VP6 Codec Released! (was: this one is for doom9)

[aicjofs]

It does irk me that they rename AVI files to .VP6, I mean, it's an AVI file... if you rename it to VP6, it implies that you're using your own file format, which you're not... annoying.

You remember the .divx extension? I agree though.

Hey SeeMoreDigital, fancy seeing you here. How was the VP6 codec in Jove player with software decoding. Did it work. I created 2 VP6 encodes for codec comparisons, haven't had time to test it on Jove yet.

I like this codec but think it is best used if you want to save space, otherwise my personal opinion is that DIVX or XVID are still better solutions for overall quality when compared with encoding time. I couldn't really notice a large differnce between the three at low bitrates and the encode time were neglegible.

[shlezman]

I'd like to share some of my impression of VP6
I like to do my tests with QCIF sources for several reasons :
1. It's MUCH faster
2. It amplifies the artifacts
3. Cleaner sources
4. Harder to control the bitrate
5. In high bitrates most codecs look good (some better)
enough apologies :rolleyes:

The VP6 Rate control isnt much accurate 3%-5% overshoots.
The artifacts look more MPEGish then WMV/RM9/H.264
VP6 is better then any MPEG4 compliant (or semi-compliant) codec, with finer details and good object extraction.
VP6 is slower then any codec (except H.264)
Any claim of On2 that VP6 is better then WMV/RM9/H.264 is a fantasy.

[SeeMoreDigital]

Originally posted by aicjofs
... Hey SeeMoreDigital, fancy seeing you here. How was the VP6 codec in Jove player with software decoding. Did it work. I created 2 VP6 encodes for codec comparisons, haven't had time to test it on Jove yet. Hi aicjofs.

I'm sorry to report that 'it's a bust' on my PC! An 800MHz P3 is just not quick enough for 'forced' software decoding with the Xcard and JovePlayer (JP) - So it looks like a new motherboard and faster processor for me.... Xmas, is a comming!!!

I recieved an email from JP saying that WMV9 VCM.avi files will be supported in the next release. And that some of the issues users had with RV9.rmvb files (and hopefully their audio streams) will be resolved too! I've managed to watch RV9 files via a friends 2.8/P4 PC (with Xcard and JP) and they looked great when fed to the TV!

Maybe I should uplink an low bitrate VP6 file to JP, so this format can be included in a future release too!

Cheers

[Sagittaire]

Perfectly of were agreement with shlezman. The comparative test with the WMV9 an VP6 reached in CBR 1 pass ...? http://www.on2.com/pdf/vp6_white_paper.pdf



October 21, 2003 15

To Build The Test Clips For Windows Media
1. Downloaded the video compression manager from Microsoft.
2. Using VirtualDub (http://www.virtualdub.org/), we opened a raw clip and selected
Windows Media 9 as our compressor. (We also ran the same file through Windows
Media Encoder using settings similar to those used above to insure that we get the same
results with their standard encoder. We did get the same results.)
3. We opened the configure screen and selected the following options:
On the first tab, we did progressive frames. We did no resizing or cropping.
Advantages of TrueMotion VP6 Technology
On the Compression tab, we set:
§ Method = One-Pass CBR. ( We found that for these short clips two-pass provided
worse PSNR results.)
§ Bit Rate = We changed this parameter for each point in the graph.
§ Buffer Size = 5000 milliseconds
§ Smoother Motion - Sharper Image = 0 ( with it set at higher numbers we frequently
got frame drops which resulted in frames with a very poor PSNR. With the number at
0 we got every frame encoded and a much better PSNR number)
§ Decoder Complexity - Complex
§ Performance - All the way to the right towards Better Quality
§ Key Frame Interval - 99999. The clips used for testing are short we didn’t want
either codec to throw unnecessary keyframes.
4. Compressed the file.
5. Opened the file we just created in VirtualDub and selected compressor uncompressed.
6. Saved the file as a raw RGB file.
7. Ran the peaksnr.exe program on the file comparing it to the raw file.
8. Recorded the PSNR we got and the file size and plotted a point in the graph.


Moreover to have the best possible PSNR one needs desactiver Post-Processing of the WMV9 which is regulated to the maximum by defaut setting and which degrades much quality and the PSNR in weak bitrates.

The VP6 is a good codec but according to my test he is inferior compared with the other codec (XviD Devapi4, DivX 5.10, WMV9, RV9 EHQ)

[MemeX]

I have been playing around with a little 320x240 clip for a while, trying div3 fast/low motion, DivX5, XviD and yesterday with VP6. It's only 91 seconds and has a few still and "normal moving around people and pans", no audio, fairly clean source. I wanted to see how much I could compress this clip and it would look awful, but watchable and the subtitles properly readable. The size i was trying to hit was about 2.4mb.

VP6 really performed best imho and it was even 350kbs smaller then the rest, took a bit of tweeking to get the size up from 700kb on the first encode. It likes to blur things to avoid blocks, but at the same time keep details at "important" places, its pretty clever.

Compered to the DivX/XviD competitioners it had less blocks and about same detail and a lot less banding and smiring. Personaly I was very pleased with VP6 at this low bitrate I dont know about high-bitrate though.

MemeX

the only ridicules thing is that the audio would
be quite big compered to the video size

[Sirber]

can you try also with RV9 please?

[Sagittaire]

Here is RV9 EHQ vs VP6 VBR Best Quality

http://jfl1974.free.fr/Video/RV9-715.rar
http://jfl1974.free.fr/Video/VP6-715.rar

LOTR Trailer 640*272 715 Kbps 165 sec (compressibility 40% with XviD quant2)

[The People's Elbow]

I have some comments...
it is obvious that you did sth. very wrong with the advanced settings in vp6, because the quality fluctuation is much too high. Plz post the settings, either, and I'll keep on commenting... rv9 looks just like its supposed to do... no clear artifacts, but sometimes very noticable smearing. There are scenes in which vp6 looks better though the overall appearance is worse due to the variance of quality.

EDIT: With pp and noise enabled even this vp6 clip is sort of competitable.

greet,
Elbow!

[Human_USB]

I get great Anime encodings with VP6. I used VMD 1.5.4.1 (Build 2178) with VP6 Anvanced Profile. Here is a list of what I did;

Good Quality Fast Encoding

Target Daterate 100
Noise Reduction 0
Autokeyframe ON
Max Frames Btw Keys 320
Undershoot 90
Adjusted Quantizer ON
Minimum 1
Maximum 15
Temporal Resampling ON
Down watermark 20
Spatial Resampling ON
Down watermark 35
Up watermark 20


Out of this I got a 138 MB avi. If you wish to try it use RahXephon DVD 1 episode 2.

[trbarry]

I still have not done any really scientific testing but I think for my purposes of making 1280x720p @30 HDTV clips this codec is more or less competitive with either Xvid or WM9.

However I'm still concerned with the playback speed, decoder complexity. It seems to play faster than WM9 but not as good as Xvid. So I think I could only use it for 720p@24 clips on my current machine.

The playback performance is one of the reasons I've stuck with Xvid over WM9 so far. But once hardware HDTV players come out then I guess all bets are off and I may switch to whatever arrives first.

- Tom

[Sagittaire]

@ trbarry

For playperformance use this reg for desactived the PP of WMV9. With 0 value playperformance is very good and quality is better ...
http://jfl1974.free.fr/Video/PP-WMV9.reg


Windows Registry Editor Version 5.00

[HKEY_CURRENT_USER\Software\Microsoft\Scrunch]
"Force Post Process Mode"=dword:00000000


@ All

Here my profil for VP6 codec.
http://jfl1974.free.fr/Video/Two%20Pass%20-%20First%20Pass%20-%20Perso.vps
http://jfl1974.free.fr/Video/Two%20Pass%20-%20Second%20Pass%20-%20Local%20File%-%20Perso.vps

[C0mPr355]

Originally posted by Blight
I think the whole point of VP6 is that they write their own code so that they don't have to pay royalty for any external patents.

That would make it their own closed format (like WMV9?)

It does irk me that they rename AVI files to .VP6, I mean, it's an AVI file... if you rename it to VP6, it implies that you're using your own file format, which you're not... annoying.

Also, someone contact gabest and tell him this format doesn't work with his AVI Splitter properly.

There is no reason why you cannot save your file as an avi. It does not have to be saved as .vp6. That is only on our website.

[Atamido]

Originally posted by C0mPr355
There is no reason why you cannot save your file as an avi. It does not have to be saved as .vp6. That is only on our website. Shouldn't you be saving it as MKV? :p

[The People's Elbow]

@Sagittaire:
Would you mind to try the lotr trailer with the following settings from the "advanced" tab?

- undershoot: sth. between "90" and "100" depends on your bitrate setting to match rv9
- adjust quantizers marked and set to min: "1" and max: "32"
- no temporal or spatial resampling
- variability: "90"
- min section: "0"
- max section: "1000"

I'm eager to know what the trailer would look like with those settings (best quality setting... though I know it's damn slow ;) )

greetz,
Elbow

[C0mPr355]

one more thing...

-for all of you compressing in VD or whatever app. please make sure that that application's form of autokeyframing is disabled with the use of 9999. In other words, VD has its own "target datarate" and "force keyframe every" dialogue box before you go into the configure menu of a codec. Set the "force keyframe every" to 9999 to ensure that VP6 is the only thing generating the keyframe list.

-"adjust quantizer" should be checked if you want the rate control to work properly. If it is not checked then the compressor will run in a fixed quantizer mode using the maximum value entered into the fields. If you uncheck the box and the highest value between the min and max is 4 then the compressor will compress everything at 4. If you want a range then highlight the box and enter two diff quant values. Lower values = better quality but harder to hit the datarate, higher values = worse quality but easier to hit datarate. Of course if your datarate is something in the area of 4Mbs or something high it really shouldnt matter all that much how low you bring the quantizer.

-remember that higher datarate clips SHOULD NOT be using spatial resampling. The reason for this is that the clip has been downsampled in order to hit the datarate and then played back upsampled. Think about taking a really nice picture at 320x240 and then playing it back at 1024x768...it would look pretty jagged and not very pleasing. SR is recommended for lower bitrate encodes such as 56K or low bandwith.

-For short clips the pre-buffer, optimal buffer, and max buffer settings will affect datarate. For CBR the data rate control attempts to insure that the buffers will never run empty provided that the connection is able to support the specified bitrate. If you set a 4 second pre-buffer and then code a 10 second clip it is possible, therefore, to significantly overshoot the target rate. On a 10 minute clip, however, an overshoot of 4 seconds worth of data is negligable.

- VBR and two pass are not specifically made for short clips (i.e a few seconds) as the VBR works on a scene by scene basis. On longer clips they should get close to the target avg bitrate unless the max Q is set too low.

- If you wish to do a two pass encode of a movie I would recommend that a 38-40 Q be set. Again do not select spatial resampling or temporal resampling.

[trbarry]

Of course if your datarate is something in the area of 4Mbs or something high it really shouldnt matter all that much how low you bring the quantizer.

ComPR355 -

Assume you mean that for DVD resolutions. For my 1280x720x30 tests I don't think 4 mbps would be considered all that high, even for VP6. With HDTV there's probably some expection it will be viewed sitting < 3.5 screen heights from a large hi-rez display and any artifacts will be more visible.


-remember that higher datarate clips SHOULD NOT be using spatial resampling. The reason for this is that the clip has been downsampled in order to hit the datarate and then played back upsampled. Think about taking a really nice picture at 320x240 and then playing it back at 1024x768...it would look pretty jagged and not very pleasing. SR is recommended for lower bitrate encodes such as 56K or low bandwith.

Can you say by what ratio the clip is internally downsized when needed? 1/4 rez?

When I'd previously considered things like this I thought I could drive it off the quantizers that would otherwise be used, say start to downscale when the quants would otherwise be > 7-10 or so, at least as used in MPEG. The idea would be for things like my HDTV football clip where sometimes the needed bit rate goes through the roof when you do a fast pan over crowds. To me that would be a good candidate for spacial resampling, even at otherwise high bit rates.

Maybe it turns out that's not best but could you describe a bit more how the values for spacial resampling are used? It seems to be driven off buffer usage instead of quants but I'm afraid I don't really understand exactly how.

- Tom

[Human_USB]

Well..... I think I found a problem. When I mux VP6 into a OGM with a srt file, the SRT file will not play. I muxed the same SRT into a OGM with a XVID file and it played, so the SRT is good. Can other people please test this.


*EDIT*
I get the same problem with MKVs too.

-Thanks
Jason

[Kast]

I'm very impressed by this codec. I encoded Lion King at 1650 Kbps one pass with Divx and then at 170 Kbytes (min 3 max 30) with VP6. Final sizes were 1015MB for VP6 and 1150 for Divx. In high motion scenes DivX simply could not be compared to VP6. I mean after the high motion scenes and the following seconds everything was pixelated with Divx. On the contrary with VP6 there was nothing! (i loaded the same avs file for both) The time with VP6 was 3 hours more but on the other hand who cares for three lousy hours when i'm encoding for 9 hours? ;)

[Sirber]

you should try 2 pass VBR, you'll get the best of both codecs then.

[Bulletproof]

I must be doing something wrong here since there are a considerable amount of people saying that VP6 is impressive. I tried it today and it gave me absolutely horrible results. I tried the One-Pass Best Quality option and set quantizers to 2-31 and left the other options off. I left the keyframe options to their defaults. The result that I got was really blurry and resembles a WMV9 or RV9 encode. I did try forcing to quantizer 2 and that made it look ok but it of course made the filesize huge and didnt offer any advantage over xvid, and the video still looked slightly blurry.

[Tommy Carrot]

Originally posted by Bulletproof
I must be doing something wrong here since there are a considerable amount of people saying that VP6 is impressive. I tried it today and it gave me absolutely horrible results. I tried the One-Pass Best Quality option and set quantizers to 2-31 and left the other options off. I left the keyframe options to their defaults. The result that I got was really blurry and resembles a WMV9 or RV9 encode. I did try forcing to quantizer 2 and that made it look ok but it of course made the filesize huge and didnt offer any advantage over xvid, and the video still looked slightly blurry.

Yupp, i also wasn't too impressed. There is simply no way to ensure that the quality is the same at all scene. It's constantly changing, somewhere better, but somewhere horrible. I would't say it's significantly better than mpeg4.

[Prettz]

Originally posted by MemeX
I have been playing around with a little 320x240 clip for a while, trying div3 fast/low motion, DivX5, XviD and yesterday with VP6.
Err, why would you use divx3 fast/low motion? Divx3 cannot even be called a real codec unless you use something like Nandub SBC. Comparing anything to the fast/low motion codecs is a waste of time.

[Kast]

Originally posted by Bulletproof
I must be doing something wrong here since there are a considerable amount of people saying that VP6 is impressive. I tried it today and it gave me absolutely horrible results. I tried the One-Pass Best Quality option and set quantizers to 2-31 and left the other options off. I left the keyframe options to their defaults. The result that I got was really blurry and resembles a WMV9 or RV9 encode. I did try forcing to quantizer 2 and that made it look ok but it of course made the filesize huge and didnt offer any advantage over xvid, and the video still looked slightly blurry.

Max keyframes to 360 maybe? It's C0mPr355's suggestion and it worked fine with me. Sirber you won't even imagine how long does it take on my computer to do 2 pass encoding. I wish I could...

[Enigmax]

FANTASTIC!
"The Two Towers" in a single CD and resolution 640 x 256.
I have used VP6 Advanced profile, with byterate of 60 KBytes per second (480 kbps). Forced activated Keyframe and a value of 360. The rest of defaults. Nothing to activate resampling.

The result is very similar to the same codification made with DivX for 2 CD whit 1000 kbps.

O.K. :)

Greetins

[Sagittaire]

.... ????!!!!

[Ramirez]

The result is very similar to the same codification made with DivX for 2 CD whit 1000 kbps.

Are you sure about that? , it's strange because my last VP6 encoding attempt turned out as a complete disaster: 850KBPS.720x329 (I've deleted this smearing \blocking blurng mess: /24 hours of encoding time) however now I'm taking into account a few valuable pointers given by compr355 and other members so I'm starting another encoding session (I hope it'll be way better then my first one. :P )

[Enigmax]

VP6 surpasses to other codecs in compressibility.
It works to little bitrate far better that DivX or XviD.
In High o normal bitrates, I prefer to XviD.

Greetings

[MemeX]

I have a question regarding VP6 and smartphone support. As VP6 seems perform well on high compression it might be well suited for playback on smartphones like the nokia series 60 mobile phones.

Would it be easy to make a VP6 player for the phones? it might be interesting for streaming videos in the futur..

[SeeMoreDigital]

Originally posted by Enigmax
VP6 surpasses to other codecs in compressibility.
It works to little bitrate far better that DivX or XviD.
In High o normal bitrates, I prefer to XviD... There's no doubt that VP6 generates very impressive looking encodes at 'constant' low bitrate speeds. Which is what it was designed for!

However, I've generated quite a few 2pass VBR encode tests using VR6 (as have many other forum members) and when compared to say RV9, it does not come close!

To my eyes VP6 produces very visible blocking. Even the VP6 versus WMV9 encodes (and the VP6 only Xmen trailer) that are available for download on the VP6 website show this to be the case. And whoever generated the WMV9 encodes, could not have done a better job of making them look bad!

Cheers

[MemeX]

Originally posted by Sirber
can you try also with RV9 please?

sorry to have taken this long, its rather by chance that I was able to make an encode with RV9 to compere it to VP6.

The quality was also impressing, for example the white subtitles were nicely readable and a little cleaner around the edges then with VP6. I did notice that the impression of faces were a little flat with RV9 tho. VP6 had a little more lets call it noise for now, which made the faces look more human and natural..then at other times RV9 gives a better overall impression. I also noticed a pumping effect of keyframes on VP6, but that might be because of my settings, force keyframe every 120 was probably the default back then instead of 9999.

i further investigated my source file, its a tv capture so its not as clean as a dvd source which both codecs had to fight with a little so I used a little noise reduction from the codecs' settings. I can't remember the exact settings for VP6 i used but it was probably what the vphelpfile on their website suggested and then changed the quants to hit my filesize.

DivX/XviD performed rather bad on this file, as i mentioned before and i think that VP6 is definatly competible to them at least on lower bitrates from my "little" experince and the price is attractive for home users ;)

MemeX

[Blueseb]

I encoded anime for my first test with vp6, advanced profile two passes. accidentally I leaved empty the target data rate field. the result was a full episode (about 23min, 640x480) compressed in 16MB !!! and the video is still good to watch (it has a lot of blockness of course, but the only fact it is watchable it's a miracle).
I mean: I'm talking of 0.016 BPP per frame, 95 kbit/sec, in xvid this should have been a 9% of the first pass size but I think it is not even possible to do with it. If I didn't saw it I wouldn't believe.

Maybe this codec sucks on high bitrate or with normal movie stuff, I dunno; but at extreme low bitrate it's impressive.

Worth to mention the encode time (1st+2nd) took ~20 hours on a duron900 :p does anyody cares??? ;)

[C0mPr355]

-a quick note to some of you that are compressing with non-multiples of 16 resolutions: Try using multiples of 16 for a better output :D

-and about the difference between 60 and 61. Basically 60 is designed to run a little bit faster (with slight downgrade in quality) as opposed to Vp61 which is more processor intensive, better looking, and slower on playback.

For the two pass encoding results I would recommend tweaking the percantages of min and max datarate variablilty.

And yes please use a auto-keyframe generation of 360 rather than 120. And make sure that this is set in the configure dialogue box of Virtual Dub NOT in the original datarate/keyframe dialogue box that virtual dub has.

[C0mPr355]

Originally posted by SeeMoreDigital
There's no doubt that VP6 generates very impressive looking encodes at 'constant' low bitrate speeds. Which is what it was designed for!

However, I've generated quite a few 2pass VBR encode tests using VR6 (as have many other forum members) and when compared to say RV9, it does not come close!

-Please give specs of all these encodes.

To my eyes VP6 produces very visible blocking. Even the VP6 versus WMV9 encodes (and the VP6 only Xmen trailer) that are available for download on the VP6 website show this to be the case. And whoever generated the WMV9 encodes, could not have done a better job of making them look bad!

-If you are seeing the blockyness then perhaps the deblocker is not on or has been turned off. Windows media, when the deblocker is not on, looks rather blocky as well. I am not saying that one is better than the other I am simply stating a fact. Also of note, the Xmen trailer that is up there right now was not encoded with the newest vp6 codec that we have released to you. That one is about three months old.

-As far as the WMV encodes, perhaps you are not getting ( as stated above) the deblocker for WMP. When creating the encodes of those clips (football, cheerleader, etc)...they were done with default style settings. No third party tool was used, only Windows Media Encoder. Perhaps this may be why they do not look so good (or the blocking issue above). If you can get a better result from the stock WME then please post your complete encode parameters.

I am curios. <---mi enternet splleing is teh gud. :D



Cheers

[Ramirez]

Well, my impression on VP6 so far (that's might change) I wouldn't recommend it for a DVD backup ATM, I agree with SMD, this codec simply producing none satisfactory blocky and most of the time blurred output. (No, I don't think that artifacts hiding Post-filter is the answer :rolleyes: )

Perhaps I'm doing something wrong, I dunno (any suggestions are welcome) I've used very clean source for my test encodes and I tried various settings.

My last encoding settings as follows:

Source: Ghost-Ship /duration: 1:26:58
Rez: 720:396
Bitrate: 825KBPS

Codec Settings
VP6 Advanced
2 Pass -VBR
AutoKey: ON/ 360
Undershoot: 90
Adjust Quant: ON
Min-4 /Max: 32
Temp R: OFF
Spatial R: OFF
Variability: 90/ Min: 0 Max: 1000

BTW: same movie encoded with Xvid (same bitrate, rez), looks simply great. ;)

[Sirber]

What about RV9? ;) (No flaming please, it was a joke :D)

[SSIXS]

My first impression of VP6...looks pretty darn good for my use. I spend quite a bit of time away from home for work and I take DVD backups with me as not to damage or lose the originals. Since these are just backups, I dont get all wrapped up in having the most pristine "DVD quality" but I still want em to look good. I dont like to spend what little time I have at home trying to figure out the "optimum settings" for each and every movie. As a rule (due to time constraints) I generally do single pass encodes...set it up just before I goto bed and have it ready to burn when I get up. I took my latest dvd purchase and made a couple test encodes with VP6 facing off against divx 5x, wmv9 and xvid. I must say, using single-pass CBR, VP6 came out looking great compared to the others...wasnt quite as blocky or blurred and handles fast motion sequences quite a bit better.

Source: My Science Project/duration: 1:34:29
Rez: 640:272
Bitrate: 88 Kilobytes/second

Codec Settings
VP6 Advanced
1 Pass - Best Quality - CBR
AutoKey: ON/ 360
Undershoot: 90
Adjust Quant: ON
Min-4 /Max: 20
Temp R: OFF
Spatial R: OFF
Peak Bitrate: 100
Prebuffer: 4
Optimal Buffer: 5
Max Buffer: 6

I just wanted to thank On2 and all the folks that had a hand in the releasing of this codec to us home users. Keep up the good work !

Derek

[Tommy Carrot]

Originally posted by Ramirez
Well, my impression on VP6 so far (that's might change) I wouldn't recommend it for a DVD backup ATM, I agree with SMD, this codec simply producing none satisfactory blocky and most of the time blurred output.

Exactly this is my experience. For higher bitrates, this codec doesn't compete well against xvid or divx. But for low bitrates (<400 kbps), it's indeed much better. Too bad i never encode at those bitrates.

[SeeMoreDigital]

Originally posted by C0mPr355
... Please give specs of all these encodes. Well, the test I think you will be interested in is my 'video only' StarWars 2 Chapter 41 test. Which is 377.24 seconds long and consists of 9431 frames!

Over this past year I have used this source to test just about all the codecs currently available. Using just about every available setting!

My goal, as with all my tests, is to generate 720x576 'anamorphic frame' and 640x272 'crop and resize frame' encodes.

The encodes are generated using 1pass CBR and 2pass VBR settings. And must be no greater in file size than 28.5MB

By following this standard proceedure I am then able to compare any new encode against any of my other encodes.

I noticed that your 'cheer leader' WMV9 encode has an image pixel frame size of 352x240. Can you confirm if this was done in one pass or two. The same goes for your VP6 encode?

I'm not knocking the codec. As I said before I think it generates very impressive looking encodes at 'constant' low bitrate speeds. Much better than DivX and WMV9, it's even very close to RV9 in this regard. But not RV9 EHQ!

Hovever, I have found that when generating 2pass VBR encodes it's a very different story!

If you would like me to send you a CD-R with some of my test files on it, please PM me with you details.

Cheers

[Sirber]

Originally posted by SeeMoreDigital
I'm not knocking the codec. As I said before I think it generates very impressive looking encodes at 'constant' low bitrate speeds. Much better than DivX and WMV9, it's even very close to RV9 in this regard. But not RV9 EHQ!Nice :D All I wanted to know :cool:

[SeeMoreDigital]

Originally posted by Sirber
Nice :D All I wanted to know :cool: Yes.....

Whenever I post a comment about RV9, I swear I can hear you cheering from Canada. Which is quite a distance from Nottinghamshire, England!

Did I tell you, that I can now view RV9 .rmvb and WMV9 .avi and .wmv encodes on my TV via my Xcard with JovePlayer..... very nice!

My favourite is still 2pass VBR DivX though!

Cheers

[C0mPr355]

I understood that you were not knocking things...just merely giving your feedback. No need to worry. Like I said I was just curious that's all.

The vp6 cheer encode was done in single pass many moons ago (back in june or earlier I believe). The wmv encode was done in 2-pass encode.

**VP6 is our first codec with two pass in it...and as such we are new to this whole arena. In that respect we would love to get feedback on how to improve said codec in the two pass mode. So please keep the feedback constructive and well thought out...not just "my results suck, or my results are crappy..." Tell us what would make two pass better.

and SMD I will get in touch with you via PM :D

[MemeX]

Originally posted by Prettz
Err, why would you use divx3 fast/low motion? Divx3 cannot even be called a real codec unless you use something like Nandub SBC. Comparing anything to the fast/low motion codecs is a waste of time. I was talking about the nandub SBC.

[hanz_giebenrath]

whats the matter guys?..no more tests and experiences? ;)

[SeeMoreDigital]

Originally posted by hanz_giebenrath
whats the matter guys?..no more tests and experiences? ;) Hi Hanz. Welcome to the forum.

You will find more posts about VP6 if you look in this thread -

http://forum.doom9.org/showthread.php?s=&threadid=64789

Cheers

[SeeMoreDigital]

Hi C0mPr355,

Can I ask you, if any 'goals' have been discussed for this codec yet - other than streaming to cell phones/PDA's?

Personally, I think it would be great to see the codec appear in hardware players (once VBR lives up to the CBR performance)!

And are there any thoughts about using other containers apart from AVI?

Cheers

[C0mPr355]

SeeMore,

Our press releases pretty much speak for themselves on our goals for the company (other than getting everyone to try out the codec ). We have worked with several things other than PDAs over the past couple of years, mainly settop box integration for things such as digital cable etc.

I would love to see this in more hardware as well as that would mean that we could compete a little better with higher end companies like microsoft, real, etc.

It is hard to compete when you are relatively new and a small fish in the sea :D But we will keep pushing to get our codec better and better.

As far as containers are concerned I believe that we will be sticking with the avi format for now, and use the .vp6 for specific projects that relate to our truecast player. Avi, however, seems to be the most valid format for our uses right now though.

[GIR]

Just found this interesting article on The Register: "China unveils 'DVD killer' video disk format" (http://www.theregister.co.uk/content/54/34097.html) China using VP6 for EVD!

EVD uses On2's latest video codec, VP6, which offers "better image quality and faster decoding performance than Windows Media 9, Real 9, H.264 and QuickTime MPEG-4," the company - formerly known as The Duck Corporation - claims.

[Sirber]

that's funny. "better image quality". LOL!!!

(personnal opinion)

[Tommy Carrot]

They said exactly the same about VP4. I think the marketing guys didn't try their own codec. :D

[SeeMoreDigital]

DAvenger has started a thread about the same subject here too: -

http://forum.doom9.org/showthread.php?s=&postid=401824#post401824

Cheers