Encoding 101 - simple encoding to upload to YT

Sooooo people don't really know how to encode. That's what this is for! \o\

What is encoding?:

Why should you listen to me?:

What these instructions are useful for:

Programs and stuff:

Step 0: Install everything:

Step 1: Download/Render video.:

Step 2: Open the video and audio in VirtualDub.:

Step 3: Add filters (Optional):

Regarding Video Resolution:

Step 4: Choose your codec:

Step 5: Save as .AVI:

Wait a bit and you're done!
Depending on the video, the encode should take anywhere from 3 to 30 minutes

I probably forgot some things I need to explain somewhere. If you have any issues just ask.

THANK YOU! I've always struggled with encoding (I use Vegas or avidemux + Nico encoder) so this is a big help and really insightful *A*

Could this thread be stickied?

This is a good guide, though I am wondering, how do I lower the file size? 15GB won't upload to NND obviously...

see "Step 4: Choose your codec"

Oh woops I had the thing set to high....

I uploaded the finished thing to NND and it was still terrible quality... I'm not sure what I did wrong

Had what thing set to high?

NND follows a whole other set of rules and requires more complicated instructions then are in this thread. I'll write it up... eventually.

Stickied the post for people's references : D

Just sharing small things I've tried

I *never* encode my video for upload

Assuming you are using existing PVs from NND without any modification or editing, there's no need for encoding. In fact, we should avoid it. We're not dealing with GBs of raw HDTV broadcast dump if we get our source video from streaming sites, they've handled it for us and the rest should be much easier.

The magic here is called 'muxing', we won't re-encode the video, a process which will degrade the quality. Let's imagine a video from NND reprinted to YT, in this process YT will re-encode that video. Then, you download the YT reprint, and using VirtualDub re-encode it to avi, upload it to YT where they'll do yet another encoding. That's a re-encode of a re-encode of a re-encode. More blocks, more color bands, uglier video
Indeed, encoding is a difficult art which takes years to master (I can't do it well either).
Instead of encoding, we'll keep the video intact, discard the original audio and then put your mixdown in place of the original audio. You'll save much time if you avoid encoding. Muxing only takes several seconds while encoding takes at least several minutes on a dual core PC/laptop with 50% or more workload for the CPU.

First, prepare your final mixdown as 160kbps AAC (that's the max for typical streaming sites, you can set it lower if you only have low quality off-vocal and/or low quality recording result). If your DAW can't export AAC, export it as .wav (or .flac) or other lossless format. We'll do it in such way that there's no lossy-to-lossy audio encoding.

You'll need to download the video from NND using any flash downloader of your choice. It has to be from NND for the best result, not YT reprint.

To 'mux' the video, set your video converter settings as explained below:
Video codec/compression: 'copy' (not x264 or other)
Audio codec: this depends on your mixdown, but we'll need AAC audio. If you already export the mixdown in AAC format, set it as 'copy', if not, set it as 'aac' 160kbps. AAC encoder in a professional DAW might be better than one in your free encoding software, so you should have your mixdown prepared as AAC if you're using AA or other pro-tools.
Output format: mp4. AVI doesn't handle x264 well and it's the preferred upload format for streaming sites.

As for detailed procedures, it depends on which video converter you are using. The easiest shortcut is using mkvmerge-GUI (just works out of the box since it can only mux without any encoding features), but you'll get an MKV file instead of MP4. YT accepts MKV files but I've never tried submitting MKV to NND.

If you want to add subtitles, overlay your artwork etc you'll have no other choice but to encode the video. It can be quite costly, depending on what stuff you're putting on the video. For example, we'll have to adjust CRF value by -2 for adding typical printed subtitle (hardsub), this significantly increases the final output file size and encoding time.

CMIIW. Needs more help and explanation from experts here