View Full Version : FOX Pre-Daytona 500 Engineering Report


mikemikeb
02-15-08, 04:31 PM
Before Sunday's Daytona 500, I want to report in with some observations and thoughts.

I was watching the Daytona 500 Pole Qualifying last Saturday (OTA on an analog TV thru WTTG analog/DC), and I noticed something very interesting with the white balance. There was an interview after the pole, I believe with Michael Waltrip, though I could be wrong, but I was looking at the "FOX Sports" logo on the microphone, which was moving in and out of sunlight due to a shadow. When sunlight was lighting it, the white printing of the "FOX Sports" logo was a very nicely pure and vivid white. However, when placed into the shadow, the white logo immediately developed a noticeable blue tint. It should have turned more a dimmer more greyish white than a dimmer, bluish white.

Otherwise, white balance was very good in sunlight conditions, but whites (and, as I remember, the skin tones on those with fairer skin) had a more bluish tint than usual in lower light. I believe I'd notice the same things on a different TV, as well as through a digital signal. I believe the problem is with the cameras.

It's kind of interesting that the Super Bowl started with amazing white balance and colors at the beginning of the game, yet diminished as the sun went down. By the end of the first half, the grass, formerly a vivid green, had dulled in color, and developed a bluish-greyish-green tint. The whites of the Giants' shirts had a bluish tint that wasn't seen on photos of coverage in the morning paper. If [the reader] didn't read the Super Bowl thread, there was a play toward 6:33 left in the 4th quarter; it was a short (<15 yards) pass or run (I don't remember) to the far side of the field, and I'm pretty sure the player went out of bounds. In this play, the grass looked more blue than green. That was about the worst I saw, though there were some pretty bad white balance issues otherwise.

However, even toward the end of the game, there were some shining moments. The "super-slow-mo" replays, always had consistently good white balance, I had no problems with them. FOX Engineering should check the records to see if the final snap of the game (with Eli Manning in victory formation) was being filmed with a Sony 3500 camera instead of a 1500 camera; if so, then I think the blue white balance issue is a defect in the 1500 alone, and not the 3500, due to some difference between the cameras (either sensor, processor, or firmware, or something else, but check the three biggies first). FOX should, if they haven't done so already, call up Sony and have them troubleshoot the issue.

I suggest a manual white balance occur on said cameras after the sun goes down and most of the lights lighting the track come from the fluorescent lights above the track. Perhaps this will alleviate the blue white balance issue, at least for the rest of the match.
______________________________________

The Daytona 500 will have very fast motion, and will be subject to even more demanding encoding than the Super Bowl, which in itself had unacceptably low bitrates (people were saying the wide shots looked more 480p due to the encoders smoothing out fine detail to avoid macroblocking. So, FOX, if you're reading this, as early on Sunday as you can, please temporarily drop any and all feeds (SD and HD) that won't be used during the Daytona 500 from the 73 Mbps mux pool to ensure the HD feed(s) do get 15.5 to preferably 17 mbps video like they did last year. Thanks.

ABCTV99
02-15-08, 09:43 PM
I don't know that this kinda of color mismatch would be a defect in the 1500s so much as a noticeable difference in ambient color temperatures. However, the Super Bowl was done (basically) indoors with the roof closed under mostly artificial lighting. On top of that it was somewhat of an overcast day so the color temperature impact from the sun through the roof should've been minimal. To boot, the last few years the Super Bowl lighting scheme has been augmented with Luminys System's SoftSun lighting rigs (5400k) which in Miami added 80,000 watts of additional light on the field specifically to make the HD broadcast more vivid and even -- not sure what the specs for Phoenix were.

I'm not surprised that a white object in the shadows (if tungsten balanced)would look blue. You're talking a temperature difference of 2 - 3,000+ kelvin (which I'm sure you know). This logic however wouldn't hold up for the Super Bowl. Even if FOX balanced their cameras early in the day accounting for the sunlight coming through the roof mesh, this would've yielded a high color temperature reading probably in the 5,000 - 6,000K range. As the artificial lighting took over and the ambient color temperature came down everything should've taken on a more reddish hue not bluer I would surmise. Not only that most stadium light fixtures hover somwhere in the 4,500 - 5,500K range anyway and are rarely flourescent. (I would certainly be shocked to see flourecent lighting at Daytona).

My logic might be a little flawed I probably need to think this through a bit further, but I've seen those Sony HDC-1500s used in a ton of applications (including NBC Sunday Night Football) and haven't noticed a color defect (though i'm no V.C.) I would imagine these huge shifts in color temperature would be more the result of human influence than electronic.If a $5,000 Panasonic HVX-200's imager can handle the color spectrum fairly accurately surely a $100,000 Sony broadcast camera should be able to excel.