GX8 v G6 long exposures at night

Henry Richardson wrote:

Yes, the dpreview E-M10II review said the high ISO is special:

http://www.dpreview.com/forums/post/56564774

Henry Richardson
http://www.bakubo.com

Which they later recanted and attributed to testing error.

Thanks for the info. Can you provide the link?

Well, following the first link, it takes you to the DPReview review of the camera:

http://www.dpreview.com/articles/4887393042/bang-for-the-buck-olympus-om-d-e-m10-ii-review/8

Courtesy of The Wayback Machine, this is the original Raw Performance section, as of September 9, 2015:

http://web.archive.org/web/20150909185044/http://www.dpreview.com/articles/4887393042/bang-for-the-buck-olympus-om-d-e-m10-ii-review/8

"Looking at the Raw files, we were quite (pleasantly) surprised: the E-M10 II appears to have less noise than the E-M5 II, significantly less noise than its predecessor, and even catches up in performance to the Canon Rebel T6S. That is, the E-M10 II appears to be punching above its weight with respect to low light performance, catching up to larger APS-C sensors. This is an intriguing result, and we verified rigorously via quantitative signal:noise ratio analyses, which confirmed the results you see in our linked studio scene comparisons against the E-M10, E-M5 II, and Canon T6S. We'll be investigating this further to rule out things like noise reduction at play, but color us impressed for now.

To expound on this low light Raw performance, the pattern we saw in JPEG noise performance is largely mirrored here: the Nikon 1 J5's sensor size means it has more noise to deal with, while the larger sensor of the Fuji X-T10 and its built-in color noise reduction gives it a slight edge. The Nikon D5500 does show a slight noise advantage, but the fact that the E-M10 II is holding up this well to a class-leading APS-C is interesting. The E-M10 II fairs remarkably well against the Sony a6000, which we find to be a bit of an under-performer with respect to low light image quality."

And this is the current version of that section, December 10, 2015:

"Raw noise performance is similar to the E-M10 II's siblings, and is along the lines of what we've come to expect from micro four-thirds sensors. This means you can expect roughly a stop of ISO performance hit compared to modern APS-C cameras like the Nikon D5500, and roughly two stops compared to well-performing full-frame cameras like the Nikon D610.

Interestingly, we were initially very impressed by the ISO performance in our daylight-balanced scene, as the E-M10 II appeared to have less noise than the E-M5 II, significantly less noise than its predecessor, and even caught up in performance to the Canon Rebel T6S. It turned out, however, that this was most likely due to shutter speed inaccuracies with the E-M10 II, with fast shutter speeds actually being slightly longer in duration than stated (and therefore letting in more total light). When we reshot the E-M10 II with a 6 stop neutral density filter - which forced the camera to use slower shutter speeds - its performance dropped to the level of its siblings. In fact, a glance at our low light, tungsten-balanced scene, which uses slower, typically more accurate shutter speeds, shows that the E-M10 II does, in fact, perform similarly to the E-M5 II.

To expound on this low light Raw performance, the pattern we saw in JPEG noise performance is largely mirrored here: the Nikon 1 J5's smaller sensor means it has more noise to deal with, while the larger sensor of the Fuji X-T10 and its built-in color noise reduction gives it an edge. The Nikon D5500 shows a significant noise advantage as expected, due to its larger sensor. Finally, the E-M10 II catches up to the larger sensor Sony a6000, probably because the a6000 tends to be a bit of an under-performer with respect to low light image quality."

In the Comments, there is a post apparently by you, Henry R, asking them about their findings. They (Rishi Sanyal) replied:

"Yes, we discovered it was most likely due to shutter speed inaccuracies. When we reshot the daylight studio scene with a 6-stop ND filter (to force slower shutter speeds), the performance fell to that of the E-M5 II, which makes far more sense.

Additionally, the low-light scene shows no increase in performance relative to other micro four-thirds cameras.

We suspect that faster shutter speeds were actually longer in duration than they should've been, which essentially meant the camera was getting more light than it should have at higher shutter speeds.

I've just updated the review with this information. Sorry for the confusion."