Personal View site logo
NEX-7 vs GH2 comparison
  • Originally, when I bought the NEX-7 I had entirely still photography in my mind. Its mighty APS-C sensor (24MP) is capable of creating fantastic photos. Considering its video recording capability, I haven't expected anything remarkable. Although, it is quite interesting- I've made this test footage last week and compared it with the GH2, which is (especially here on the P-V site) very well-known to most of the visitors:

    Test footage details:
    The choice of lenses was done very carefully. Considering different crops of the APS-C and micro-four-thirds sensors (1,5x vs 1,86x recpectivelly) I've used the lenses that have similar view angle, same speed and extremely similar optical performance. All lenses had proper lens hood at all times. The aperture was most of the time stopped down for one step (day at f/4.0, night at f/2.8) except in very dark scenes, where both Nikkors had fully open diaphragm. In the NEX-7 there is no possibility of shooting in 24fps mode so I set it to 25fps and adjusted the GH2 to HBR and 25fps.

    NEX-7 with stock firmware (PAL) // GH2 with the patch Cake 2.3 by @balazer
    AVCHD 25fps 1080 // AVCHD HBR 25fps 1080
    The GH2 camera files were quite exactly double-sized compared to the NEX-7

    Daylight footage settings:
    Lens: Canon FD 35/2.8 // FD 24/2.8
    Lens aperture: (always) 4.0, no filters
    Crop: 1.5x35=52,5mm equivalent // 1.86x24=44,6mm equivalent
    ISO: 200 // 200 (always)
    AWB // AWB
    EV 0.0 // 0.0
    Metering: multi // frame+center
    Colors progr.: standard // standard
    In-camera setting: default // default (0,0,0,0)
    Mode: aperture priority (both)

    Night footage settings:
    Lens: Nikkor 35/2.0 Ai // Nikkor 28/2.0 Ai
    Crop: 1.5x35=52,5mm equivalent // 1.86x28=52,1mm equivalent
    Lens aperture: 2.0-2.8
    ISO: 1600-1600
    AWB // AWB (one take manually set to 2500K both)
    EV: at first 0.0 both, later -0.3 / +0.7
    Metering: multi // field+center
    Colors progr.: standard // standard
    In-camera setting: default-default (0,0,0,0)
    Mode: manual (both)

    Comments and bit of my subjective impressions:

    Exposure: by default settings (EV 0.0), NEX-7 tends to overexpose EV+2/3, GH2 to underexpose EV-2/3
    Colors: NEX-7 tends to reddish, GH2 to greenish colors. The NEX-7 has possibility of adjusting the color tones for every color mode, also for AWB.
    White balance: NEX-7 is capable of getting much cooler AWB, extremely weird situation occurs at 2m45" when the WB in both cameras was set manually to 2500K
    Moiré: two tipical situations by 1m40" but generally, the NEX-7 has obviously less moiré
    ISO: NEX-7 has feature of changing the ISO sensitivity during the recording, the GH3 offers this but GH2 and GH1 not. In the daylight ISO was set manually to 200, in the night to 1600 in both cameras.
    Sharpness: I was expecting clear winner GH2 and much worse NEX-7. Yet, the result is very similar. Please mind that in one scene I forgot to focus the lens (by 2m35") mounted to the NEX-7
    Contrasts: in all test situations the NEX-7 shows more details in the shadows, but it would need bit more testing because of the camera's propensity to overexpose what I didn't notice at first.

    The NEX-7 is even smaller than GH1 or GH2 and it has bigger sensor.... I like its concept, very minimalistic- pure idea of the "mirrorlessity". Its big advantage is the still photography, such a comparison wouldn't make much sence because it simply blows off not only the GH2 but even most of the DSLRs with APS-C sensor.... it's newer generation. The big drawback of the NEX-7 considering its video functionality is its sensor overheating. On the room temperature, some 20min are maximum one can expect, regardless if the display is lifted or left in its position. I am kind of very pleased to see that the NEX-7 renders the shadows very smooth and in cinematic spirit, with lots of details. It is a feature that we are struggling to get with our GHx cameras. This cinematic image gets more obvious by using a bit less sharp lenses than Canon FD or Nikkor that I used in this test.

  • 8 Replies sorted by
  • UPDATE: Sony NEX-7 firmware update v1.01 loaded today.

    Overheating is not an issue, the temperature warning appears after some 15-20min but after I blowed on the camera back just a bit, the temperature warning turned off again for several minutes. I must admit, I didn't try to solve it so easily before updating the firmware. I really can't belive that the sensor warms up less after the firmware update. So, no big deal.

    NEX-7 records 30min and instantly after it stops, one can press the button again and it goes for another 30min.

  • What is going on at 2.48 with the white balance? Sodium Vapour lights? The Nex-7 is doing a very Sony thing almost as if it detected difficult subject matter and turned on a specific post-processing feature. I have noticed over the years that Sony cameras aren't always accurate but they can fake a good picture when other cameras fail.

  • @vicx
    Good question: yes, I've mentioned that in the opening post, isn't it strange? I've set there WB in both cameras manually to 2500K. Color of the street lights was indeed quite reddish. The Nex-7 was able to completely neutralize it and really reproduce the white color of the leftmost car. The GH2 wasn't. Weird thing is only, that the WB in both of the cameras was set to the same value.

    Manual selecting WB in the NEX works two ways:
    - by choosing the color temperature value
    - by shooting a test photo on the relevant spot in the field (white paper or so) and the camera is after measuring the colors setting the WB itself...so it is half automatic.

    I did the second solution and measured the WB on the first car from left.

    As the NEX-7 provides full in-camera color correction which can be set individually in any mode, I suppose that after ordering it to measure WB on the white spot, the camera uses also color corrections very flexible way, according to the situation. I mean, one can make even 2500K pre-setting even much cooler if pushing the colors towards blue. It is quite amazing that this way even that reddish street light was usable for the video and looked almost neutral. For my own eyes, all was quite red when I was shooting that situation, even more reddish than it looks on the GH2.

  • Great news:
    Stream Parser is able to analyse good part of the NEX-7 files! Few more valuable information about video capabilities of the NEX-7 that the devices sold in NTSC regions are capable of shooting in 24p (instead of 25p) and 60p (instead of 50p) with significantly higher bitrates. (see first attachment on th ebottom)

    Btw: Sony's video encoding is obviously excellent, so I see significantly higher video quality by the same bitrates like Panasonic.

    Two great video features:
    Besides the possibility to record in 50/60 fps AVCHD 1920x1080 (for PAL/NTSC models, respectively) the NEX-7 is also capable of changing the ISO sensitivity during the footage:

    Those functions are not available in GH1 and GH2, the new GH3 features them.

    My first conclusion of the comparison between the GH2 and NEX-7 after some more analysing:
    - NEX-7 has excellent low-light image quality and retains more details than the hacked GH2 in dark areas
    - The sharpness of the NEX-7 is excellent, although obviously less sharp than the GH2. The result is less moiré, though sharpness is ceirtanly not missing (in my video on Vimeo in the first postage especially visible by 01:50-02:06)
    - Interesting feature by the NEX-7 is that besides AWB (which works just the same way like on the GH2) the camera offers something like grey-card-metering with a test (still) shot. It is stunning, how easily the camera "ignored" strongly reddish street light (Vimeo video 02:45-03:14), it can make really bad light colors very usable for the video footage.

    So, this camera definitely has not only the potential for still photography (which is meanwhile well known) but it has already in unhacked version excellent video performance.

    Since my first postage in this thread I experienced that Vitaliy works hardly on hacking Sony cameras.... so this one here may turn into the video monster if sensor heat doesn't get out of control after the hack. I will be praying for the success of this new and exciting hack project!


    Btw, after first firmware update by Sony I have feeling that the camera handles much better with the heat: no one stop due to overheating within time limit (30 minutes).

    Attachments in order:
    NEX-7 video features and formats list
    Frames of the NEX-7 seen in Stream Parser
    StreamParser analyse of the same take from test footage done by GH2 and NEX-7, respectively.

    NEX-7 video options.JPG
    1024 x 768 - 120K
    NEX-7 frames.JPG
    1024 x 768 - 113K
    GH2 frames mode.JPG
    1024 x 768 - 148K
    NEX-7 frames mode.JPG
    1024 x 768 - 135K
    GH2 time mode.JPG
    1024 x 768 - 150K
    NEX-7 time mode.JPG
    1024 x 768 - 149K
  • And just a little comparison of the JPEGs (although the Nex-7 is clear, superior winner there):

    I've done with those shots the of one flower (Hybiscus Syriacus "Boule de feu") by attaching the same lens (Minolta Rokkor 50mm f/1.4 at f/4) within two minutes on the GH2 and on the Nex-7.

    To avoid uploading big files here, this is just print screen shot of out-of-camera JPEGs. Both cameras were on neutral settings, lowest ISO and AWB. Sony's colors are quite realistic, the JPEG is definitely usable while GH2's JPEGs simply suck, whoever wants good stills out of GH2 needs to spend lot of unnecessary work on the RAW files in the post. IMHO, the GH1 produces much more usable JPEGs than the GH2.

    Photos shot with the NEX-7 retain much more details, contrast, clear defined (and naturally rendered) colors and they are much sharper. Biggest problem with GH2's pics is color blur and rendering, but also its quite low DR in more contrasty scenes. Another problem is the diffraction, which occurs obviously already after 3 stopps. The pixel density of the sensors is comparable. But while the image quality on the NEX-7 with this lens increases up to apperture f/8 where it gets razor sharp and contrasty, the GH2 drops obviously by f/5.6 and isn't comparable anymore by f/8. It is BTW not particularly GH2's problem, but rather physical limitation of the micro-four-thirds sensors.

    GH2's RAWs are surprizingly good compared with its JPEGs, so the converting process is kind of cheap and doesn't use whole potenitial of the GH2. Again, GH2's RAW stills are obviously not better than Nex-7's, but here we'regetting at least half way closer and usable after some work.

    GH2 (left) vs Sony NEX-7 (right), click to enlarge:

    PrtScr GH2 vs Nex7.jpg
    2361 x 1271 - 614K
    PrtScr GH2 vs Nex7 crop.jpg
    2373 x 1275 - 509K
  • @tetakpatak in that close-up comparison on the right I actually thought the gh2 looks better than the nex7 and that surprised me.

  • I actually thought the gh2 looks better than the nex7

    @johnnymossville it could be matter of the printscreen, but it really doesn't. GH2's colors are totally faked, I hardly can restore real colors of the flower from GH2's RAWs in photoshop. It is for me kind of important when I send pics for botanical magazines...

    ....the flower you see on the screenshot is completely new cultivation, you can't see it anywhere else :-)

    I don't give damn about cameras, I just need them to save the pictures, so I don't "like" any camera, just appreciate the ones which serve me. GH2 does it for video, but doesn't for stills, that's all.

  • @johnnymossville @tetakpatak I think the focus-plane of the GH2 stills is a bit farther away than the one of the NEX-7 stills, so that's very difficult to actually judge the quality and sharpness difference. At least for me the GH2 stills seem "cleaner" concerning the defocused areas, I see more, well I'd call it "even grain" while the Nex-7 stills seem more "blotchy" and noise-reduced in these regions. BUT since my experisnce with GH2 stills in general was not very good (I really only use it as a superb video cam) and since I don't own the NEX-7 and I don't plan to buy one - I can't do a comparison how much detail rendition there actually is and how the lowlight performance is, so I will rely on tetakpatak's judgement ;)