Announcement

Collapse
No announcement yet.

OM-D E-M5 image samples developed from RAW

Collapse
X
  • Filter
  • Time
  • Show
Clear All
new posts

  • OM-D E-M5 image samples developed from RAW

    Samples developed from RAW with the Noise Filter set to OFF

    It has been pointed out that the OM-D E-M5 pre-production camera sample images I have already published here:

    ..are camera JPEGs produced with the camera's noise filter set to 'Standard'. This does mean some smoothing is applied and so some detail is sacrificed. But these shots were also recorded in RAW files and I have reproduced the same shots using Olympus Viewer 2.3 developed from RAW with the noise filter set to OFF.

    I must emphasise that although I was able to process the E-M5 RAW files using Viewer 2.3 I don't believe that Viewer 2.3 - even though it was updated just a few days ago - is yet optimised for E-M5 RAW files. But I find that the results are surprisingly good. Nevertheless, I think a later update to Viewer will produce even better results.

    Once again, I have complied with Olympus request not to actually publish the RAW file, and what I have published is reduced to 1600x1200 pixels because the camera used is not a finished production model. For each ISO setting you can choose a link to open the version without a noise filter and one with the noise filter (NF) set to standard. The version processed from RAW with no NF was exported from Viewer to 16-bit TIFF and imported to Lighthroom 3.6 before being re-exported as a 1600x1200 JPEG, just like the original JPEG samples.


    My view is that even at ISO 25600 there is surprisingly little luminance noise smoothing with the noise filter at Standard. But judge for yourself - and please post your feedback here

    Ian
    Founder/editor
    Four Thirds User (http://fourthirds-user.com)
    Digital Photography Now (http://dpnow.com)
    Olympus UK E-System User Group (http://e-group.uk.net)
    Olympus camera, lens, and accessory hire (http://e-group.uk.net/hire)
    Twitter: www.twitter.com/ian_burley
    Flickr: www.flickr.com/photos/dpnow/
    Pinterest: www.pinterest.com/ianburley/

  • #2
    Re: OM-D E-M5 image samples developed from RAW

    Quite impressive. Using higher ISO's will not have the concern as it was with using ealier Oly DSLR's.
    Ross
    I fiddle with violins (when I'm not fiddling with a camera).
    Cameras: OM-D E-M1 & Mk II, Olympus Stylus 1, OM-D E-M5.
    Lenses: M.ZD12-40mm f2.8 PRO Lens, M.ZD40-150mm f2.8 PRO Lens with MC-14, M.ZD12-50, M.ZD60 Macro, M.ZD75-300 Mk II, MMF-3, ZD14-54 II, Sigma 150mm F2.8 APO Macro DG HSM.
    Flashes: FL36R X2, FL50R, FL50.
    Software: Capture One Pro 10 (& Olympus Viewer 3).

    Comment


    • #3
      Re: OM-D E-M5 image samples developed from RAW

      Is it just me or is there a huge jump in noise between ISO 12800 and ISO 25600? To me, the unfiltered 12800 has less noise (and obviously less smoothing) than filtered 25800.

      I won't say I'd never use 25800, but I'd be reluctant. However, on the strength of these examples I'd happily use 12800, either filtered in camera or in pp from raw. At ISO 6400, to be honest I cannot really see much difference between filtered and unfiltered! On this showing I'd be tempted to shoot unfiltered and reduce just a little in pp if really needed.

      What are your views, having seen the full res files, Ian?

      Edit: Not that I can forsee many situations when I'd need to go above ISO 6400 anyway!
      View my ebook, The Light Fantastic, at: http://store.blurb.co.uk/ebooks/3026...ight-fantastic

      John

      Comment


      • #4
        Re: OM-D E-M5 image samples developed from RAW

        Thanks Ian for posting the images.
        Looks like noise filters could become a thing of the past, really impressive.
        I`ve just been scanning some old negatives on my Imacon Flextight Precision II scanner at 3150 and 6300 DPI and as a comparison the image at 21500 ISO is comparable to film at 100 ISO. In fact I`d say the Olympus is sharper (the lens and camera is Minolta 7000 with 50mm F1.7 and films, AGFA XRG 100 & Kodak Gold neg films).
        I want an OM-D E-M5.
        Andrew.

        Comment


        • #5
          Re: OM-D E-M5 image samples developed from RAW

          Originally posted by John Perriment View Post
          Is it just me or is there a huge jump in noise between ISO 12800 and ISO 25600? To me, the unfiltered 12800 has less noise (and obviously less smoothing) than filtered 25800.

          I won't say I'd never use 25800, but I'd be reluctant. However, on the strength of these examples I'd happily use 12800, either filtered in camera or in pp from raw. At ISO 6400, to be honest I cannot really see much difference between filtered and unfiltered! On this showing I'd be tempted to shoot unfiltered and reduce just a little in pp if really needed.

          What are your views, having seen the full res files, Ian?

          Edit: Not that I can forsee many situations when I'd need to go above ISO 6400 anyway!
          Relax. You won't ever have to worry about using 25,800 on this camera, but you might be tempted (or left with no option) to use 25,600. Sorry, I couldn't help it. :\ That would be the top most ISO value that would not normally get used & I would think that if the highest value didn't have some degradation in it, then users would be wanting to see what a higher value (that wasn't included) might be like if the highest setting was good. I'm impressed with what this camera can do in the ISO range.
          Ross
          I fiddle with violins (when I'm not fiddling with a camera).
          Cameras: OM-D E-M1 & Mk II, Olympus Stylus 1, OM-D E-M5.
          Lenses: M.ZD12-40mm f2.8 PRO Lens, M.ZD40-150mm f2.8 PRO Lens with MC-14, M.ZD12-50, M.ZD60 Macro, M.ZD75-300 Mk II, MMF-3, ZD14-54 II, Sigma 150mm F2.8 APO Macro DG HSM.
          Flashes: FL36R X2, FL50R, FL50.
          Software: Capture One Pro 10 (& Olympus Viewer 3).

          Comment


          • #6
            Re: OM-D E-M5 image samples developed from RAW

            Originally posted by AndyBeeson View Post
            Ive just been scanning some old negatives on my Imacon Flextight Precision II scanner at 3150 and 6300 DPI and as a comparison the image at 21500 ISO is comparable to film at 100 ISO. In fact I`d say the Olympus is sharper (the lens and camera is Minolta 7000 with 50mm F1.7 and films, AGFA XRG 100 & Kodak Gold neg films).
            I want an OM-D E-M5.
            Andrew.
            I've read in the past that the Flextight Precision II has problems scanning 35mm colour neg but produces great scans from slides for some reason.

            I borrowed an Imacon some years ago but can't remember what model it was and only scanned slides with it.
            It's the image that's important, not the tools used to make it.

            The Grumpy Snapper blog or follow me on Instagram.

            Comment


            • #7
              Re: OM-D E-M5 image samples developed from RAW

              I am very impressed - when looking at this sort of comparison I usually write off the top ISO setting as "emergencies only" but here I think it is available for normal usage.

              Roger
              Last edited by RogerMac; 19th February 2012, 03:55 AM.

              Comment


              • #8
                Re: OM-D E-M5 image samples developed from RAW

                I am surprised how clean is ISO 3200 with NF OFF. I know it was downsampled, but I can hardly see a difference between pictures with and without NF (most visible is some color shift). There is a little detail loss, but almost no noise, which is strange.
                That leads me to some questions:
                -is it possible, that Oly now do the same as Nikon, ie. altering/denoising RAW before saving it?
                -is it due to used converter, that may apply denoising on import without telling it to user?
                -am I paranoid and the new sensor/processing combo is simply so great, that it can beat even D4 image samples?

                I think the ISO 25600 won't be usable due to low-frequency noise maps.. maybe after some advanced denoising or in B/W.
                What is more interesting, is amount of detail preserved - most of the texts are readable even at highest ISO!
                Regards, Pavel.

                E-1(x2) | ZD 14-45 | ZD 14-54 | ZD 40-150 Mk.1 | ZD 70-300 | FL-50 | Velbon Sherpa 750R
                M42(Pancolar 50/1.8, Pentacon 200/4)+ M42->4/3 reduction

                Comment


                • #9
                  Re: OM-D E-M5 image samples developed from RAW

                  Hi,
                  I`ve had no problems with the Imacon other than getting the colour right which I can correct in Photoshop anyway. It pics out lots of detail from the negatives and slides. If a negative or slide is perfectly exposed and developed properly then you`ll get good results but still no match for the OM-D I think especially at equal ISOs. The OM-D looks a vast improvement over my E-PL1 by the looks of the results also.
                  Andrew.

                  Comment


                  • #10
                    Re: OM-D E-M5 image samples developed from RAW

                    Originally posted by David M View Post
                    I've read in the past that the Flextight Precision II has problems scanning 35mm colour neg but produces great scans from slides for some reason.
                    You may have seen this (Review /1):

                    http://www.photographyreview.com/cat...3_3126crx.aspx

                    Those comments were written a long time ago. Don`t know how useful they are today.

                    Comment


                    • #11
                      Re: OM-D E-M5 image samples developed from RAW

                      12,800 is pretty impressive. Welp, I might just have to change my mind on that GF1. Still prefer the design of the GF1, but frankly don't give a damn if the picture quality is significantly better from this guy.

                      I will be at B&H in March, hopefully I can try them both when I am there.
                      ~Reggie

                      Nikon D3s | D500 | Nikkor 300mm f/4 PF | Tamron 10-24 f/3.5-4.5 | 70-200 f/2.8 G2 | Sigma 50 f/1.4 Art | 500mm f/4 Sport

                      Panasonic GX8 | GF3 | 20 f/1.7 II | 12-35 f/2.8

                      Comment


                      • #12
                        Re: OM-D E-M5 image samples developed from RAW

                        The image quality is impressive. I'm wondering how reliable the AF is at those high ISO's and low shutter speeds.

                        Comment


                        • #13
                          Re: OM-D E-M5 image samples developed from RAW

                          Incredible, just incredible. Not much more to say really Ian.
                          John
                          Olympus UK Forum
                          E-3 | E-510 |
                          E-400 | 7-14mm | 11-22mm | 14-54mm |35mm | 40-150mm | 14-42mm | 70-300mm | 1.4x | FL-50 | My Gallery - www.reflectingme.com

                          Comment


                          • #14
                            Still to wait and see

                            Usually, it's too early to tell, but some improvement it is. Especially I find the shadows much cleaner than what I've seen in older generation cameras. I will not bother with the high ISO settings, I rather see that I have cleaner shadows and an improved dynamic range. Hopefully the new sensor generation if it's a such will deliver on that.

                            Comment


                            • #15
                              Re: OM-D E-M5 image samples developed from RAW

                              Originally posted by Ross View Post
                              Relax. You won't ever have to worry about using 25,800 on this camera, but you might be tempted (or left with no option) to use 25,600. Sorry, I couldn't help it. :\ That would be the top most ISO value that would not normally get used & I would think that if the highest value didn't have some degradation in it, then users would be wanting to see what a higher value (that wasn't included) might be like if the highest setting was good. I'm impressed with what this camera can do in the ISO range.
                              Haha! I'm all at sixes and eights when talking about high ISO!
                              View my ebook, The Light Fantastic, at: http://store.blurb.co.uk/ebooks/3026...ight-fantastic

                              John

                              Comment

                              Working...
                              X