Is my exposure off? Noisy FX9 in SLOG3

[Doug Jensen]

To each its own I suppose... The S-LOG2 and S-LOG3 LUTs that come with Resolve are simple 1D LUTs so of course they can't look great.

Serriously, you're talking about 1D LUTs? Why would anyone care about those. Thanks for the confusing information.

BTW, last time I did a clean install on a client's Mac, Resolve did come with some S-LOG3 3D LUTs, so that information of yours might be wrong as well.

 

[Doug Jensen]

Sorry, I stand corrected: there is an slog2 LUT. For some reason it's not displayed in my LUT sidebar, though—it only appears if I right click on a node and select the "LUT" submenu. I haven't gone that route to apply a LUT in ages, hence my overlooking it.

Is that a 1D or 3D LUT?
BTW, right-clicking on node #2 and applying a chosen LUT is how I do it because the LUT needs to come at the right stage of the flow chart.

 

[bobtronic]

Serriously, you're talking about 1D LUTs? Why would anyone care about those. Thanks for the confusing information.

Thanks for being vitriol. I never use these LUTs because they looked bad and only found out they are 1D because I double checked and opened them in a text editor.

 

[Doug Jensen]

Thanks for being vitriol. I never use these LUTs because they looked bad and only found out they are 1D because I double checked and opened them in a text editor.

Have you tried any of the 3D LUTs that are specifically made for S-LOG3 -- as a starting point for a grade? If so, what don't you like compared to the other methods you mentioned?

 

[bobtronic]

Have you tried any of the 3D LUTs that are specifically made for S-LOG3 -- as a starting point for a grade? If so, what don't you like compared to the other methods you mentioned?

The S-LOG3 LUTs seem to be fine if you only deal with S-LOG3/Cine. There seems to be no official S-Gamut3/S-LOG3 LUT. I actually only chimed in because we went on a tangent of S-LOG2 and Resolve. Anyways, I will not add more to the confusion

 

[drboffa]

Is that a 1D or 3D LUT?
BTW, right-clicking on node #2 and applying a chosen LUT is how I do it because the LUT needs to come at the right stage of the flow chart.

Is this still a factor if the LUT is the only correction applied? I generally only do one operation per node.

 

[Doug Jensen]

Is this still a factor if the LUT is the only correction applied? I generally only do one operation per node.

Yes. Exposure is fine-tuned on Node 1 and the chosen LUT is applied on Node 2 with no other adjustments on that Node . Everything else you want to do to the clip comes after Node 2. The flow of the nodes matters and is a powerful feature of Resolve you just don't get in an NLE.

 

[Doug Jensen]

The S-LOG3 LUTs seem to be fine if you only deal with S-LOG3/Cine. There seems to be no official S-Gamut3/S-LOG3 LUT. I

I know there is an official Sony S-LOG3 to Rec709 LUT because I've had it on my computers for years. I think it came as part of a normal install of Resolve, but, just like with the S-LOG2 LUTs, I can't say for sure. Doesn't really matter anyway because I can't see any argument for not shooting with the .cine version of S-LOG3 that is more compatible.

 

[drboffa]

Yes. Exposure is fine-tuned on Node 1 and the chosen LUT is applied on Node 2 with no other adjustments on that Node . Everything else you want to do to the clip comes after Node 2. The flow of the nodes matters and is a powerful feature of Resolve you just don't get in an NLE.

Okay, yes, this is how I do it as well (technically LUT on node 3 after WB). I was wondering if there was a difference between applying a LUT via right-click vs applying a LUT via double-clicking (or dragging) from the LUT library in the left panel. I know that nodes themselves have an internal flow of operations so I wasn't sure if the two methods meant different *internal* orders of the image pipeline within a single node.

 

[ahalpert]

Very nice videos. I love the overall look and the editing. It is the kind of video I'd be no good at producing, so I admire those who can.

Thanks, Doug. Incidentally, the 2nd video (https://f.io/Kljid9v1) features a lot of full-frame 12mm shots. I remember we had a discussion about why most zooms don't go that wide - and while I admit that I use that lens infrequently and mostly in order to exaggerate perspective and scale, there are a couple of shots in there that more or less look like normal wides because there's nothing in the foreground. Which is all just to say that while it might not be a necessary focal length for everyone, it's also not a totally crazy look

 

[ahalpert]

It's my understanding that if you set your camera for S-LOG3.cine, you are essentially recording footage that has very similar gamma and color gamut as Arri LogC. In post, you can treat S-LOG3.cine the same as LogC.

Here's my understanding of the evolution of S-LOG. This isn't written down anywhere in any official document, but it is the conclusion I came to after a lot of face-to-face conversations with the right people and reading between the lines of Sony's documents. I have mentioned this theory to the powers that be several times in casual conversations, and nobody has told me I'm wrong:

S-LOG2 is the superior version of S-LOG. I don't think anyone who has serious experience working with S-LOG2 and S-LOG3 would dispute that. One of the big differences is that it handles highlights better and won't clip whites at 94 IRE like S-LOG3. However, S-LOG2 was not well-received by high-end experienced colorists in Hollywood and elsewhere. They didn't care that it was technically better. What they cared about is that Sony footage shot with S-LOG2 had to be treated differently in post than the Arri footage that dominated (and continues to dominate) the industry. Some people are very resistant to change and doing things differenty, so there was a lot of push back at high levels. Even for lower-end producers (like us) S-LOG2 was a little tricky because Resolve didn't include any S-LOG2 compatible LUTs, which meant people had to track them down and install them on their own. Which, apparently, is a big hassle for some people. I haven't done a fresh install of Resolve for many years, but as far as I know, there still aren't any S-LOG2 LUTs included with Resolve. But I might be wrong on that point. Perhaps someone who has recently installed Resolve for the first time could check to see if there are any S-LOG2 LUTs.

Sony's solution to the slow accpetance of S-LOG2 (and therefore slow acceptance of their cinema cameras) by mainstream Hollywood was to create the shooting mode called S-LOG3/S-Gamut3.cine. The whole point of shooting with S-LOG3.cine is the the file is very similar to Arri LogC. Whatever LUTs or other grading you apply to an Arri clip can be directly applied to a Sony clip and you'll see similar results.

I still prefer to use S-LOG2 on my cameras that offer it. But going forward, we'll be limited more an more to having only an S-LOG3 option.

BTW, one advantage of shooting RAW with my F55 is that I can bring that footage into post and decode it as S-LOG2 or S-LOG3, just as if it had been shot that way in the first palce. In every side-by-side test I've done, I always preferred working with the S-LOG2 version.

Now, this is interesting. It would appear that I've been laboring under a misconception. I had thought that slog3 could provide greater dynamic range than slog2. But what you're saying totally makes sense with my experience of the two and why I have a harder time grading the slog3 shoulder

You've led me to revisit this article by Alistair Chapman comparing slog2 and slog3: https://www.xdcam-user.com/2016/12/t...-debate/?amp=1

Chapman says here

the captured dynamic range, brightness range and shadow range does not change between S-Log2 and S-log3, both of which take the entire sensor range (they just store that same range using slightly different levels or code values). After applying a LUT or other conversion to your normal viewing gamma both S-Log2 and S-log3 will have the same brightness, same highlight and same shadow range...

S-Log3 is based on the industry standard Cineon log gamma curve. This curve was developed in the 1980’s to allow the digitising of film using 10 bit data... Arri’s Log-C is also based on Cineon as are many other log curves...

S-Log2 was specifically designed by Sony to work with an electronic sensor with 14 stops of dynamic range and is optimised to match the performance characteristics of video sensors. By using a 14 stop curve with a 14 stop camera almost every bit of available data is utilised, there is no wastage. So S-Log2 makes better use of the data you have available to you...

However, if you are recording 10 bit or greater you have a lot of data whichever curve you use, so in practice it will be rare to see any difference in the final result.

I'm a little embarrassed that I had previously read this article and yet developed the idea that slog3 could give me more dynamic range (which i think is only true if the sensor has more than 14 stops of DR). I must have gotten that idea from unreliable sources online...

BTW, I think that fcp has the slog2 lut built-in now

 

[ahalpert]

Yes. Exposure is fine-tuned on Node 1 and the chosen LUT is applied on Node 2 with no other adjustments on that Node . Everything else you want to do to the clip comes after Node 2. The flow of the nodes matters and is a powerful feature of Resolve you just don't get in an NLE.

One thing I've wondered - although I indeed use this workflow of exposure adjustments before LUTting - is if a pre-LUT exposure adjustment in post is equivalent to an in-camera exposure adjustment since the latter will be mapped onto a specific gamma curve whereas the former is...I dunno, more linear? Or does it not matter? It has worked out for me, but I do wonder.

Also, if I can apply effects in a sequential order in fcp, and thus control my order of operations of color corrections, is it different from the node ordering you describe?

 

[cyvideo]

Now, this is interesting. It would appear that I've been laboring under a misconception. I had thought that slog3 could provide greater dynamic range than slog2. But what you're saying totally makes sense with my experience of the two and why I have a harder time grading the slog3 shoulder

I'm a little embarrassed that I had previously read this article and yet developed the idea that slog3 could give me more dynamic range (which i think is only true if the sensor has more than 14 stops of DR). I must have gotten that idea from unreliable sources online...

You are not amiss, Abe. Not very often that one can say that Alister is incorrect, but if he did in fact say "the captured dynamic range, brightness range and shadow range does not change between S-Log2 and S-log3." then his statement is in direct conflict with Sony's own published data regarding S-LOG2 and S-LOG3 with regard to dynamic range. There is a 1.5 stop greater dynamic range capable with S-LOG3. The info can be found in Sony's white paper titled "Technical Summary for S-Gamut3.Cine/S-Log3 and S-Gamut3/S-Log3" See pix below: Where Sony states:

"S-Log3 is based on Cineon Digital Negative revised in 2007. S-Log3 tone curve has no shoulder and is reduced the toe (non-linear curve in shadow area). It is more like a pure log encoding than S-Log2 to provide better log based grading (See chart below). Mid-tone value; 18% grey is brighter than S-Log2. S-Log3 has 1.5 stops wider dynamic rage than S-Log2, covering the full EI range. It is not necessary to change the conversion formula for each EI setting."

Chris Young

 

[morgan_moore]

Making adjustments before lut aplicatio - you can use try the log wheels which may get arount the log/lin adjustment situation

-/-

persoanlly I apply a lut to node 2 and play woth 1 and 3 to make a pretty pic and 4 to add grain

just make pretty pictures - its simple time and sensitivity - mot really theory?

 

[Doug Jensen]

You are not amiss, Abe. Not very often that one can say that Alister is incorrect, but if he did in fact say "the captured dynamic range, brightness range and shadow range does not change between S-Log2 and S-log3." then his statement is in direct conflict with Sony's own published data regarding S-LOG2 and S-LOG3 with regard to dynamic range.

Yes, exactly why I never mentioned anything about dynamic range differences. Sony is ocrrect, but I think Alister is kind of correct too because it is my understanding that none of the the cameras today are able to use the full capability that S-LOG3 offers when the cameras themselve have better DR. So technically there might a different in the future, but not yet. What matters to me in the mean time is how S-LOG2 handles highlghts better right now (exactly the best reason to be using LOG in the first place) and it won't clip at 94 IRE.

 

[Doug Jensen]

One thing I've wondered - although I indeed use this workflow of exposure adjustments before LUTting - is if a pre-LUT exposure adjustment in post is equivalent to an in-camera exposure adjustment since the latter will be mapped onto a specific gamma curve whereas the former is...I dunno, more linear? Or does it not matter? It has worked out for me, but I do wonder.

Also, if I can apply effects in a sequential order in fcp, and thus control my order of operations of color corrections, is it different from the node ordering you describe?

Sorry, I do not know the answer to those questions so I will take the advice I have given others, and not say anything that would just be a guess.

 

[Doug Jensen]

BTW, I think that fcp has the slog2 lut built-in now

Maybe, but grading is soooooo much more than just dumping on a LUT and walking away. FCP falls way short of having the tools to do it right and efficiently. If you learn to grade in Resolve you would never go back to anything else.

 

[ahalpert]

Maybe, but grading is soooooo much more than just dumping on a LUT and walking away. FCP falls way short of having the tools to do it right and efficiently. If you learn to grade in Resolve you would never go back to anything else.

oh, I'm not trying to convert you to fcp or claim that it's better for grading or something. I'm just registering a data point about the state of slog2 acceptance

 

[ahalpert]

Making adjustments before lut aplicatio - you can use try the log wheels which may get arount the log/lin adjustment situation

-/-

persoanlly I apply a lut to node 2 and play woth 1 and 3 to make a pretty pic and 4 to add grain

just make pretty pictures - its simple time and sensitivity - mot really theory?

what are the log wheels? is this a resolve thing?

I agree with you about good taste being the only truly necessary ingredient for grading. but understanding the underlying theory sure speeds it along!

 

[morgan_moore]

Yep in resolve you have liear and log wheels which have different response

persoanally i use bars and curves not wheels