1. Reminder: Please user our affiliate links to get to your favorite stores for holiday shopping!

Long Exposure Comparison

Discussion in 'Olympus Cameras' started by ExCoderAtWork, Jul 3, 2015.

  1. ExCoderAtWork

    ExCoderAtWork Mu-43 Rookie

    15
    Jan 20, 2014
    I recently got a Lee filter set with a Big Stopper to do some long exposure photography during the day. I own an E-M1 and knew it had challenges with long exposures. What I didn't realize was how bad it was. Even with dark frame subtraction I thought the images were unusable. As a result I bought an E-PL5, knowing it had the same sensor as the E-M5. With both cameras in hand I decided to do some comparison shots. I did not turn on dark frame subtraction so I could see the sensor characteristics clearly. I wanted to do this because while helpful, by its definition, dark frame subtraction removes information from a photo thus degrading the image.

    All shots were taken with the same settings and lens and the lighting in the scene was fully controlled. The only post-processing done was to standardize the white balance between shots (I forgot to turn off Auto WB). I really can't believe that the E-M1 is this bad. Is there something that I am doing wrong?

    Here's the info for the shots:

    ISO 800
    F 2.8
    150 s exposure
    IS turned Off
    40-150 Pro at 90mm
    No Dark Frame Subtraction

    E-PL5 (No Noise Filter)
    19352740096_2a45b3c3a0_z.
    E-PL5 No Noise Filter
    by ExCoderAtWork, on Flickr

    E-PL5 (Standard Noise Filter)
    19352703246_0de8dfa574_z.
    E-PL5 Standard Noise Filter
    by ExCoderAtWork, on Flickr

    E-M1 (No Noise Filter)
    19382830021_ece73f754e_z.
    E-M1 No Noise Filter
    by ExCoderAtWork, on Flickr

    E-M1 (Standard Noise Filter)
    18758274133_317e4e8eb5_z.
    E-M1 Standard Noise Filter
    by ExCoderAtWork, on Flickr
     
    • Informative Informative x 1
  2. wjiang

    wjiang Mu-43 Hall of Famer

    Hmm, interesting test. I don't have my old E-M5 to compare with the E-M1 any more, but I can't say I've run into this much trouble (but then I haven't gone longer than 30 s before). A multiple minute exposure from the E-M1 does look a bit disappointing though, especially when my GM5 (which has a Panasonic sensor as well, go figure) gives better results (it's limited to something like 64s though).
     
  3. AL904

    AL904 Mu-43 Regular

    42
    Mar 29, 2012
    Jacksonville Beach, Florida
    Allen Forrest
    Very interesting (and disappointing) test!

    Does the Olympus EM-5 Mark II have the same Sony sensor as the EM-5 Mark I and the E-PL5 ? or a Panasonic sensor like the EM-1?
     
  4. fsuscotphoto

    fsuscotphoto Mu-43 Top Veteran

    819
    Feb 15, 2013
    St. Cloud, FL
    Ron
    The II has the same sensor, but it is supposed to be improved.

    I am stunned by the EM-1 results. I have both 5 models and have seriously been considering some ND filters.
     
  5. Lcrunyon

    Lcrunyon Mu-43 Top Veteran

    755
    Jun 4, 2014
    Maryland
    Loren
    I use the E-M1 and the Big Stopper often. I know that the E-M1 has been criticized in this area (generally in astrophotography), but I've never had results like that. In fact, I don't see noise in my long exposure shots. However, I keep my ISO at 200, I don't tend to do such long exposures (usually don't have to go over 60 secs) and if I am using an ND filter, I'm most likely shooting in the day, so conditions were always much more conducive to less/no noise.

    Were you shooting in a darkened room with the Big Stopper on as well? Maybe you were just taking the E-M1 beyond its (admittedly lower than the E-P5) capabilities. But I don't think this is indicative of normal uses with an ND filter and E-M1.
     
  6. barry13

    barry13 Super Moderator; Photon Wrangler

    Mar 7, 2014
    Southern California
    Barry
    Fwiw, Luminance noise can be reduced (on any sensor) by overexposing a bit.

    Barry
     
  7. lightmonkey

    lightmonkey Mu-43 Veteran

    480
    Dec 22, 2013
    only speaking for myself here, but theres no scenario id ever need a 150s exposure so while i dont deny the validity of your results, its all but irrelevant for most?
     
  8. PeHa

    PeHa Mu-43 Veteran

    206
    Nov 12, 2013
    Uppsala, Sweden
    The E-M1 is very dependant on dark frame subtraction, if you turn that on (menu option Noise Reduction on or auto) the difference to E-M5 will nearly disappear. However, with real long exposures and high ISO (like 5 minute exposure at ISO 3200), the E-M5/PL5 will be noticeable better. But E-M1 is usable for astrophotography, the sample below was shot with E-M1, 3 minutes exposure and ISO 1600.

    [​IMG]
     
    • Like Like x 8
  9. maritan

    maritan Mu-43 Veteran

    388
    Oct 30, 2014
    Holy crap, @PeHa@PeHa. Nicely done!
     
    • Agree Agree x 1
  10. PeHa

    PeHa Mu-43 Veteran

    206
    Nov 12, 2013
    Uppsala, Sweden
    Thanks @maritan@maritan! (it's the Orion nebula shot through a 500mm Skywatcher Equinox refractor telescope).
    It did take some postprocessing including extra noise filtering but that would be the routine with the E-M5 as well. I guess a reason why the E-M1 is more sensitive to noise is because it gets more easily heated, possibly due to features like the phase detection sensors that the other m43 cameras dont have.
     
    • Like Like x 1
  11. AL904

    AL904 Mu-43 Regular

    42
    Mar 29, 2012
    Jacksonville Beach, Florida
    Allen Forrest
    That's interesting! If that is the cause, couldn't a firmware update turn off the phase detection sensors for long-exposure shooting?
     
    • Agree Agree x 1
  12. PeHa

    PeHa Mu-43 Veteran

    206
    Nov 12, 2013
    Uppsala, Sweden
    Sounds wise but I wonder if its possible to turn part of the sensor off with software. It might also be the surrounding chip architecture just to allow the features that causes heating in long exposures. And the E-M1 sensor is Panasonic while the E-M5 is Sony (from what I've read) so the base design might be real different. There's quite a lot of info/theories on this issue to be found by googling on e-m1 long exposure hot pixels noise etc
     
  13. LovinTheEP2

    LovinTheEP2 Mu-43 Top Veteran

    619
    Feb 15, 2011
    Toronto
    • Like Like x 1
  14. LovinTheEP2

    LovinTheEP2 Mu-43 Top Veteran

    619
    Feb 15, 2011
    Toronto
    I haven't seen any long exposure pictures of the em5ii, let alone with dark frame subtraction action off.. Have been able to track some down.. Just curious to see what the noise pattern is like.
     
  15. wjiang

    wjiang Mu-43 Hall of Famer

    Gordon at CameraLabs has comparisons between E-M1, E-M5 and E-M5 Mk2 long exposure noise in the E-M5 Mk2 review.
     
  16. LovinTheEP2

    LovinTheEP2 Mu-43 Top Veteran

    619
    Feb 15, 2011
    Toronto
    Just checked out camera labs. Saw one pic of a crop with and without noise filtering .. The pier one. Maybe I'm missing something but hardly any comparison between the 3 bodies done at the same time with same settings for all three bodies with and with dark frame subtraction no any written observations of how the mark 2 behaves in comparison to the mark 1 on long exposures without a bias. Maybe it's there, just didn't see it,
     
  17. PeHa

    PeHa Mu-43 Veteran

    206
    Nov 12, 2013
    Uppsala, Sweden
    I couldnt find the three cameras compared in the same test (at Cameralabs at least) but the separate EM5 Mark II review tries long exposures and says noise is handled more like EM5 than EM1 which means better and not so dependant on Noise Reduction/Frame Subtraction.
    See http://www.cameralabs.com/reviews/Olympus_OMD_EM5_Mark_II/ (search "long exposure" on page).
    The pier testing is here http://cameralabs.com/reviews/Olympus_OMD_EM1/EM1_long_exposure_noise_vs_EM5.shtml

    Personally, I use the E-M1 for 90% of my shooting but also have an E-PM2 for astrophotography because of this issue with the E-M1 (I've also dared to remove the internal cut filter of the E-PM2 to make it even more adapted for astro work).
     
  18. ExCoderAtWork

    ExCoderAtWork Mu-43 Rookie

    15
    Jan 20, 2014
    In general shooting, I totally agree. But as mentioned earlier, astro-photography is a scenario where an exposure this long may be beneficial. I could also see benefit for this long of an exposure in architectural photography in order to remove non-stationary people from the shot.
     
    • Agree Agree x 1
  19. Lcrunyon

    Lcrunyon Mu-43 Top Veteran

    755
    Jun 4, 2014
    Maryland
    Loren
    True. Occasionally longer exposure times will be needed where dark frame subtraction is not possible.

    Still, I have to wonder about the OP's results. Those are by far the worse that I've seen, and I use the E-M1 and Big Stopper on a regular basis. I also would have to test if 2.5 minutes is really necessary for removing people. Seems like a lot, unless they were just standing around.

    In that kind of scenario, I wonder if it is possible to implement something similar to dark frame subtraction, but with a normal shot. It could fill in the pixels, so to speak... Just an idle thought..
     
  20. wjiang

    wjiang Mu-43 Hall of Famer

    For removing people, you might be able to get away with stacking half a dozen shots using median blending of moderately long exposures, provided the base exposure is long enough to do a decent job of blurring already. For guided astro shots, stacking moderately long exposures works too. For star trails Live Composite might work well.