Hi,
Thanks for the observation!
I think that enclosed details explain part of the question. In the Phase One histogram you can that there are counts down to channel one while the Sony histogram has zero counts below 15 (or so), that is because the P45+ distributes the darks over many channels. it has something like 15 electron charges of readout noise. The Sony has something like 3 electron charges of readout noise (I guess).
What the histogram shows is not really that the P45+ has a one stop advantage but it indicates a wider subject illumination range. The probable explanation for that may be that light has changed, but my guess is that lens flare plays a major role. Some of he light bounces around in the lens and will lighten up the darks. The Zeiss Distagon 50/4 used has 8 lens groups while the Sony 24-70/2.8 has 13 groups. Both lenses are T* coated, so I guess per lens reflections are similar. With more lens groups there will be more light bouncing around in the lens. Also the 24-70/2.8 is a zoom lens and I would suggest that the lens barrel may be less optimal for containing reflections than on the prime lens. So the lens flare reduces overall contrast.
I have seen this pretty clear on my unpublished Stouffer wedge shots. The Stouffer wedge has 14 stop range. The 24-70/2.8 showed excessive flare, so I used my 100/2.8 macro for those shots. The reason I did not publish the results from the Stouffer wedge was that they were hard to evaluate. There is a lot of information in the image, but it is hard to extract. Also, DxO-mark makes similar tests, in a scientific way.
The comparison here was not a scientific one. It was shot on a sunny day with clouds. What I did was to try to find a couple of reasonably ETTR exposed images and compare shadow detail. So it is an ad hoc experiment. The shadow areas in sunlit areas were quite constant, I would guess.
What I see is that I normally don't have problems with DR on either P45+ or Alpha 99, but darks are cleaner on the Alpha 99 (if you are really pushing darks). The Alpha 99 has wider dynamic range than the Alpha 900 I had before, but I could not observe it on real world shots. The only case I could see was duping Velvia in a "totally" dark room with all light leaks masked of. My guess is that flare dominates normal shots.
The situation may be slightly different if you are shooting in large dark area with some small illuminated area.
Another observation I made is that I have shot quite a few HDR images, but in general I just prefer pulling shadows and compressing highlights on an ETTR image.
These two articles may contain some interesting insight:
http://echophoto.dnsalias.net/ekr/index.php/photoarticles/61-hdr-tone-mapping-on-ordinary-imagehttp://echophoto.dnsalias.net/ekr/index.php/photoarticles/63-lot-of-info-in-a-digital-imageThis image is a three exposure HDR:
While this one is made from the -2EV exposure by pushing shadows an pulling highlights (entirely processed in LR4):
The HDR image has less shadow noise, but it will always (or almost always) have some artifacts. I still shoot HDRs, it is just that I don't really use them that often.
Best regards
Erik
Erik, thanks for the link. It seems to say in the histogram that the P45 has about 1 ev advantage over the Sony A99. +3 to -8 vs +3 to -7. It's also interesting that the slope of the darkest evs are completely different for the same shot.
On the picture, the P45 blocks up to black much faster than the A99 on the darkest tree trunks despite the 1 ev advantage. Any idea why that might be?