During the summer of ‘99 my friend, Gary Graver, offered me an opportunity to work on a digital video feature. I accepted right away. Working with Gary is always a pleasure and I was eager to learn more about digital video. "The Blair Witch Project" was very much in the news and many people feel this is the future of low budget filmmaking.

The genesis of the project was Ray Mazarec, the keyboardist and organizing force behind The Doors. He and Jim Morrison were both film students when they met and formed the band. Ray has made several documentaries about the band: Doors Live at the Hollywood Bowl and The Doors: Dance on Fire.

Rick Schmidlin, best known for his re-edited revivals of Orson Welles’ Touch of Evil and Erich Von Stroheim’s Greed, was also producer of Ray’s documentary films. From the start they were thinking that digital production would make it possible to shoot a top quality project for $500,000. Also with video it is possible to shoot long takes and experiment with the actors. During production Ray did take advantage of this, and takes often ran longer than five minutes, a considerable challenge for Chuck Homyak, my boom operator.

Love Her Madly had more resources behind it than most lower budget shows. The performances by veteran actors Madison Mason, Richard Danielson and T.J. Thyne were very good, and although Gary’s filmography includes many titles he would probably wish forgotten, he is in fact a cameraman of considerable ability. For many years he was Orson Welles’ cameraman and was responsible for shooting F Is For Fake and The Other Side of the Wind as well as some of the infamous Paul Masson wine commercials.

The script, by Rick Valentine, is an ingenious but conventional narrative. It opens with a news bulletin of the discovery of a murder on a college campus. Before the victim is identified we meet the various players in flashback sequences. The principals include a professor of literature (Madison Mason) who enjoys some renown for a play he wrote many years ago but who has now devolved into a older campus lothario. He holds court with two of his students in a local tavern. A provocative young lady (played by Jennifer Lothrop) transfers into the school, enrolls in his classes and comes to have relationships with all of them. A number of events are viewed, Roshomon-like, from the perspective of different participants. We don’t learn the identity of the victim or the murderer until the very end of the story.

Gary brought the camera, a Canon XL-1, by my home a few days before we started production. He had acquired an interface box from Canon that permits connecting line-level XLR cables to the camera. I had borrowed a similar unit made by Beachtek from Location Sound Corp and we proceeded to make some tests. Navigating the various menu options to achieve the proper set-up is hardly intuitive but we did, finally, manage to get everything working together. Theoretically the camera has the capability to record very high quality audio; its specifications rival those of professional DAT recorders. Our initial tests seemed to bear this out when we monitored recorded tracks by connecting headsets to the camera’s audio out. However, when I connected the audio out of the camera to my Cooper mixing panel and monitored the track through the Cooper, the noise floor was much more apparent and we also noticed that unexpectedly hot signals were prone to distortion.

Much of the new digital gear boasts audio specifications that are not actually realized because the cheap pre-amps or other analog components hamper the digital recording. I advised Gary that the recording I could monitor off video was not sufficient for this project. It's always possible that recorded audio was actually high quality and limited by the playback capabilities of the camera, but determining that required more testing than we had time for. Without hesitation Gary agreed that I would record all the tracks on DAT and we would proceed, as do all features, in double system.

Use of the Canon XL-1 as a production camera is somewhat restricted by its inability to record time code in a normal fashion. Time code can be recorded on the tapes but there is no access to the software driving it to make settings. Worse, if the camera is turned off at any time it automatically resumes the time code count from zero so a single roll of tape may have multiple instances of duplicate time code. Not a good idea. It is possible to work around this by pre-striping tape cassettes. Gary would take the camera home each evening and record a tape cassette with the lens cap on. This would record a continuous code on the tape but no picture. The camera would recognize the presence of code and not record over it when shooting picture the next day. This, like formatting tapes for the DA-98, etc., is a problem during production and there were several occasions when we ran out of pre-striped tape and had to shoot with blank. The rumor I heard is that Canon declined to pay the licensing fee to SMPTE for the use of professional time code. Apparently the concept of numbering each frame cannot be protected by patent but the precise protocols of SMPTE time code can be licensed and protected.

The XL-1 also proved to be somewhat cantankerous. On several occasions when we tried to change the settings to enable recording an audio feed from my mixer there was a bizarre anomaly in the picture. Large horizontal bars would appear in the viewfinder, obscuring much of the image. We confirmed that these bars were actually being recorded. Lengthy calls to customer service at Canon did nothing to resolve the problem. Once the image problem appeared, it was necessary to shut the camera down and let it rest for at least 30 minutes before normal operation was possible. We quickly decided to avoid the whole audio section of the camera.

The lens supplied with the camera is another limitation to the system. Auto-exposure and auto-focus are built-in and not easily defeated. Fortunately, it is possible to use other lenses and Gary acquired a Taylor Hobsen lens that was sharper and permitted manual exposure settings and conventional follow focus. The one continuing annoyance was the flashing exposure warning in the viewfinder when using this lens. We couldn’t find a way to turn that off, and eventually resorted to putting a piece of tape over that spot on the director’s monitor.

Working double system, we used a standard Denecke time code slate to match picture and sound. This considerably simplified work on-set since no connection needed to be made between the camera and audio. However, it also meant that dailies would not have the production sound unless the tapes were sent to telecine. The cost of such a procedure would wipe out some of the advantages of working in digital video so we did not use it. Theoretically, the tapes might at least have audio from the onboard mic so the director and editor would be able to evaluate performances. We turned that mic on and subsequent dailies had the benefit of audio from the camera’s mic.

No effort was made to sync dailies with the production audio for about ten days. This was largely because no editor was yet assigned to the project. Gary and I were concerned that so much material was shot without any confirmation that everything fit together properly and we urged Rick to make conformation of dailies a higher priority.

Eventually Andrew Pierce assumed the editing duties and began work on matching sound and picture. This proved to be more problematic for him than we had expected. Unfortunately Andrew came on the picture late and we had no notification of the problem until we were almost done with the shoot. It appears that he had been working with an older model Avid and the software in that unit had difficulty coping with discontinuous time code. This revelation was followed by the sound of my hand slapping my forehead. Of course, if we had only known, we would have arranged for record-run code and transmitted the code to the slate with a Comtek. The cost of an additional Comtek set would have been insignificant compared to the extra expenditure of time hand syncing. Ultimately everything synced-up properly and all the audio was working well for the picture.

Working with Ray proved a special pleasure. There was a piano on nearly every set and he would noodle around whenever we were between set-ups. We enjoyed our own private Doors concerts with a history lesson in rock and roll thrown in. We had a lot of fun making "Love Her Madly". Everyone got along well, taping proceeded smoothly (for the most part) and the quality of the results was satisfyingly high.

I haven’t seen the completed film but I have seen many of the scenes on the Avid and everything looks very good. The state of "prosumer" digital cameras today is markedly more "sumer" than "pro" but the quality of the image is remarkably good.