Video Encoding for DVD

     Okay sports fans.  Here's the deal.  DVD players have a 9.4mbps limit on combined audio-video for DVD.  Blu-ray on DVD can get 15mbps.  Balance that out between audio quality and video quality.  If your audio is 256kbps (decent audio for most purposes), thats  0.25mbps and will leave you with 9 as the max setting for DVD (you might say 9.2, but its more like 9.14, and I just round it to keep some distance to prevent the errors that will undoubtedly appear). 

When you encode for DVD:

Your TARGET (or average value over the whole video) should be .75 to 1mbps lower than the max value you set, while your lows can be anywhere, but are best at .75 lower than your target (if you have this minimum option).  This is an average of your non-motion frames, which will keep clear quality within 2mbps of your max, for slower action (some stage performances and such will work great with this; for sports, keep your minimum within 4, and it will actually keep base clarity with your faster motion, keep your target closer to your max, and all motion will keep general look and feel).

The whole point here is that you can set whatever bitrate you want and get great quality files, but no DVD player will push past 9.4mbps.  It's a hardware limitation of EVERY MODEL. 

    The best way to get great results with encore is a 2 step encoding.  First, output an extremely high quality file in any format that has a high quality, like 30-40mbps max with a split of 5 for the other sliders.  If you use mpeg2, you'll get a HUGE file.  However, if you divide the data rate, you'll divide the file size.  Check your average data rate.  This will tell you if your file can be compressed well.

    IF you set a 40mbps max and a 30 min, and your average rate is between 30 and 35, you can compress this file well down to 8-9mbps at 1\5th quality of data.  Now you can do the same rip, but use the highest quality file (if that's your source, which is unlikely, as it's compressed to a lot lower data rate), go from your sequence; if your new file has a higher data rate than your source, it will look the same quality, but compress with BETTER quality, as the compression will be more accurate to each frame, and should take less time.

    There are 2 definitions to contend with for the term quality.  The first is DATA Quality, the second is VISUAL Quality.

     Why? With compressed data coming into your sequences (most consumer cameras compress the capture to proprietary formats), you're seeing a 94-98% accurate file (just checked an online listing for AVCHD qualities).  When you blow that data back up, you get closer to 94-98% per frame (There is some distinction you'll see in a minute).  Since each frame in your source (camera data) probably has less data built, the rest needs to be guessed, and reformatted, so it will very likely be a less accurate calculation.  When re-encoding between formats, the decompression step will USUALLY do bare minimum, and will cut decimal data down severely as well.  This results in lower precision and accuracy of data to format to the new file.

    IF you expand that data first you cut down on the inaccuracy.  Also, moving to a compression within the same file format reduces the inaccuracy in the compressing stage, mainly because you're not rebuilding the frame, or reformatting--you're just dumping more of the less relevant data (pixels that generally don't change between frames need only be in one frame and left out of the next).  Because the math of the compression math and other formatting are kept the same, they are inherently more PRECISE.  Thus the file shouldn't APPEAR to lose as much quality as it would if you went from your source. 

    Just note, I said APPEAR.  It will generally Appear to have better quality VISUALLY, but be poorer "quality" in the accuracy of what data was actually captured by the sensor in the camera (there will be differences, but you should have more precise values for the data between frames, rather than a more "Perfect" match to capture, and the visual will be smoother in small details, like skin, but sharper at edges where things change more). 

    It will lose a little more data quality, mathematically; but 98% of something at 98% is above 97%, Do that again, but compress down to maybe 90%, an you get more like a 90-91% overall either way (source or decompression).  The excess loss may seem unnecessary, but if you have compressed source, you're almost always better off decompressing coming out of edits or going into them; it's not about whether or not you lose anything, it will happen no matter what. It's about where you lose it.  If you lose it in the rebuild of each frame, you have a low grade frame going to lower grade frame.  If you have a more complete set of data, you get a higher grade frame being compressed to lower grade.

Lastly, only because I've pontificated about DVD playback, and you may want to encode Blu-ray or high-def video:

To encode Blu-ray data to DVD, you can build a 15mbps blu-ray quality in h.264 (combined audiovideo data rate, by the way), create the Disk image in encore, don't burn, and then use software like TOAST to burn it to a DVD.  That disc will play in many blu-ray players.  Why only 15mbps?  The read spead of a DVD in a blu-ray player is limited to 15mbps (about half normal blu-ray rate due to reflective properties and track distances); what this means is that it can play any dvd at normal rate, while blowing it up to size, but play blu-ray data from a dvd slower than normal blu-ray due to laser movement time being about double.

PS

     Set top playback does the REVERSE of any PC youtube screenplay.  A tv has one input going in, meaning you can remap the output without affecting the playback of other inputs.  Not so on a PC.  Look up PPI, pixels per inch.  Generally, video conforms to its pixel height x pixel width \ Pixels\inch where pixels\inch is a value of 72 on playback.  If you adjust the 72 down to say 24, you've blown everything else up 3 steps.  That turns DVD resolutions into fullscreen height size for a TV, without really losing much.  Some things will be smoothed out too much but the TV usually adds some noise to clarify the image.  Visually, it's difficult to tell the difference; you only see that it's bigger, and maybe has a bit more noise added to sharpen things a bit. 

     On a pc, you have many other apps to contend with, though some create a virtual screen in gfx memory to overlay onto the main screen, shutting out all other inputs, which is how they can play a dvd at a similar quality to TV sets and it functions much the same.

Okay sports fans.  Here's the deal.  DVD players have a 9.4mbps limit on combined audio-video for DVD.  Blu-ray on DVD can get 15mbps.  Balance that out.  If your audio is 256kbps, thats  0.25mbps and will leave you with 9 as the max setting for DVD (you might say 9.2, but its more like 9.14, and I just round it to keep some distance to prevent the errors that will undoubtedly appear).  Your target should be .75 to 1mbps lower, while your lowes can be anywhere, but is best at .75 lower than your target (if you have this minimum option).  The whole point here is that you can set whatever bitrate you want, but no player will push past 9.4mbps.  The best way to get great results with encore is a 2 step encoding.  First, output an extremely high quality file in any format that has a high quality, like 30-40mbps max with a split of 5 for the other sliders.  If you use mpeg2, you'll get a HUGE file.  However, if you divide the data rate, you'll divide the file size.  Check your average data rate.  This will tell you if your file can be compressed well.  IF you set a 40mbps max and a 30 min, and your average rate is between 30 and 35, you can compress this file well down to 8-9mbps at 1\5th quality of data.  Now you can do the same rip, but use the highest quality file (if that's your source, which is unlikely, as it's compressed to a lot lower data rate), go from your sequence; if your new file has a higher data rate than your source, it will look the same quality, but compress with BETTER quality, as the compression will be more accurate to each frame, and should take less time.  With compressed data coming into your sequences, you're seeing a 97-98% accurate file.  When you blow that data back up, you get closer to 97-98% per frame, rather than trying to decompress and recompress on the fly.  Since each frame in your source probably has less data built, the rest needs to be guessed, and reformatted, and it will be a less accurate calculation.  Moving within the same file format reduces the inaccuracy, and using a file that's less compressed (more data built into each frame) takes less time to calculate the rest of each frame for recalculating the corresponding new compression; in fact, it only need examine existing data, and reduce it for each frame by the factor necessary.  Thus the file shouldn't Appear to lose as much quality.

To encode Blu-ray data to DVD, you can build a 15mbps blu-ray quality in h.264 (combined audiovideo), create the Disk image in encore, and then use software like TOAST to burn it to a DVD that will play in many blu-ray players.  Why only 15mbps?  The read spead of a DVD in a blu-ray player is limited to 15mbps (about half normal blu-ray rate due to reflective properties and track distances); what this means is that it can play any dvd at normal rate, while blowing it up to size, or play blu-ray data from a dvd slower than normal due to laser movement time being about double.

DVDs you buy for movies and TV shows are double-layered which has 8.5GB capacity.  Most of DVD-R/RW disks for the home burning is a sinble-layer which has 4.7GB capacity.  There is double-layer DVD-R disk with 8.5GB.  But, your driver need to support the double-layder to use this media.

Another way to put more video into DVD is to reduce the size of the video file.  For example of AME with MPEG2 DVD format, you can do this by reducing the bitrate under the Video tab.  But, this results in lower video quality.  You may want to experiment and see which settiings are acceptable for you.