FlashComGuru Home Imediasee Influxis Uvault Akamai
                                                                                       Forum Index | Active Topics | Register
                                                                                                          List Overview | List Archives
                                                                                                                           About this site | Advertise
 

home

Adobe AIR (8)
Applications (32)
Books & Training (9)
Collaboration (9)
Components (6)
Events (50)
Flash Player (10)
Flex (29)
FMS (70)
General (109)
Hosting (5)
Jobs (13)
Off topic (31)
Press Releases (9)
Site Check (8)
Tools (41)
Videos & Players (54)

Flash On the Beach

Alltop, all the cool kids (and me)

 
As you probably already know, Macromedia today announced Studio8 which includes a massive new release of Flash. One feature which has already been discussed widely is the support for a new video codec: ON2's VP6 codec.
To give you an idea of the improvements that this codec will bring I have put together a little sample. The video below consists of 2 parts, one part is encoded with the existing Sorenson Spark codec of Flash MX 2004, the other one uses the new VP6 codec. Turn the mask off to see VP6 only or enable the mask to see VP6 side by side with Spark. This application uses progressive download.
The demo works best once all files are cached to disk and you may want to let it download before viewing it again to get a better sync between the clips.

NOTE: you need the Flash 8 Player to view this video. If you view this using Player 7 you will only see the Spark version!


The filesizes between the 2 versions do differ a bit with the VP6 version being a little heavier.
Considering the quality improvement I think this is no big issue and if you had a target filesize in mind you could easily meet it and still exceed the Spark quality by a mile.

Comments
[Add Comment]
wow yup that looks indeed a lot better.

Hummm ok you are aware of the bigger filesizes, so let's look at the relative difference in filesize:
56 kbps: sparc (496 KB) vs VP6 (676 KB) = ~36% bigger
330 kbps: sparc (3,187 KB) vs. VP6 (3,849 KB) = ~21% bigger

Now that learns that the VP6 versions are ~30% bigge ron average.
And this is hard to say, but is the VP6 version on 'visually' about ~30% bigger -at least- ?

What encoding settings did you use?
# Posted By Rob Coenen | 8/8/05 1:45 PM
Interesting choice of a video clip but it's a little too dark for me to make a quality judgement. I wasn't able to see any difference between the masked and unmasked settings.
# Posted By mark mcglinch | 8/8/05 2:08 PM
What do you mean by visually bigger? If you got Player 8 you can see for yourself what each version looks like... that's the whole point of this demo. If you would adjust the encoding bitrate of the VP6 version in order to achieve the same filesize as the Sparc version then the VP6 would qualitatively still be *a lot* better.

If I find time I will post another sample using target filesizes. For the time being I think this one demos the new codec pretty well.

Stefan
# Posted By Stefan | 8/8/05 2:09 PM
Okay...reloading your page helped see the split screen
# Posted By mark mcglinch | 8/8/05 2:10 PM
uhmm.. I by ' visually bigger' I mean 'visually *better*'
you know.. as in.. looks better!
# Posted By Rob Coenen | 8/8/05 2:18 PM
ok, so don't you agree that the VP6 version is at least 30% better? In any case, I feel a target filesize comparison coming up...
# Posted By Stefan | 8/8/05 2:28 PM
I do agree with that. However if I have my files 30% bigger using some new hitech codec, I would expect nothing but at least 30% better visually imporvement.

OK - I know that increasing the filesize with 30% doesn't have to mean the video will look 30% better too.
First of all you can't judge something like 'visually better' in exact percentages, and secondly there are many other encoding factors that influence the end result / size.
But that is why I aksed for which exact encoding parameters were used.

It looks good for sure, can't wait to get my hands on the encoder myself :-)
# Posted By Rob Coenen | 8/8/05 2:36 PM
yeah I am unable to comment on the tools used right now but what I can say is that I used identical settings for both formats, all I did change was the codec.
# Posted By Stefan | 8/8/05 3:31 PM
While the VP6 video has much less artifacting compared to the Spark video, the Spark shows more surface detail.
( VP6 looks almost "smeared" compared to Spark )
# Posted By TheDingo | 8/9/05 3:24 PM
Breakdown:
Spark vs. VP6.2
same quality: VP6.2 will be 1/2 the file size
same file size: VP6.2 will be 2x the quality

obviously that's not exact. There is some cost to higher quality, 2x is a loose number.
# Posted By ed | 8/11/05 8:09 PM
Hi,
fun yours videos
where you recovered them?

Yaya
# Posted By Yaya | 8/24/05 12:26 PM
I have to add my bit.. you can't really compare these 2 videos in my mind..
the different file sizes makes all the difference... more data better quality!

there are some really interesting points pertaining to the VP encoder here..

http://www.kaourantin.net/
- read the article on...
"Performance traps in Flash Player 8" (relates to the new codec)

written by the Principal Engineer on the Macromedia Flash Player.

Mike
# Posted By Mike (wainui) | 8/25/05 1:05 AM
This is not a scientific experiment... But to put your mind at ease, check this example which has near identical filesizes: http://www.flashcomguru.com/index.cfm/2005/8/9/One...
# Posted By Stefan | 8/25/05 8:51 AM