7 Replies Latest reply on Oct 14, 2012 3:00 PM by AlcoholicBear

    AVCHD Decompression?

    AlcoholicBear Level 1

      Hey guys,


         I recently bought a Panasonic Lumix DMC-ZS19K for $220.00 (awesome deal!) and it records AVCHD 1920x1080 60p at 28-30 mbps bit rate. The quailty isn't bad, I just like working with uncompressed files like YUV 4:2:2 etc. Since I upload to YouTube, which compresses like crazy, I want to keep the crisp image I get on my monitor on YT as well.. Now, I understand that AVCHD uses a proprietary algorithm to compress data so it can write it on a standard SD card. I want to decompress the video to a uncompressed format( reverse the compression), but various google searches have yielded old 2008-ish answers.


         If there are any of you who knows how to decompress the compression on CS6 (I bought the master collection... pricey), please share any info.


      As for my hardware, I am an avid gamer, so my PC serves a dual purpose when editing   PS. I am using windows 7


      Here are my specs


      Intel 3770K @ 5.2 Ghz watercooled by XSPC 360mm rad ( over 24 hrs of O.C to get here)


      16 GB DDR3 @ 1866


      Asrock OC Formula Board


      4 x 1000 GB Carviar Black drives in RAID 0 @ 420 read / 380 Write


      Asus GTX 560 Ti ( in Mercury Playback Engine config)


      LSI 9207-8i RAID card


      4 X Samsung 128 GB 830's RAID 0    Read seq: 2014 mb/s  Write Seq: 1228 mb/s


        • 1. Re: AVCHD Decompression?
          Stephen_Spider Level 3

          The AVCHD is decompressed every time a frame of video is displayed. You would not gain anything by converting the video to an uncompressed or less compressed format. Information is discarded for compression and you can not retrieve it. Premiere works with the AVCHD natively, so you won't incur any generational loss until you export your final video. In simple terms, the video is what it is until you export it, an intermediate won't help your quality, in fact, there is mathematical loss in the conversion process to even a better format.


          The only valid reason to convert your AVCHD to a less compressed codec would be if your system could not handle the native editing. With your processor and your GTX card, native editing is what you want.

          • 2. Re: AVCHD Decompression?
            AlcoholicBear Level 1

            Well, it looks like I will revert to recording on my Black Magic Intensity Pro and Shuttle 2. I thought that compression worked like a .zip or .rar file. Thinking about it now, that would reqire serious processor power on a camera.


            I have read about H.264 video having the ability to decompress, is that true?


            Also I read that this user was able to decompress AVCHD http://forums.adobe.com/thread/745684

            • 3. Re: AVCHD Decompression?
              josephs51576386 Level 3

              You can't gain detail back that isn't there. Video doesn't work this way. So by converting to uncompressed format you aren't going to gain anything. The only advantage to converting to a less compressed format would be to help put less strain on your CPU. But taking AVCHD video and converting to ProRes or DNxHD or a uncompressed format after it's already been recorded won't gain any quality.

              • 4. Re: AVCHD Decompression?
                Stephen_Spider Level 3

                You're asking the same question again, the answer is the same.


                The link you provided was by somebody asking the same question as you. One problem here is the way you both are using the term decompression when you are talking about conversions.

                • 5. Re: AVCHD Decompression?
                  AlcoholicBear Level 1

                  I see, once you compress, the original quality is gone for good. Well, I think I know the answer now. Thanks Stephen_Spider and ComputerNovice25 for answering.   If only YouTube offered YUV 4:2:2 format for uploading....

                  • 6. Re: AVCHD Decompression?
                    Jim_Simon Level 8

                    I thought that compression worked like a .zip or .rar file.


                    That depends on the compression scheme used.  Zip and RAR are lossless, meaning you can reproduce the original exactly as it was.  There are lossless video compression methods, like the Lagarith and UT codecs, but AVCHD (or anything MPEG based) isn't one of them.  Even ProRes and DNxHD aren't lossless.


                    Now the reason this is important to understand is that you can't see a compressed video signal.  It has to be decompressed for you to see it.  PP will decompress each frame as it shows it.  Your DVD or Blu-ray player will do the same.  Windows Media Player and QuickTime players, the Flash player on the YouTube web page, all will decompress the video so you can see it.  And despite some claims to the contrary, you not only can put back what was thrown out during compression, you have to put it back in order to see the image.


                    The issue is how accurately it gets put back.  With lossless compression codecs, it gets put back exactly as it was originally, without alteration.  With lossy compression codecs like AVCHD, there are always errors recreating what was thrown out during compression.  Those errors are called artifacts.


                    Here's the point I'm getting to.  Once the image is compressed using a lossy codec like AVCHD, those decompression artifacts are going to show up whether PP decompresses it for viewing or another program decompresses it to a new Uncompressed file.  You can't avoid it.  There just isn't enough data to restore the image without errors, no matter what hardware or software does the work of decompression.  This is why people are saying there's no point turning your AVCHD into Uncompressed for editing.  The errors will show up regardless.  All you do by converting is waste time and disk space.


                    Working with Uncompressed YUV 4:2:2 is a nice idea, but unfortunately there isn't a camera on the market that can do this.  The only 'exception' might be the new Black Magic Cinema Camera, which can record to lossless CinemaDNG RAW files.  But that's a pricey proposition with a headache of a work flow.  Other than that, every camera out there will compresses the video in some way, throwing out data that causes errors to show up when decompressed.  Even the RED cameras Peter Jackson is using for The Hobbit aren't perfectly lossless.  Of course, the errors are likely to be so small and insignificant that no one will ever see them.


                    While the AVCHD codec will generate more errors than Redcode RAW, it's still a pretty clean codec overall.  The errors are likely to go unnoticed by most people in most situations.  So, for your YouTube uploads, you're fine just using the AVCHD media from the camera.

                    • 7. Re: AVCHD Decompression?
                      AlcoholicBear Level 1

                      Thanks for the articulate and through explanation, I think I finally understand the concept and use of compression. I use a Blackmagic Intensity Pro and Shuttle 2 to record directly off my camera's sensor via integrated  HDMI output using the YUV codec and DNxHD on the intensity pro or shuttle 2. The conparision between the two is significant. The stock compression s fairly agressive on image quality. After sharpening the frames in After Affects, the "uncompressed"(aka the more accurate YUV) video looks superb, a BIG upgrade from the stock config. I can't wait to see the "uncompressed" video of a T4i or Pentax K-30. Again, thank you all for your great responses! This is a very helpful forum.