F55: F5/F55 firmware v4.1 released. Sony says "a more natural look image"

Thanks Dennis, and thanks for the post on the 4.1 upgrade, I've been sort of waiting for the 4.1 thread to shake-out over on the Sony forum before makinging the upgrade, looks like it is safe to go ahead.
 
found out from Doug @ VortexMedia that the "more natural look" did not make any change to the overall look but rather, "The improvemens seem to have been limited to fixing some ringing issues (digital artifacts) in the highlights"

This is good to know.
 
found out from Doug @ VortexMedia that the "more natural look" did not make any change to the overall look but rather, "The improvemens seem to have been limited to fixing some ringing issues (digital artifacts) in the highlights"

This is good to know.

I had a devil of a time shooting on a football field where metal halide stadium lights were in the shot. Just couldn't handle those highlights. Hopefully it's improved.
 
Please let us know if after the upgrade you notice a better response to that type of lighting. Its good information.
 
Today I shot some interviews with very nice lighting using slog3-cine.gamut + 709A LUT baked into 4K XAVC 10-bit 422 and it really does look nice, thanks for the encouragement to try it Bill.

The executive producer said he liked the image better than Alexa, it was less muddy.
 
Today I shot some interviews with very nice lighting using slog3-cine.gamut + 709A LUT baked into 4K XAVC 10-bit 422 and it really does look nice, thanks for the encouragement to try it Bill.

The executive producer said he liked the image better than Alexa, it was less muddy.

Nice!! There ya go. Having a EP see/say that is good biz for all of us. I used "Admit One's" two-up still grab of the girl where one is Alexa and the other F5, showed it to a Producer and asked him to tell me which was which. He guessed wrong. :)
 
...I will say for some reason or maybe it's my DP6 monitor, the final image looks better then my MLUTed preview on my monitor.

FYI previewing the images today I also noticed they looked slightly better on my ikan D7W monitor live via SDI.

Only way to verify any differences between int. and ext. would be to use an external SDI recorder and compare against internal in post.

I wonder if this is some bug or something to do with XAVC codec.
 
I've obtained permission to post this frame from some work I did this week.

I baked in LC-709A to the internal recording using the "look" built into the camera. This is more or less straight out of the camera. I did have to correct the black input to pull a bit of a white wash out of the image.

Let me know what you guys think.

14405979522_69cb2ff7f7.jpg


14405979522_69cb2ff7f7_c.jpg


14405979522_fbc34d589d_o.jpg
 
Last edited:
Thanks Sam.

Seriously, you know what I have been going in post to "de-log" -- this was de-logged in real-time in-camera and I really can't complain about it. I would pull some warmth out of it, but its not bad for "in-camera".

Want to know my lighting kit/setup?
 
A reflector? Looks good. Natural. I wouldnt pull any warmth out of it though. Maybe just a touch more contrast.

ian
 
Thanks Sam.

Seriously, you know what I have been going in post to "de-log" -- this was de-logged in real-time in-camera and I really can't complain about it. I would pull some warmth out of it, but its not bad for "in-camera".

Want to know my lighting kit/setup?

The only advantage of log is you could rescue the highlights - but to me it is natural that they blow anyway.

Thats the thing about my stills workflow (and what I have been seeking in motion) is that you get a good look out of the box but can.. if you choose work on recovering those highs or shadows.

Doing work to get the image in the ball park is my (and your?) 'problem' with log - as you say it should happen 'under the hood..' (until you want to open the hood and start tweaking..)

--

Lights? Could be nothing or a reflector - or maybe something huge and costly!

S
 
OK so after thinking more about the observations made that images in monitor look slightly better than in recording I think I figured it out.


Data levels vs Video Levels.

The images recorded to the cards are likely using 0-1024 code value space same as slog would be, whereas through SDI either they conform to 64-940 code value or the monitors know how to automatically deal with the different code value representation.

The solution in post is to pull the pedestal back down to where it should be, or possibly set the clip attributes to "video levels" (64-960) but you'll need to lower your highlights.

This likely explains it - thoughts?
 
Back
Top