h a l f b a k e r yGet half a life.
add, search, annotate, link, view, overview, recent, by name, random
news, help, about, links, report a problem
browse anonymously,
or get an account
and write.
register,
|
|
|
Please log in.
Before you can vote, you need to register.
Please log in or create an account.
|
Sorry for the length of this, but it will require some explanation:
Currently, the DVD standard allows for two different frame aspect ratios, 1.33:1, commonly called "standard", and 1.77:1, called "anamorphic" or "enhanced for wide screen televisions." Films are made in many different aspect
ratios, and if a film doesn't match the ratio of the DVD frame, it will need to be cropped, paned and scanned, or letterboxed. The former two are completely unacceptable for most true film fans, so we are left with letterboxing. Now, letterboxing is good, but it does require that some of the available bandwidth of each DVD frame go unused, resulting in lower resolution. For example, 480 lines of picture resolution can be represented in a single frame on DVD. In order to represent a 2.35:1 ratio film frame on an "anamorphic" DVD frame, only 360 of those lines end up being used, the remaining are black.
Now, one would think that DVDs formatted with the "anamorphic" method would only be viewable with a widescreen television. Fortunately, this is not the case. Every DVD player is capable of "down converting" each 1.77:1 frame to a letterboxed 1.33:1 frame for display on a standard television. Essentially, this means taking the 480 lines of picture information recorded on the disc, reducing that to 360 lines and then adding 60 black lines to the top and 60 to the bottom of each frame. Even more fortunately, it is possible to get a fairly inexpensive 1.33:1 TV that can perform an "anamorphic squeeze", which means drawing the 480 scan lines in the middle part of the screen so the DVD player need not down convert the content of an anamorphic DVD. My 27" 1.33:1 TV can do this. Now this brings me to the whole point of this idea: at some point, 1.77:1 TV sets will become standard, and 1.33:1 sets will be unavailable. I think that that state of affairs will occur before movie studios will be willing to risk releasing their content on a high definition successor to DVD.
So my idea is: redefine the "standard" DVD frame to be 1.77:1, and redefine the "anamorphic" frame as 2.35:1. This would require no hardware changes to work, in fact, if such a disc were burned right now, I could watch it properly on my existing equipment. I would put the TV into squeeze mode, and have the DVD player down convert at the same time. Now the real benefit comes when I upgrade to a 1.77:1 television. Hopefully, distribution of DVDs using my scheme would prompt TV makers to include a squeeze mode in 1.77:1 televisions. That way, once I have such a TV, I could watch 2.35:1 ratio films from my (theoretically) existing DVD collection, at the full 480 line resolution offered by the DVD specification. Those with 1.77:1 televisions but no squeeze mode would be able to have their DVD players perform the down conversion step.
Oh, and since those with regular 1.33:1 televisions that lack a squeeze mode would be unable to wach such a disc properly, the film could be encoded in the regular old way on one side of the DVD and in my way on the other.
(?) Chart
http://www.uvm.edu/.../aspect/ratios.html
Maybe this chart will clarify just what in the heck I am talking about [JakePatterson, Dec 21 2001, last modified Oct 04 2004]
(?) Thread from Usenet
http://groups.googl...l%3Den%26filter%3D0
Previous discussion of this idea on usenet [JakePatterson, Dec 21 2001, last modified Oct 04 2004]
(?) How Video Formatting Works
http://www.howstuff...om/video-format.htm
From howstuffworks. Lots of information on aspect ratios, soft matting (what bristolz discusses) and convertion to TV. [pottedstu, Dec 21 2001, last modified Oct 04 2004]
(?) 2.4:1
http://www.plastecs.com/fiber_optics.htm
Read about 1/4 way down -- Great for burning your dvd onto a 5.25 pancake [reensure, Dec 24 2001]
[link]
|
| |
Might be a good solution for 'scope and even for flat academy (1.85:1) but what about the other aspect ratios that films are shot in? Some biggies are shot with Super 35 ("Titanic" comes to mind) which is a 1.33:1 (4x3) frame but, in the theatre you get to see only a 2.35:1 slice out of the "sweet spot" of the picture. It isn't until it is presented later, on 4x3 telecine that you realize the full-frame--which means 4:3 is (can be)superior in those cases. 70mm production creates yet more problems as it is generally 2.20:1 (sometimes 1.66:1). I guess those formats could be presented in the center portion of the 16:9 screen. Just food for thought. |
|
| |
The wonderful thing about standards is that there are so many to choose from. |
|
| |
[bristolz] IIRC, Academy ratio is 1.33:1 [correction, thanks pottedstu, Academy ratio is 1.37:1], so films shot that way (Casablanca and anything else from that era) would just use the standard encoding method. For other ratios, it would be possible to arbitrarally define a frame on a DVD to be any aspect ratio, but there exists a limitation having to do with the way that DVD players down convert "anamorphic" frames for display on regular 1.33:1 televisions. The players are only capable of taking the frame and either passing it raw to the display, which must be the correct aspect ratio itself, or be capable of putting itself into the right aspect ratio by squeezing, or the player may do a 25% down conversion first. My scheme would take advantage of the fact that the existing hardware would be able to down convert a 2.35:1 frame for display on a 1.77:1 display, or on a 1.33:1 display that is capable of doing its own 25% squeeze. I hope that makes sense. Current practice for films composed for 1.66:1 could be adapted for films that are composed for 2.20:1, that is use the "anamorphic" frame, and slightly "pillorbox" which would cause the loss of a little horizontal resolution, but not too bad, and largely irrelivant due to vertical blanking on most CRTs. |
|
| |
I just added a chart (see the link) that shows common film aspect ratios, and a diagram of how a frame of each aspect ratio would be stored (letterboxed, full frame, or pillorboxed) on each of the three DVD frame aspect ratios. The rightmost column is my proposed addition to the standard. The black portion of each image (except for the insides of the "o" characters) represent wasted bandwidth. If you had to encode a film onto DVD, you might look at the chart, and select the column that minimizes the black part of the frame for your ratio. My scheme allows all of these common aspect ratios to be encoded on DVD with minimal wastage of bandwidth. |
|
| |
[Pertinent Oak] No no no I am not proposing a new aspect ratio for film, 2.35:1 is a common ratio that is used today. Star Wars, The Matrix, Rushmore, 2001, (off the top of my head) were all 2.35:1. The problem is, currently, it is necessary to waste significant resolution/bandwidth on the DVD storage medium in order to put a film of that ratio on a DVD. The key to my scheme is that the existing hardware that can down convert a 1.77:1 frame for letterboxed display on a 1.33:1 screen can also happily down convert a 2.37:1 frame for letterboxed display on a 1.77:1 screen. Since 2.37:1 is close enough to 2.35:1 (a commonly used film aspect ratio) this will work. |
|
| |
[UnaBubba] I would really like to see this baked, but the problem is, very few people that I describe this to understand what the heck I am talking about. Also, I am assuming that DVD publishers will be somewhat nervous about listening to unsolicited crakpot ideas. |
|
| |
...Which is why there are 4,632 ideas here related to getting every other driver off the road... |
|
| |
Rant: Why are stamps so godd*** expensive. *grumbles* |
|
| |
This isn't as much of a problem as you might think, because in MPEG-2, the compression standard for DVDs, it requires very little bandwidth to encode the black strips at the top and bottom compared to the moving bits in the middle of the frame. |
|
| |
[pottedstu] You are correct about the black bars not taking up much raw storage space in bytes, but I am talking about the bandwidth in the logical domain of analog lines of resolution that ultimately get drawn on a cathode ray tube. The DVD specification calls for each frame to be 720 by 480 pixels, which is a digital domain, but the DVD player must output an analog signal to the CRT. For example, my DVD copy of annie hall is not "enhanced for widescreen televisions" therefore, each of its 1.77:1 frames are letterboxed in a logical 1.33:1 frame, and only has 720 by 360 pixels in the actual picture area. The top and bottom black bars don't take up many bytes on the digital encoding of the disc, but they still have to be translated into the analog signal in order to have the image be the right aspect ratio on the display device. On the other hand, my DVD copy of Run Lola Run is "enhanced for widescreen televisions", and each of its 1.77:1 frames are stored in a logical 1.77:1 frame on the actual disc, with the full 720 x 480 pixel resolution. This gets translated to an analog signal with 480 lines of resolution, which I put into my Sony Wega TV and I have the TV "squeeze" the signal so that all 480 lines get displayed in the 16:9 portion of the screen. I hope that makes sense. |
|
| |
[pottedstu] That link you put up has some good information, but it also includes at least one error, it says (wrt anamorphic DVDs): "When you play the DVD on a widescreen television, the player unsqueezes it so that it fills the screen." Actually, the /player/ performs no squeezing, streaching, or unsqueezing in that case, the television simply happens to be wide, resulting in the correct ratio. You can actually tell the DVD player that you have a widescreen TV even if you don't, and you can see that the objects in the film look elongated. Many 1.33:1 televisions have a mode where all of the scan lines are drawn more closely together (by 25%) resulting in the correction of such an anamorphic image, and higher resolution then you get when you make the player down convert. |
|
| |