Jump to content

DaVinci Resolve S-Log2 LUT Blowing Out Excessivly


Recommended Posts

I have posted a few times before but in error posted to the camera forum so I'm starting as a newbie here. But not new to Sony video as i shot a lot with the EX1 a few years ago, Just moved to A7ii.

 

 Experimenting with S-Log2 as this is all I intend to shoot but I have two problems.

 

FIRST I am going on week long remote desert shoot where I will not have computor to apply a LUT so I'm apprehensive about coming back with a bunch of bad footage. Until I feel more confident in this camera what non s-Log profile is recommended?

 

SECONDLY (and more importantly to me) when I have applied the S-Log2 LUT to my footage in Resolve I get total blow-out. Like an almost white screen with the bottom at 50 IRE and the top about 6 inches above my monitor. What's happening here/ Surely this is not right as training videos of applying LUTS give a fair image off the bat.

Link to post
Share on other sites

Hi Mike.  I found the same thing when I did a couple of tests using the supplied slog 2 inversion LUT in Resolve..... so just to get the footage passed on to the person that needed it I did a custom curve that looked pretty good on a calibrated monitor, and retained all the data in the waveform so it was all useable.

I might look into it a bit more if I get time.

You can also go into the "ACES" working space... but that might be a bit much for you to delve into.  I rely on our colorists for guidance when it comes to screwing around with ACES settings in Resolve.    :blink:

Link to post
Share on other sites

Yep. Applying a standard grade of my own to each clip gives better results and is easier than using the Sony SLog2 LUT, YET I've been criticized for doing this and told few have the skill to do it and they say I can't be consistent, so something is wrong somewhere. The camera SLog2 and the Resolve anti-SLog2 can't be matched.

Link to post
Share on other sites

After some famililization with the A7 I want to update my post. I now believe from my experience that the exposing 3.3 stops over as preached by Wolfcrow and many others is wrong. I am getting best results AND the Sony S-Log2 LUT (in Resolve) is working right with best color when I expose to the 18% grey card. 

 

Please looking forward to as many comments as possible.

Link to post
Share on other sites

Ran some tests (A7R2) on colour/tonescale charts under locked off conditions and cycled through each of the PP's just to see what they were doing technically, both on-camera and in Resolve's scopes...

interesting watching the histogram on the camera as it starts compressing the information from a standard tonescale into less and less bits as you roll through the PP's.
What I'm starting to see is that for any (controlled?) lighting situation where there may not be extreme dynamic range you want to retain, it's probably better to stay well away from S-log2.

Using Resolve's Slog2>rec.709 1D lut shows massive 8-bit stepping in the histogram once it stretches out the range, and the image looks rather noisy compared to say the normal rec.709 (PP3/PP4) profiles or Cine profiles.

 

2c

 

 

edit:  I mentioned ACES earlier.... if anyone's interested there's whitepapers & forums online that explain what ACES is ... but here's a brief whitepaper that also talks about Sony's s-log (and ACES) as used in Vegas Pro (explained in Sony's own way, of course)...   
http://dspcdn.sonycreativesoftware.com/whitepapers/vegaspro12_s-log_and_aces_workflow.pdf

 

There's other whitepapers on s-log and its variations if you search for them, it's been around for quite a while in Sony's pro/cine cameras.

Link to post
Share on other sites

My personal issue is that I do not shoot in studio but no nature wildlife landscape outdoor shoots usually in very bright conditions where cloud detail is often an issue. I am going on a shoot in the Mojave desert next week and I think I will triple take in PP4 and PP7 and PP7 + 2 or 3 stops. With PP7 NOT over exposing the Sony out works better than I can do by eye, but with ETTR I find the out is useless and my eye better.

Link to post
Share on other sites

  • 2 weeks later...

FOLLOW UP: (This is the same follow up I posted in a similar thread)

 

After some shooting and grading I now

1) Don't ETTR to the max. Come back at least a stop after 100+ zebras gone, and

2) Grade using ACES as suggested by CWS. 

 

My iMac can't quite cope with the ACES, I need render to truly review result, but much quicker and easier and better more constant results than applying a LUT to each clip.

Link to post
Share on other sites

  • 2 weeks later...

ANOTHER UPDATE. Did a lot of testing today and I now understand why the Sony S-Log2 in Resolve blows out. It blows out because it expects the footage to be UNDER-EXPOSED as the A7 records when set to Programed Auto. It's not for ETTR!

 

Set your A7 to fully auto and shoot PP7 S-Log2 and you will just how wrong the cameras exposure is. the histogram on auto-exposure is like at 45%. This is just mental, but put the footage into Resolve and applying the Sony S=Log2 LUT and resultant exposure is spot on, but the benefits of shooting Log are gone. It's just like PP4.

Link to post
Share on other sites

my general thought is that since the A7's are still recording an 8-bit codec in video, slog isn't all that useful.... its just a different way of mapping the tone-scale to that of a rec.709 gamma.  Using Slog is a bit different if you are recording 12-14bit RAW and it gives you a lot of room to play with the light...  I seem to be sticking to PP3 as a preference.

Link to post
Share on other sites

CWS - why your like PP3? Philip Bloom say PP2 for low light and PP4 for bright, I read yesterday of someone who prefers the PP2 "Still" setting for video. The tests I have done have drawn me towards PP6. It would be nice IF THE SONY PEOPLE WHO DESIGNED AND CHOOSES THESE PROFILES TOLD US JUST WHAT SITUATION THEY WERE DESIGNED FOR. Just too many choices.

 

I personally DO get befit from S-Log2 because of the wide altitude over exposure allowing a choice between noise and highlight detail depending on situation and the more natural colors. The big down side is monitoring what you are getting.

Link to post
Share on other sites

Mainly because I'm using it for in-house training videos, so sticking to one of the more technically correct tone-scale rendering profiles (PP3, PP4) makes it simpler for the various people using the camera to record reasonable video without making terribly bad exposure choices.   Chose PP3 based on running exposure tests on grey-scale and Macbeth charts under basic internal lighting conditions, it seemed to reproduce the grey-scale more evenly visually than what Sony describes as the "technically correct" PP4 setting for rec.709, which is a formulaic profile.  I think they may have put a slight rolloff on top end, rather than clipping (theoretically what PP4 would do)... but not certain as I deleted the test frames.

Link to post
Share on other sites

Create an account or sign in to comment

You need to be a member in order to leave a comment

Create an account

Sign up for a new account in our community. It's easy!

Register a new account

Sign in

Already have an account? Sign in here.

Sign In Now
 Share

×
×
  • Create New...