VP7: On2 Just Won’t Quit

I was just thinking the other day about whether On2 would release VP7. This is the organization that brought us TrueMotion 1, TrueMotion 2, VP 3.0, VP 3.1, VP 4.0, VP 5.0, VP 6.0, and VP 6.1, not to mention VP 6.2. VP70 just seems like a natural progression.

What does ‘VP’ stand for anyway? I have never read any definitive answer. However, if On2’s technology focus is any clue, it stands for “Video Predictor”, owing to On2’s almost religious devotion to prediction-based compression algorithms.

Michael Roitzsch tipped me off that On2 has, indeed, unleashed their VP7 codec upon the world. True to form, their literature asserts that VP7 “rUleZ!!1!” and that “MicRo$$oFt is teh suck!1!!!” (slightly paraphrased, they state their points using graphs).

Anyway, VP7 sample files are here. Do not be fooled by the .tc8 extensions; they are just plain AVI files (they have even dropped the special start-of-file signature that the .tc7 files used to have). Further, they apparently have a 10-day trial decoder available.

And they also have their obligatory whitepaper (PDF file). Some of the highlights:

  • TrueMotion VP7 is the best video compression technology currently available
  • Unmatched image quality at all image sizes and for all types of material
  • VP7’s visual quality bests even the outstanding VP6 codec
  • VP7 surpasses all other video compression technology
  • VP7 uses a set of highly-tuned heuristics to produce the best possible results
  • Currently the best video codec on the market– by 50% or more
  • metrics show better compression than all competing technologies
  • better than Microsoft Windows Media 9 running in its fastest mode in both frame count and frame quality

Goodness! I haven’t read this much self-promoting pap since I perused On2’s VP6 whitepaper. Okay, that’s enough of the “we rule” rhetoric. Some quasi-technical details:

  • can play back 1920×1080 HD material on 2.5 GHz machine
  • specially designed SIMD friendly loop filtering
  • new 4×4 DCT transform technology (aha! a clue– 4×4 DCTs just like H.264)
  • sophisticated second-order DC encoding
  • block adaptive quantization and filtering (so the quantizer matrix and filter coefficient set can be selected on a per-block basis, rather than, say, a per-frame, per-plane, or per-slice basis)
  • comes in 3 different profiles: hi-def, simple, and advanced; this likely means that there are VP70, VP71, and VP72 identifiers for the different levels, similar to the VP6 series
  • another prominent marketing point is that VP7 allegedly requires less code to implement than, say, H.264; this is encouraging for reverse engineering effort

There is plenty more in the whitepaper. Nothing terribly ground-breaking, just the same family of algorithms as always. Our job as multimedia hackers is to figure out how the same algorithms were mixed and matched to encode a particular bitstream format.