Personal View site logo
Make sure to join PV on Telegram or Facebook! Perfect to keep up with community on your smartphone.
25mm m4/3 fast lenses comparison topic
  • At present, the fastest lenses to watch are the existing Voigtlander Nokton 25 f/0.95 and the proposed SLR Magic f/0.95

    Full list of contenders: (will be amended according to contributions)

    • Leica Panasonic DG Summilux 25mm f/1.4
    • Olympus 25mm f/1.8
    • SLR Magic f/0.95
    • Voigtlander Nokton 25 f/0.95

    (This original post revised, thanks Vitaliy)

  • 23 Replies sorted by
  • Please make title specific, if you describe content as specific.

  • [Moved from 25mm f/0.95 Voigtländer Nokton topic]

    Those two comparison shots speak for themselves. I appreciate the way the tweeter did not extrapolate or embellish.

    The shots compare an identical subject of a technical nature. They look like similar distance from lens to subject.. Maybe not.


    This is an excellent comparison for technical subjects, good fine lines and limited colours. Full marks to the tester! If I want to photograph anything like another lens, close up and in low light, I'd choose an SLR Magic 25 over a Voigtlander. however does extrapolate, using the old question mark trick, "Is the SLR Magic.. better?" to get a headline and sell some ad clicks. That's media biz. Stay tuned for blow-by-blow comparisons at 1, and 2, 5 metre distances, then skin tones right up to circles of confusion. A page for each and an ad on each page; hyperbole in the copy serving to incite a frenzy of contributors' opinions . That's the state of science in 2013.

  • Slr Magic one is sharper. No question about it. I want more thoughtful comparison.

    Bokeh look and feel, flares, CA, distortions, vignetting, maximum sharpness at f4 5.6 8, build quality, focus play and feel, etc.

  • @stonebat

    I am starting to believe the actual apertures vary, as some lenses do - and that this clears up the differing DOFs in some of the pics we've been given.

    Somebody recently quoted the SLR-Magic 25mm as really a "T/0.95" camera; and that could easily explain the sharpness. It could be like this: The SLR-M has glass elements that let in more light than the Voigtlander. Its aperture could then be mathematically or optically smaller, focusing sharper whilst still letting in the same amount of light as an f/0.95. That would explain a whole lot about the obviously different depths of field! Below, Hugo of Beeldlab Blog says he focused on the letter "P" with the Voigtlander and on the "R" with the SLR_M. But the latter has nearly all the letters in focus, while the Voigt's lens markings are sharper. The right-hand pic is even darker, dammit!

    So ignore Hugo's sloppy comparison work, tighten up the aperture on the Voigt and they'd look much more alike.

    [Edit: added a close-up of my own test from some time ago: tried to guess distance and light so as to make it comparable. But still a genuine 0.95 aperture]


    490 x 577 - 52K
  • If this was a "test" I have no idea why the tester would focus at different points. The Voigt's lens markings "appear" to be sharper because the SLR's lens markings are out of it's DOF, not because it is less sharp. When stopped down, I'm sure the they will be comparable, but more "tests" need to be done.

  • @gonna I'm back, but not to haunt Nok lovers like yourself. Not too long ago, we got into some argy-bargy here, but I'm glad you've tested the lens more extensively, and discovered some things for yourself. The halo effect was one reason why i sold it. To be fair, most fast lenses especially those at f1,4 would tend to glow in this way, some Contax Zeiss and early Leica R glass comes to mind. Just didn't like the Nok's hazy look enough for me to keep it.

  • @stonebat

    The SLR magic is your lens for sharpness at low light. If you want shallow DOF, the Voigtlander is the one.

    Once again, take a squiz at the video link re T-stops.

    [ If we can believe the tests so far], I'd question whether the SLR-M lens iris opens up fully. Shall we call it a widest mechanical, measurable opening of around f/2 in order to explain it's wide DOF and sharpness, as observed in the sample pics? .

    Did they lie to us by putting f-stop markings on the barrel? No. They have to put them there - by convention and for exposure calculations.

    But a lens which has better glass* which lets in more light can give impressive sharpness stopped down. (Please open up that iris, I say, give me that blurry f/0.8 or whatever; I'll take care of the stopping down myself).

    *Maybe it's not the glass. It could be coating or lower reflectivity sending all that extra light at the iris. [Hell, they may even have some way of concentrating more light onto the iris. (singing)~ Birds do it, bees do it, Meta-bones & magni-fying glass do it].. .

    I see the day coming where T-stops will make up a bigger part of our language.

  • Focus test. Repeatable focus seems to work fine. Ring not as smooth as the Voigtländer though.

  • Thanks for the test.

  • Has anyone else compared the SLRM 25mm to the CV 25mm ?

  • Yes. Basically, the SLRM is sharper in the center WO, while the CV is a bit softer in the center and sharper in the corners. Stopping down, the CV is overtaking the SLRM, but the it's bokeh is a bit harsher.

  • @nomad does it look like the SLRM is 1/3 stop darker WO or it just the way it renders color? Can you comment how does your focus ring feel?

  • I get exactly the same readout on both WO.

    The focus ring on the CV feels buttery smooth and has no play whatsoever, the SLRM has some play and feels a tad stiffer and not fully constant in friction, but still quite useable. The CV has about 270 degree focus throw, the SLRM only about 180, but the additional throw is only used for the closer NFD (17cm vs. 26).

    From the mechanical side of things the CV is superior, but lens aesthetics are an important factor, and highly subjective…

  • I saw this a while ago, thought it was pretty interesting how each performed:

    [This little shootout was more a kind of spontaneous incident as we had some daylight remaining at the end of an "Adorable 35s" session planned for that day in Hong Kong with Andrew Chan and some friends. Andrew is the driving force of SLR Magic and very passionate about great lenses - not only those he builds by himself. But let me state that I am not affiliated with SLR Magic and payed for all lenses I got from them. We got some kind support from Bertille Tabourot, who is a stylist and fashion designer from France, living currently in Hong Kong. For this comparison shooting she took the chance to present one of her fancy creations acting as a freelance model.]

  • I hesitate between SLR Magic f/0.95 and Voigtlander Nokton 25 f/0.95... As S Kubrick, I want to shoot some cheap budget footage lighted only with candles. So I just need more light in my sensor :-)

    i did some tests with my smc takumar 50 1.4 : GH2, A mode, base iso, still photo, same exposition gives :

    F 1.4 -> 1/160 seconds

    F 2 -> 1/100

    F 2.8 -> 1/50

    F 4 -> 1/25

    Then you notice that from F 2 to F 1.4 some light just escaped rrrrr.

    For comparison, could some of you do same test with same gear and SLR Magic f/0.95 and Voigtlander Nokton 25 f/0.95 ?

    Voigtlander Nokton 25 :

    F 0.95 -> ? seconds

    F 1.4 -> ?

    F 2 -> ?

    F 2.8 -> ?

    SLR Magic f/0.95 :

    T 0.95 -> ? seconds

    T 1.4 -> ?

    T 2 -> ?

    T 2.8 -> ?


    From tests it is common for light to "escape" when faster than f/4. You get shallower dof but not necessery more light due to limitaion of sensor technology.

  • thx for your answer :-)

    In fact I just need a comparisons of your beautiful lense with the beautiful nokton. Hey, Dxomark did comparisons with a lot of lenses about this "loss" of light...

    Nota: as far as i have understood dxomark article F-stop-blues shouldnt be T-stop-blues no?

  • @slrmagic may be you can give us your test for the SLR Magic T cine/0.95 ! it is quite easy and quick to do it. So : SLR Magic T cine/0.95, GH2, A mode, base iso, still photo, same exposition gives : :

    T 0.95 -> ? seconds

    T 1.4 -> ?

    T 2 -> ?

    T 2.8 -> ?

  • @magnifico the results may vary depending on camera sensor and camera firmware. For example, many learnt the Pocket camera can perform very differently to light depending on sensor calibration or camera firmware so there is no fixed value.

  • T-stops should give you quite correct steps, while f-stops don't wide open. The transmission losses in relation to the aperture contribute a higher percentage WO. When comparing lenses without T scale for real stops (unfortunately not the ones you are asking for), I constantly observed any significant differences at stops below 1:2.

  • @slrmagic I agree with you :-) btw I just need a comparison of the 2 lenses with one "observer" lets say my GH2... Yesterday I had the chance to try nokton 25mm .95 stuck in my GH2. Test was done, but please consider it as an non official quickie, (I did nt check luminance in raw picture, etc).

    So: Nokton 25mm .95, GH2, A mode, base iso, still photo, same exposition gives :

    F 0.95 -> 1300 seconds

    F 1.4 -> 800

    F 2 -> 400

    F 2.8 -> 200

  • That means my sensor with Nokton 25mm gathers 500/800=0.625 lights units more light from F 1.4 to F.95. We are here at a sqrt(2/1.625)= 1.1 level of light reaching and captured by the sensor.

    That means Nokton 25mm should behave like a T 1.1 (no light lost to the sensor) with my gh2 at full aperture.


    What about SLR 25mm T.95? I read today from dpreview in the "SLR Magic 25mm -light gathering capabilities" thread : "...when turning the aperture from T1.4 to T0.95, it only gathers about 0.3 light units more light, when it should be 1EV in theory..."

    I dont even know if the numbers are ok, so be careful here! So with these numbers we find Sensor (Olympus OMD) level for SLR 25mm is : sqrt(2/1.3) = 1.24

    That means SLR 25mm should behave like a T 1.24 (no light lost to the sensor) with an Olympus OMD at full aperture.

    My 2 cents here, as we need a more precise experiment to tell the right numbers.