Second Annual MSU MPEG-4 AVC/H.264 Video Codec Comparison

For professional users and codec authors

MSU Graphics & Media Lab (Video Group)

Return to second H.264/AVC comparison home page!

Part 3. "Best speed" preset

Take a look at this article on the new site! Follow the link
https://videoprocessing.ai/codecs/mpeg4-avc-h264-2005-part3.html

Contents

  • Part1. Introduction
  • Part2. "Best quality" preset
  • Part3. "Best speed" preset
  • Part4. Additional information and conclusions

    • Preset "Best speed"

    Main codecs' objective in this mode is the fastest video compression with satisfactory quality. Therefore main graphs in this section are graphs that show compression speed.

    Y-PSNR

    However, let's consider Y-PSNR graphs at first.

    Graph type

    Total number of graph

    Inserted in this document

    Average metric

    196

    7 (3.5%)

    Picture 1.   Y-PSNR. Sequence "foreman"

     

    Picture 2.   Y-PSNR. Sequence "susi"

     

    Picture 3.   Y-PSNR. Sequence "bbc"

     

    Picture 4.   Y-PSNR. Sequence "battle"

     

    Picture 5.   Y-PSNR. Sequence "simpsons"

     

    Picture 6.   Y-PSNR. Sequence "matrix"

     

    Picture 7.   Y-PSNR. Sequence "concert"

     

    Speed/Quality Tradeoff

    Here are the most important graphs for this preset. They show codec's work time (abscissa axis) and quality of sequence, compressed with it (axis of ordinates). If codec A is to the left of codec B and above it then codec A is better than codec B; if to the right and below it then on the contrary, codec B is better than codec A. In other cases nothing can be said about definite advantage of one's codec compared to another's.

    Bit rates in cutlines inside this section are target bit rates (not real codec bit rates).

    Graph type

    Total number of graph

    Inserted in this document

    Speed/quality tradeoff

    1960

    8 (0.4%)

    Picture 8.   Codec comparison with at speed/quality tradeoff graphs

     

    Picture 9.   Legend for Speed/quality tradeoff graphs

     

    Sequence "susi", Y-PSNR

    Picture 10. Bitrate 100 kbps

     

    Picture 11. Bitrate 340 kbps

     

    Picture 12. Bitrate 938 kbps

     

    Picture 13. Bitrate 2340 kbps

     

    Sequence "battle", Y-PSNR

    Picture 14. Bitrate 100 kbps

     

    Picture 15. Bitrate 340 kbps

     

    Picture 16. Bitrate 938 kbps

     

    Picture 17. Bitrate 2340 kbps

     

    Bitrate handling

    Consider graphs showing how accurately every codec kept requested bit rate. Grades on x-axis show target bitrates (point 0 is 100 kbps, point 9 - 2340 kbps). Grades on y-axis show how many times codec exceeded declared bit rate (real bit rate to predetermined bit rate ratio).

    Picture 18. Bitrate handling. Preset "Best speed". Sequence "foreman"

     

    Picture 19. Bitrate handling. Preset "Best speed". Sequence "susi"

     

    Picture 20. Bitrate handling. Preset "Best speed". Sequence "bbc"

     

    Picture 21. Bitrate handling. Preset "Best speed". Sequence "battle"

     

    Picture 22. Bitrate handling. Preset "Best speed". Sequence "simpsons"

     

    Picture 23. Bitrate handling. Preset "Best speed". Sequence "matrix"

     

    Picture 24. Bitrate handling. Preset "Best speed". Sequence "concert"

     

    Conclusions:

  • Situation here is similar to "Best Quality" preset - DivX keeps bit rate very bad, raising it too high on low bit rates.

  • The best among H.264 codecs in bitrate handling is Fraunhofer IIS, as before.

    •

    Time

    Now let's consider graphs that show time spent for each video sequence encoding in "Best Speed" preset and average normalized time for this preset.

    Picture 25. Encoding time. Preset "Best speed". Sequence "foreman"

     

    Picture 26. Encoding time. Preset "Best speed". Sequence "susi"

     

    Picture 27. Encoding time. Preset "Best speed". Sequence "bbc"

     

    Picture 28. Encoding time. Preset "Best speed". Sequence "battle"

     

    Picture 29. Encoding time. Preset "Best speed". Sequence "simpsons"

     

    Picture 30. Encoding time. Preset "Best speed". Sequence "matrix "

     

    Picture 31. Encoding time. Preset "Best speed". Sequence "concert"

     

    Picture 32. Normalized average encoding time. Preset "Best speed"

     

    Conclusions:

  • Absolute leader by speed is codec from ATI company. At the average, it worked several times faster than its nearest competitor - VSS codec. And at the same time it is not worst one by the quality, though it is one of trailing codecs.

  • The slowest codec in this mode again is Fraunhofer IIS. It lags behind its nearest competitor more than 3 times.

  • Encoding time for Fraunhofer IIS codec increases noticeably when bit rate is increased.

    •

    Contents

  • Introduction
  • Thanks
  • Overview
  • Goal and testing rules
  • Sequences
  • Codecs
  • "Best quality" preset
  • Y-PSNR
  • U-PSNR, V-PSNR
  • SSIM, VQM
  • Blurring measure, blocking measure
  • Bitrate handling
  • Time
  • "Best speed" preset
  • Y-PSNR
  • Speed/Quality Tradeoff
  • Bitrate handling
  • Time
  • Final part
  • 2-pass mode and High profile
  • Per frame comparison
  • Visual comparison
  • Informal codec comparison
  • General conclusions

    • Download

    Contact Information

    Additional information

    See all MSU Video Codecs Comparisons

    MSU video codecs comparisons resources:

    Other materials

    Video resources:

    3D and stereo video
    Projects on 3D and stereo video processing and analysis
    MSU Video Quality Measurement tools
    Programs with different objective and subjective video quality metrics implementation     		Codecs comparisons
    Objective and subjective quality evaluation
    tests for video and image codecs
    Ext. link: x264 parameters efficiency comparison
    Public MSU video filters
    Here are available VirtualDub and AviSynth filters. For a given type of digital video filtration we typically develop a family of different algorithms and implementations. Generally there are also versions optimized for PC and hardware implementations (ASIC/FPGA/DSP). These optimized versions can be licensed to companies. Please contact us for details via video(at)graphics.cs.msu_ru.     		Filters for companies
    We are working with Intel, Samsung, RealNetworks and other companies on adapting our filters other video processing algorithms for specific video streams, applications and hardware like TV-sets, graphics cards, etc. Some of such projects are non-exclusive. Also we have internal researches. Please let us know via video(at)graphics.cs.msu_ru if you are interested in acquiring a license for such filters or making a custom R&D project on video processing, compression, computer vision.
    Video codecs projects
    Different research and development
    projects on video codecs     		Other
    Other information
    Last updated: 12-May-2022

    Server size: 8069 files, 1215Mb (Server statistics)

    Project updated by
    Server Team and MSU Video Group

    Project sponsored by YUVsoft Corp.

    Project supported by MSU Graphics & Media Lab