Would you kindly expand on a "Vegas HDV M2t" timeline to "Flash as represented in that spot" workflow??

thnx,
v

I don't know how they fully did it, but if you render as mp4 and deliver at a high bit rate, you can get compression like that. The Flash bitrate for that spot is probably waaaaay higher than youtube.

The On2 codec for Flash is suppossed to be extra-spiffy.

http://www.flashvideofaq.com/

Basically I've been de-interlacing using Mike Crash'e Smart de-interlacer then encoding using on2's 2 pass flash encoder. Really shouldn't make any difference what the source is in terms of workflow. Judging by how the clip played out down here I think they're using a fairly high bitrate. Having a very clean source no doubt helps no end too. I know of at least one local production shot on 35mm and the only delivery will be on Youtube.

It's when you get into high compression ratios that the measurabator stuff starts to count.

Bob.

The Hollywood Film Chorale

Official Website

I would guess they are using hardware scaling as well. I do this stuff on a daily basis and I can tell you that there is a clear and sometimes striking difference between a clip transcoded and scaled using only software vs. a true hardware scaler.

On2 is head and shoulders above the other Flash video codecs visually, but it is also very complex, which can make realtime streaming difficult for all but the fastest processors (or two).

Farss, concerning VIc's general question about detail, this is outside the realm of vegas' workflow. Very High-end encoders, permit sub-fields/zones/areas of a larger field to be encoded at a higher rate,(yielding greater detail in a small region). Adobe PS CS2/3 is now offering it for web graphics. This requires a lot of work vs just using a template for rendering. This can create a "wow" effect if used wisely, like DOF, a powerful tool in post.

"I know of at least one local production shot on 35mm and the only delivery will be on Youtube."

Thats jsut plain crazy bob.. at least try to convince them to havea HQ encode on their site..

as for this i found a neat lil free flash encoder which doesnt require any flipping etc etc

thing abotu flash is that depsite the colour and sharpness, it requires much more bandwidth than WMV or any other web compression.
MP4 is just huge and WMV makes too many compromises.
Flash is a good go between and with good source material, will offer a truer representation of the work itself

My browser cache says it is only 7.2 MB. Huge compression with outstanding quality.

2 minutes, 5 sec long. Do the math.

I think that was an extended version of a commercial I saw in the Super Bowl or the Academy Awards. I've seen this group before on TV around that timeframe.

Correct in the question of how they were able to deliver. Notice the frame size was fairly small, 320x240 widescreen, about? Compare to Patryk Rebisz (in my memory) larger frame size and quality. I don't have file size for Patryk's to compare, though.

I'm aware of these kinds of fancy encoders and oh yeah, compression is something one can build a career out of.
But I've recently found out something new. I've always been a noise nazi because of how it challenges intraframe compression but especially with wavelet codecs that's not the whole story. Scene complexity also has a major impact on bandwidth, more detail needs more bandwidth, even with a static shot.

So looking at this footage and the way it's shot there's very little detail, no EE, no sudden jumps in color, mostly very smooth gradients. There's heaps of resolution though and I suspect that's part of the trick and why well shot, well lit, carefully graded 35mm holds up so well.
I've seen much the same with mpeg-2 compression, I've scrunched football matches at 3Mb/sec without visible artifacting (apart from one dissolve I stupidly used), I've made a DVD from a BBC DigiBeta copy of Blue Planet, again it looks just stunning and yet I've had footage fall apart woefully at 8Mb/sec. The one factor that seems to me to have the greatest single impact at high compression ratios is the content. I'm not saying better tools don't do a better job but it always seems to me even the basic tools that Vegas ships with can do an excellent job with very well shot footage from the big cameras in competant hands. Conversely nothing can save crud from the effects of high compression ratios. I've even noticed this with audio, really clean audio seems to survive conversion to mp3 much better than poorly recorded stuff.

Bob.

Geez, Vic, you made it sound like the second coming of Christ. I was expecting so much more. It's your average, good quality video.

Farss, "DVD from a BBC DigiBeta copy of Blue Planet", that had be fun ;-) .

"content" and a clean shoot is always King!

"fall apart woefully at 8Mb/sec" colorspace issues? Like bright colors that are moving in the colorspace spectrum with motion?

One of the "secrets" of high-end encoding, besides cutting scenes into smaller sections for encoding, is like using "layers" or tracks, where backgrounds are encoded at low bitrates and foreground objects are encoded at high rates.

"fall apart woefully at 8Mb/sec" colorspace issues? Like bright colors that are moving in the colorspace spectrum with motion?

You got it, blue lighting, dancers in satin, fog and stobes. Impossible to encode.
About you're other suggestion, one simple trick is wherever possible CK static backgrounds, even a locked off camera can move one pixel from vibrations and thermal effects. Then there's the possibility of shadows moving on the background.

Ah, I just tweaked, flash supports multiple layers with alpha, must look into this more.

Bob.