I'm looking to set up a site where people can watch movies (I'm working with a small film company). It strikes me that Flash streaming is the way to go because it prevents the movie from caching etc. (and 90%+ have flash installed blah blah blah)
The question I have is about disk space. I've searched on google for streaming hosts and the disk space offered by all of them seems a bit small. An example:
100 Simultaneous Connections
50 mbps burstable streaming bandwidth
(Up to 1.5 mbps per connection)
2 GB Disk Space
75 GB transfer / month
$99.95 a month
As the packages increase, you get more transfer, bigger bandwidth... But tiny little disk space increases. Given that streaming video is about... Well, streaming video... I'd expect more generous disk space. 2 gig isn't really going to allow for much movie when encoded at a good/high quality bit rate.
So my question is: What am I missing here? Why is disk space so expensive when looking to stream flash movies?
Even with the deal above, I'm better off getting a dedicated and installing Wowza. I'm not keen to go down that route because I'm not desperate to jump into server administration on this particular project - but I'm left scratching my head at the numbers and thinking "there must be a reason which I don't know about".
Any advice would be much appreciated. I don't want to set off along the wrong path!
Yeah... I'm beging to think you're right and I'll have to go that way. I was looking at uk2.net who seem to have a huge amount of bandwidth... Will have to poke around to find the standard bit which says something vague about "fair use"
You might want ot check influxis if you haven't already, they have a custom configurator so you can alter all the diiferent stuff e.g disk space. You'll still probably end up being cheaper with the wowza + dedicated route. You could also use haxevideo or red5 but the learning curve is a lot steeper.
It is interesting to see this discussion, especially considering the only solution anyone seems to have is to go dedicated.
I've been working on a plan to create a custom niche flash server for hosting sites just like this. We've modified apache and been testing a few different things to make a faster more reliable server specifically for hosting sites created solely in flash. This would work very well for what you're trying to do, with the right setup.
Mike ECI -
As I said, it isn't up yet, at this point it's just a concept with some programming and work behind it. There's no room for 'perfectly well'. What we're working on is a replacement for the current web server structure, specifically designed for media. If/when we launch this, it will not read html at all, will cut down loading time, and bring many new possibilities.
It's not solely for video or streaming media.
Be interesting to see what you can come up with. My ideal media streaming server application would do this:
- stream native formats without any transcoding, processing or large memory use (like icecast does for mp3)
- handle both live and on demand rtmp flash video/audio.
- but able to interact with serverside logic (like FMS/Wowza)
- would use an existing language to handle events/interaction (like FMS which uses as1 on the server) but not need to compile runtimes for applications (like wowza/red5)
- can handle contemporary high quality codecs like H264 and he-aac and common ones such as flv,mp3.
- can retransmit streams to edge or geographically dispersed servers for a mini-CDN
- has inbuilt metrics system (i.e can log stats and useage data)
- can record live video/audio
FMS does pretty much it all but is expensive both in terms of licence and in hosting requirements. And their regional ripoff pricing is rubbish. Wowza handles incoming icecast->rtmp but is too java-centric and hard to use and program. Red5 is again too java-centric. Haxevideo is too 'alpha' and feature light. They all have their own advantages but none of them covers all bases.
You obviously don't know much about FMS, please don't pretend you do.
I've used WMS systems and they're great for basic video streaming and are a cheap solution for live broadcasts, but old fashioned.
Which is why pretty much why all the major online broadcasters don't use it any more.
Youtube, BBC, Myspace, Facebook, Bebo, Google video, Abc News, CNN, Reuters, Amazon, New York Times,Washington Post, Yahoo, CNET, Fox, Discovery channel, Motionbox, Jumpcut, Brightcove, video.msn.com (yes video.msn.com!!!) all are now switched to mostly flash video.
Perhaps you know something they don't..?
Apart from ease of use, customizability, cross-platform adoption, brandability etc other things WMS cannot that FMS can:
- record video and audio from a webpage
- broadcast live video or audio that is being transmitted from a user on a webpage
- share data across mutiple users (e.g whiteboard/chat)
- check the authenticity of external playback device (checksum on swf)
- windows media player doesn't have h.264 playback or aac codec. h.264 is now the defacto standard in the broadcast and movie industry for reduced bandwidth transmission.
- play back content in a custom player or integrate easily with other interactive content
Most mac/linux people won't view WM content. If you have to ask your viewers to download 20 meg of stuff before they can play it .. they won't. Why would you make them jump through that hoop when the platform offers no advantage over solutions that have better features, better adoption, better quality?
Mike ECI: It's not really in development, as I don't have the resources, and the programmer I'm working with on this and I are not able to put any money or large amounts of time into, which would be needed... but if we end up meeting somebody who wants to partner in a business venture, win the lottery or something like that, we'd go full on in developing it.
It's a great idea / plan though...
Thank you for the market research tips though, will definitely be helpful if we manage to get it going.