DV Info Net

DV Info Net (https://www.dvinfo.net/forum/)
-   Sony XDCAM PMW-F3 CineAlta (https://www.dvinfo.net/forum/sony-xdcam-pmw-f3-cinealta/)
-   -   Data for S-log correction (https://www.dvinfo.net/forum/sony-xdcam-pmw-f3-cinealta/512503-data-s-log-correction.html)

Ron Aerts December 4th, 2012 01:11 AM

Data for S-log correction
 
1 Attachment(s)
I'm using a Transvideo CineMonitor Evolution SBL8 monitor, Really great tool!
The screen has a tool which provides curves for log C, log F and Panalog. I asked Transvideo to provide S-log correction for viewable (rec 709ish) in these presets. They are very helpful and are willing to implement this. I was asked if I have the 'S-log description'. Does anybody know what this is? I assume some values to work out in a gamma formula.
Can some provide the correct data or a place where I can find this? In the attachment what the manual says about curves.
Maybe Allister?

Thanks a lot

Nate Weaver December 4th, 2012 09:22 AM

Re: Data for S-log correction
 
Sony has a white paper they released a little while back:

The formula describing the slope is in there for sure.

http://pro.sony.com/bbsccms/assets/f...log_manual.pdf

While they're at it, they should do S-Log 2 as well, but I haven't seen a white paper for that yet.

Ron Aerts December 4th, 2012 11:13 PM

Re: Data for S-log correction
 
thanks Nate, I forwarded it to Transvideo. I hope it works out.

Douglas Villalba December 5th, 2012 08:22 AM

Re: Data for S-log correction
 
Quote:

Originally Posted by Nate Weaver (Post 1766299)
Sony has a white paper they released a little while back:

The formula describing the slope is in there for sure.

http://pro.sony.com/bbsccms/assets/f...log_manual.pdf

While they're at it, they should do S-Log 2 as well, but I haven't seen a white paper for that yet.

They never published one for the F3 that I know about. The one that Nate posted is for the F35.

Alister Chapman December 5th, 2012 09:14 AM

Re: Data for S-log correction
 
Doesn't matter whether it's for the F3, F35, F55 etc. It's the same curve for all cameras, so the same corrections would be applied. Arri Log C is so close to Sony S-log that for general monitoring and viewing applications Log C would work fine.

Bill Thomas December 5th, 2012 10:53 AM

Re: Data for S-log correction
 
To Allister's point. I just recently loaded some "de-logging" LUT's in to Resolve Lite that Andy Shipsides posted on the Abel blog. He had made replica 3D luts for S-Log like the ones that come in the camera (after upgrade). Most had a very subtle look/change except for the SLog to 709 LUT that instantly clipped highlights. Not Andy's fault, that LUT in camera does the same thing. For the fun of it I tried the Arri Log-C to 709 LUT and it was very close to perfect. Had to increase exposure and de-saturate a little but it gave me a great starting point. I've never done it but for rough monitoring on set I'd assume the Log-C lut with a little monitor adjustment would work ok.

Alister Chapman December 5th, 2012 01:04 PM

Re: Data for S-log correction
 
There is a big difference between a viewing LUT that converts S-log to 709 or similar to having the correct gamma curve in the monitor. A viewing LUT takes the full log range and converts it to a different range such as 709 or another gamma compatible with most conventional monitors (which if they are HD will most likely be Rec-709). As 709 can not correctly display the full tonal range of S-Log, then something has to go. Often highlights are clipped or crushed or contrast is reduced. So the LUT adjusted video will never truly represent what the camera is shooting. If you have a monitor with alternative gammas, then the monitor can skip the Rec-709 gamma curve altogether and directly convert the S-Log input to the on screen view. This is much better and more accurate than having to go in to the monitor via a restricted range gamma like 709. However for the very best results you need an OLED monitor to show the full dynamic range of the camera.

Richard Crook December 9th, 2012 07:56 AM

Re: Data for S-log correction
 
Quote:

Originally Posted by Alister Chapman (Post 1766444)
Doesn't matter whether it's for the F3, F35, F55 etc. It's the same curve for all cameras, so the same corrections would be applied. Arri Log C is so close to Sony S-log that for general monitoring and viewing applications Log C would work fine.

No its not all the same curve. The new F's have S-Log 2, which is different from S-Log. The S-Log2 Gamma Curve was designed specifically for the extended dynamic range to support full Sony F65 latitude at 1300%, and now they expanded to the F5 and F55.

Alister Chapman December 9th, 2012 10:46 AM

Re: Data for S-log correction
 
S-Log2 is a different curve, but S-Log is S-Log no matter which camera and the curve data for S-Log will be the same for all cameras and what the OP is after is the curve data for S-Log, not S-Log2. Admittedly I shouldn't have included the F55 in may camera list, although it is possible that you will have the option for both S-Log and S-log2.

Ron Aerts December 9th, 2012 12:08 PM

Re: Data for S-log correction
 
Quote:

Originally Posted by Alister Chapman (Post 1766485)
If you have a monitor with alternative gammas, then the monitor can skip the Rec-709 gamma curve altogether and directly convert the S-Log input to the on screen view. This is much better and more accurate than having to go in to the monitor via a restricted range gamma like 709. However for the very best results you need an OLED monitor to show the full dynamic range of the camera.

Am I right here?:
The Transvideo monitor has a great measuring tool for exposure. I think it is important not to use the SDI mon output as it can show the converted lut and therefore give inaccurate information. I use the SDI through on the outboard recorder connected to the A and/or B output of the F3, Then I know for sure to measure exposure on the right levels.
I use the measuring tool on the F3 a lot. Is the measuring still correct when one of the LUT's is chosen? I didn't experiment on that yet

David C. Williams December 9th, 2012 05:16 PM

Re: Data for S-log correction
 
Middle grey in Slog is 38%, Slog2 it's 32.3%. A different LUT is needed.

Alister Chapman December 10th, 2012 02:13 AM

Re: Data for S-log correction
 
The F3 centre spot meter or histogram etc is only correct when no LUT is selected. The Transvideo tools should remain accurate when you use the Log C gamma curve as the monitor is measuring the input level, provided you are measuring the cameras clean output, ie no LUT applied in camera.

Ron Aerts December 13th, 2012 10:10 PM

Re: Data for S-log correction
 
Maybe it would be better to let the spotmeter/histogram always measure the sensor value when S-log is on...

Alister Chapman December 22nd, 2012 11:50 AM

Re: Data for S-log correction
 
I agree that it would be better for the center spot meter to read sensor values as opposed to post gamma correction values. If you are using the HG LUTs then the last two digits of the LUT name tells you where mid grey should be. So if you select HG8009G40 you should have middle grey at 40 when using the centre spot meter while the LUT is enabled.

Leonard Levy December 24th, 2012 02:31 PM

Re: Data for S-log correction
 
"So if you select HG8009G40 you should have middle grey at 40 when using the centre spot meter while the LUT is enabled."

That is extremely close to 38% is it not? Pretty much as close as I ever get with that center meter anyway. Am I guessing right that other levels like flesh tone would be further off the mark.


All times are GMT -6. The time now is 06:55 PM.

DV Info Net -- Real Names, Real People, Real Info!
1998-2024 The Digital Video Information Network