logo Sign In

Non-DVD transfers?

Author
Date
Author
Time
Post link
Has anyone thought about or even done a non-DVD transfer of their LDs?

I don't mean just making a 640x272 700mb DivX/XviD file to upload somewhere, but a full (or scaled higher than) DVD resolution 4.3gb DivX/XviD/wmv etc. file for playing back on a DVD-Rom drive in a PC connected to a large screen/projector.
http://www.haku.co.uk/pics/LukeCruise.gif http://www.haku.co.uk/pics/dontcare.gif
***Citizen's NTSC DVD/PAL DVD/XviD Info and Feedback Thread***
Author
Time
Post link
What would be the point of that? You can already play the DVD on a DVD-ROM drive to watch on a projector. You certainly will not gain quality by going to a higher resolution than 720x480. You can never gain quality by going to a higher resolution than the source; in fact, you will never gain quality over the source (in this case, 528x480 laserdisc); period. The fact that LD is an analog video stream that will always suffer loss just getting to its destination; added to the encoding to MPEG-2 or MPEG-4; guarantees a loss here. Increasing the resolution over the source only magnifies this loss (much like zooming in on the source itself would make it look worse because there is not enough information there for a higher resolution than it is intended for). It is in the acceptable range to increase the resolution of LD material to DVD standards but that is already pushing it; to saying nothing about trying to go even bigger. It would be little different than playing the DVD in Media Player Classic and zooming it to a higher resolution on the fly.

One more thing; MPEG-4 would be saturated for a 2 hour movie at something under 3 GB; probably around 2.5 GB, so a 4.3 GB MPEG-4 file (XviD/DivX/WMV/MOV/ect) for Star Wars is not likely even if you tried nor would it be of any benefit if you succeeded.
Author
Time
Post link
Yes I know all the arguments against upscaling an image, the reason everyone's all doing letterbox or upscaling anamorphic pictures is because of the native resolution of DVD and tvs, but when you're dealing with a projector the image has to be upscaled on the fly (I don't know of any PAL or NTSC native resolution projectors) so what I'm proposing is doing the upscaling beforehand with some slight sharpening so the file you're playing back is in the native resolution of your projector.
In my case my projector's native resolution is 1024x768, upscaling a letterbox/anamorphic PAL/NTSC image doesn't give as crisp an image without additional on-the-fly sharpening, ie I've done some tests with creating 1024x436 XviD files from my LD's using a 2x edge-directed resampler to increase the resolution two fold by edges rather than simply scaling, the difference between Lanczos3 and this edge-directed 2x resizer can be seen here, scaling the image back down to whatever you need improves the picture I've found, this is a quick grab from one of my tests:

http://www.haku.co.uk/pics/StarWarsLD1024.jpg

because the width is already 1024 there's no rescaling when playing back and it looks clearer than the same scene encoded to mpeg2 and scaled on-the-fly.
http://www.haku.co.uk/pics/LukeCruise.gif http://www.haku.co.uk/pics/dontcare.gif
***Citizen's NTSC DVD/PAL DVD/XviD Info and Feedback Thread***
Author
Time
Post link
I happen to think you (MaximRecoil) are almost completely wrong.

It is certainly possible to make an LD transfer look far better than the original LD, and you can improve the quality substantially on an upsample. I mean, if you're going to scale, you'd may as well do it at less than real time with competent filters, rather than let some budget TV try to do it on the fly.

I don't know what you're looking at to say otherwise.
Author
Time
Post link
I've been considering taking a PAL source and changing it's frame rate to 24fps, then using the uncompressed PCM audio from an NTSC set as a soundtrack (or possibly audio from a higher quality source, see below).
The video would look closer to the way it was originally intended as PAL doesn't mess up the video nearly as much as NTSC does (once it's been slowed down anyway, I can't stand PAL speedup personally).
It is likely that a higher resolution soundtrack could be created for the entire film using the VHS tapes that were mastered from analog sources, as VHS has good sound reproduction (better then CD quality PCM at least) if the tape and the player are in good condition.
I've also consider using the LP records to create a higher quality 96,000 Hz, 24 bit isolated score, as I've always felt that the Star Wars score suffered quite a bit in CD standard 44,100 Hz, 16 bit as a result of the heavy use of the brass section in the score (the quality of the reproduction of brass sounds on CDs compared to that of LPs is often considered a prime example of why LP is the superior medium). A 96,000 Hz, 24 bit isolated score made from a decent condition LP would sound much better. I'm not certain that the entire score is included on the LPs, but if anything isn't included it could likely be pulled from an analog master sourced VHS.
Author
Time
Post link
Actually I am planning on doing exactly this with all of the SW-movies. Since I've seen the HDTV-DivX of "Attack of the Clones" on my new 32" HDTV-LCD (via DVI input) I am looking deparately to get the HDTV-version of "Phantom Menace" as well. As for the OT (from which I assume no HDTV-versions exist), the pristine 2004-DVDs are simply perfect for upscaling them to 720p. I've done a 5-minute test-clip of "Empire" upscaling it with several AviSynth-filters, and the quality just blows the DVDs away easily.

I know, people keep saying that upscaling doesn't add anything to the image, but that's not true. AviSynth-filters like "LimitedSharpen" add artificial detail, and really make the image look outstanding (especially when fed to the TV via DVI instead of RGB-scart). Here some samples:

Empire Sample 1
Empire Sample 2
Empire Sample 3
Empire Sample 4
Empire Sample 5
Empire Sample 6
Empire Sample 7
Empire Sample 8
Empire Sample 9
Empire Sample 10

On closer observation there are some artifacts visible, which make the image look like it's painted... these are only visible on stills. When in motion, they get pretty much compensated.
Author
Time
Post link
Originally posted by: MaximRecoil
One more thing; MPEG-4 would be saturated for a 2 hour movie at something under 3 GB; probably around 2.5 GB, so a 4.3 GB MPEG-4 file (XviD/DivX/WMV/MOV/ect) for Star Wars is not likely even if you tried nor would it be of any benefit if you succeeded.

About that, I did a quick test converting captured LD footage in DV format to XviD format, the only thing I did to the LD footage was undo the 3:2 pulldown, the XviD was encoded at 100% image quality with each frame being a keyframe, had I gone ahead and encoded a 2 hour capture the avi would be an estimated 9-10gb in size, so it's perfectly possible to have a 2 hour 4.3gb XviD avi by having 1/2 or 1/3 the number of keyframes.



The main reason I started this topic is because whilst waiting for my PAL LDs to arrive I'm still trying to work out what format(s) I eventually end up putting my set onto, anamorphic PAL is a very high consideration for playing on my 28" widescreen tv and giving a copy to my friend who has a 42" plasma, both setups have the same DVD player and are multiregion so NTSC playback is not a problem except PAL has a smoother playback. But there's also my PC powered projector (5 foot wide image) so I'm toying with the idea of producing a 1024 width XviD for playing back on that.
http://www.haku.co.uk/pics/LukeCruise.gif http://www.haku.co.uk/pics/dontcare.gif
***Citizen's NTSC DVD/PAL DVD/XviD Info and Feedback Thread***
Author
Time
Post link
Since I've seen the HDTV-DivX of "Attack of the Clones"


Was that from a broadcast? Or is it available, like Terminator 2, on some HD-DVD format?

DE
Author
Time
Post link
Originally posted by: Darth Editous
Since I've seen the HDTV-DivX of "Attack of the Clones"


Was that from a broadcast? Or is it available, like Terminator 2, on some HD-DVD format?

DE


It was from the broadcast... unfortunately it was cropped on the sides to fit a 16:9 aspect ratio, also I don't mind too much, consindering the image quality.

Was there a HDTV-broadcast of "Menace"? Does anybody know?
Author
Time
Post link
Wasn't it cropped to 4x3?

<span class=“Italics”>MeBeJedi: Sadly, I believe the prequels are beyond repair.
<span class=“Bold”>JediRandy: They’re certainly beyond any repair you’re capable of making.</span></span>

<span class=“Italics”>MeBeJedi: You aren’t one of us.
<span class=“Bold”>Go-Mer-Tonic: I can’t say I find that very disappointing.</span></span>

<span class=“Italics”>JediRandy: I won’t suck as much as a fan edit.</span>

Author
Time
Post link
don't you mean 16.9??
"A Jedi can feel the force flow through him".
Author
Time
Post link
@ MeBeJedi: No. It's 16:9. I have it, of course.

@ Jedikev: No. It's 16:9. If that was humour on your part, you forgot your smiley!

@ Laserschwert: Was there a HDTV-broadcast of "Menace"? Does anybody know?

I've never heard of an HD Ep I. That doesn't mean it doesn't exist, but since I've been collecting HD for more than three years, I'd like to think I'm nearly as authoritative on this point as anyone.

Then again, I'd have said there was never any OT HD stuff, either, and now I have some. So maybe it's, "I've never heard of an HD Ep I yet."

There's a thread on another board about upsampling the SW OT to HD that pre-dates this one by several months. I might have to go find it.
Author
Time
Post link
" @ MeBeJedi: No. It's 16:9. I have it, of course."

I remember you saying that, but for some reason, I thought you said it was 4x3. Am I confusing it with another SW video you have?

<span class=“Italics”>MeBeJedi: Sadly, I believe the prequels are beyond repair.
<span class=“Bold”>JediRandy: They’re certainly beyond any repair you’re capable of making.</span></span>

<span class=“Italics”>MeBeJedi: You aren’t one of us.
<span class=“Bold”>Go-Mer-Tonic: I can’t say I find that very disappointing.</span></span>

<span class=“Italics”>JediRandy: I won’t suck as much as a fan edit.</span>

Author
Time
Post link
There are dvd players that have special chips for realtime upscaling and the results are actually pretty good. The one I have expereince with is the Denon 1910 . It has a faroudja chip that's able to do 720p and 1080i. My projector is 720p and it gives the dvds a real film like look... no jaggies and it brings out details that were not visible with my cheepo pioneer. Its about $200 US (its Denon's budget level player )... but in my opinion well worth it since the quality difference is fairly noticeable.... even my wife noticed the difference Check out some AV forums if your interested.

I suspect you could eek out a little more detail and perfomance with going to the trouble of trial and error with virtualdub and the likes, but the faroudja upscaler pretty good...

Author
Time
Post link
Originally posted by: KaryudoThen again, I'd have said there was never any OT HD stuff, either, and now I have some.


Well, THAT sounds interesting... are those the complete movies (even if just the SEs)? The only HD I've seen from the OT was the HD-trailer for the 2004 DVD release.

Author
Time
Post link
Originally posted by: Karyudo
I happen to think you (MaximRecoil) are almost completely wrong.

It is certainly possible to make an LD transfer look far better than the original LD, and you can improve the quality substantially on an upsample. I mean, if you're going to scale, you'd may as well do it at less than real time with competent filters, rather than let some budget TV try to do it on the fly.

I don't know what you're looking at to say otherwise.
It is impossible to end up with more information than you started with (i.e. your source). "Looks better" = subjective and is not synonymous with higher quality. For example, a 1400 MB MPEG-4 encoded from a 5 or 6 GB MPEG-2 DVD source often looks "smoother" than the original. It is not higher quality than the source however. The source contains all the information you need to make all sorts of different looking versions via processing. The resultant 1400 MPEG-4 does not because of loss. Even if you re-encode to the same bitrate, format and resolution there will be some loss (not much); it is inevitable.

On the other hand, now that I have read the original poster's reply to my post, I can see the potential benefits of upscaling the image in the initial encode rather than doing it on the fly via hardware stretching.

Author
Time
Post link
About that, I did a quick test converting captured LD footage in DV format to XviD format, the only thing I did to the LD footage was undo the 3:2 pulldown, the XviD was encoded at 100% image quality with each frame being a keyframe, had I gone ahead and encoded a 2 hour capture the avi would be an estimated 9-10gb in size, so it's perfectly possible to have a 2 hour 4.3gb XviD avi by having 1/2 or 1/3 the number of keyframes.

With default settings in XviD at 100% quality, I have never gotten much over 2.5 GB for a movie. Increasing the number of keyframes would certainly bring that up but that always seemed to me as though it was defeating the purpose of MPEG-4. Either way, your idea sounds interesting now that I have read more of this thread than just the initial post that I replied to last night.
Author
Time
Post link
Originally posted by: MaximRecoil
It is impossible to end up with more information than you started with (i.e. your source).


That's not true -- but it's also not what you mean. Of course it is possible to end up with more information than you started with. That's pretty much the whole idea. What I know you're saying is that it's impossible to end up with more useful or accurate information than you started with, and you do have a point there. I still disagree, but only in specific instances. I would agree with you that it is very easy to end up with several million more pixels of junk. However, it doesn't have to be so.

Imagine a very low-sampling frequency rendering of a sine wave. Something that looks like it came from an Atari 2600 game, for example. Say that's the original data. Now, imagine that you take a look at that rendering. It would be pretty clear by looking at it that it's supposed to represent a sine wave, even though it's blocky as all hell. From the 'original data' -- i.e., the blocky rendering -- you can still immediately see the frequency and the amplitude. Good enough, but is that the end of the line, quality-wise?

Now, let's say you wanted to make a better version of that sine wave, but you start from the blocky version. It doesn't take that much thought to see it's possible (and maybe even relatively easy) to draw a much smoother curve that fits the data. Same frequency, same amplitude -- better resolution. In fact, you could probably downsample and get the same exact original blocky curve back -- or you could downsample less and still get something smoother and more pleasing than the original. Clearly, in this case, you have ended up with a result that has more information than you started with, and that information is useful and accurate.

I will be the first to admit that a full-frame, 24-bit colour DVD image is not a sine wave. It's zillions of times more complex (1 zillion = 10^a lot), and to teach a computer how to recognize where and how to make changes to improve things isn't easy. Still, the principle is the same as in my simple example. It is possible to make improvements to the original data that will give a result that is better than what you started with. It takes a watchful eye and lots of clock cycles, but it can be done. I think to say otherwise (absolutely, no less) is overly simplistic... and incorrect.

That's my take, anyway.

Author
Time
Post link
That's not true -- but it's also not what you mean. Of course it is possible to end up with more information than you started with. That's pretty much the whole idea. What I know you're saying is that it's impossible to end up with more useful or accurate information than you started with, and you do have a point there. No, it is not possible to end up with more information than you started with; and your qualifiers are not necessary ("useful or accurate information"). Increasing the quantity of redundant information is not the same as new information. For example, a single text file that says "hello" = the same amount of information as a million copies of the text file that says "hello". Any possible version you could make of a movie starting from the source is already a potential of the information that is already contained in the source. When you talk about adding things intelligently, then that is different, as you are adding more sources by default when you do that; for example, Lucas' additions in the SE and DVD versions. But simply encoding and filtering and whatnot does not add information. It is only working with what is there.
Now, let's say you wanted to make a better version of that sine wave, but you start from the blocky version. It doesn't take that much thought to see it's possible (and maybe even relatively easy) to draw a much smoother curve that fits the data. Same frequency, same amplitude -- better resolution. In fact, you could probably downsample and get the same exact original blocky curve back -- or you could downsample less and still get something smoother and more pleasing than the original. Clearly, in this case, you have ended up with a result that has more information than you started with, and that information is useful and accurate.
Are you saying that you can look at a blocky sine wave and manually draw a much smoother curve? That is using more than just one source. If you are using your own drawing skills and knowledge of since waves to draw it then you are adding yourself as a source. If you are running it through a program that can interpret the sine wave into numerical data and then draw out a nicer looking wave to represent the same data then you are adding the programming of the software as a source. That is a different breed of cat altogether than encoding and filtering. Maybe one day we will have software that can look at a fuzzy image of a tree and extrapolate that into a picture that appears identical to a high quality image of the same tree; but we aren't quite there yet. This would still be a case of adding sources, rather than gaining information from a simple encoding or filtering process.
Author
Time
Post link
In a nutshell, it would appear that the difference is "Finding more information", as opposed to "Creating more information." This makes me think of digital versus optical zoom on camcorders. Optical zoom allows you to see more true detail, while digital zoom interpolates exisiting information to create more detail.

<span class=“Italics”>MeBeJedi: Sadly, I believe the prequels are beyond repair.
<span class=“Bold”>JediRandy: They’re certainly beyond any repair you’re capable of making.</span></span>

<span class=“Italics”>MeBeJedi: You aren’t one of us.
<span class=“Bold”>Go-Mer-Tonic: I can’t say I find that very disappointing.</span></span>

<span class=“Italics”>JediRandy: I won’t suck as much as a fan edit.</span>

Author
Time
Post link
From what I can see you're both looking at the same 'object' from different angles, no you can't increase the information without a better source but you can present that information better (filtering etc.) and that's what we're doing when we filter our captured video, presenting the video in a visually better way than it currently stands (the laserdisc and the plain capture).

As for the sine wave thing, the edge-directed resampler filter I'm using for VirtualDub is along the same lines, it doubles the image size by looking at the edges rather than blindly scaling the image, the results on hard edges is noticable when compared to Lanczos3 for example, sure it's not increasing the information but it's presenting it in a new (and what I feel to be a better) way.
http://www.haku.co.uk/pics/LukeCruise.gif http://www.haku.co.uk/pics/dontcare.gif
***Citizen's NTSC DVD/PAL DVD/XviD Info and Feedback Thread***