4 Replies Latest reply on Nov 4, 2010 5:32 AM by jkosmicki

    H.264 VBR, 1 pass vs. 2

    Am_I_Lame? Level 1

      Is there any way to quantify the advantages of 1 pass vs. 2?  I realize it depends, in part, on the source material, but I've run some tests on 15 second clips from a project that's 30 minutes long, and I don't see any noticeable difference in image quality.  Also, the output file is only about 2% smaller with 2 passes than it is with 1.  On the other hand, it takes me 30 minutes to render the 15 second clip (it's HD) with 2 passes, which would translate into 60 hours for the whole project, vs. 30 for 1 pass.  Doesn't seem worth it to me, but I'm a newb.  Is there something I'm missing?

        • 1. Re: H.264 VBR, 1 pass vs. 2
          jkosmicki Level 2

          I notice the difference in crossfades or other areas were the image composition is changing. With 1 pass there is usually noticeable artifacting during crossfades, with 2 pass the quality level is much more even through out the video.

          But more importantly, the question is, why is it taking 30 min. to render a 15 sec. clip?  I know rendering times can vary a lot depending what's on the timeline, but it takes my system typically just a few seconds to render a 15 sec. HD clip.

          -Jeff

          • 2. Re: H.264 VBR, 1 pass vs. 2
            Am_I_Lame? Level 1

            > But more importantly, the question is, why is it taking 30 min. to render a 15 sec. clip?

             

            I was wondering that, too, but having nothing to compare it to I just figured that's the way it was.  How's your system configured?  I'm running Windows 7 Ultimate x64 on a 2.33 GHz dual core Xeon with 16GB of RAM and dual WD VelociRaptor 600GB HDDs (10K RPM, 32MB cache) configured as a RAID 1 (for security, not speed) on a SATA I controller.  My source is 1440 x 1080, 30 fps, AVCHD, and I'm outputting to 1920 x 1080, 30 fps, H.264.  Both cores max out during the render.

             

            It would be great if there were something relatively simple I could do to substantially reduce this time as I'm going to be doing more of this sort of thing and this experience has me looking at new PCs.

            • 3. Re: H.264 VBR, 1 pass vs. 2
              Powered by Design Level 4

              You really need to run the PPBM test to see why your system is so slow.

               

              http://ppbm5.com/

               

              Once you run the test then you will know how your system is running compared to others.

               

              After running it im sure Harm and Bill can make suggestions to make your system faster.

               

               

              GLenn

              • 4. Re: H.264 VBR, 1 pass vs. 2
                jkosmicki Level 2

                Until earlier this year, I was editing on a very expensive dual-quad core Xeon system with 32GB RAM. I replaced that system with a comparatively inexpensive core i7 system and difference is amazing. Rendering and encoding times dropped anywhere from half up to a tenth of the time. Xeons used to be the champs, but I think i7s have really replaced them, especially for editing. And if you compare costs, i7 really gets the advantage.  Also, an Adobe certified Nvidia card can make a huge difference too.

                -Jeff