Haarm-Pieter Duiker

[Brian Karis @BrianKaris]

Sigh. Apparently I need to weigh in on this TAA topic. MSAA is an antiquated technique designed to generate more samples on polygon edges as that was the only area that was undersampled at the time. 1/12

[Haarm-Pieter Duiker @hpduiker]

@troy_s @cloudsinmytea @biovf @s3ththompson @renderwonk I’m still hoping for side by sides with supporting equations that show a better way forward. Per channel curves have been a useful tool for as long as digital imaging has been around. Always room for improvement though. Show us the way!

[Haarm-Pieter Duiker @hpduiker]

@troy_s @biovf @s3ththompson Yep. Just getting people to work with HDR values was the challenge at that point. Pretty sure Richard Kirk and the FilmLight team were already thinking at the level of spectral responses though.

[Haarm-Pieter Duiker @hpduiker]

@troy_s @biovf @s3ththompson Gamut mapping wasn’t a big part of the conversation at that point. The main film stock of a project was the target to match.

[Haarm-Pieter Duiker @hpduiker]

@troy_s @biovf @s3ththompson When I left ESC for EA, film for games, there wasn’t as much knowledge of HDR processing, so I implemented what was a pretty standard film display pipe - Lin to log + Kodak Vision-ish film curve just to get the conversation started.

[Haarm-Pieter Duiker @hpduiker]

@troy_s @biovf @s3ththompson The compositing was primarily in linear in Shake. Encoding to disk was either in float (.iff) or log (.cin). A good number of the tech leads had worked on Cineon so there was no shortage of knowledge about the mapping between linear intensity and density.

[Haarm-Pieter Duiker @hpduiker]

@troy_s @biovf @s3ththompson The log encoding I used for the filmic tonemapping at EA was directly based on the density-based Cineon log format / transfer function that was used for the compositing, HDR and IBL pipelines at ESC for the Matrix sequels. Only so many ways to skin that cat.

[Haarm-Pieter Duiker @hpduiker]

@biovf @s3ththompson @troy_s As an aside, a proper log-space S curve was a big step forward compared to earlier approaches. Worth acknowledging the work that it took to get where we are, while thinking about better / newer approaches.

[Haarm-Pieter Duiker @hpduiker]

@biovf @s3ththompson @troy_s Side by sides was my hope for with this thread. @troy_s I think I see your point about chromaticity skew with different per-channel curves, but would love to see examples of your proposed approach.

[Haarm-Pieter Duiker @hpduiker]

@troy_s Tweets don’t give you a lot of room, but can you get a little more detailed? What does max(RGB) mean in this context?

[Massimo @Rainmaker1973]

The eerie appearance of the human eye under the macro photography by Suren Manvelyan [read more: https://t.co/CcBCR3TPZV]

[Rod Bogart @RodBogart]

Spectral measurement of California sky at 11am Sep 9.

The spectroradiometer was pointing directly at the portion of the sky where the sun should be. The overall brightness was only 125cd/m², which is similar to a medium brightness computer screen.

[Haarm-Pieter Duiker @hpduiker]

@jh_strom The curve, ie specific choice of parameters, isn’t copyrighted / trademarked. The formula was patented by EA.

[Haarm-Pieter Duiker @hpduiker]

@Jiayin_Cao Ha. That wasn’t the distinction I was aiming for originally, but we’ll take it.

[Haarm-Pieter Duiker @hpduiker]

@Jiayin_Cao Glad to hear it. Happy to answer questions about that setup.

[Dan Hanzus @DanHanzus]

This was the greatest camera angle in the history of the NFL.

[Haarm-Pieter Duiker @hpduiker]

@ivanski @drwave Stop at G&B for coffee after! Wexlers is in the same spot if you’re feeling like a bagel and lox.

[Haarm-Pieter Duiker @hpduiker]

@Peter_shirley Additional background on the technique https://t.co/P61Yo4F7GB