39124 Camp Drive
Prairieville, LA 70769-4522
USA
voice/fax: 504/673-3627
To: ITSMail <ITSMail@itsnet.org>, OpenDTV Mail <OpenDTV@pcube.com>
From: "Charles R. Caillouet, Jr." <cailloue@advtel.net>
Subject: [STS-95] launch notes
Cc: "Turner/John Turner" <turnereng@attmail.com>,
"NHK/Philip Hack" <philiphack@aol.com>, "NHK/Elke
Titus" <ElkeMT@aol.com>
Thanks to all on the net who commented (good and bad) on the launch coverage from Kennedy Space Center (KSC). Feedback in the process is always useful. I would like to offer a few observations from the field. Sorry this has taken so long but we have been pretty busy digging out from under the project and from covering the landing as well.
I was the engineer in charge of getting the production system in place for NHK. NHK provided the high definition mobile unit and the support vehicle for the event. Harris provided the production organization for the live broadcast and through Turner Engineering, the encoding and transmission of the event. WRAL provided an edit suite and additional HDCam camcorders and decks.
Two separate signals were encoded and uplinked, one for commercial stations and one for non-commercial. NHK also delivered another signal (using a 45 Mb KDD service) to Japan for live release. The complication of building three slightly different programs for simultaneous release taxed the capabilities of all of us and some of your comments indicate that we were not entirely sucessful in making the process invisible. I think that we met our basic objectives although we always know that we could have done some things better.
The main body of the program was essentially in video/audio sync as were the commercials. At least, it looked that way to all of us at the site. Our audio guys took pains to even deal with the DVE delays in the production switcher (Yes we had two channels of DVE and a digital switcher.) Most of the problems that you reported with sync seem to be the result of some decoders slipping out of sync for periods of time and some even needing to be rebooted to fix the problem, however, there may have been some problems imbedded in some of the pre-recorded packages. The editing system brought in for the event was brand new and some problems surfaced in system configuration, comm protocol and machine control interfaces. These obstacles, combined with a tight editing schedule resulted in some compromises in the products. Given more time, all of those could have been cleaned up. Such are the trials and tribulations of live television. The audio was ProLogic encoded so that we could easily record it, switch it and monitor it and so that it would transparently pass as simple stereo. We decided that trying to do 5.1 for three different feeds would be more than we could handle. As it turned out, just dealing with the audio that we had was hard enough.
The comments concerning the stage cameras are well taken. I have looked at the tapes and the camera images are not as bad as some of you reported but they do contain high average picture levels, because of the choice of wardrobe and because of the lighting decisions. Some projectors, probably being pushed for max brightness, may have reproduced those images as washed out. Again, given more time to work on the show look, those could have been improved.
One important thing to understand was that this operation was not part of an established network operation with ongoing and well understood standards for "look" and show format. The producers were putting it together as they went along. The ad hoc network that produced this event will likely not exist again in the future so there is no corporate or production crew "memory" that improves with each show.
The equipment that we used was not "experimental" as some of you surmised so any weaknesses in the performance can be chalked up to lack of organization on my part or lack of experience with this particular configuration. The cameras were Sony HDC-700 and 750 with full camera control on all. Three of the cameras were located from 15,000 to 32,000 feet from the unit, running on existing KSC single mode fiber. These cameras are designed to run directly on fiber and we built camera fiber to raw fiber interfaces to make the long runs. The cameras were locally powered but all other functions operated normally. The other six cameras were cabled directly to the unit using standard fiber camera cable. Three were used on the stage, one for the countdown clock and the other two for launch tracking and color shots. The long lenses were Canon 65x and the tracking heads were O'Connor 2575. The closeup shot of the shuttle on the pad was provided by a Sony 750 mounted on a remote controlled pan and tilt head provided by Vinten and controlled through a separate fiber circuit using a Telecast data link.
The recorders were a mix of Panasonic HD D5 (2000) and Sony HDW-500 HDCams. Both are HD SDI in and out and are generally compatibile with the possible exception of some machine control issues between the two.
The upconversions that some of you noted were composite NTSC straight off the video "stumps" that the networks use, conditioned with GVG clamping EQ DAs and converted by a Snell 5100 upconverter. Some of the images originated on low quality cameras that are placed into the oribiter while the crew is being strapped in. Others came from various permanent cameras in the firing room and the "White room" outside the hatch. In any case, they are the same cameras that you saw on the network NTSC coverage.
We also set up an HDCam playback station for the KSC Public Affairs office and many of the shots on the day prior to and the morning of launch (not the live shots) that you saw on news coverage might actually have been down converted from the video shot by a KSC cameraman shooting with the Sony HDW-700 HDCam camcorder. We also provided live downconversions of our three remote cameras to KSC-TV for use during the launch coverage.
The landing coverage was lower key. We again provided a camcorder to the PAO photographer for his landing shots and we set up another camcorder with a Fujinon 36x lens at a location near the touchdown point. These tapes will be used by NHK as part of a documentary on the mission, which included the flight of an HD camcorder and a Japanese astronaut. The completed program is slated for broadcast on "Hi-Vision Day" (11/25) in Japan.
I hope that this information is useful as you try to process what you actually saw. If you need more information, drop me a note and i will try to answer your questions. If you have transmission questions, John Turner is the guy to call. He is at <turnereng@attmail.com>.
Roll Credits...Thanks to Elke Titus and Kohei Nakae at NHK for this opportunity and to Philip Hack, the NHK Producer, who put the crew and the equipment together. Thanks also to our technical team of Barry Minnerly, Abby Levine, Jim Lucas, Dave Carr, Hiroyuki Takematsu and our associates from NHK Japan and our excellent camera guys. Our vendor partners included Gary Rotondelli and Dobie Borovecki of Vinten for the remote camera control, Steve Nelson of Telecast for data and SMPTE 292M fiber links, Snell and Wilcox for converters and Chuck Lee of Fujinon who bailed us out at the last minute with a camcorder lens. The KSC folks were just great, technically and otherwise. And of course, John Turner made the transmissions work for Harris, who came up with the cash.
C. R. Caillouet, Jr.
Vision Unlimited/LA for
NHK Enterprises America
Technical Producer