x crop factor it seems like you should be using a 28mm or 35mm lens on the NEX’s when using a 50mm lens on the M9. Keep up the interesting comparo’s though. That’s one of my favorite things you do on this site.

  16. The NEX raw comparison will be interesting, but probably not improved as much as the M9, since the M9 is known for its weak JPGs. I have both an M9 and a Pentax K-5, and the K-5 JPGs are much closer to the RAW quality than is the Leica. (Also obvious from the JPG file sizes – the K-5 JPG files are much larger than the M9.)

    • This is one shot. Not to defend the poor 5N but I think Steve’s shots from the other day showed a much closer comparison with the 5N looking better in some shots, worse in others but overall they looked VERY close with just some minor color cast differences and the 5N exhibiting CA badly in one shot.

      And before you say, “but the M9 shot is a processed RAW”, I would say, a better test would have been to show both the M9 RAW and JPEG against the 5N JPEG. If the M9 JPEG looks as bad as this 5N JPEG does, then we can say that RAW is the difference.

  17. When the glass is identical and AA filters and lenses are aligned as best they can for their respective sensors all that’s left is, of course, the sensor. And it makes no difference whether that sensor is in a $700 body or a $7000 one.

  18. What a difference, much bigger than I’d have expected. The Sony shoot looks horrible, not like from a decent aps-c senor camera. Was the same lens used on both?

  19. It looks like the Leica shot was made wide open. My Summilux is certainly not that weak off-center at 2.8 and the M9 always measures the used aperture-value too high (never too low, e.g. 80% of the time it says 1.7 instead of 1.4 that was actually used). But that doesn’t matter, in the center it’s a good response to the JPGs we saw a few days ago 🙂

  20. I think that this comparison defines a new standard method for comparing Leica M-mount lenses, be they from Leica or from other manufacturers. IMO, the ground-breaking feature of this comparison is that it was performed on a camera with high IQ that also allows for live-view focusing. I feel that sharpness tests of M-mount lenses on Leica’s current equipment are useless, as there is no way to know if the differences observed are due to incorrect focusing of the rangefinder (varies lens to lens), or due to the lens itself.

    • Remember the M9’s strengths are in its RAW files and the posted NEX image was an OOC JPEG, no enhancements at all. Ill do some raw vs raw as soon as I can!

    • If you look at the enlarged images there is a clear difference in focus point between both – which, with the greatest of respect to Steve – is ‘operator error’ rather than sensor or lens specific issues.

      If you look at both images – where the focus is correctly aligned – then there is little in it.

      Probably the greatest difference is at the edges, where the M9 no doubt auto-corrects for lens aberrations, putting the Sony at a greater disadvantage with non-OEM glass.

      From personal experience with the Nex I find lenses with a longer register (i.e. most SLR lenses) provide superior results – producing a more ‘perpendicular’ light path from their outer edges. Like the M9 with Voigtlander superwides, modern sensor edge results can go funny without in built correction algorithms. Sensors ‘aint Film!

      • Not sure about it, but in all image to image comparisons I’ve seen here, the focus is always a problem.

        I guess if you spend 10k$ in a camera you kind of want to make a point somehow, even unconsciously hehe ! (which all the complaints so far on the previous article are confirming IMO…)

        So here, not judging on the sharpness, the colors are still awesome on the Leica but I still have a question. In real, the situation is closer to Leica or to NEX ? Because the dreamy effect is cool, but the NEX looks more like what I can see in some fields around my place … Just out of curiosity !

        • Yes stopped down is more like what you see in real life. The less DOF you have, the further from reality it will look. Very shallow DOF like with an f/1.4 is nothing like what your eyes will see. f/4 on a crop sensor will have much more DOF than f/2.8 on a FF sensor. We are just stating the obvious here…

  21. Had a close look at both shots; didn’t take long. That’s settled then. What a difference; much larger then I expected.

  22. These two shots highlight the conundrum with the Sony NEX cameras (for me). The Sony NEX 5N shot is fine. The RAW will be much better. But it looks nothing like the Leica shot. I felt this way about the other comparison shots. I don’t see the lens signature in the NEX images, which to me seems to be the main reason for mounting these lenses on digital bodies.

    I’m willing to bet the upcoming Zeiss primes will look as good or better on the Sony as the adapted Leica glass looks (I’m not claiming they will look as good or better than the M9 images – have to wait to see). Plus it will be AF, give you full EXIF data, take advantage of all the shooting modes, etc.

    This is pushing me away from the considering the NEX cameras as a true, affordable digital solution for my Leica lenses.

  23. Is it just me or the NEX jpeg is over compressed? Also, NEX colors don’t really conform to my definition of “vivid”…

  24. Understood that it is a OOC JPG, but some curve matching might help to give a better example of the comparative strengths.

    I’d imagine I couldnt pick 5N image from the 5 image since they are so close in color response. Obviously the vibrant but controlled M9 image wins in this regard but the sterile Sony image is probably closer to real life..??

  25. Shows what the Nex system can achieve with good glass …. not sure why Sony/Zeiss are dragging their heels in providing quality lenses. The initial success of 4/3rds camera was probably due to the lens quality/availability as much as their portability, etc.

  26. The sony is a much brighter exposure. At least a stop and a half. And its because of this that I find the sony to be a lesser quality image, mainly due to the blow out around the trees in the upper left. bring that exposure down and bump the fill light alittle in the raw conversion and I think these two will be very very close. just my 2 cents

  27. If anything maybe the Sony file could have been exposed a bit darker. but its still going to be flatter than the M9.. But do it really matter, they both take fine images, and they both are completely different cameras. 🙂

  28. Clearly the Sony is having problems with the edges and corners. The M9 shot also shows some softness at the edges, but not nearly as much as the Sony.

Comments are closed.