[Chris Wright]

what are you exporting for?

[Chris Wright]

narrow down the problem to either audio settings, codec, or tv. try a normal bs1770 preset in audition first.

[Chris Wright]

is it changing every 2 seconds or every 5 minutes? something slower you could keyframe white balance and fade between the clips so its gradual or cut adjustment layer and use dissolve.

something changing every 2 seconds would require AE unless there is an auto color stabilizer for premiere.

[Chris Wright]

yes, AE has ‘motion sketch’ that will record your mouse movements in real time and turn it into keyframes

[Chris Wright]

if project is same fps as footage, it doesn’t matter, but if they are different numbers, the engine won’t see all the frames
and stabilizing will suffer.

for example:
if you slow it down, the whole frames will have duplicate information and tracking will take longer, if you speed it up, it will lose quality tracking. best to keep 1:1.

[Chris Wright]

you have too many nested comps. render them out.
do you have hold shift- prefs enable secret purge every 5 frames?
render out to tiff not png. png is not thread safe. don’t import/export h.264 directly. use tiff as intermediate first.

[Chris Wright]

this only works if your room is acoustically built for high, low freqs, reverb and echo; otherwise headphones are better.

[Chris Wright]

4k with pro results usually means a pro 4k camera. what is your budget? I recommend testing the lighting and exposure, shutter speed, white balance, audio levels all the way through production and post production to get any
issues nailed down.

[Chris Wright]

the more drivers, the better. some headphones have tweeters called a 2-way.
reason being, a single driver can’t play all freqs at the same time with zero distortion, creates a muffling power loss effect.

having a more flat response curve is ideal +-3db across the range, although you can tweak it with a 20 band to flatten out anyway.

so it all comes down really to three things, how many drivers, freq range, and how much distortion after flattening the response curve. this is sometimes listed as ohms or higher impedance for better bass driving. you usually get what you pay for with speakers/headphones.

[Chris Wright]

i think the problem is your initial assumption that raising saturation till it gets to a nice point is going to work for all items in the frame. resolve and all other grading programs use a dumb saturation model called L*A*B* which will break apart like glass in a cement truck at the first sign of trouble.

AE kinda had the right idea with vibrance which weighted the reds a little less, but the main problem still is, push all till something breaks or looks garish. Some people have tried using masks or inverted saturation track mattes, but they will never look quite right if not perfect. i.e. a lot of vimeo.

In my grading, like in photoshop for instance, i put a levels effect set to color mode then lower output whites. this compresses the chroma underneath a regular hue/saturation effect 128 is the max rgb value that the subtractive world can reflect pure colors. you can even separate the white output of each rgb channel so they respond uniquely.

with this, you don’t need silly plugins like filmconvert.