Some of you may recall that when the original Olympus E-M1 was released 3 years ago, quite a controversy was stirred up by its handling of long exposures. Basically, the noise level for the E-M1 was much higher than that for it’s much cheaper predecessor, the E-M5. At the time, I was able to compare an E-M1 and E-M5 side-by-side, and wrote up the results for public scrutiny.
So, now that the E-M1II is available (if only in limited quantities so far), I thought it’d be interesting to compare my copy of it to my E-M1 that helped make such a stir (while I still own it). I’ve also got an E-M5II on hand, so thought I should throw it into the mix as well.
As before, I took dark frame images with each body — ISO 200, 60 seconds, noise filter set to “Standard.” For each body, I took one image with noise reduction (dark frame subtraction) off, and one with it on.
As you can see, after 4 years and many firmware updates, the E-M1 has visibly improved — but still needs noise reduction to yield a truly black frame (essential for astrophotography). It’s easier to see what’s going on with statistics, though — here, via RawDigger (pay particular attention to the standard deviation, σ, that spells out how broadly values are scattered):
The bottom line: the E-M5II is much better w.r.t. noise in long exposures than is the E-M1, and the E-M1II is better than both of them. The E-M1II is nearly good enough to not need Noise Reduction — and this, with a 60 second exposure!
Finally, a performance worthy of a camera body being positioned as Olympus’ “flagship” model.
The E-M1II also has a nice, new feature for long exposures using Noise Reduction. For both the E-M1 and E-M5II, the user is left waiting on the camera while it takes the exposures (regular, and dark frame) with no real indication of where the camera is in its process. But the E-M1II gives you a count-down timer when it’s performing Noise Reduction. I can imagine this will be a welcome sight for people doing long exposure work!
Thank you!!
You’re welcome! After the shock I had with my original E-M1, this came as a happy surprise. Makes the price of the E-M1II a *bit* less painful…
I’m trying to some night/astro photography with both the E-M5 mk II, and the E-M1 Mk II what would you suggest as the best way to process the files using Lightroom/Photoshop or Luminar. I”m trying to get some good results and don’t think I have issues capturing the images but the editing I’m still not sure about, especialy the sharpening and noise reduction.
Thanks
Vlast
I’ve only done a little astrophotography with the E-M5II (none yet with the E-M1II), but have found that a light touch in Lightroom is all that’s needed. Depending on your settings in the camera (long duration? live composite? noise reduction on?), you might just need to touch the contrast / dehaze a bit to deal with light pollution. When I’ve had NR turned on, I’ve never had to do anything with noise reduction in-camera — but this is after using low ISOs and relatively short exposures (<30 seconds). Your mileage may vary...
Great piece of analysis: I do a fair bit of astrophotography with my Mk 1, and you may just have tipped my decision to get a Mk II. Results from the I could be pretty decent, but with a fair bit of PP. not to mention freezing one’s butt off on a dark mountainside waiting for the NR 🙂
Thanks! I know just what you mean — the E-M1 has wonderful ergonomics (at least for working with my oversized hands), but its low-light performance leaves much to be desired. I was relieved to see the E-M1II’s behavior (at least in a test scenario), so I can get both. Now if only my local skies will stay clear at night, so I can try it out on actual stars…
Pingback: Olympus E-M1 Mark II High Resolution with flash | XTJ7.de
Having just received my E-M1 Mk II, I am really pleased to read this article.
I remember when I upgraded from my lovely E-M5 to the E-M1 and was ‘PHYSICALLY SHOCKED’ to find that, whereas you could take most shots on the E-M5 without NR, on the E-M1 it was always required. I did so many tests and showed my dealer so I know exactly what you mean.
After changing from the Sony sensor in the E-M5 to the Panasonic sensor in the E-M1, the change back to the Sony sensor on the E-M1 Mk II is the best thing that Olympus could have done.
Michaelfromoz
Thanks for spending the effort in performing this study. I only offer criticism to improve what we actually can learn from this data. I would suggest performing the test at a realistic ISO like ISO 1600, at ISO200 it takes several photon hits to generate 1 adu value so dark noise is naturally suppressed. Why does the EM5 have hot pixels and why are they not removed when NR is on?? And statistically are the values of the EM1 really any different from the EM1(II)?
Valid points, but I think some of your questions could only be answered by somebody at Olympus.
As to my testing, I did it at a low ISO because that’s what I used to compare the original E-M1 and E-M5; since I no longer own the E-M5, I wanted to be consistent with the old data’s approach. And realistically, the original E-M1 was far behind the E-M5 even at ISO 200.
That being said, I’ve still got my original E-M1, so will re-run the more-recent tests at higher ISO this weekend.
I would run both E-M1 and E-M1(II) at ISO 1600 where it’s likely that 1 photon is recorded as 1 adu with the body cap on. Take a two minute exposure (if you can) and take several so you can average the results since the internal temperature of the camera will get increasingly hotter with time and hopefully you can reach a constant plateau temperature. Then let things cool down (an hour?) and repeat with the long exposure NR set to on to see how good it works. This should definitely show if the (II) is indeed better than the (I).
Jim 8)
Very interesting results. Some time ago I was also comparing different OM-D sensors with RAWDigger.
Anyhow, the EM5 ii data seems strange as there is really no significant difference between the NR=on/off shot.
Either the camera failed to enable NR or NR is not needed for this camera type?? What do you think?