Posts: 1,225
Threads: 51
Joined: 2019 Oct
Thanks: 943
Given 654 thank(s) in 384 post(s)
Country:
2020-11-30, 12:59 AM
(This post was last modified: 2020-11-30, 01:58 AM by pipefan413.)
This one's kind of sad, because it's a notoriously bad LaserDisc for "disc rot" (a.k.a. "laser rot"). The image is a bit iffy because there are lots of dropouts from age-related deterioration, basically.
There is a DVD transfer of this version of the film, however it's more cropped than the original CAV LaserDisc, from a less than perfect source, and has what I reckon are almost certainly very badly wrong colours. The LD one also seems to reveal more detail in backgrounds because the contrast is cranked too high on the DVD which means some faint pencils are lost in clipped-off bright whites. Also, several dissolves are not proper dissolves as they are on the LD, but instead have a scratchy white mark moving from left to right and then a hard pop-in when it switches from black and white pencil drawing to coloured not-quite-final animation.
Some quick side-by-sides to give you an idea of what I mean:
Here's one example of both the rot and the radically different colours:
I'm wondering if maybe the original LaserDisc transfer was from negative whereas the DVD is from a print or something, because only the DVD has things like messy-as-hell cue marks:
This one's a good showcase of how messed up the contrast is on the DVD version:
Whilst I think the LD probably has more correct colours overall, it definitely has the ol' magenta bias that @ PDB and others have noticed on older CRT-graded masters:
Posts: 45
Threads: 0
Joined: 2020 Oct
Thanks: 1
Given 26 thank(s) in 22 post(s)
Country:
Quote:because it's a notoriously bad LaserDisc for "disc rot" (a.k.a. "laser rot")
Many moons ago I read about an avs filter called TooT. I've never capped LDs, but it worked like this IIRC: Because laser rot specks are random enough in the frame, ya do three (or more) captures and run them through TooT. The filter does a compare on each pixel and leaves out the pixel that's the most different of the three, the end result being a wiping out of massive amounts of the rot specks. (TooT="Two out of Three")
No one seems to ever talk about this approach, which I would've thought would've been rendered quite solid after all this time. But mostly I hear/read, "The LD is rotted so it's useless" or similar.
Could this LD here not be salvaged using this (or more modern version of) approach for a nice capture result? I'm guessing the reply will be along the lines of "in practice the two out of three approach doesn't yield good enough results on a well-rotted disc" but I'm curious to hear from someone with experience, so I'm bringing this up here if that's okay.
Posts: 1,225
Threads: 51
Joined: 2019 Oct
Thanks: 943
Given 654 thank(s) in 384 post(s)
Country:
2020-11-30, 02:47 AM
(This post was last modified: 2020-11-30, 02:52 AM by pipefan413.)
(2020-11-30, 02:01 AM)WXM Wrote: Many moons ago I read about an avs filter called TooT. I've never capped LDs, but it worked like this IIRC: Because laser rot specks are random enough in the frame, ya do three (or more) captures and run them through TooT. The filter does a compare on each pixel and leaves out the pixel that's the most different of the three, the end result being a wiping out of massive amounts of the rot specks. (TooT="Two out of Three")
No one seems to ever talk about this approach, which I would've thought would've been rendered quite solid after all this time. But mostly I hear/read, "The LD is rotted so it's useless" or similar.
Could this LD here not be salvaged using this (or more modern version of) approach for a nice capture result? I'm guessing the reply will be along the lines of "in practice the two out of three approach doesn't yield good enough results on a well-rotted disc" but I'm curious to hear from someone with experience, so I'm bringing this up here if that's okay. ![Smile Smile](https://fanrestore.com/images/smilies/smile.png)
I'm familiar with TooT but as far as I'm aware it won't do a whole lot for disc rot because...
1. As far as I can tell the dropouts occur at the same positions in the image each time, so you'd just be comparing pretty much identical frames with dropouts in the same places. (I've captured discs multiple times and compared them manually and found them to contain identical dropouts in several cases. It could be that specific types of deterioration cause less predictable dropouts in the video, whereas others are consistent, but I haven't found it to be improved or changed by repeat captures.)
2. The video isn't the only thing affected by rot: it's causing some distortion in the audio as well.
Posts: 1,640
Threads: 26
Joined: 2015 Apr
Thanks: 39
Given 203 thank(s) in 155 post(s)
Country:
2020-11-30, 04:52 AM
(This post was last modified: 2020-11-30, 05:28 AM by SilverWook.)
It was a notorious rotter. Image did a massive recall where one sent in their rotted set for a shiny new repressing, so good copies are out there. Alas, my entire LD collection is in storage for the forseeable future.
Can the PIP version from the Blu-Ray be extracted, or is it the same as the PAL DVD?
The crossfades on the LD master from WIP footage to finished animation were done for home video. The PAL edit points are likely what the workprint that was screened publicly actually looked like. (I once had a bootleg tape of one other Disney animated workprint, and it was similar in this respect.)
One reason the WIP came out on LD when it did was that Disney was worried about bootlegs made from the LD eating into VHS sales, so the release of the "finished" version on Lasersdisc was not until a year later!
Posts: 45
Threads: 0
Joined: 2020 Oct
Thanks: 1
Given 26 thank(s) in 22 post(s)
Country:
pipefan413 Wrote:As far as I can tell the dropouts occur at the same positions in the image each time
Ah. Well, that takes care of that. ![Sad Sad](https://fanrestore.com/images/smilies/sad.png) What was the purpose of the filter/script then I wonder? I specifically read it was for LD rot cases. Maybe, as you kinda said, there was one kind of rot out there on which TooT would work. Oh well. Thanks for your reply of experience.
Posts: 1,640
Threads: 26
Joined: 2015 Apr
Thanks: 39
Given 203 thank(s) in 155 post(s)
Country:
(2020-11-30, 05:22 AM)WXM Wrote: pipefan413 Wrote:As far as I can tell the dropouts occur at the same positions in the image each time
Ah. Well, that takes care of that. What was the purpose of the filter/script then I wonder? I specifically read it was for LD rot cases. Maybe, as you kinda said, there was one kind of rot out there on which TooT would work. Oh well. Thanks for your reply of experience.
There are different degrees of Laser rot. BATB WIP is probably the worst I've ever seen. When even the audio track is affected, it's pretty severe. My oldest player (made in 1983) can read through mild rot, but BATB it could not cope with.
In theory, one could capture two different copies of the same movie and combine frames?
Posts: 428
Threads: 27
Joined: 2015 May
Thanks: 61
Given 81 thank(s) in 67 post(s)
Country:
IIRC, there have been two different versions of the BatB WiP. The version that was shown at the New York Film Festival prior to the films release, is not the same version that is on the 2-disc Special Edition which was a cleaned up/enhanced version.
You would think the 1992 LD would be the original version, but the 2002 PAL DVD definitely looks to be the worst of the two sources. We need to compare those frames to the NTSC SE version. The original version is pretty rare and I think it's your DVD.
Posts: 7,153
Threads: 601
Joined: 2015 Jan
Thanks: 1081
Given 1463 thank(s) in 963 post(s)
Country:
To "save" notorious rot-prone laserdisc (video wise), the only two ways are:
- find a non-rotted disc (difficult)
- TooT or median or average three different copies
last technique improves also noise, hence it could be used with beneficial effects also using non-rotted discs.
Posts: 1,225
Threads: 51
Joined: 2019 Oct
Thanks: 943
Given 654 thank(s) in 384 post(s)
Country:
(2020-11-30, 10:12 AM)spoRv Wrote: To "save" notorious rot-prone laserdisc (video wise), the only two ways are:
- find a non-rotted disc (difficult)
- TooT or median or average three different copies
last technique improves also noise, hence it could be used with beneficial effects also using non-rotted discs.
Aye *that* makes more sense, capturing different copies. Might be what I end up doing.
One thing I'm not sure on is I've seen two people - but only those two people - claiming that capturing in 10-bit instead of 8 means one is better positioned to adjust levels without clipping in post, but I'm not certain that necessarily makes complete sense and the captures are both larger and more annoying to work with. Do we reckon that's sensible or just daft?
Posts: 1,225
Threads: 51
Joined: 2019 Oct
Thanks: 943
Given 654 thank(s) in 384 post(s)
Country:
2020-11-30, 01:47 PM
(This post was last modified: 2020-11-30, 02:06 PM by pipefan413.)
(2020-11-30, 09:48 AM)Doctor M Wrote: IIRC, there have been two different versions of the BatB WiP. The version that was shown at the New York Film Festival prior to the films release, is not the same version that is on the 2-disc Special Edition which was a cleaned up/enhanced version.
You would think the 1992 LD would be the original version, but the 2002 PAL DVD definitely looks to be the worst of the two sources. We need to compare those frames to the NTSC SE version. The original version is pretty rare and I think it's your DVD.
You know what? It looks like the NTSC DVD uses something of a mix of the two! The colours look more like the LaserDisc, but the dissolves (or rather the lack thereof) look more like the PAL DVD. And the logo is the fancy one from the PAL DVD rather than the old-timey looking one from the LaserDisc:
The NTSC DVD however has the logo fade out into a whole new colour image behind it, which then abruptly cuts (with one of those scratchy vertical line transitions seen elsewhere on the PAL DVD) to a darkened B&W drawing that fades up to colour, before then cutting to B&W pencils again with the musical cues leading up to the "Little town..." lyric. On the LD, the logo fades into the dark B&W image, dissolves into the colour image quite quickly, then does the hard cuts in time with the music thing. On the PAL DVD, the logo does that scratchy vertical wipe thing (which I assume is meant as a cue for "put a dissolve/wipe here") into the dark B&W image then does the dissolve to colour followed by the hard cuts.
In other words: all three of these transfers are different from each other not just in terms of image presentation but also editing. It looks like the PAL DVD might be the closest to the original film fest presentation at least in terms of the editing (because of the lack of dissolves, though it still has some), but the levels look f***ed as shown in some of the screenshots above, and it's cropped down to widescreen instead of being more like 1.66:1. Also, that logo seems like it must've been added more recently because the 1992 LaserDisc one looks much older than the logo on either DVD version.
I want to look into the Blu-ray's picture-in-picture one out of curiosity now but I don't even own the Blu-ray of this one because I know I won't ever want to watch it because of all the changes to it. Hell, even the colours are horribly different.
|