I'm trying to optimize workflow for deflicker refinement with 10,000+ images. Sometimes when applying deflicker, after Visual Preview has made a few frames I can see immediately that the whole sequence will need to be refined. However it seems LRT requires finishing all previews before that option is available again. Is there a way around this, or am I missing something?

Also, what exactly does "refine" do? Does it just add a multiplier to the deflicker factor? 

And the refinement applies only to selected frames, correct? Testing and previewing is obviously faster with just a subset of frames. So then if you find a level of refinement you like that you want to apply to the whole sequence, you need to reset all deflicker values and then be sure to de-select, or select all frames, and re-apply deflicker and refinement?

Refine will need the developed deflickered images to be able to calculate the next flicker step. So yes, you have to wait until the previews are generated. But refine will only change images, where flicker has to be redone, such that the amount of images will get less.

Sent mobile with Tapatalk.

It might be cool in the future to have a deflicker "strength" control, that you could preview with selected images only...

There was one in the past but it caused more confusion then helped, that's why I removed it. The way it's implemented now sets the strength automatically. Normally one or two passes are enough, if you consider the tips for editing and use a reasonable reference area.

Sent mobile with Tapatalk.