no_connection

I involved a few more nodes than that. Some corrections done in linear space, some in log, some by lut to add some film simulation to give colors a push.. Same grading for all shots above. I'm just a hobbyist with an interest in a lot of things, a pro would probably put hand to forehead. *edit* I'm not saying simple is wrong ether, if all it takes is a slight adjustment, then that is the right one.

I don't really know what the video is about or what you want to say with the picture mood. Make the exposure work for the mood and add character.

I wish they would stop using video levels. That would end confusion like this. I for one like that they use the whole range instead of throwing away more starved bits. And if ppl blindly clip the footage to video levels, that is user error. The real question now is how much can you lift blacks in profile to get a tiny bit more leverage.

It seems that they are making an effort to keep artifacts and banding in check. Here is a saturation map of XT2 compared to a clip buy fuzzynormal from the original files thread of the GX85. I could not fine the same amount of contrast/punch tho, so comparison is a little stretched. To make a saturation map you shift hue by 180*(0.5 in fusion) with the color corrector node, then combine with difference node to the original. You then desaturate the result to grayscale. I gained it by 2 here to make it more visible. The XT2 clip appears to be dithered and a bit smoother compared to the GX85 which is showing a bit more blocks. It is not without blocks as something have to go when compressing. But how these artifacts behave and how much effort goes in to handle them determines how much the file can be pushed around before it breaks up. This is also why flat profiles are not always a good thing. btw saturation map works great as a mask for things.

I did check both the video samples for download above with ffdshow internal reporting to .csv file. And it did report an I frame every 15 frames. 448 P frames and 32 I frames. For the clip with water. The magic happens when I open the clip in fusion. I expected to see mush with huge blocks at best. But this? Not sure if fusion is responsible for grain or not but I can't get anywhere near as good of a result by converting it to another format.