Better LUT for good skin tones - for Odyssey 7Q+ & FS700 ?

Very good points Nick, thanks. I will definitely pull out the less contrasty LUTS in HDR situations, but I've just been wanting something that looks like what the camera is pointing at.
I always expose to the waveform with the LUT off on the Odyssey 7Q, so I know I'm never going too far in slog2 ;)

I wish my eyes could take a photo, because that a7rii photo is almost exactly how it looks in real life!
To be clear, I didn't match the fs700 image to the a7rii. I took the a7rii photo AFTER I had finalised the LUT (out of pure interest,) and was delighted to see they match very closely :)
 
In the course of my research for Convergent Design, I have created some LUTs which may be suitable for use with FS700 S-Log2 material. They can be downloaded here.

This is a set of six LUTs for use with daylight and tungsten white balanced material, with versions of each at normal exposure and at minus one and minus two stops. The names should be self-explanatory.

Please note, these LUTs are a work in progress, and are not officially endorsed by either Convergent Design or Sony. I would like feedback on how people find the results of using them on a variety of material.
 
One thing to note about the LUTs. They follow the format of the preset LUTs from the Odyssey, that is to say they are "EE" LUTs in the terminology of the ARRI LUT Generator – they are intended for S-Log2 material with the range in Clip Attributes of Resolve set to Video. This is the format to which I converted all the LUTs which ship with the Odyssey. This is not the same as the Sony S-Log LUTs which are "LE" – they expect the Clip Attributes to be set to Data.
 
I've spent the better part of the last day and a half trying to make my own colour accurate LUT.
My only goal was to try and match the 7Q screen to what I saw with my own eyes in real life.

Using LUTcalc and the xrite colorchecker passport, I got pretty darn close I think.
Cyan was the trickiest, but red's look like reds and blue's don't look like turquoise anymore which was my main goal.

Here's the untouched Prores Slog2 for comparison
(overexposed by 2 stops approximately)
View attachment 115603


Here's the Convergent Design LUT:
SONY_EE_SL2_L709A-2
View attachment 115604


And below is the LUT I've developed:
Ky_Fs700_SL2_Rec709_v1.57_-2
View attachment 115605

And to put it into perspective here's a photo taken with a Sony A7rii.
This is a useful comparison as picture profiles were disabled (Just standard JPG)
It represents what I saw with my eyes fairly closely.
View attachment 115606


Let me know if you find it useful or not or if you have any suggestions.
I'm just a dabbler with LUTS, not an expert!

The zip contains a regular exposure, a -1 and a -2 for those of you like me who like to expose to the right.
I've also converted them for use on the 7Q via the CD LUT converter utility as well.

Download here
This is really awesome! They look very accurate. I just applied them to some outside shots with lots of snow. The snow looks white now, not tourquise like the Odyssey standard luts.
I'm using the FS5 with raw output.
 
Thank you for the great LUTs, they really made me appreciate the FS700/Odyssey7q+ combination.
I was wondering if it is somehow possible to make colors match as much as possible to the LUTs presented in this thread when recording in ITU 709 (800%) in low light situations.
Could my thinking be correct to record a desaturated image using the picture profile settings and then to apply a LUT that approaches the colors of the REC 709 image to the Slog2/LUTed image?
Thank you for any thoughts on this!
 
I've spent the better part of the last day and a half trying to make my own colour accurate LUT.
My only goal was to try and match the 7Q screen to what I saw with my own eyes in real life.

Using LUTcalc and the xrite colorchecker passport, I got pretty darn close I think.
Cyan was the trickiest, but red's look like reds and blue's don't look like turquoise anymore which was my main goal.

Here's the untouched Prores Slog2 for comparison
(overexposed by 2 stops approximately)
View attachment 115603


Here's the Convergent Design LUT:
SONY_EE_SL2_L709A-2
View attachment 115604


And below is the LUT I've developed:
Ky_Fs700_SL2_Rec709_v1.57_-2
View attachment 115605

And to put it into perspective here's a photo taken with a Sony A7rii.
This is a useful comparison as picture profiles were disabled (Just standard JPG)
It represents what I saw with my eyes fairly closely.
View attachment 115606


Let me know if you find it useful or not or if you have any suggestions.
I'm just a dabbler with LUTS, not an expert!

The zip contains a regular exposure, a -1 and a -2 for those of you like me who like to expose to the right.
I've also converted them for use on the 7Q via the CD LUT converter utility as well.

Download here

Great! This made my day! Thanks!
 
I may have to try these!

recently it's not skin tones or any colour but GREEN that's giving me fits. I've been shooting cannabis growers and have a crazy setup of a gh4, 5diii and my fs700/odyssey. Not the most ideal combination, but I have found that the green is really muted in comparison to other colours! The canon's greens are GREEN!!!!!
I'm shooting slog2, but beyond doing a colour correction, I'm also doing a secondary one just to get the greens more lush - not fun!
 
I was wondering if it is somehow possible to make colors match as much as possible to the LUTs presented in this thread when recording in ITU 709 (800%) in low light situations.
Could my thinking be correct to record a desaturated image using the picture profile settings and then to apply a LUT that approaches the colors of the REC 709 image to the Slog2/LUTed image

could anyone please tell me if this is complete nonsense or if there is way to match the rec709 image with the natural looking colors you made available when recording in slog2?
 
I'm sure it's been said before in this thread, but it bears repeating. It is not "7Q S-Log2". It is FS700 S-Log2. Internal S-Log2 recordings and 7Q recordings are the same, bar any noise-reduction type processing which the camera does and the 7Q does not.
On what do you base this? When you send a raw signal from the FS700R to the Odyssey 7Q, it's the 7Q that's doing the processing from raw to S-Log2/S-Gamut. That processing needn't necessarily match what the FS700R does internally. Was Convergent Design provided by Sony with the transformations used to go from raw to S-Log2/S-Gamut? Or did Convergent Design try to mimic what the camera does internally? Any idea if Atomos used the same method? I'm thinking about ways to compensate for the non-standard gamut.
 
Last edited:
On what do you base this? When you send a raw signal from the FS700R to the Odyssey 7Q, it's the 7Q that's doing the processing from raw to S-Log2/S-Gamut. That processing needn't necessarily match what the FS700R does internally. Was Convergent Design provided by Sony with the transformations used to go from raw to S-Log2/S-Gamut? Or did Convergent Design try to mimic what the camera does internally? Any idea if Atomos used the same method? I'm thinking about ways to compensate for the non-standard gamut.

Not all cameras output the same "RAW". Sony's raw output already has a lot of processing done in camera. (Like debayering) So I don't think the Odyssey is doing as much image processing as you might think.
 
Just out of curiosity....have you tried NOT using SLog2? You can record in Rec709, RAW to ProResHQ on your Odyssey 7Q.

If you're not in a HDR scene, I wouldn't bother with Slog2.

Even though this is a super old post - I wanted to chime in and say that I personally don't shoot in Rec709 on the fs700/7q because the colours aren't right. I'd rather go slog2, then do a quick LUT correction and get better results than shooting in Rec709 where I'd need a big correction anyway. The red's aren't red and the whole thing looks off to me. So to me if you need to do CC anyway w/ some tweaking, better to have more control w/ slog2
 
Looking for a consensus among FS700 users: do any of the built-in color modes render colors correctly? I mean Standard, Still, Cinema, Pro, and ITU709 Matrix, or just turning the picture profile off. A couple of people mentioned that colors were good with the picture profile turned off.
 
Even though this is a super old post - I wanted to chime in and say that I personally don't shoot in Rec709 on the fs700/7q because the colours aren't right. I'd rather go slog2, then do a quick LUT correction and get better results than shooting in Rec709 where I'd need a big correction anyway. The red's aren't red and the whole thing looks off to me. So to me if you need to do CC anyway w/ some tweaking, better to have more control w/ slog2

There are a ton of REC709 LUTs available as well, better yet, make your own in 5 minutes using a color chart.

Point is, using SLOG2 in dark scenes is going to give you horrible noise. It gives you highlight latitude, at the expense of shadow detail.
 
Dear Friends,

I would like to add a little to this conversation.

The Odyssey7Q, Odyssey7Q+ can accept 2K and 4K Raw from the FS700 (all models that can output Raw).

It is important to set the Sony FS700 to use Picture Profile 7, and this Picture Profile must be set to Factory Defaults.
If you change Picture Profile 7, then the colors will be off.

I hope this helps.


Respectfully,


Dan Keaton
Convergent Design
 
could someone please help me understand what the best practice would be when i shoot in 4k raw to prores hd in a low light situation (for example picture profile settings) to get correct colors and best noise behavior?
 
could someone please help me understand what the best practice would be when i shoot in 4k raw to prores hd in a low light situation (for example picture profile settings) to get correct colors and best noise behavior?

Hi, marczzzn, you should probably start a new thread to make your question visible to more people.
That being said: Mitch Gross' advice – rating the FS700 at ISO 800 – has worked very well for me.
 
I made an experimental FS700 S-Log2 LUT to try to fix its color problems. PM me if you'd like to test it. It's made on the assumption that the problem with colors in FS700 S-Log2 is just a bad or unknown gamut. I calculated a matrix for gamut remapping using measurements of FS700 ITU709 Matrix as the xy colorimetry basis. I don't have enough test footage to test this assumption, so it might be wrong.
 
Back
Top