Keeping my community as "Independent" as possible?

[Gimpymoo]

I am in the process of starting up a community and so far, it has been a very long road with 2 steps forward, 1 back type scenario, BUT, progress is being made none the less.

I plan to be producing original video content for the community and also allow users to upload content HOWEVER, I have no desire to utilise Youtube for hosting. I want to create a self sustaining eco-system in that regard if that makes sense? Anyone remember web 1.0 when websites were ALL like this, I liked it, it added character I think.

I cannot understand why those creating original content combined with a community would wish to host their content on Youtube, essentially, you do the hard work of getting a community but then shove them off to another community to view your content and allow someone else to make revenue from your videos (Youtube) and the additional hassle of all the blue tape which comes with 3rd party hosting.

I have amateur experience with hosting, servers etc so understand the basics but would appreciate some guidance with what backend services (if any) are needed to allow me to fulfill my goal.

Does one simply store the video via FTP then allow uses to call them up, essentially "streaming" with their browsers doing the donkey work or do I need a method of encoding on demand (if that is even an option?). How would I then integrate this with the community.

With multimedia files being so large, what is the most effective/value driven method.

Thanks for any guidance given, it is appreciated.

[Makoto]

It's not as easy as you think. The first concern with hosting videos yourself is bandwidth. Video streaming can quickly become extremely bandwidth intensive. How bandwidth intensive depends on your specific traffic requirements, but think in the Terabytes.

Proper video streaming is not anywhere near as simple as just throwing your video up over FTP and linking to it. You have to be sure your video is properly encoded for web streaming. If you don't know how to properly encode your video, quality will suffer and you'll be wasting bandwidth, possibly leading to constant video buffering on the clients end. You'll also ideally need multiple encoded copies of each video made, higher quality versions for those with good bandwith plans and lower quality versions for those more constrained.

Naturally, your users are probably not going to be capable of doing this encoding themselves, which means you're going to need a server that is able to handle the task of video encoding on hand, as well as the web software capable of handling and processing the video uploads themselves.

The reason most people use YouTube is because self hosting and processing user uploaded videos is expensive and difficult.

[Gimpymoo]

Thank you for the reply Kirito.

I am aware of the storage/bandwidth overheads and am already allowing for a server to encode video on demand. We will have a server or two which the majority of the time, will be idling so would be good to use. To put in perspective, the videos are not primary to the sites function.

The important thing for the community we are building is branding. We want full control over the member experience. You could say we cheated by using an "off the shelf" community which is a fair shout but we can still heavily customise it as we see fit.

[Makoto]

As long as you understand the costs and hardware needs and you can meet that demand, then go for it. I've built a few applications like this myself in the past and have been considering another project related to video hosting in the near future. I fully understand wanting full control over the user experience, just want to be sure you understand it is a big undertaking, not something that's feasible for most.

If you want to integrate something like this into IP.Board, you'll probably want to have a custom application built to handle the process of uploading, converting and playing the uploaded videos.

[Grumpy]

Kirito already mentioned most of the important topics. But I just want to double stress the cost element. To stream HD content with (reasonable) ads is often not viable. With most ad contracts, it's very hard to offset the bandwidth costs. So you end up losing money every time they view. You'll want to have additional source of income to support it. Or grow to youtube size where scale makes things cheaper while catering to the most expensive ads out there.

I don't know the scale of your intention, but given that I'm big enough to make my own solution and ditch 3rd party video services, I'd go for something like... 1 server with GPU (if possible) or integrated graphics cpu (if gpu not possible) to encode. Then it passes off to distribution node(s) starting off with one. Something like raid 50 is something I find to be a good balance for download files/video/images. That would be on a 100mbps/500mbps/1gbps unmetered or 100TB package and watch it really carefully.

You might also consider other services like youtube if it's youtube that you don't like rather than wanting "independence". There are corporate solutions as well that allows white-label video hosting.

[Gimpymoo]

Thanks Grumpy.

A "White Label" solution would work, I just do not want video embeds to have the stinking Youtube logo on them.

I do not expect anything for free so I expect to pay and if a white label service can do that, I would be OK if it means I have 100% control over delivery, if that makes sense? The user experience should be seamless.

[Grumpy]

Then you might wanna check out Vimeo Pro. Last I checked, it was the only price wise affordable one. You can put your own logo if you want. It's very clutter free. You can get rid of like/share buttons too. https://vimeo.com/pro

If you're willing to get your hands dirty, I suggest making your own solution. There are many many libraries & players to make the process easier.

If you're not willing to get your hands dirty, other 3rd party video solutions may work for you. But they only remove the hassle of encoding and delivery complexity. For them doing that, there's is a quite a bit of added cost. I've seen past big clients look at their price tag and immediately say no. Though, the market may have changed a bit since.

[Makoto]

I haven't looked in quite a long time, so I honestly can't even remember the names of any of the reputable providers out there for video transcoding and delivery, but I do remember that all the ones I was familiar with when I was investing in a (smaller) infrastructure of my own were all very expensive solutions.

Doing a quick search now, I see even Amazon has a video transcoding service of their own.
http://aws.amazon.com/elastictranscoder/pricing/

Even that is quite pricey though, for a 60 minute video you'd be paying ~$7.45 for three deliverable copies.

It would be much more cost efficient in the long run to roll your own solution. There are libraries such as PHP-FFMPEG out there for the software end, but honestly it probably would be perfectly fine to just use shell_exec() with FFMPEG for the transcoding application itself as well.

JWPlayer is one of the best web players that I know of. There's also Sublime Video but.. they are very proprietary with their pricing model. Last time I checked, you're not even allowed to host the video player script locally on your servers and you have to pay based on your traffic requirements. Despite how nice the player itself looks that's always greatly put me off from using it.

[John 3:16]

I have used VimeoPro, Wistia and hosted my own.  VimeoPro is always the winner for me.  I think with what you want to do, you start with a solution like VimeoPro then as you see more and more success, your need will truly show itself.  There is major video streaming companies like Brightcove.  Though what you are talking about is really VimeoPro, the best of both worlds and the ability to protect a rogue member(competition) from sharing your material outside of your own site.