Creative Communities of the World Forums

The peer to peer support community for media production professionals.

Activity Forums Compression Techniques MPEG-2 problem

  • MPEG-2 problem

    Posted by Tom Brooks on September 22, 2006 at 7:28 pm

    I have a DV sequence consisting of mostly a talking head in front of a dark-ish cyc that was lit with a subtle wash of blues and purples. When compressed for DVD I get a pretty noticeable ‘bump’ based on the GOP length of 15 frames. It’s bad in the background, but not apparent at all on the talent’s face. Is there anything I should be doing in post (including compression) to avoid this, or is it inevitable with this type of source? I’m using all latest PPC versions of FC Studio and Quicktime. I’ve tried various settings in Compressor 2.1, but the ‘bump’ remains in all of them. Currently, I’m using the standard 90 Min Best Encode 4:3 with everything left on automatic.
    Thanks for any revelations you send my way.
    -Tom

    Final Cut Studio, FCP 5.1.1, After Effects 6.5 Pro, Quicktime 7, G5 Quad 2.5, Kona-LHe V3, 4.5 GB RAM, Nvidia GeForce 7800-GT, G-RAID 1TB FW800.

    Rafael Amador replied 19 years, 7 months ago 3 Members · 2 Replies
  • 2 Replies
  • Odetoacow

    September 26, 2006 at 9:02 pm

    You can set compression markers in FCP and export those markers to Compressor as markers to force I-frames in the Compressor MPEG2 encoder. Try setting markers in the problem area to force I-frames across the sequence. I’m not sure you can change GOP length or structure, but forcing I-frames may smooth out the ‘bump’ in you background.

  • Rafael Amador

    September 28, 2006 at 6:07 am

    Hi Tom,
    Even you are working in DV, I sugest you this procedure. Chage the setting of your sequence to 8B Uncompress keeping the same field order (if you are in NTSC, you have not problem because DV and 8b Unc are both Lower-first. In PAL is not like that). Set all the rendering parameters to the higher quality. And send to Compressor.
    Normally I worck in DVCam, bit once my time-line is ready to export, I change the seting to 8b Unc (some times 10b Unc) to keep all what I improved in the image with the color correction and all the best of thr graphics that I made with AE.
    If you keep your time-line in DV, you are recompresing to DV before to compress again to MPG2.
    Make a test and you will see the difference.
    salud,
    Rafael

We use anonymous cookies to give you the best experience we can.
Our Privacy policy | GDPR Policy