Noise test: EM1 vs EM5 vs GX7

Discussion in 'This or That? (MFT only)' started by Armanius, Sep 28, 2013.

  1. Armanius

    Armanius Mu-43 All-Pro

    Feb 23, 2010
    Houston
    Muttley
  2. Repp

    Repp Mu-43 Top Veteran

    501
    Jan 27, 2011
    Oak Harbor, WA
    I'd really like to get my hands on raw files for both the GX7 and OM1, but both look much cleaner than my GX1.
     
  3. LovinTheEP2

    LovinTheEP2 Mu-43 Top Veteran

    619
    Feb 15, 2011
    Toronto
    S it just me or does anyone else think the gx7 shares the same sensor as the em-1???????? Or is this simply evidence of the law of diminishing returns and the tech is pretty much maxed out without an evolution of sensor design ie organic leap etc.

    Very similar noise pattern, very similar Ido performance, very similar ra rendering.

    Would been nice if he added a full auto mode to see how all there cameras exposure as a last test.

    At least, people looking at all 3 can now remove sensor performance as a reason to choose between the bodies.
     
  4. Al.

    Al. Mu-43 Veteran

    372
    Jul 3, 2010
    Hull, East Yorkshire, UK
    Alan
    This is the test I have been waiting for, comparing RAW files . Looking at sensors, seems like the GX7 has made a huge stride forward catching EM5 and the EM1 has stood still. Maybe we have now reached the maximum of a 16 mp sensor.
    So now its just a question of form factor, brand preference, loyalty and affordability
     
  5. CiaranCReilly

    CiaranCReilly Mu-43 Veteran

    481
    Oct 18, 2012
    Dublin
    Ciaran Reilly
    Great comparison! I'm a little surprised at how much better the E-M1 JPGs look than the others, they really leave the GX7 especially in the dirt! Good to see similar RAW performance, levelling the playing field between Pansonic & Olympus.
     
  6. Armanius

    Armanius Mu-43 All-Pro

    Feb 23, 2010
    Houston
    Muttley
    Oly has historically produced cameras with better JPG engines than anyone else. Although arguably, IMO, Fuji's X cameras have the best JPGs now.
     
  7. Al.

    Al. Mu-43 Veteran

    372
    Jul 3, 2010
    Hull, East Yorkshire, UK
    Alan
    I have never had any problems with Panny jpegs, they and very natural looking are customiseable in camera, and with the inbuilt curves in the GX7 even more so.

    As the reviewer says, panny jpegs can be given more punch and Oly's toned down
     
  8. CiaranCReilly

    CiaranCReilly Mu-43 Veteran

    481
    Oct 18, 2012
    Dublin
    Ciaran Reilly
    I wasn't having a go at Panasonic JPGs in general, just the noise performance in this test. At 800 and especially 1600 ISO and up the JPGs from the GX7 look much more noisy than those from the E-M1 in this test, although RAW banishes this difference.
     
  9. drd1135

    drd1135 Zen Snapshooter

    Mar 17, 2011
    Southwest Virginia
    Steve
    Three things I notice:

    1. It appears the EM1 and EM5 really do have the same sensor with the PDAF added to the former.
    2. Whatever sensor the GX7 uses is essentially equivalent to what the OM-Ds are using.
    3. I'm really surprised that removing the AA filter from the EM1 has had that little effect. I guess the EM5's filter was pretty weak. I simply don't know the AA status of the GX7, except that it doesn't appear to matter. Of course, if the older GX1 had one and the GX7 weakened or removed it, that might explain some of the similarities to the Sony sensor.
     
  10. Armanius

    Armanius Mu-43 All-Pro

    Feb 23, 2010
    Houston
    Muttley
    That's what I read in a number of reports in the past. Personally, I never felt that my EM5 was failing to render enough detail for my taste.

    I will likely get the GX7 simply because of the form, wifi, tilt EVF, and 1/8000.
     
  11. meyerweb

    meyerweb Mu-43 Hall of Famer

    Sep 5, 2011
    All of the fairly recent m43s (at least from the GH2 forward for Panasonic) have had much weaker AA filters than competing DSLRs. All you need to do is look at DPRs image comparator to see that the GH2 resolved noticeably more detail than 16MP DSLRs.

    As for comparing noise in jpegs, without knowing how the NR settings were set (even if both were at default values, that doesn't mean they're the same), you can't really make much of a judgement. EVERY modern camera gives you a wide range of control over the jpeg conversion process. Based on these results, though, it's pretty clear that Oly applies more NR at every ISO setting than does Panasonic. Hence the GX7 preserves more detail at the expense of noise, the Oly's less noise at the expense of detail. Personally, I prefer Panasonic's approach. You can always get rid of noise after shooting, but there's no way to bring back detail lost of in camera NR.

    I think we are at the point, until there's a significant advance in technology, sensor performance isn't going to be the key factor in choosing one camera over another. Handling, features, and flexibility matter more.
     
    • Like Like x 3
  12. tanngrisnir3

    tanngrisnir3 Mu-43 Top Veteran

    594
    Oct 18, 2011
    Yeah, I'm really torn. The GH3 is, for me, the ideal form factor, esp. w/the swivel LCD, but that 1/8000th is something I've needed for sometime.

    Sigh. Decisions, decisions.

    Maybe if the GH3 price falls this holiday season....
     
  13. dougjgreen

    dougjgreen Mu-43 All-Pro

    Jun 5, 2013
    San Diego
    Doug Green
    Really clean minty GH3s have been going about $900 on ebay recently
     
  14. Serhan

    Serhan Mu-43 Top Veteran

    533
    May 7, 2011
    NYC
    I prefer the gx7 body form also. One more adv is the eshutter which is silent & prevents the shutter shock also:
    http://www.dpreview.com/forums/thread/3552403

     
    • Like Like x 1
  15. pdk42

    pdk42 One of the "Eh?" team

    Jan 11, 2013
    Leamington Spa, UK
    This is a good review. I've read several E-M1 "hands-on reports" from people at the Oly launch event; the reviewer in many of them was saying all sorts of things about how much better the E-M1 sensor is compared to any previous u43 sensor. I think these results put some more scientific basis to the comparison and it's clear that there's very little difference, esp with raw files.
     
  16. Dalton

    Dalton Mu-43 Veteran

    329
    Jan 24, 2010
    Portland, Oregon USA
    Dan Ferrall
    Missing comparisons

    Some early reviewers of the GX7 compared with the E-M5 are indicating that elthough the GX7 has slightly better noise levels, the GX-7 dynamic range fades far more rapidly than the E-M5 as ISO is increased. It appears in a least one review that by ISO 1600, the E-M5 has an almost two stop advantage over the GX-7 in DR. This results in lower color saturation and poorer contrast when compared against the E-M5.

    There's a lot more to sensor IQ capability than noise levels.

    Dan
     
  17. flash

    flash Mu-43 Hall of Famer

    Apr 29, 2010
    1 hour from Sydney Australia.
    Gordon
    Agreed. And it looks like Olympus think so as well. Lot's of talk about everything except the sensor, which seems to be a side note. Nice for a change.

    Gordon
     
    • Like Like x 1
  18. tanngrisnir3

    tanngrisnir3 Mu-43 Top Veteran

    594
    Oct 18, 2011
    I'd really have to see that to believe it.
     
    • Like Like x 3
  19. stripedrex

    stripedrex Do or do not. There is no try.

    374
    Jun 8, 2012
    Long Island, NY
    Alex
    Before I sold my GH2 for the OMD I tried testing exposure and ISOs on both and found that with the same exposure the GH2 shot 3200 ISO while the OMD shot 1600 ISO with the same lens, shutter speed and aperture. So for me comparing at least those 2 cameras was a little odd. How does one compare ISO properly if these cameras are so poorly calibrated or not producing the same exposure at the same ISO?
     
  20. Armanius

    Armanius Mu-43 All-Pro

    Feb 23, 2010
    Houston
    Muttley
    Some manufacturers underrate or overrate their cameras' ISO values. Consequently, when doing testing, I assume that testers have to equalize all other aspects of exposure conditions. If I recall correctly, another forum member's personal testing led to his conclusion that Fuji's values are off.