Some shots with E-M1+40-150 & D810+70-200

whumber wrote:

dpreviewreader wrote:

What matters is noise levels at higher ISOs, not at ISO 200 or 800. I processed DPR's studio shots for 5D3 and EM1 with DxO Optic Prime NR and default settings. 5D3's ISO 12800 image doesn't look better than the ISO 6400 image from EM1. I see similar results for 5D3 ISO6400 vs EM1 ISO3200, 5D3 ISO3200 vs EM1 ISO1600. IOW, in practice it is like a 1 stop difference.

I disagree with you on both counts. First, noise at ISO 200 definitely matters. This is really my only concern at the moment with switching my hiking kit completely over to m43. Even at base ISO you still have a fair amount of grain in the image. Some may not be bothered by it but it's definitely there and noticeable when printing large.

As for high ISO, I can only say that I actually own both cameras and I completely disagree with your assessment. I just took some quick shots as a sanity check and I see pretty much exactly 2 stops between the 5D3 and E-M1. The attached shot shows the E-M1 at ISO3200 | f/2.8 | 1/320s and the 5D3 at ISO12800 | f/5.6 | 1/320s with the 45 1.8 and 100L respectively. The EM1 is actually very slightly more exposed due to the better transmission of the 45 1.8 vs the 100L but it's a very small difference. The LR luma NR set to 0 and chroma NR +25. The EM1 shot is cropped to 3:2 and the 5D3 is downsampled to match the resolution of the EM1. I see almost no difference between the two. If I look really hard then I would say the 5D3 is very very slightly better but not by any signficant amount. I'd be happy to provide the RAW files if you want to check it yourself.

EM1 @ ISO 3200 [Left] and 5D3 @ ISO 12,800 [Right]

My results are given below

EM1 ISO6400

5D3 ISO12800