OM-D vs 5D2 unscientific testing

Discussion in 'Olympus Cameras' started by Foster2380, Jun 19, 2012.

  1. Foster2380

    Foster2380 Mu-43 Regular

    47
    Jun 6, 2012
    So I got my OM-D w/12-50 lens and have been trying to compare the images to my 5D2 with 24-105L. I figure each lens is my "walk around" lens so that's how I should compare them.

    I took this shot below to try and evaluate how much detail I could pull out of the shadows compared to my 5D2 which I've found to be a very nice file to work with in the past. This was a quick and dirty test, I didn't put a lot of thought into it. Each camera was set to Aperture priority f4, auto ISO, auto WB; shot RAW processed through Lightroom 4 with only slider adjustments being sharpness to 25 and shadows to +67. My focus point was on the exterior frame by the handle.
    100% crop in CS5

    DRcropOM by bnfarnsworth, on Flickr

    DRcrop5D2 by bnfarnsworth, on Flickr

    Was my "test" flawed? Am I crazy here or does the OM look like it did better retaining details in the shadows? I fully expected the 5D2 to blow the OM out of the water here not the other way around. Has anyone done a DR comparison between the 5D2 and OM?

    Here's the original file from the OM:

    P6190001-2 by bnfarnsworth, on Flickr
     
  2. fooddude

    fooddude Mu-43 Regular

    131
    May 22, 2012
    Well...in DPR, the DR page does shows a graph, where you can match up the OMD to a bunch of diff cameras, and it shows the OMD beats ALL m43 and APSC cameras.. and even beats all the old FF (5d, 5d2, d700) cameras in DR, by a lot. The only cameras beating or tying with the OMD (judging from the DPR site) are the very newest and most expensive FF cameras - the 5D3, D4 and D800.

    So I guess the DPR DR data-graph page holds some truth... even though many people are skeptical and don't believe DPR.

    Looks like the OMD is a Full-Frame killer! ;)
     
  3. WT21

    WT21 Mu-43 Hall of Famer

    Feb 19, 2010
    Boston
    The Depth of field on the OMD is much deeper than on the 5D. Some of the background in the 5D is actually OOF, which is what you would expect. I think they actually look equivalent. The other surprise is the DR looks the same.
     
  4. fooddude

    fooddude Mu-43 Regular

    131
    May 22, 2012
    I think you factor the DOF like you do the FL (x2, or basically double)... so if the 5D2 is at 24mm at f/4...you would want the OMD at 12mm and f/2
     
  5. rdearth53

    rdearth53 Mu-43 Regular

    93
    Feb 4, 2012
    Wadsworth, Ohio
    I don't think you're crazy at all! To my eyes, the OM-D file looks noticeably cleaner than the 5D2's. The performance of the OM-D is astonishing!

    Sent from my Galaxy Nexus using Mu-43 App
     
  6. Jonathan F/2

    Jonathan F/2 Mu-43 Hall of Famer

    Aug 10, 2011
    Los Angeles, CA
    I have a D800E for work, and to be honest the OM-D output in many situations look better to me.
     
    • Like Like x 2
  7. cmpatti

    cmpatti Mu-43 Veteran

    263
    May 8, 2011
    Berkeley, CA
    The EM5 noise pattern is much more pleasing, but I don't think you can say anthing about resolution in the shadows because of the DOF issues others have mentioned. Another example of the crummy DOF characteristics of FF.
     
    • Like Like x 1
  8. fredlong

    fredlong Just this guy...

    Apr 18, 2011
    Massachusetts USA
    Fred
    I don't think the additional noise and vertical banding are a result of the depth of field difference.

    Fred

    Edited to say: Maybe not noisier but the 5D2 noise is "sloppier".
     
    • Like Like x 1
  9. With_Eyes_Unclouded

    With_Eyes_Unclouded Mu-43 All-Pro

    Apr 17, 2012
    Vassilios
    What is the ISO in each photo? You mention Auto ISO but what exactly each camera actually exposed for?

    The banding in the 5D photo is quite ugly to tell you the truth.
     
  10. bongestrella

    bongestrella Mu-43 Veteran

    404
    Sep 2, 2011
    Mechanicsburg, PA
    Based on exif, em-5 is at ISO 800 while the 5d2 is at ISO 200. Wait what?
     
  11. bitmatt

    bitmatt Mu-43 Veteran

    271
    Apr 7, 2012
    Cleveland
    Matt
    That seems like it shouldn't make sense, but I also have both of these cameras and these results are no longer shocking to me. Generally my E-M5 images are nearly as clean as the Canon's at anything less than ISO 800, if not more so. I haven't found it worth my time at all to carry the bigger camera lately.
     
  12. FlyPenFly

    FlyPenFly Mu-43 Veteran

    448
    Feb 15, 2011
    Make sure the exposures are identical on a cloud less day.
     
  13. With_Eyes_Unclouded

    With_Eyes_Unclouded Mu-43 All-Pro

    Apr 17, 2012
    Vassilios
    This is virtually impossible. We cannot know the specific settings for each picture though.

    About the banding, it is somewhat of a characteristic on Canon's APS-C sensors, but I haven't seen it coming from the 5D... weird...

    Edit to give a coupla links:

    Sharpness, lens specific:
    http://colingriffiths.blogspot.gr/2012/05/olympus-om-d-v-canon-5d-mkii-lens-tests.html

    IQ and noise:
    http://dslr-check.at.webry.info/201205/article_4.html
     
  14. flash

    flash Mu-43 Hall of Famer

    Apr 29, 2010
    1 hour from Sydney Australia.
    Gordon
    The 5D2 is known to have significant banding in the shadows on some samples. There seems to be a lot of sample variation. It is much better at pulling back the highlights than pushing the shadows, as are most CMOS cameras (one of the reasons behind ETTR). Nikon and Sony sensors are better than Canon in the shadows but not as good as many CCD sensors. The OMD appears to be much more even in its highlight/shadow recovery.

    If you were do do the same test against a D700 you'd see a different result.

    I'm not saying the E-M5 isn't as good as it appears to be here. It is. But you've picked the 5D2's weak spot. However I personally prefer the roll off at both ends from the E-M5 over anything Canon currently have, except the 1Ds2 (not including the 1Dx). This little camera has a lovely smooth transition.

    Gordon
     
  15. bongestrella

    bongestrella Mu-43 Veteran

    404
    Sep 2, 2011
    Mechanicsburg, PA
    Again, based on the exif, exposure is different as well. 1/60 on the em-5 and 1/30 on the 5d2. So not exactly apples-to-apples comparison. It's a good gauge on em-5's, and the m43's sensor as a whole, improvement nonetheless.
     
  16. With_Eyes_Unclouded

    With_Eyes_Unclouded Mu-43 All-Pro

    Apr 17, 2012
    Vassilios
    This is one of the reasons I sold my APS-C Canon gear for the OM-D.

    It seems to me the all the sample photos I've seen confirm the measured DR lattitude. Furthermor, in my eyes, the way the OM-D renders noise is "film like" to a large degree. Also, color contrast (lens dependent also, naturally) seems to persist even when noise creeps in.

    I have to say though, that I have a few reservations about Lightroom. I use LR but I have also tried DxO Optics 7. I've found that DxO does a vastly better job at pulling info out of highlights and esp. shadows, in the majority of "difficult" cases. Perhaps they have more "inside info" on each vendor's RAW format? Dunno...
     
  17. Foster2380

    Foster2380 Mu-43 Regular

    47
    Jun 6, 2012
    In regards to the rest of the settings; the OM-D was 15mm, 1/60th, f4, ISO 800 and the 5D2 was 24mm, 1/30th, f4, ISO 200. Again Auto ISO was on for each camera in Aperture Priority which makes this very unscientific. It is however representative of how I leave the camera set in case I don't have time to make all exposure adjustments on my own before taking a shot.

    Maybe I'll take another shot with identical settings.

    If anything though, shouldn't the lower ISO actually favor the 5D2 while pulling shadows in LR4?
     
  18. With_Eyes_Unclouded

    With_Eyes_Unclouded Mu-43 All-Pro

    Apr 17, 2012
    Vassilios
    Please do! If possible in Manual mode, trying to "mimic" settings of one camera on the other. And with "Neutral" image settings (everything at zero).

    In my experience, this is NOT the case in LR4. Actually most of the time it's lower ISO that causes problems if you try and mess with a difficult exposure too much. In particular, the Clarity slider (which you didn't mention using) is a disaster sometimes....
     
  19. dko22

    dko22 Mu-43 Regular

    163
    Jul 26, 2010
    Stuttgart, Germany
    if the E-M5 is ISO 200 and the Canon ISO 800 it is an astonishing result!

    For what it's worth, I'm finding the Olympus highlight recovery rivals my (ex) Nikon D700 and shadows not far behind though the Nikon pulls ahead a little at high ISO. But if we're shooting static subjects then the generally larger aperture available on m43 together with all lenses stabilized makes for a surprisingly close contest with full frame cameras though if I shot action, which I don't, I'd still find a good DSLR useful.

    David
     
  20. fredlong

    fredlong Just this guy...

    Apr 18, 2011
    Massachusetts USA
    Fred
    The oddest thing to me in this comparison is the Canon sensor got three stops more exposure than the Olympus. Even taking the iso difference into account it got twice the exposure at its rated iso.

    Fred