p.2 #13 · Sony a7s Topic [DR + HIGH ISO + RAW + compare 1dx] - REVIEW
sebboh wrote:
awesome tests, thanks! we need to find someone to loan you a few rangefinder wides.
I should ping my Sony contact, though due to other commitments, it wouldn't happen for at least another week or two.
I suspect if sensor topping specs are similar for the 'S' relative to the other a7 cameras, edge color shift may be better but I doubt there will be significant difference in edge smearing. I believe edge color shift is influenced by pixel density whereas smearing is very dependent on sensor topping thickness. Both are influenced by lens exit pupil distance.
As for the 1DX... yeah, Sony's sensors sure are some of the best at the moment. Too bad they don't make a camera that can rival the 1DX (or D4) in other categories, along with the full stable of lenses. Then they'd seriously have my attention. But I guess the pro action camera segment is a niche they may ignore, and is a niche that may be further eroded by other technologies soon.
As with the discussion in Seb's other thread, everyone has different needs and priorities. For some it's the sensor, others it will be the sensor balanced with AF, UI, build quality, feel. Having shot a couple hundred thousand images with the 1DX, there's only a handful of times I've pushed the shadows that hard. So for me at least, I can't say I've felt cheated by Canon. But then, the kind of work I'm doing with it doesn't often call for that capability.
In the normalized images (1DX at 12MP), I can't say I see much of a difference other than the Sony having perhaps a bit wider dynamic range. A point Seb made and I think is important is if the light is crappy quality, even at high ISO, the light is going to look crappy. There's no substitute for good, broad spectrum light.
p.2 #14 · Sony a7s Topic [DR + HIGH ISO + RAW + compare 1dx] - REVIEW
rscheffler wrote:
In the normalized images (1DX at 12MP), I can't say I see much of a difference other than the Sony having perhaps a bit wider dynamic range. A point Seb made and I think is important is if the light is crappy quality, even at high ISO, the light is going to look crappy. There's no substitute for good, broad spectrum light.
This comports with the DxoMark data - nearly identical midtone performance but the A7s having at least 1EV better High ISO DR. All current FF cameras have about the same midtone performance - it's sorta a cap in the efficiency possible for this size sensor. The High ISO DR makes a big difference though since it translates to noise/detail in the shadows, esp at the ultra High ISOs starting at 25.6k.
p.2 #16 · Sony a7s Topic [DR + HIGH ISO + RAW + compare 1dx] - REVIEW
Jim Kasson has what looks to be the start of a series of posts for his a7S evaluation: http://blog.kasson.com/?p=6127
Somewhat technical, but nevertheless, informative.
Observations so far:
Camera is 13 bit set to single frame advance; 12 bit set to continuous.
ISO 1600 is somewhat noisier than ISO 3200.
Snapsy - thanks for the info. I'd be curious to try the a7S at a night football game at around ISO 12800... With the 1DX I generally try to stay under 6400, where I feel dynamic range takes a bit of a hit, but have used it at 10000 for some events and it was OK... Just checking the 1DX on DxO and ISO 6400 and higher seems to be where the dynamic range drop off is around a stop for each stop of increase... so in other words, probably not worth shooting higher than 3200-6400 and just pushing in post. Maybe that's what's happening at that point inside the 1DX - digital gain. In the above link Jim indicates it looks like the a7S uses analog gain through the entire ISO range and the DxO numbers show the crossover with the 1DX is after ISO 6400 in favor of the Sony.
p.2 #18 · Sony a7s Topic [DR + HIGH ISO + RAW + compare 1dx] - REVIEW
snapsy wrote: Here's a test with a 28mm Summircon.
The 28 Cron is a tricky lens for this kind of test because it has gentle field curvature and needs stopping down to f/4 at least, to get good across frame performance. That said, my impression of his test image is that there is edge smearing because the focus falloff seems to sudden and nervous. Kind of wish he posted something with the 21/3.4 as well because it's a great performer across the frame already from wide open and any smearing would certainly be caused by the camera.
Thanks for making the raws available too.