SCREEN-L Archives

July 2005, Week 2


Options: Use Monospaced Font
Show Text Part by Default
Show All Mail Headers

Message: [<< First] [< Prev] [Next >] [Last >>]
Topic: [<< First] [< Prev] [Next >] [Last >>]
Author: [<< First] [< Prev] [Next >] [Last >>]

Print Reply
Leo Enticknap <[log in to unmask]>
Reply To:
Film and TV Studies Discussion List <[log in to unmask]>
Wed, 6 Jul 2005 22:19:18 +0100
text/plain (109 lines)
Scott Hutchins writes:

>Miles Tudor, one of the authors of Decca's DVD of _The Death of 
>Klinghoffer_, claimed that my assertion of his and Philip Rowands's 
>incompetence authoring this DVD is libel.  He also claimed that what I 
>desire is a pan and scan version of the film, which I turned around abd 
>told him was libellous, since all I want is the film to play in its 
>correct ratio, rather than squeezing.  The settings of my player have been 
>checked and double checked, but I've never gotten the disc, which has a 
>SRP of $29.99, to play properly.  I had the same problem with _Monsters, 
>Inc._ and a Disney technician admitted that the problem was 
>incompatibility.  As far as I'm concerned, if compatibility is an issue, 
>it should state as such on the box.  The vast majority of 16x9 enhanced 
>discs play properly on my player.  I can only conclude that the few that 
>do not are incompetent.

In my opinion (as someone who has authored around 100 DVDs over the last 
few years, some for teaching, but many of them retail sale compilations of 
archive material), I don't think that conclusion is necessarily a safe 
one.  It could be in this particular instance, but to make that 
generalisation is to bat off a bit of a sticky wicket, IMHO.

By 'squeezing', I presume you mean that pixels which should be displayed 
anamorphically are not, resulting in a 4 x 3 image on your display monitor 
or TV which looks like a 35mm CinemaScope print shown without the 
anamorphic lens (i.e. people look too thin and tall).  If so, then there 
are four sets of variables which could - individually or in combination - 
be causing this.

Note - I'm using the specifications for the PAL DVD standard in the 
following text.  The principles also apply to NTSC - I just can't be 
bothered to look up the specs.

Whatever image the picture is supposed to be displayed in, it's digitally 
encoded as frames which are 720 pixels wide by 526 tall.  The 'red book' 
DVD video implementation of the MPEG-2 standard recognises three aspect 
ratios: 4x3, 16:9 (both used extensively) and 2.2:1 (used very rarely in 
commercial releases).  The difference between them is the width to height 
ratio with which each pixel is displayed.  The MPEG-2 standard itself 
recognises a number of options, e.g. square pixels (1:1), 1:1.067 (PAL 4x3) 
and so on and so forth.  The third variable is that each 'asset' (be that a 
menu, programme stream or video file) has to be assigned an aspect ratio as 
part of the DVD authoring process.  These three settings all have to be 
consistent.  For example, if you want to be sure that your 16:9 video will 
display correctly when played back from the DVD, then the MPEG stream 
itself has to be encoded with the correct pixel ratio flag, and that 
stream's assignment within the DVD file system also has to be set 
correctly.  Both of those settings have to be the correct pixel ratio under 
the DVD 'red book' rules for the 16:9 ratio as it is defined either in PAL 
or NTSC.

But even if you get all that right - and ESPECIALLY if either of those 
three parameters are inconsistent with the other two - there's a fourth 
potential bogie in the bathtub.  That is the combination of the software in 
the set-top DVD player or playback software and the monitor being used for 
the display.  For example, if the MPEG stream is correctly flagged as 16:9 
ratio pixels but the DVD video flag (the .IFO file as distinct from the 
.VOB) is not, some players might incorrectly determine the whole disc to be 
4:3, and output a full-frame, non-letterboxed image to a 4:3 monitor.  It 
may be possible to 'cheat' a widescreen TV or monitor by setting it to 
anamorphically expand the pixels, regardless of what the player is outputting.

If I had to make a guess, it would be that a combination of an obscure 
authoring error plus an obscure combination of hardware and software at the 
playback end is causing this problem.  That puts you into very murky waters 
- I agree that Decca's testing procedures should perhaps have been more 
rigorous, but I can also see them using two possible defences - (i) that 
you're using a weird and atypical playback setup, and/or (ii) that the 
problem is in the software they used to author the DVD (I bet it's Adobe 
Audition, which is the most unstable, bug-ridden package I've ever come 
across) and that therefore the software manufacturer is at fault, not them.

>Am I in the wrong wrong for broadcasting this incompatibility?

I don't think you're wrong in broadcasting it, but I do think that you 
COULD be wrong in attributing the cause solely to incompetence in the 
technicalities of the authoring process.

>I believe it's important for consumers to be informed of it so they don't 
>waste money on an unwatchable DVD like I did.

Agreed entirely there, and I think there needs to be flexibility and 
understanding on both sides.  In that sprit, I think it should be pointed 
out that this compatibility issue may only affect a tiny minority of 
playback systems, and that stores should take back DVDs for refunds with no 
questions asked.  But the interlocking sets of technical standards involved 
in DVD production and playback are so horrendously complex that I'd shy 
away from making out-and-out accusations of incompetence where aspect 
ratios are concerned.

Of course the fullproof solution would be to encode widescreen images 'hard 
letterboxed' within a 4:3 frame, as in the 'pillarboxing' process for 35mm 
rerelease prints of 1:1.38 titles, whereby the Academy ratio image is 
placed within a 1:1.85 frame, thereby enabling it to be shown in theatres 
which don't have a lens and aperture plate for the old ratio without
cropping.  But in both these scenarios, people who do have fully compatible 
kit will take a huge quality hit.


Leo Enticknap
Curator, Northern Region Film & Television Archive
Middlesbrough, UK  

To sign off Screen-L, e-mail [log in to unmask] and put SIGNOFF Screen-L
in the message.  Problems?  Contact [log in to unmask]