Pages

Thursday, September 5, 2013

UPDATE: Eye-fi & BitTorrent Sync (#DropBox#) for events with live (almost) previews

A little while ago I outlined my workflow for shooting an event and having a semi-live feed of the images being shot coming over to displays at the event.

The biggest bottleneck for me was the DropBox component - specifically, we need a connection to the outside world for this to function properly.

A couple of months back I've begun testing BitTorrent Sync and have now I've since changed over to this technology for most syncing tasks between devices.  I say devices, because it works great not only on Mac and PC, but also now with Android and iOS devices.

BitTorrent Sync works on a peer to peer basis - meaning we do not go to "the cloud".  We don't have space limitations imposed on us by DropBox basic accounts.  And, on a local network - it is lightning fast.

All transmissions are encrypted, and BitTorrent Sync is optimized for large files.  It is fast.

On a recent project, I was syncing tens of video files, thousands of audio files, documents, and images, between my working "on the road' notebook, a studio workstation, and two backup servers - across three locations, and two continents.

The notebook media was on an encrypted volume.  As soon as it was mounted, and data that had been added at the studio location would propagate down to my machine.  If the files (or part's thereof) were already on the other machines, I was pulling files from the other three locations at the same time.  By passing any upload bottlenecks which may have existed in any of the above mentioned machines.

And, naturally, all updates I made on the road, were synced back to ALL of the machines as soon as I had some sort of connection.

By the end of two months, the project was at over 80Gb on disk, and my "traveling" machine had burned its video subsystem to a crisp.  Not having time for repairs, the machine was replaced, and the project was synced back to the new machine within a day.

In short - the technology proved itself, and on a simple task of integrating it into an Eye-Fi / display workflow, it performs flawlessly.

No comments:

Post a Comment