|
Forum List
>
Café LA
>
Topic
re: My turn to get my tuchus kicked on YouTubePosted by derekmok
I need some input on this one.
A music video I shot, edited and directed is almost complete. My friend (the singer) now wants to post it on YouTube. Knowing the pitfalls of the double-compression nightmare, I looked up Brian Gary's guide on Ken Stone as well as Philip Hodgetts' guide in LAFCPUG. Basically, both versions look like garbage. Now, obviously it doesn't help that the original format is DV. But following the formulae in the two guides above, I still got heart-witheringly bad results. So I wanted to get your eyes on these YouTube versions...are they going to be this bad no matter what I do? Or am I bungling something that can make them look more decent? As a systematic comparison: 1. This file follows the Brian Gary formula, except I used H.264 instead of Photo JPEG. This seems to have netted the worst results of all -- corruption on the right side, horrible motion -- despite the fact that this is the largest file (99.1MB) at the highest data rate (5000kbps before submitting to YouTube): 2. This file follows the Philip Hodgetts formula. Smaller size (320x240), lower data rate and overall file size (2000kbps and 47.6MB). But it appears to have been less messed up than the other one. 3. And this is a 24.8MB MPEG-4 of the video that gives a closer idea of what the master's quality is supposed to be like... [www.derekmok.com] So, what am I doing wrong in the YouTube process? Or should I just get used to this? Opinions would be greatly appreciated. www.derekmok.com
I dunno Derek
Don't see that much difference between Gary version and Hodgetts version. It might be that you are too close to it. I'm guessing that you just have to get used to it. You are relying on youtube a bit here and not much you can do about that except what you have done. That or keep experimenting and if you find Nirvana then write up a tutorial. Michael Horton -------------------
My two cents worth. The Hodgetts version looks better. I know that I have had to play with the Gamma Correction, and BlackWhite Restore filters. He gives a good starting point, but I played with them even more, and it helped. I went to I think it's "Better" on the de-interlacing, and it made the compression time longer, but the file size was still small, and the quality was better.
Cameron Young
> Derek, It's YouTube deal with it ;-) Personally they both look fine. I'm with Mike, you're too
> close to the project at this point and I think are expecting TOO much. From someone who > doesn't know what it's SUPPOSED to look like I think it's fine. And now my editor gets to laugh at me for seeing differences that aren't really there. Oh, wait...my editor is me. That makes me feel better, though. Thanks. www.derekmok.com
<<<Personally they both look fine. >>>
Remember round-tube color TVs? Of course not. Those were popular in the Time of the Pharoahs. They took a bit of work to get them properly aligned, but the instructions were clear and if you followed the steps you could usually produce a very nice picture, but there was a trick to it. Knowing when to stop. "I bet I can get the blue convergence a little better in the upper left-hand corner. Oh, wait. That messed up the red purity in the middle. I just need the smallest adjustment in this magnet......" It's time to shoot the technician and watch the ball game. Koz
I think YouTube is a lot to blame, but also it's the old thing of the content making the big difference. Have a look at this 10 minute part I posted a while ago. Even though it's long, for the most part it looks pretty good, and I think that's mostly due to the fact that a lot of it is animation without a lot of fast movement or bazillions of colours.
What I do is just make the file as close as I can to 100MB, using H.264. It's abviously not anywhere near broadcast quality, but it's generally pretty clean. Also - make your text bigger than normal where possible. Small text is a goner on Youtube.
Sorry, only registered users may post in this forum.
|
|