Sony A7SIII = $3500, 10-bit 4K120p internal, 16-bit 4K60p RAW out

Thanks for your feedback Abe. Great to hear about the green bias. Since both the A7sIII and the R5 are difficult to find in stock, I am waiting until the R5/R6 might come out with C-LOG3 to do a real DR comparison. And purchase decision. Sadly, the R6 just seems like it is in a gray area unless it gets C-LOG3 and some other improvements.
 
jbregar said:
I have a pre-order in on the C70 as a replacement b-cam for my C200... but everything I read about this camera has me second-guessing that a bit.

Big thing for me is good AF and 4K 120fps... and colors that don't require a lot of pushing around.
Click to expand...

Right there with you.

Plus, even with the sony audio module, it is still $1500 less than the C70.
 
mcgeedigital said:
Right there with you.

Plus, even with the sony audio module, it is still $1500 less than the C70.
Click to expand...

don't forget about lenses. if you already own canon lenses you're happy with, then the C70 can equal out in terms of price. Personally, I needed to replace my 24-70 anyway...
 
Bassman2003 said:
Thanks for your feedback Abe. Great to hear about the green bias. Since both the A7sIII and the R5 are difficult to find in stock, I am waiting until the R5/R6 might come out with C-LOG3 to do a real DR comparison. And purchase decision. Sadly, the R6 just seems like it is in a gray area unless it gets C-LOG3 and some other improvements.
Click to expand...

frame3.jpg
frame3cc.jpg

Here's just a sample of a shot where I overexposed, and you can see how much I could bring the clouds down. (Not much, but this is the Standard profile with reduced contrast, not a log profile)
 
ReelFilmGeek said:
Looking good would love to hear about your thoughts on how intense of the footages to edit and how's the battery life as those are my main two fears about the camera
Click to expand...

Battery life is great. I rolled 2 hours 50 minutes of HD 24 and 60p over an 8 hour shoot with 2-3 hours of camera downtime. I went through 2 $20 off-brand batteries (https://www.amazon.com/gp/product/B07WZSBNJD/ref=ppx_yo_dt_b_asin_title_o02_s00?ie=UTF8&psc=1) and didn't even get to the Sony battery that came with it. I swapped one out at about 20% power and the other ran til it died at the end of the night. So, I'd say that shooting HD you get 2.5 hrs or so of on-and-off shooting on these batteries.

I couldn't speak to the intensity of the editing yet. I've only shot this one job in HD 50Mbps, so it edits effortlessly. (It was a low-stakes client, perfect for a test drive.) I will know more soon on that front.
 
CineD shared some of their infamous testing about the DR of the camera today, and of course it's nowhere near 15 stops as claimed (at least from their testing), but nonetheless a solid camera.

And so DR continues to be the Achilles' heel for almost anything and everything...which is kind of nice as it provides some camera hope for the future.
 
NorBro said:
CineD shared some of their infamous testing about the DR of the camera today, and of course it's nowhere near 15 stops as claimed (at least from their testing), but nonetheless a solid camera.

And so DR continues to be the Achilles' heel for almost anything and everything...which is kind of nice as it provides some camera hope for the future.
Click to expand...

Why are they testing slog3? I thought slog2 has more DR...
 
Honestly, I don't know. I haven't followed every dynamic range test out there.

But I personally always used S-Log3 (even on the 8-bit cameras) as I felt it looked better.

(You would think each iteration is an improvement.)
 
Did you say the future, Conan? All the way to the year 2,000.

Insofar as the DR goes, it'll probably be a dual gain/dual readout rather than a new light sensing material or a combination thereof for a long time.
 
scorsesefan said:
Why are they testing slog3? I thought slog2 has more DR...
Click to expand...

Slog3 has more potential DR

"The S-log3 curve could be used in a future camera to capture up to 16 stops and in fact the new Venice camera records over 15 stops. But as all of Sony’s other cameras sensors can only see 14 stops and the S-Log3 curve is designed to go beyond 14 stops, stop No. 14 is not recorded all the way at the top of the recording range. S-Log2 is a 14 stop maximum curve, so the peak level is recorded right at the top of the recording range."

http://www.xdcam-user.com/2016/12/t...mentations both,range, there is no difference.
 
NorBro said:
CineD shared some of their infamous testing about the DR of the camera today, and of course it's nowhere near 15 stops as claimed (at least from their testing), but nonetheless a solid camera.

And so DR continues to be the Achilles' heel for almost anything and everything...which is kind of nice as it provides some camera hope for the future.
Click to expand...

It would seem that I was completely vindicated in my discussion with James about larger photosites capturing greater DR. I'll take my payment in upvotes, James.

The headline feature is really the high DR at high ISO. Although I'm not sure they tested the DR at the correct ISO because I think the production version may have the higher native ISO a notch lower than the preproduction model (based on what Bloom or someone else said, I haven't actually tested it myself yet).

I'm curious to see results of head-to-head ProRes raw from A7SIII and s1h/s5. Certainly it should increase under exposure latitude on the a7siii

But honestly who's gonna buy a different camera because of a fraction of a stop difference in DR when the greater differences are features and design?
 
I wouldn't complain about the DR of the a7S III based on the cined results: they put it ahead of the S5, the FS7 and the FX9, and almost two stops ahead of the R5.

But also looking at their own recovered underexposure shots, it seems to be on par with the R5, so, either the Sony has two stops more headroom in the highlights, or the numeric measurements just don't translate to real-world footage. Which wouldn't surprise me because not all noise is equal: what I see in the raised shadows in the a7S III at -4 looks a lot like the stuff I had on the original a7S, with big blocks that are not easy to clean with NR (because they generates big dancing patches whose nastyness can't be appreciated in a static image), whereas the noise on the R5 -4 image is strong but finer, and would clean a lot better.

I just wish they did overexposure tests, similar to their underexposure ones. Lensprotogo does those, but they take more time to run their tests.
 
Oh, I'm not complaining about the DR results. Although it's hardly an apples-to-apples comparison because you can't turn off the NR on the A7SIII, plus the R5 is limited by its log choice (which may change).

I wonder what raw will show about the underexposure... isn't it available already? couldn't they have tested that?
 
As much as the A7sIII is presented as the low light king, in most of the testing videos I have seen, the R5 is pretty close from a noise/recovery perspective. The Sony does look better on the spot when going into very high ISOs but I see this as sort of a party trick to an extent. Great if you really need it but not all that applicable for most paid work. The dual gain(ish) feature at 12800 is useful for demanding environments where you want more shutter speed like slo-mo or more DOF.

Bottom line for me is that once the Canon's get their C-LOG3, these two products will be about even in my eyes from a DR and noise perspective. With a win for Sony on ISO. I am impressed that Sony made their picture profiles 10bit. The R5/R6 with 8bit profiles kind of limits the camera imho.
 
scorsesefan said:
Why are they testing slog3? I thought slog2 has more DR...
Click to expand...
I am assuming you are saying that because you know that most Sony cameras on the market today, Slog2 may still be a little better, as in slog3, some of the "bits" get wasted. But Abe is correct in that slog3 is the wider dynamic range curve.... but most cameras don't have that much usable DR. This is especially true when shooting in 8bit, as we've all experienced.
 
You must log in or register to reply here.
Back
Top