Is this the best quality for the web?

[www.releasing.net]

This is a 2:21:17 trailer of my movie.

BUTTERFLIES IN THE WIND was shot in 35mm and edited on my flat bed.

I edited the trailer on my Power Mac G5 dual processor 2.7 GHz and FCP5.04 Studio with 4MB of RAM.

It was telecined on and captured from DVCAM tapes.

Now, my question: are these the best choices?

I exported the timeline via QT Conversion to:

H264 at current Frame Rate (29.97),

Key Frames: 120, Frame Reordering checked,

Data Rate 2000, Streaming selected,

size: 400 by 300 Fit in Frame,

Sound PCM Linear 44.1 Mono Little Endian selected

Well, the question is about compression and quality. Looks much better than your previous incarnations, especially when it comes to frame rate and motion. A larger frame size couldn't have hurt.

But Vic, if you're serious about this film, get a sound mixer and designer. Even if you feel you want to stick to overdubbing all your dialogue, you don't have the skill to manipulate the sound to make it fit the onscreen motion. That alone will negate any improvements you may make to image and sound quality. You need colour correction as well; the look of the piece is all over the map.

But at least content-wise, it's making more sense so that those issues can actually be noticeable. They couldn't be seen in the older trailers because those weren't in good enough shape to be judged on that level.


www.derekmok.com

im confused as to the "h.264" situation. the video you linked to is a swf???

I'm seeing a Flash (Registered Trademark). Did you convert it twice?

Koz

No, I'm sorry, I'm in this state of mind that ... well, you know I'm assuming you know what I'm thinking :-)

I submitted an H 264, 93 MBs to Google. They of course spit out Flash.

My goal is to come as close to the 100 MB limit they're allowing me.

So the question is: is this the best 93MB encoding I could do?

Oh, and, Derek, thank you for the compliments. Frankly, I'm very pleased. I've made a dent :-)

Okay, regarding color correction. You know, I do make color corrections but I don't have the monitor to know how far to go. So I'm a bit reticent, but I agree, the colors are bothering me too.

As for sound mixer and designer, I'm not completely sure I understand. Of course, it's hypothetical; I have no budget for this. I'd still like to improve the sound quality -- mix and maybe even work more withing Soundtrack Pro in learning and improving the sound in general.

Thanks for your constructive criticism.

why would you lossy-compress it at all? if you want best quality via google or youtube, send them a MOV file animation (lossless) compressed at the target dimension and save yourself an extra hit...

Yeah, that's so great!

But what is the target dimension? How can I find out?

And also, if I make a lossless .mov, how can I determine what the size of the file will be?

> And also, if I make a lossless .mov, how can I determine what the size of the file will be?

Well, since you're trying to use YouTube, you won't get to use a lossless file given the 100MB limit. We've mentioned many times not to go that route, but you seem to be adamant about doing that, so you'll have to live with the limitation.

Wayne is possibly suggesting you use a different codec while reducing the file size by manipulating the frame size. But I don't think a 2:30 file in Animation codec, even at a reduced frame size, would be under 100MB.


www.derekmok.com

yeah derek thats what i was thinking. to export it at the final size it is now in a lossless codec.

but yeah vic, for the life of me i simply do not understand why you insist on letting google/youtube compress your media. as you suggested to someone in another thread - get the flix exporter

If you plan on submitting to YouTube I have this new article by Brian Gary.

[www.kenstone.net]

--ken

Hi,

Sound design/editorial tip #1- Don't use music with lyric/vocals underneath dialog.

tc

> Sound design/editorial tip #1- Don't use music with lyric/vocals underneath dialog.

Actually, that's done all the time. But you do have to de-emphasize the vocals, and the mix bias of the original song does have something to do with it. For example, hip-hop works poorly because vocals are always very up front, dry, and prominent, but Cocteau Twins and Lush work very well because the vocals are blurred into the mix, with lots of reverb. You can also scoop out some of the midrange in the song to make room for the dialogue. That sounds like Sarah McLachlan, though -- do you have the rights to use this?

Not to hijack this thread too much...the sync problems will trump any sound-quality issues. Gotta fix that before you even worry about EQing, sweetening, and music.


www.derekmok.com

First of all, thanks, Ken, for that wonderful article by Brian Gary. I made notes of everything and will definitely use it for uploading movie clips to Youtube. For better or worse, I've joined the Youtube revolution :-)

Wayne, is the flix exporter a free download? Anyway, the Brian Gary article makes it possible to use Compressor 2 as the encoder.

Derek, I didn't try to fix the lip synch in the trailer. Some of the takes I used, I dropped their sound and placed another dialogue line from the take I wanted over them, so, in other words, it was a narrative technique, not an out of synch problem.

But even if I misunderstood your point, I appreciate your comprehensive approach to criticism. Thank you for your comments.

> Some of the takes I used, I dropped their sound and placed another dialogue line from the
> take I wanted over them, so, in other words, it was a narrative technique, not an out of
> synch problem.

Not true. I replace lines of dialogue all the time, using the visuals from one take and audio from another. Sometimes it's to improve the flow of the sound to prevent a clunky edit; sometimes it's to mesh the best performances. But if you don't match the lips, it just looks like bad lip-synching. And it's doable. Just try it -- mark the opening consonants on the waveforms for both the video take and the audio take, line them up as well as possible. Actors often fall into certain verbal rhythms and it just takes a little bit of work to make the Frankensteining better. I worked with audio mixers from Skywalker Sound -- they do even more magic, to the point where even I could barely tell an ADR line apart, and I had the whole film's dialogue memorized.

Unlike your old "party" clip, this one seems acceptable quality-wise to me. Get professional help to do colour correction and fix the sound.


www.derekmok.com

no - for christs sake it's NOT FREE! and almost nothing of quality ever is!

you can get the flix standard version for $40. IMO, if a person cant generate $40 to get a quality end product, they have no business posting video to the web in the first place!

Commonly (since the invention of stereo devices), the lead vocal will be straight up the middle of the stereo image - so by lowering this part of the mix, you will generally get a softer vocal from a standard audio mix.

As for the compression, I am refusing to deal with Flash encoding for anything like this. I can't see the point.

why ar eyou refusing to deal with flash? did it kick your sister or something?

Thanks, Derek. I will do that.

I didn't know that, Justin. Thanks for the tip.

What do you mean by "stereo image" ?

Is that a feature of the mix window in FCP?

Or does it refer to the mid range of the EQ frequencies?

I actually don't know if you can do it in FCP. I'm not working today, so i can't check for you.

By stereo image I mean a combined stereo file - ie 2 mono tracks panned left and right.

It has nothing to do with EQ, purely about audio level.