Uploading to YouTube

I personally use xvid avi, and have never had a problem with it.

If money is an issue, you can use VirtualDub

What res are you uploading at? The highest I'm wanting to go is maybe 960 x 540 widescreen.

Is Xvid avi a program?

Does Virtual Dub render to .flv file?

Thanks for the tips by the way

I usually upload at 1280 x 720 (720p). Xvid is a compression algorithm, and Virtualdub will only export to avi (but Xvid will compress good quality videos well without losing much quality).

- Today, YouTube re-encodes every upload with no exception. That upload seems to be one of those from back when you could modify the duration reported in the FLV headers to make YouTube allow you a higher bitrate.
- Had that video been uploaded normally, converted by YouTube in HQ, it would've looked better. HD videos uploaded at a high quality look much better. Garbage in, garbage out. YouTube videos look crap if your source video is crap. Here's an example of what you can get if you upload high quality video:



- Does your cam really shoot at that aspect ratio? Maybe you mean pixel aspect ratio? It's best to leave that alone if you aren't a video fanatic. I've no idea where you got the idea you lose 'half of your quality' by resizing. When you say quality, maybe you mean resolution?

How to get a high quality: Deinterlace your video. Try to use the slower, more advanced options like motion-compensated deinterlacing. Avoid any kind of blending at all costs. Don't change the frame rate. Don't change the resolution. Don't change the aspect ratio. Don't add black bars. Use 15Mbps or more for the bitrate -- you're uploading to YouTube, a giant file is okay.

Quote:
It wouldn't. It would cause you even more trouble, since you seem to want quality...

Apple's implementation of the SVQ format (some .mov files):

(the encoder used all its allocated bits on the static shots, and left none for motion -- this is what it looks like when there's any motion)

The open source, libavcodec, implementation of the same format:


When it comes to video compression, Apple has not even the slightest hint. Any flavor of .mov being high quality is a myth.

Quote:
Yeah I always import my DVD discs straight into Vegas for editing/effects work and whatnot. (My cam shoots mini DVDs, and they are of excellent quality for such a plain little cam) I don't really understand containers and all that. I used pre-sets from a guy called guerillabill on You Tube for all my software and it worked great until You Tube changed the upload requirements...now nothing works.

So it's MPEG-2 to lossless .avi and follow the guide here in a while. Okay, I'll try that.

Uploading to You Tube is just such a pain in the A that if I do it, I'd rather get the best quality I can and do a thorough video, then leave it there for a while you know what I mean? I'm trying to make some videos to help show the basics to potential gpu volt modders and really need to be able to show detail on the pcb as much as possible.

Thanks for the tips guys. Believe me, they are very much appreciated.

There really is no relevance to the actual format you upload to YouTube, as long as it's not something extremely proprietary that it can't convert. Basically, everything ffmpeg can handle, YouTube can handle. The only thing that matters is the quality. Or in your case, preserving what you already have. Just do your thing, then export at some gigantic bitrate. The encoder/codec really doesn't matter, just use some ginormous bitrate. 15Mbps or higher.

Doing it right only matters when other people are going to be downloading it directly (get the same quality for less size), or if you're trying to conserve bandwidth of time. But if you have an OK upload speed and a sane bandwidth limit, you really shouldn't care.

If you want to do it right (not recommended, just do it the easy mode way): you can deinterlace it in Vegas and export it to a Lagarith AVI (google Lagarith), then use this to convert to mp4, and upload that. I can't be bothered explaining everything so just search around. I've probably said everything you need to know already, just search for the pieces. They're all in this sub forum.

To Coma: Here's a screenie of my Cam's disc settings when I use the "match settings tab in Vegas. My cam shoots shoots 4:3 and 16:9 and the pixel aspect ratio...well I will let you see it.

These are what Vegas shows simply importing a video shot at 4:3 MPEG-2 and matching the project settings to the original video:

This is all new to me because I usually just go MPEG-2>edit/effects, chapter markers>DVD Architect file>burn DVD.

Quote:
Okay, got it.

I can imagine how much trouble it'd be explaining all of this to someone not schooled in video work but you've got me started in the right direction. Thanks.

I'm used to just asking for pre-sets and setting them and going from there.

I'll get to reading.

Quote:
That's why I mentioned the pixel aspect. I had always thought it should be 1.

*Whew, this cam is about 6 years old.... but it still works.*

Okay, I'll try these settings and do some reading on this. Thanks for the tips. I really appreciate it.

Thanks Coma. Really. You've shed some light into my noob darkness.

Then keep the pixel aspect and bit depth that way. It doesn't really matter... YouTube might correct for it, and it might not, but at the end of the day, it doesn't matter that much for most videos (like... any video that doesn't show planets or balloons all the time). Nobody will notice.

Trivia: Older (pre-HD era) TVs had pixels that were shaped more like triangles than squares. This is the pixel aspect ratio. There is no way to know which ratio video on your DVDs (or even your camera) was made for... other than sizing up a perfectly round object which just happens to appear in the video from the right angle so that it isn't distorted... assuming the DVD you're watching was made "properly", then you're watching it on your modern TV (or computer monitor) in the wrong aspect ratio. And if it wasn't, then you were watching it in the wrong aspect ratio before. And not only that, but there are multiple ratios for each region, and then completely different ratios for other regions...

Oh, and TVs also crop video on the left and right. Arbitrarily. There is no standard. The aspect ratio changes based on that too (the X axis gets skewed).

And nobody ever noticed.

tl;dr none of your viewers have/will read this page - http://lipas.uwasa.fi/~f76998/video/conversion/, so don't bother doing tons of reading about it and just take the ezmode route.

I linked to it mainly for shock value It's confusing even for people who have been dealing with video for a living for years. If you want to learn something useful (and knowing what's on that page is no longer useful), try Avisynth and x264. Learn how x264 works, how each of the options affect quality (and I mean what they do exactly that makes quality better or worse, not "quality goes up if you use this" ). http://forum.doom9.org

If you wanted to improve quality further, you would learn about the different deinterlacers available for Avisynth. Almost all of them are tons better than commercial solutions, but remember: if you can't see the difference anymore, then it's good enough. If you can't tell where the gain is (even if someone else might), then better quality usually involves using obnoxiously large numbers for every option and waiting 4 days for 30 minutes of video.

Here's the option you showed that kept my quality excellent going from MPEG-2 as the source video using Vegas Pro 9:
Render As:
Sony AVC (.mp4)
Template: Internet 4:3 SD 30p (Hit Custom)
Then the Screenies.

Yeah, that page was pretty hardcore but, every little bit learned helps.... May take me ages but I'll learn something here!!!!

I'm going to tinker with the motion blur types and see how they all work...

Thanks again.

Switch Profile to Main and Entropy encoding to CABAC.

By the way -- you should check YouTube's help for tips on improving your video quality that aren't computer-related in nature. For example, they suggest to avoid a shakycam -- use a tripod. Better motion estimation means less bits required to encode. Also, make sure everything is well-lit. Check your video and look for dark spots anywhere, and then try to make them better lit.

Try shooting your video in a scene-based format. If you're making a tutorial on how to mount a heatsink, for example, then write down the process in steps, and put several smaller steps together into a group - a number of actions that can be done without moving the camera. Record these a group at a time, each a scene. When done with one scene, stop recording and reposition your camera for the next scene.

Later, edit the video so the view cuts from one point to another, like a real movie (except in your case the audio will cut too, which is why you will be doing things in grouped steps, narrating each scene separately). Avoid moving the camera while recording. Position it so it can remain still during the entire scene.