[James Ewart]"The game is up surely with Adobe Anywhere?"
Care to expand upon this. Your statement (or question) isn't clear. Adobe Anywhere requires a significant investment into a shred storage and server infrastructure. It's a product targeted at larger enterprises, not individual users or small/med shops. So what were you getting at?
Saw a demo of it last year. It's a hefty priced server backend and storage you'll need to get it to work. Lots of hurdles to overcome, but they are exploring new workflows and I give them kudos for trying.
Look what broadcasters are thinking of doing with production workflows and the cloud hub going forward.
Cloud hub…that will be a marketing term we are sure to see more at NAB 2014.
What happens when all your media is tied up in the cloud and you can't get to it through a bad connection. Oh so many "what if scenarios" to over come to make this a reality. I wish them luck and I hope to see it in my lifetime.
on Feb 11, 2014 at 12:20:09 pm Last Edited By Keith Koby on Feb 11, 2014 at 12:27:18 pm
There's a major push among IT types to get the datacenter out of their facility and house all data in the cloud (amazon). They have transcoding there. This is different though because it requires a major investment in house. Post infrastructure guys like and dislike this Anywhere idea. I think most post guys recognize we aren't ready for the cloud as well.
The pros are obvious and for us, with the weather and transit breakdowns, it's interesting.
The cons are the price tag and that us post guys like to see our actual clips playing to a reference monitor and not flash like proxies on a laptop monitor.
edit: prematurely posted... it happens to everybody. don't deny it.
[Andy Edwards]"Saw a demo of it last year. It's a hefty priced server backend and storage you'll need to get it to work. Lots of hurdles to overcome, but they are exploring new workflows and I give them kudos for trying."
What are some of the hurdles? Are they mainly just technological (fast enough server + storage, faster enough internet, etc.,)?
Yes, I'd say it's the base server + large enough storage for all sizes of media and proxy generation space + fast enough bandwidth costs to make it operational. It was only running on a mac mini talking back to adobe servers, which served out the proxy files for the demo. We have a very fast connection, so not something an editor would see connected to hotel clamped wifi or ethernet connection. No 4K, just standard HD proxies in a timeline so you can chop your edit together. This NAB we should see a new version coming after a year of testing. CNN was one of the largest beta-testers, so we will most likely see a presentation from them again at the booth like last year…(pure speculation on my part)
The essential architecture is this. The centralized Windows server handles all streaming of full-res media from a SAN. This server has to have beefy NVIDIA cards in order to composite and add effects on-the-fly in real-time. This only happens at the "home base". So basically the server is running a satellite version of Premiere for every connected editor.
Then, this full-res stream is served up as a single-stream, low-res proxy file that is streamed to the editor who is connected long-distance. Whenever the stream is paused, the server adapts the bandwidth and feeds a single frame of full-res video, instead of the playback of a real-time proxy file. So the server is throttling the image quality based on demands and network load. Feeding a single frame of full-res video in pause, permits the editor to remotely adjust effects with some accuracy. Note that most (if not all) third party plug-ins are probably excluded for now.
When the cut is locked, a "publish" command (via AAF?) - back at the central server - encodes any master files and deliverables. So no full-res video actually travels between the Anywhere server and a remotely connected editor.
there's a place in london that has invested heavily in anywhere to allow it to function for commercial material across multiple sites on different continents. They're on for it, and the workflow financial knock on benefits at scale are sizeable - but its a bit like buying a mclaren, in that you get mclaren/adobe engineering workflow people turning up in force to discuss and break down implementation. right now its super expensive big iron stuff. hence the adobe butlers. They're going after top end systems media production cash.
for all that adobe pooh pooh generating yet another intermediate codec, there is a custom adobe codec sitting very privately in the middle of anywhere.
its funny because people investing in it then become incredibly curious about whether adobe has a fit for purpose video legaliser in the actual edit system itself for instance.