Check this out
From that test, I cannot see 1 stop over 5d3, at most half a stop, maybe still not as good as D3s or D4
Yes, someone just pointed this out to me. I was about to post here that I'm no longer sure about the whole all-time SNR king stuff. Not sure what to say. One guy's test with the sunflowers showed an easy 1DX SNR advantage. Stan's test does not. Both tests seemed to be carried out well. Anyway it shows that using real world images and programs that process RAW introduce lots of variables. Not that you can use them for SNR, but the much sneered at 'fake' 'ridiculous' black frames for DR sure make things easy and results repeatable.
May as well just wait for DxO for the SNR results and we'll see whether the sunflower guy or stan's test tells the truth better. I almost wonder if it won't be closer to Stan's. The results from the other guy's files jsut seemed like such stunnigly, shockingly amazing news maybe they were too good to be true.
(EDIT: or maybe copy to copy variation?? http://www.canonrumors.com/forum/index.php?topic=8028.0
Along these lines I've noted that my 5D3 has a bit worse vertical banding at low ISO than some otehr copies
OTOH it seems solid for high iso banding while Stan's copy seems to get more high iso banding than other copies I've seen. Maybe his 1DX is extra noisy? who knows)
SOmeone else tested DR at high iso and the 1DX definitely beats the 5D3 for that, at high iso, though. That test uses black frames so it should hold up well.