Burned VS Pressed DVDs...

Fleshpainter wrote on 6/13/2005, 12:47 AM
If you pause through frame by frame, on a hi-definition screen, your typical rental DVD has a certain quality level. Now do the same with an AVI clip of similar subject matter, it’s STILL NOT BAD. View an MPEG 2 of the same AVI source clip and obviously you can see the compression, that’s to be expected.
I know all about the “film look” arguments but that’s not what I’m talking about here. I’m talking strictly about compression. I also realize that film, and/or hi-def cameras have more information to begin with but my point is that the standard definition AVI looks almost as good as the film-source rental MPEG 2 on the same screen, when you pause them one frame at a time.
Obviously when you render and burn you loose some. My question however is this... is anything salvaged when DVD’s are pressed or is the difference due strictly and exclusively to the source?

Another way to view the question is this:
If two DVD’s were created, one pressed and another burned from the same 35mm photographic film clip as the original source, would there be a noticeable difference between the two?
Now do the same thing using AVI’s from DV, side by side, any difference here? I would expect a difference between film and DV, but is there also a difference between pressing and burning?

(I’m not talking about the financial economics of pressing vs burning, just the compression quality.)

Comments

Liam_Vegas wrote on 6/13/2005, 12:51 AM
It's all just one's and zeros on the DVD - irrespective of how they got there (pressed or burned).

The primary difference will be the amount of time (money) spent in creating the MPEG2 file before it made it onto the disc.

Other thing is that you need to take into account the bit-rate the MPEG files were created at. Most hollywood DVD's have 9GB of space to play around with... and we've only recently been able to achieve that with dual layer discs.

With 9GB to play around with the resulting MPEG file can have much larger bit-rate than would be possible with single-DVD's.
jaegersing wrote on 6/13/2005, 12:54 AM
Think of it as just 2 different storage media. The data is exactly the same so there should not be any difference in quality.

Richard Hunter
FrigidNDEditing wrote on 6/13/2005, 1:54 AM
the thing to watch out for is longjevity - they say that their supposed to last so long, blah blah blah. inks will fade - especially with usage. or direct sunlight, etc...

Stamped is there till the disc surface is not able to let light pass through properly.

Dave
B_JM wrote on 6/13/2005, 6:44 AM
the difference is in the source files - no mater what you think ..

DV avi or the like doesnt have the quality of scanned or better telecine files from film source.

Commercial dvds bit rate average is often pretty low - around 6000 is normal ...

baysidebas wrote on 6/13/2005, 11:10 AM
All that said, the primary reason most commercially pressed DVDs exhibit higher video quality is due to the hardware (vs. software) multi-pass (vs. single pass) encoding used.
riredale wrote on 6/13/2005, 11:30 AM
Couple of thoughts:

(1) Pressed or burned makes no difference, since it's the same data, though a typical DVD player may have a harder time reading the burned disk due to several factors, such as a lower contrast ratio.

(2) A hardware MPEG2 encoder is not necessarily better than a software codec, especially if the job does not have to be done in real time. Ten years ago: different story.

(3) I've read that Hollywood can deliver gorgeous compressions primarily because the source has very high resolution, it runs at a lower frame rate (more bits for each image), and it's noise-free (video noise just clobbers MPEG2 encoding).

(4) As pointed out above, Hollywood doesn't use bitrates that are out of the ordinary. Use your PowerDVD or WinDVD players to look at the bitrate on a movie--it's typically around 6Mb/sec or so.


Edit: One more:

(5) Hollywood deals with a progressively-scanned source, not interlaced.
B_JM wrote on 6/13/2005, 1:45 PM
no - not true

software encoder or hardware - no diff. really - with the right software or hardware .. if fact i can often get better quality with a software encoder and believe me, ive used plenty of both ..

commercial dvds are done either way nowadays ..

a hardware encoder is just a software encoder with embedded software

\\ edit -
riredale said the same thing:
(2) A hardware MPEG2 encoder is not necessarily better than a software codec, especially if the job does not have to be done in real time.

B_JM wrote on 6/13/2005, 1:46 PM
no - not true

software encoder or hardware - no diff. really - with the right software or hardware .. if fact i can often get better quality with a software encoder and believe me, ive used plenty of both ..

commercial dvds are done either way nowadays ..

a hardware encoder is just a software encoder with embedded software

\\ edit -
riredale said the same thing:
(2) A hardware MPEG2 encoder is not necessarily better than a software codec, especially if the job does not have to be done in real time.

B_JM wrote on 6/13/2005, 1:46 PM
no - not true

software encoder or hardware - no diff. really - with the right software or hardware .. if fact i can often get better quality with a software encoder and believe me, ive used plenty of both ..

commercial dvds are done either way nowadays ..

a hardware encoder is just a software encoder with embedded software

\\ edit -
riredale said the same thing:
(2) A hardware MPEG2 encoder is not necessarily better than a software codec, especially if the job does not have to be done in real time.

B_JM wrote on 6/13/2005, 1:56 PM
no - not true

software encoder or hardware - no diff. really - with the right software or hardware .. if fact i can often get better quality with a software encoder and believe me, ive used plenty of both ..

commercial dvds are done either way nowadays ..

a hardware encoder is just a software encoder with embedded software

\\ edit -
riredale said the same thing:
(2) A hardware MPEG2 encoder is not necessarily better than a software codec, especially if the job does not have to be done in real time.

B_JM wrote on 6/13/2005, 1:57 PM
sorry about the multiple posts - no idea what happened ..

johnmeyer wrote on 6/13/2005, 2:07 PM
Many Hollywood movies are recorded on DVD-5 (4.7 GByte, single layer DVD), the same as what most of us use when we record to DVD-R and DVD+R. They certainly use better encoders, and have the tools to work on small sections and improve the quality. However, I think the real reason why the quality is better -- to answer the original post -- has to do primarily with three things:

1. 24 vs. 30 fps.
2. Progressive vs. interlaced
3. Source quality

I have encoded movies that I have captured off the air. When I have encoded them directly from the capture, the quality is similar to what was described in the original post, especially when you view them frame by frame. On the other hand, if I first do an "inverse telecine" pass to change the capture back to 24 fps, progressive, and then encode to DVD with the film flag set (so it plays back at 30 fps interlaced), the quality is dramatically better.

And, when you start with 35mm film, which has a "resolution of at least 4000 x 3000 vs. DV NTSC AVI which is 720x480 pixels, the encode is obviously going to be a lot cleaner.
B_JM wrote on 6/13/2005, 2:18 PM
most 35mm is only scanned at 2k res. if going to be used for HD or DVD .

and sometimes less -- but the point being about a higher quality to start with is "the point"

Scans at higher resolution are done though - 4k and up for 35 and 70mm for delivery for d-cinema conversion and for use in cgi and composites - also the color depth is much better than DV of course and the DV to Mpeg2 ntsc dvd color space conversion is a rough one ..





Fleshpainter wrote on 6/15/2005, 1:39 AM
Can I safely conclude, from the above, that if the end product is strictly a DVD release, with DV as the source, to be viewed on TV's and computers, that the MPEG 2 render from Vegas is pretty much the bottom line?