Forums
Add Reply
http://i52.tinypic.com/98c2g0.png
So this is what your Procaster preferences tab should display.
Starting from the top, let's learn a bit about Digital Video and how using it for the web works!
Source:
Source is just what camera you want to use. Assuming you're streaming a game from your PC, keep it on No Camera.
Encoder:
The encoder is exactly what it says it is, the method used to compress/package your video for the web. People may know this, but most people get lost when they look at the options, no matter how short the list. So let me help clarify a small amount of each Encoding option.
H264
H264 is a relatively new video compression standard that was to bring the M-PEG format to the web while keeping its quality intact.
What you need to know: H264 is a fancy way of saying .mp4 video. It can support high quality video, but might have a longer encode time, causing more delay to the stream. (Larger in size)
On2 VP6
VP6 is a video format and codec (mostly associated with Flash videos). It's probably mostly used for Podcasts and other web videos.
What you need to know: VP6 is the equivalent to a mid-to-low quality Youtube video. Its quality is limited, but it should cause less strain on your PC while taking less time to package for the web. (Smaller in size)
If you're looking for high quality, try H264 after reading through the rest of the guide. If you're struggling to keep your stream from looking like a slideshow, start with On2 VP6.
Output Resolution:
It gives you the option of changing the amount of pixels in any given scan line of the video and how many rows of pixels there will be. The smaller the resolution, the faster and smaller your video will be. Resolution does NOT effect the quality of the video, just the size at which it will be delivered. If it needs to be scaled up once it reaches the server you're streaming to then you will start to lose quality. Try to find the standard dimensions of what you're streaming to.
Output Aspect Ratio:
Used to define the shape your video will be displayed. If your resolution isn't the same ratio as your Aspect Ratio, it will be rescaled and might lose quality or 'squish' the video to look skinny or fat. Check your game/sources Video Settings to find the aspect ratio.
Target Frame Rate:
Digital Video is just a series of images played back to trick our brains into seeing a moving picture. The target frame rate is just the desired amount of images per second. The more frames (images) per second the larger the video will be and the longer it'll take to encode. Most people don't realize that smooth video is perceived at around 23.97/24 frame per second. You can lower this to about 12~15fps and people will still find it tolerable for web streaming. Although if your stream will have a lot of action or movement, you'd be wise to have a higher frame rate.
Target Bitrate:
The Bitrate is the amount data that the video can put into each image (Not an exact definition, but it's just so you understand the very basic of it). The higher the Bitrate, the better quality each image will be, the better quality your video will be. However the higher the Bitrate, the longer it will take to encode and the larger the video will be.
I have to add, people think because Livestream has it capped at 500Kb/s that making it higher won't help, and that is partially wrong. Typically starting out with a high quality image or video and then compressing it will still output a slightly better image. Whether or not it's noticeable is questionable.
Keyframe Intervals:
Have you ever tried going to an exact time in a video, let's say 52 seconds and the playhead keeps jumping forward to 56 seconds or 48 seconds? That's because you can only search to a Keyframe in a Digital Video. Think of Keyframes as "reference frames". Without going too in-depth, there are three kinds of frames.
- I-frames (a Keyframe or Intra-coded frame) - A fully rendered image.
- P-frames (Prediction frame) - Uses the prior Keyframe or P frame(s) to predict the next frame, adding only the additional data needed from the previous frames.
- B-frames (Interpolated frame) - Same as the above but uses both frames before and after it.
So what all that sums up is, if there's a lot of motion in your video then you need more Keyframes for P and B frames to reference, otherwise you'll lose a lot of quality in-between. For fast paced games, 1 second per Keyframe should be suitable. But be careful, the more Keyframes there are, the larger in size your video will be.
Deinterlacing:
A lot of people don't know the difference between 1080i and 1080p, or even why 720p is HD but 720i is not.
The p on 1080p stands for Progressive Video and the i on 1080i stands for Interlaced Video. Progressive video is like film, it displays full images each frame. Interlaced video uses fields (Even and odd numbered scan lines), much like your computer screen, to display an image. There are a few methods interlaced video uses, just know that it doesn't display the from top to bottom without skipping lines.
I won't get into why Interlaced video exists, the only thing you need to know about deinterlacing is this:
If you deinterlace a video, you will have a smoother video (useful for a lot of motion) but this makes the encoding take slightly longer.
Anyways, I hope this helps some of you streamers/editors out there. Given some effort, you should be able to start streaming better video with this Procaster.
Back to top
MultiQuote
