Working through a 4,000-seat show court in Austin with 4000K LEDs and PV shade canopies, I’m comparing AGi32 versus Radiance/Ladybug to predict ball visibility and spectator comfort using GR and veiling luminance, but the sims aren’t matching our on‑site mockups at 2000 lux. Anyone have a reliable workflow for outdoor tennis that ties glare to acrylic color/gloss and seat finishes so we hit sustainable materials without washing out contrast?
Anchor with a luminance‑calibrated HDR pano from center court: run evalglare to grab veiling luminance and cross‑check glare metrics, then tweak the Ward specular/roughness for the acrylic and seat finishes until the sim matches the mockup and carry those values across (evalglare: https://www.radiance-online.org/learning/documentation/manual-pages/pdfs/evalglare.pdf). Small caveat: AGi32’s material model is simpler, so pull the coating’s spectral data (or a quick 60° gloss read) to set RGB and a modest specular lobe — “paint chips lie” — and your ball/background contrast will fall in line.
Quick win from our Austin court: in Radiance, feed the real 4000K LED SPD and make the ball fluorescent (or at least about 0.6–0.7 Y) while giving the PV canopy and seat acrylic a rough BSDF — after that, evalglare’s veiling luminance landed within about 10% of our 2000 lux mockups; AGi32 stayed rosy unless we toned down seat specular. +1 to @mira_lopez1973’s HDR anchor — try a short “pan with a tossed ball” clip and run evalglare per frame to catch the spikes; want the material settings?
Had the same mismatch on a 5k-seat bowl at “2000 lux” — — what finally closed the gap was measuring the actual seat acrylic gloss at 60° on-site and swapping the generic plastic for a tuned Ward BRDF in both tools. With your 4000K LEDs and PV canopies, a tiny drop in gloss (or a matte clearcoat) plus a slightly warmer canopy underside finish cut the blue spec hits and brought the sim within about 5% of mockups. If you can’t grab gloss data, fudge it by ~−25% and sanity-check against a quick gloss reference like https://en.wikipedia.org/wiki/Specular_gloss — does that fit your workflow?