A pickup channel is the ultimate chicken and egg problem. People will decide to play there because it is active because people have decided to play there. Years ago it was easy enough to build up a following and long-standing community in a channel simply by having a decent server, a couple of mature admins, and a bot that did the job. Unfortunately with the advent of TF2Lobby and the vast amount of mixes/doublemixes that are organised that isn’t really enough anymore.
An idea is the least valuable thing you will ever create.
I started work on the Pixel Pickup bot back in Summer 2010, mainly to occupy my mind after a break up. Many of you reading this won’t be aware, but back then the big boy in the world of European pickups was a channel named #mpuktf2.pickup, the non-invite little brother of #mpuktf2.pickup2 that still maintains its infamy for incredibly high invite requirements to this day. Unfortunately after a dip in activity (mainly due to pickup2 making a noticeable drop in the invite criteria), the channel was shutdown leaving those in div4 and below with no where to play to casual games of 6v6 TF2. I believe it was around this time that double mixes started to become a regular occurrence.
Last Updated: 18/02/2013
So you’re Dendi? That’s pretty cool. Well for years if you’d wanted to streaming yourself feeding as drow ranger or failing every hook, the go-to solution would have been XSplit, but no longer. A fine gentlemen going by the alias of “Jim” has spent the last few months of life writing an amazing piece of kit called “Open Broadcaster Software”, or OBS for short. Essentially what he’s managed to do is take the same product that a team of full time developers backed by great company have taken 3 years to create, and produce something that works much better in just a few short months. It’s still lacking a little the way of features (although this has improved significantly since I wrote my TF2 streaming tutorial just 3 weeks ago), but the core of the software itself is infinitely better and the missing features are things that shouldn’t ever be a concern for player streams. It puts out better quality, the UI is cleaner, it handles high FPS (i.e. streaming at 45fps+) far more easily, and it does all this while have much less of an effect on your game’s performance.
So you want to stream your TF2 gamings? Great! Historically if you wanted to stream a game you would use XSplit. No More. A few weeks ago a guy released a piece of software he’d written entirely in his spare time called “Open Broadcaster Software”, or “OBS” for short, and to be honest it’s blown away everything SplitMediaLabs have accomplised in 3 years of having a team of full time developers working on XSplit. The performance is astronomically better, the UI is far cleaner, it has a much lower memory footprint, and has a far higher degree of customization. It’s a little lacking in features right now, but they’re being added a blistering pace so I doubt this will be the case for long. Hopefully this guide should be everything you need to know. The focus of this guide will be on achieving the best quality possible while minimizing FPS drops, input lag and micro stutter, i.e. how to keep your game playable.
Streaming is complicated. I’m not going to pretend it isn’t. There are far too many guides around which follow the formula of “These settings worked for me, so they are obviously perfect for everyone, follow this guide and copy them verbatim”. I’ve spent the last year (somewhat successfully) trying to push streaming within the TF2 scene, I’ve dealt with dozens and dozens of streamers from all over the world who all have wildly different computer specs, internet connections, and there are so many variables that all I can do is to explain – in plain english – what the pros/cons of different software, and give you a rough guide as to which settings are sane.
First however I’m going to lay out a few truths.