HDR workflow

[Paul Sage]

If your intention with HDR is to just make everything brighter, then sure, simply capture a non clipped image.
But there are more dynamic ways to expose and light for HDR.
I am sure we can agree that adjusting lighting on set will always look better than trying to re-light in post with power windows and qualifiers.

 

[Marc Wielage]

If your intention with HDR is to just make everything brighter, then sure, simply capture a non clipped image. But there are more dynamic ways to expose and light for HDR. I am sure we can agree that adjusting lighting on set will always look better than trying to re-light in post with power windows and qualifiers.

Nowadays, it's often a combination of the two: the cinematographer gets it about 75-80% of the way there, and then the colorist (assumedly under the DP's supervision) takes it the rest of the way. A lot of my work nowadays is adding a little fill to faces and also flagging down backgrounds, particularly in cases where this would've been impossible on-set. It's not so much re-lighting as it is touching up, which is not quite the same thing.

There is always the danger of modern DPs expecting a "Revanant-like level" of lighting change in post, and that can't be done in a short amount of time and at a standard indie budget level. It can be done if you have four or five people doing the work and 12-16 weeks on the session.

HDR requires a lot more thought, and I think the o.p. needs to ask the questions, "what's our available budget? How much time do we have? Who is the intended market for this show? What specific HDR standard should we use? What does our distributor need?" These are important questions that shape any post project.

 

[Stacey Spears]

Here is a series of articles I consulted on. Probably the best article(s) on HDR yet.

 

[Mike Finegan]

Here is a series of articles I consulted on. Probably the best article(s) on HDR yet.

Great but author seems to be unfamiliar with LUT usage in post to remap color space A to color space B.
The claim that studios are delivering P3 color graded content dumped unchanged into BT2020 seems inaccurate.
(BBC shows ffmpeg excample here: http://www.bbc.co.uk/rd/sites/50335...3e0efb118297/BBC_HDRTV_Use_of_LUTs_FFmpeg.pdf)

 

[Stacey Spears]

Great but author seems to be unfamiliar with LUT usage in post to remap color space A to color space B.
The claim that studios are delivering P3 color graded content dumped unchanged into BT2020 seems inaccurate.
(BBC shows ffmpeg excample here: http://www.bbc.co.uk/rd/sites/50335...3e0efb118297/BBC_HDRTV_Use_of_LUTs_FFmpeg.pdf)

What he is saying is that P3 content is converted to 2020 via a 3x3 matrix. P3D65 -> XYZ -> 2020 is what is happening. The color volume is not changing, it still only as big as it was. Using a 3D LUT for this conversion would be the wrong tool. ColorFront Trankoder does a real transform and is one of the primary tools used for this. It also does tone and gamut mapping if they want to target something smaller than P3 since consumer displays tend to clip the gamut, which results in a hue shift and they have poor tone mapping. (outside of Dolby and Sony).