The following notes are from a client project, being shared in generic format so that other site members/readers can benefit.
This project was considered long-term with no rush, to allow for careful work. It is not a typical project, most projects are in and out within a week or two. Keep this in mind when reading. Know that it is not uncommon for difficult projects to take many months when attacked slowly, in tiny pieces, at a relaxed pace.
Since this is being written for the customer, it will be using "you" and "your" quite a bit.....
___________________________________________
This was a special video for you, and neither of us wanted to rush it. I went back to it maybe once per month, did a little work. Often I ran into an issue that needed more thought. Not necessarily because it was hard or that I could not fix, but it was very often a question of SHOULD it be fixed. And if so, by how much?
By their very nature, compilation videos are a nasty creature to convert, restore and edit. The pieces can vary from a few frames to a few seconds to a few minutes long. When different cameras shot the video at different times, it gets harder. When the tape is compiled from sources of varying age and generation, it gets harder yet again.
On the other hand, this was truly a "home movie" (at least, most of the time?) with a scripted plot (plots?) and thought went into its creation. It wasn't a camcorder tape that sat in a camera over the period of 2-3 years, pulled out and added to at each family holiday or gathering. A home movie like this has a genre and style all of its own. The noise and quality sometimes adds to the quirkiness, camp and appeal of the video. I kept this in mind, to BOTH preserve and restore.
The final video is BETTER than the original tape, but it has not been STERILIZED as digital restoration can sometimes do. It's not over filtered.
The video was capturing in no less than 37 separate segments, restored piece by piece in hardware and/or software, and then re-compiled in an editor. The editing was done in the high-bitrate MPEG domain, to allow for non-hard drive return of the footage. It is max DVD specs (9000+ kbps bitrate), returned on
- DVD-Video discs, ready to play
- data DVDs, with two M2V files (two halves of video), and the restored WAV audio file
- data DVDs, with two muxed MPG files (audio+video), 384kbps AC3 and again high bitrate MPEG. These are author-ready files.
Editing performed with Womble MPEG Video Wizard DVD software, exporting and authored in another app to menu-less discs, 3-minute chapter intervals.
More work can probably be done to these, but I did not want to risk further degradation of quality. It is sourced from grainy SLP VHS, and that does help degrade quality more each time it's re-compressed digitally. Since I know you're planning to re-edit these again, I've supplied the data DVDs. The DVD-Video dsics are your preview to watch on TV.
When it comes time to edit these in Premiere, you'll want to mask the overscan with a black color (use the "clip" filter in most versions of Premiere, do not "crop"), Some of the clips can probably have more color tweaking done, on an even smaller level. These corrections should all be minor, within the abilities of Premiere for sure.
These videos can handle one more MPEG re-encode okay. I don't think they could handle two more.
Prior to editing in Premiere, consider extracting the video back out to a large uncompressed or lossless-compression file (
HuffYUV, for example), and then using this intermediary format in Premiere. There won't be any quality loss, but it will be easier to handle in Premiere than MPEG is. The timeline will scrub faster, for example, preview of filters will work better.
Chroma noise, IRE, color balance, color shifts -- each of these was addressed and corrected, as best as possible, per clip/segment on the tape.
The audio has a lot of inherent noise that is actually "good" for the production -- outdoor wind noise, for example, other ambient sounds that actually add to the mood of the video. Tape-based hiss, buzz and hum was removed. Anything "unnatural" and due to the aging media was taken out. In a few cases, audio that could have been recorded better at filming was finessed some.
Analog-based editing with lower-end equipment --
be it consumer gear or even something at the local cable company at the time -- tends to leave a lot of ugly jerks and video squiggles and other noise in the visual signal. I went through and edited out all bad frames that I could find. given how the tape was made, there were quite a few of them.
The tape also appeared to have been re-used or built from re-used tape sources. There were duplicate cuts (partials) in a few places, and I went through and removed all the "junk recordings" that were unintentional by the analog edits when it was made. While that may seem like extra work, it was actually helpful to some latter video and audio filtering passes that were done.
Everything was pre-processed in hardware, and then post-processed in software once, before hitting the MPEG encode.
On a video like this, it won't magically look better. The improvements are in the details. Keep this in mind if you do an A/B comparison against the tape.
You'll notice that your video has been returned in a new clamshell. The original VHS tape was damaged internally, and the reels were moved into a new shell. Because the old shell had notes on the labels, it has been returned, too. Cases have been provided for both tapes, since it was case-less when it arrived.
Hopefully the work was worth the funds and wait.
If you have any questions about the editing processes, once you get the project back, feel free to sign up for the forum, and we can go over your questions and troubleshooting in some new posts.
Thanks again.
