Olympus OM1 equals Nikon Z7 in long exposure photography noise tests

In addition to birds in flight, I do a lot of long exposure photography of the sea and rivers. A good way to assess the long exposure performance of a camera is to take long exposure dark frame shots. I recently tested the dark frame performance of the new OM1, and as a matter of interest compared it to the same tests on my Nikon Z7. To my amazement, at exposures of 2, 4, and 6 minutes, the OM1 at ISO 200 without noise reduction was equal to or better than the Z7 at ISO 64.

This means that the time taken for a long exposure with the OM1 is halved, and it is now my default camera for this kind of work.

Long Exposure photography noise

The smaller Olympus 4/3rds sensors of the EM1 iii and the EM1 x always had the downside of high noise during long exposure shots, which meant that noise reduction (NR) had to be used. This process takes an additional dark frame image after the photograph is taken, and subtracts the dark frame noise from the image. This doubles the time taken for a 6 minute shot to 12 minutes for example. Full frame cameras usually have sufficiently good noise characteristics to mean the NR process is not needed, and this gave the larger format a considerable advantage for long exposure photography.

Typical LE exposure times

The exposure time used for LE of moving water depends on the speed of movement. This is my usual set of times:

• Fast moving rivers and sea flowing over rocks to give ‘lines’ of water: 1/6-1/8s
• Heavy seas on rocks to give the ‘mist’ effect: 2 minutes
• Slow rivers and flat seas to give a mirror effect: 6 minutes

Dark Frame tests

The best way to see the extent of the noise problem in camera is to take a dark frame, as this is what is used for noise reduction processing.

At speeds of 1/6 s there is no noise issue with either camera. I therefore tested the Olympus OM1 at it’s base ISO of 200, and the Nikon Z7 (which I also own) at it’s base ISO of 64, for times of 2m, 4m, and 6m, which are the only LE noise times of interest for me. For the Olympus I took shots with and without NR. For the Z7, I took shots only without NR.

Results

I have taken centre crops of the test images equal to around 200%, and show them below as JPGs at 100% quality. Click to get larger lightbox images. You will see that at 2m, both cameras show no noise. At 4m there are two luminance spots visible in the Z7 (lower left quadrant) and none in the OM1. At 6m, the luminance spots are slightly brighter in the Z7 and there are still none in the OM1. Also elsewhere in the Z7 frame at 6m there are a few red colour noise spots, and none in the OM1.

It’s worth noting that I did not test for the very long exposures that astro photographers sometimes need, and the results may well be very different there. However, one of the best YouTube astro experts is Ben at the Narrowband Channel (already a confirmed Olympus user), and he will no doubt soon be doing his customary detailed testing on the camera.

I would also note that if the dark frames are pushed from 2 to 5 stops, a great deal more noise is revealed in the OM1 sensor. However, pushing a pure black tone 5 stops is not not a scenario I will encounter, so again, it’s not of interest to me.

For the true enthusiast, the full size JPG files are here. If anyone is interested, the RAW files can be made available (email me), but as they take up over 180Mb, I am not putting them on this site.

Conclusions

For me, these results mean that for the kind of long exposure photography of water that I do, the OM1 needs no NR processing, and as a result the time of the exposure is halved. For rivers and slow seas, 6m + 6m or a 12 minute delay before seeing the image was intolerable. Now, it’s readily achievable. And the higher pixel count of the Z7 counts for little for NR scenes which are by definition low resolution at least in the areas of water. And this plus the massive benefits of the OM1 over the Nikon in actually shooting long exposures (see this post) make it a no brainer for all the LE work that I do.