logo Sign In

Doctor M

User Group
Members
Join date
1-Feb-2005
Last activity
5-Sep-2025
Posts
2,545

Post History

Post
#275173
Topic
Info: Hold onto your old Little Mermaid discs!
Link
https://originaltrilogy.com/post/id/275173/action/topic#275173
Time
EDIT: I noticed what looked like blended frames too, but I'm pretty sure it's from when they originally photographed the cells. You can see localized blurring, but sometimes what looks like shadows of a cell buckling and such. In the 2 seconds I worked with any blurring was present in both releases not just the new one.
In fact the restoration fixes some rotation in the cells that is present in the original release.

It's not a matter that the new release is GOOD, it's a matter that the last release was so BAD.

I compared the two and there is no way I would prefer the picture of the old letterboxed, artifact ridden edition.
I still hope for a better release in the future, but since this is what the restoration looks like, expect it to look this bad on Blu-Ray.

And yes, the knees were intact in the featurette, but I believe it was letterboxed (and possibly from the unrestored original release). Funny thing is I never checked before I did the edit. (I'm hoping it wasn't restored or I'll feel like an idiot.)

Edit:

Yes the featurette is letterboxed. But the color timing and framing (there was different cropping in the 2 versions) indicate it is from the restored source AND before they removed the knees. The downside is quality.

Featurette Screenshot:
http://img232.imageshack.us/img232/1381/featuretterm3.jpg

Dr. M's Restored Theatrical Edition:
http://img507.imageshack.us/img507/2884/restoredjr4.jpg

The manual correction might have been more work, but the quality difference was worth it. Even if it is literally 2 seconds of film.
Post
#275111
Topic
Superman by Taolar (Formally: Son of Jorel) WORKPRINT AVAILABLE
Link
https://originaltrilogy.com/post/id/275111/action/topic#275111
Time
Wow, this is one of those instances that shows that the editor is an important aspect of any film.

If I had just seen the Donner footage I would have said there was nothing wrong with it. And then the SoJ version has a real economy to it and makes it a much better action beat. Kudos.

I disagree with a lot of what Tweaker said. I don't see a problem with the pace or the lifting.
I do agree that Reeves has a stupid grin on his face. It looks like you could fix that by cutting away from him sooner (keep the audio but put it over other footage).
Post
#275107
Topic
Song Of The South - many projects, much info & discussion thread (Released)
Link
https://originaltrilogy.com/post/id/275107/action/topic#275107
Time
Hey, I'm not saying I'm right, it's just a best guess scenario. I'm pleased because I couldn't figure out why the LD transfer I had ran short (and I believe it was from the Korean release).
Since there was so much talk of a new official DVD release in the UK, perhaps this is the work that was done and they just decided to go to TV with it instead.
Post
#275072
Topic
Song Of The South - many projects, much info & discussion thread (Released)
Link
https://originaltrilogy.com/post/id/275072/action/topic#275072
Time
What makes us believe this is from a different source is that the runtime is actually correct for the movie and NOT sped up.

The filter used for cleanup is specifically designed to remove the merged fields that occur from an NTSC 29.97fps to 25fps PAL conversion and not the other way around.

The laserdisc transfer I have is indeed shorter than this new release. For whatever reason the source is obviously not a 24 frame film to 24 frame PAL telecine, hence the belief there was a 29.97 intermediate conversion.
Post
#274531
Topic
Song Of The South - many projects, much info & discussion thread (Released)
Link
https://originaltrilogy.com/post/id/274531/action/topic#274531
Time
Before encoding I did a 5% encoding of quality based 1 pass VBR (Q=1) to test for an estimated average bitrate.
The upshot is I had to come down several hundred kbps to fit the movie to the disc.
I'd rather not go to dual layer or compromise the video anymore than necessary.

That being said, I see no problem with someone building a second DVD of extras as a companion since I know there is a lot of cool stuff floating around out there.

Edit: I just got your PM. Ok, maybe I'LL be building a second disc.
Post
#274452
Topic
Song Of The South - many projects, much info & discussion thread (Released)
Link
https://originaltrilogy.com/post/id/274452/action/topic#274452
Time
Ok, my release has been dubbed the "NTSC Restored Mentor-BBC Edition". Any concerns or complaints about that?

Also, screwing around trying to decide on a menu I amused myself with something:

http://img341.imageshack.us/img341/5292/vaultdisneyju7.jpg

I was considering using the above for the menu (without the Disney Vault text) and having the word Play with some pliers... but that's a little too preachy for my taste. :-)

I'm having a hard time finding clean art to make a menu from. PM me if you can help out.

Finally... I'm thinking of using some music from Splash Mountain for the menu since it's cleaner and has a better soundfield. I know it's a bit of a sacrilege.
If anyone has any serious issue with that, speak now.


Post
#274359
Topic
Song Of The South - many projects, much info & discussion thread (Released)
Link
https://originaltrilogy.com/post/id/274359/action/topic#274359
Time
Thanks for the info Moth3r. That's exactly what I did. AviSynth was outputting YV12, and I set Lagarith to YV12 to avoid loss from colorspace change (which is theoretical and nothing I've ever been able to see myself either).

The ghosting in the animated sections are quite odd. Although it seems to only occur about every other field in live action portions, there's something like 50% more blended fields in any section with animation. Sometimes many consecutive.

I don't know if it has to do with the original film, or something else. Since if I threw out 100% of the blended fields there would be large jumps in scenes, it's a necessary evil.

Restore24 is quite intelligent. It compares fields and determines (based on the number of edges in a field) the most likely blended field (and in our case the worst of the available ones) and keeps the best. So although there is still _some_ blending, you can definitely see smoother motion.
The original has ghosts that lead motion and is really ugly to my eye. That issue is largely absent in the final video.

Between that and the sharpening, I'm going to grudgingly say that my final video might be superior to the original...

I've tried to contact Mentor to get his blessing on this project and see how he feels about the use of his name in my final release... but to no avail.
I believe this might become another in the "RtW" series. :-)

(Oh and the posted test clip is not sharpened. It was just a test of restore24's function.)
Post
#274019
Topic
Song Of The South - many projects, much info & discussion thread (Released)
Link
https://originaltrilogy.com/post/id/274019/action/topic#274019
Time
So after much testing, I've decided to drop in the asharp filter to bring up some of the detail lost in the clean up.
I'm usually opposed to edge enhancement, but in this case it looks quite good.

Now I have a tech question for the crowd.
The restore24 filter is VERY SLOOOOOOW.

To do a 5 pass encoding will probably take me something in the neighborhood of 30 hours (give or take)(estimate is based on an OPV encoding of 5% of the film).

Since the filter is the limiting factor is there anything wrong with me instead taking the avisynth output and encoding to a lossless format (HuffYUV or Lagarith) and then doing a mulipass CCE encode from the lossless version? I have to figure that'll cut the encoding time in half or better.

Is there anything wrong with that idea?
Post
#273379
Topic
Song Of The South - many projects, much info & discussion thread (Released)
Link
https://originaltrilogy.com/post/id/273379/action/topic#273379
Time
Yes, there are some sections with large numbers of blended fields all together (especially during the animated scenes) that make it unavoidable that some will be kept.
Also, the software uses a threshold to determine if a frame is blended or not so that'll mean will some show up... but theoretically the best of the source is preserved and the worst discarded.

Not ideal, but pretty darn good considering what we've got to work with.
In fact I'd say it almost looks better than the PAL version because the large number of blended frames can show discernible trails during motion.