Monday, October 15, 2007

Workflow for editing "Chalk"

Since no one has yet written a playbook for multimedia production, it seemed worthwhile to keep notes on the work flow that evolved in the production of “Chalk.”

By now I've viewed dozens upon dozens of sound/slide shows on the Internet and most of them are forgettable. I didn't even finish viewing a lot of them, brief as they were. Mostly it came down to the difference between an audio-driven slide show and slide shows with sound. The difference depended upon good editing, starting with audio. In fact, that seems to be the primary difference between newspaper work and multimedia work. With the slide show you start by thinking how you can tell the story, and THEN you think about what you are showing.

Rough cuts – I began my audio editing by organizing the converted files into separate interviews (one per artist and one long interview with the chairman) and supporting files (registration and two different guitar numbers). The basic idea was to break up the long narrative interview with brief interludes from some of the artists. I then opened the chairman interview in Audacity and cut out everything I knew I would not go into the production, leaving silent gaps in between the usable material. I made similar cuts on the artist interviews and supporting files, eliminating some of the less promising interviews.

Content tags – To the several Audacity files I added Label Tracks, noting the content topic under each audio segment. Once I figured out the editorial value of tagging, I stopped wasting time trying to transcribe audio during the editing process. Instead, I started tagging audio segments as part of the rough-cut process.

Project file – Once all the audio resource material was prepared, I opened a blank project file and saved it with a working title.

Assembly of story segments – I started the production by importing the registration segment. Even before beginning to edit the chalkfest material, I had both my opening and ending in mind, as I had for “School of Rock.” Knowing where you’re going to start and where you’re going to wind up makes story editing so much easier! I recorded several people registering for their sidewalk space as the obvious place to begin coverage on the day of the event, but not until the next day's rain did I have my ending. I knew the irony of the rain had to be covered in the follow-up interview, so I planned the interview questions accordingly. Switching from staff photographer to multimedia producer meant learning to think as an editor from the start of coverage to conclusion of the production.

Story tags – I added a Label Track to the project file to keep track of the story development until the basic storyline was completed. Without the story tags in the development of “School of Rock,” I was constantly having to relisten to segments or refer to handwritten notes to remind myself where I was in the storyline development.

Additional audio tracks – In the “School of Rock” audio story, I added a parallel track fading additional segments in and out of the main storyline, rolling from one to another. In “Chalk” I wanted more separation between the separate components. For continuity I used music recorded at the festival, which went on a separate Audio Track. Rather than use an identifiable song, I used melody-free music that could serve as background for the entire length of the production, fading down beneath the various speakers and back up between the segments.

Trim & fine tuning – With “Chalk” it proved more expedient to wait until the basic storyline was assembled before making any serious effort to trim and fine tune individual audio segments. It was the trimming and fine-tuning decisions that proved most problematic in terms of the journalistic protocol involved in audio editing. As with the audio notes edited from the seminar, I had no firm guidelines for what was allowable within journalistic standards. (Perhaps specific instances can be addressed in a future blog entry.)

Transcription – Unless the success of the production depends on precise coordination of visual content with the audio storyline, it shouldn’t be necessary to fully transcribe the audio before assembling images for a slide show. I felt this was necessary for “Chalk.” The best means I found was adding the transcription to the project story tags in the Label Track. I exported the expanded Label Track to a text file, which included the exact timing of each entry in the audio track, and then transferred the text file into an Excel spreadsheet. The spreadsheet gave me the advantage of calculating time segments. With this precise audio log I could start assembling my images.

Assembly of images – Editing images for a slide show is quite different than for a photo layout. Instead of developing a visual story in terms of spatial relationships, you are assembling a linear progression complete with transitions. And then there is the sheer volume of images involved in producing a slide show. For “Chalk” I had nearly 300 images to edit. I started by separating the images into categories to correspond with the story segments, each divided by a blackout. I composed the title and credit frames in Photoshop, selecting a suitable typeface. The last step in assembly was to number the individual frames in their order of appearance and place them in one folder.

Soundslides – The real value in this $40 program is its simplicity: Import the Audacity project saved as an mp3 file and then import the folder of numbered jpegs. The program presents a storyboard of your production that allows you to adjust the order and timing of individual images, as well as the type of transition between the images. For an additional $30 you can go “Pro” and add the spiffy Ken Burns movements, among other features.

Final production format – Instead, I opted to buy the additional Video Plugin ($20) that allows you to Export a Soundslides production in the Quicktime format. Ironically, the video plugin can’t capture any of the additional Ken Burns movements you pay for in the Pro version of Soundslides, but the advantage of converting to one self-contained file in a widely popular format outweighed that limitation. [UPDATE: Since producing "Chalk" I actually read the entire manual for Soundslides. Not only does Exporting to the Soundslides publish_to_web preserve the integrity of your production (including the Pro features), the total files amount to less than half the size of the one Quicktime file and typically runs much more smoothly online.]

No comments: