The previous post detailed how Rclone can reliably upload large files with their checksums to Backblaze unlike other programs. This post will outline the workflow and some gotchas to keep in mind when doing massive data loads over the internet.
Backblaze B2 is an incredibly cost-effective cloud-based archival platform. I had a few TBs of large file video footage stored on a Synology NAS that I wanted to archive to B2, in case anything happened to my local array.
Sometimes it’s a live gig and I’m doing multiple things, like camera and sound. Throw in some stage delays and out of a 56 minute clip, half is not needed. On a Sony AX100 shooting at 1080p XAVCS 30fps, that’s 11 out of 22 GB extra.
I knew that normal iPhone videos are variable frame rate. Learned that the hard way, once when I was given a concert recorded on an iPhone to sync up to mastered audio. #superfail #ididntrecordthis
I’ve recently been experimenting with realtime editing of live events. That way I can speed up turnaround times for multicam edits, especially for 2+ hour events like typical Indian classical dances and concerts.
Maybe I’m too used to rewriting history in Git, but sometimes I modify the create date of my pictures and videos. Metadata.
Recently, I found myself deinterlacing footage from two Blu-ray discs using some Red Giant After Effects plugins. The two compositions were HD 1080 59.94i files, 51 minutes and 94 minutes each. My jaw dropped when I saw the estimated time for just the first file start at 50 hours, and keep increase. Plenty of CPU was still available on the mach...
With over twenty years of Indian classical music concert footage, I’ve always dreamt of tagging each file by raga, tala, artist, audio quality, etc. to quickly pull up footage based on smart bins. Looking for Śivarañjanī? No problem. Dhrupad? Got that too. How about all files that need audio synced up? Coming right up.
Earlier this year, I was livestreaming a music concert where there was no access to the soundboard’s mic output. It was a low-key, unlisted stream for family and friends, but I still wanted a find a way to improve the audio quality rather than using the camera’s built-in mic.
This May, I tried something new and recorded a HD Wirecast livestream to disk as ProRes 422 @ 480p to help save on disk space/throughput.
Apparently I forgot my own advice. Use fast external disks as the destination.
I often use Finder labels colors to manage media, specifically colors. These can easily indicate statuses: green for complete, blue for “cold storage” projects, red for abandoned projects, etc. These labels however are not supported on many popular bucket storage, e.g. Amazon S3 and Backblaze B2.
Sometimes, when collaborating on projects virtually, file storage becomes a problem. Professional studios can probably run highly performant SFTP servers or bucket storage systems, but when you’re working with other freelancers on tight budgets—who may not be as tech savvy—a 30 GB file starts becoming a big problem.
We’ll start with a quote from Larry Jordan: X.264 and H.264 should only be used when creating files for the web. If you plan to edit the resulting file, convert it to ProRes instead. AVCHD files compressed into H.264 for editing will look just awful.
Last May, for Chhandayan’s All-Night Concert 2015 in NYC, I cut between two cameras in the livestream for the first time. Little did I know, there was a lot to learn.
Newer, prosumer codecs these days like Sony XAVC-S don’t seem to record video with timecode. It’s one of the chief drawbacks that knocks the otherwise brilliant Sony AX100 from solid professional use. Timecode tracks can save mountains of time when editing multicamera shots, relinking media files, trimming clips precisely and so on—which, osten...
There is quite a bit of literature on the net how Adobe applications can embed XMP metadata and modify original footage on import—wreaking havoc for other NLEs like FCP. Posts range as far back as 2011, the year FCP X was released, to even one in mid 2014. However, for those that roundtrip between FCP X and Adobe with XMP metadata, turns ou...
For a long time, Sony’s XAVC-S codec was completely a mystery to me. It was the new format the Sony CX900 and AX100 use, and when those cameras originally shipped, the files could not be edited natively with FCP X. (Version 10.2 added that functionality.)
Many new Sony video cameras (like the FDR-AX100 and HDR-CX900) record into the XAVC-S format. XAVC-S writes to MP4 containers, and alongside them sit nice little XML files with metadata. The following is what I’ve unearthed from poking around the SD card.