BmB said:
The man has no talent. Franchise dead.
Better to deliver the eulogy over a decade late than never at all, I suppose.
Han: I'm calling because I'm planning on asking Leia to marry me and... I was hoping to get your approval.
Bail: Han, let me ask you a question.
Han: Sure.
Bail: How well do you know my daughter?
Han: Um, we've been dating for two years...
Bail: Because if you feel the need to ask me about this scenario, I have a sense you don't know Leia at all.
Han: Sir, I love your daughter and I want to marry her. That's why I'm calling.
Bail: First of all, Han, the truth is this is just a courtesy call. Like when you say to your neighbor, "We're having a loud party on Saturday night if that's all right with you." What you really mean is, "We're having a loud party on Saturday night."
Han: Mr. Organa...
Bail: Leia doesn't give a damn what my opinion is. What interests me is that you do.
Han: It's just a custom to call the father, that's all this is...
Bail: Well, then, I'll tell you what. I may become your father-in-law, that's just fine. But I will not be used as part of a charming little anecdote you tell your friends at cocktail parties so they can see what a quaint, old-fashioned guy Han really is. Are we clear?
Han: Yes, sir...
Bail: Good. Then welcome to the family.
ProgMetalMan said:
So, uh, on topic, I wonder what this guy thought of TPM.
FWIW, "this guy" is Miss Manners. If she saw it, she most certainly hated it. If you take her complaints about Empire and substitute TPM, they all actually apply much better (no plot structure, check, no character studies let alone character development, check, no emotional or philosophical point, check). If you take her compliments about Empire and substitute TPM, they make no sense at all (Yoda is touted as a successful addition--in TPM, yikes, not so much).
Also, I think I'd like to add there's a pretty apparent bias here when you think about who's writing. Her objection to the Force centers on its apparent acceptance of rudeness (she's Miss Manners, remember?). Read through that lens.
I think she's quite right to imply Empire is clearly more directly descended from popular pulp serial works than high culture (score notwithstanding--I do find it odd that this Wagner fan didn't give a little credit for the score). I think at the time she wrote this, a lot of critics would have shared the value judgment she put on that distinction. It's easy to look back on this and laugh because the very idea of high culture has now thoroughly collapsed in the modern world. She didn't know this--she was at the tail end of a different era. She thought Cultured People would still be attending the rarefied opera and shunning the common cinema. That's really where she was ultimately wrong.
Is there any way to let an account keep posting all they want, but just auto-add it to everyone else's ignore list by default? Might cut down on overhead, just sayin'.
It also explains his beautiful smile. He was becoming a millionaire by selling himself toothpaste! Such a cunning plan!
Humby said:
That being said, if he can keep the story focused and character driven, then we're already looking at a film that should be better than the PT and almost on par with the OT.
The ST is already better than the PT.
Is anyone working on a bonus DVD for this release? You know, a non-anamorphic, washed-out, shaky, dirty, smeared version of the Special Edition? Maybe the 97 version but with the 2011 soundtrack? So people could see what the Special Edition really was like, for objective comparison.
That is so far beyond the sort of thing I'd ever have noticed without a side-by-side comparison... and it's great now!
That last script did it! It may not have done perfect IVTC on those problem sections, but I don't see anything obvious anymore. Thanks to everyone for the help!
Rather than re-encode, you could also just demux/remux and burn onto a Blu-ray disc. I believe the problem you're having can't be due to the encoding, and is more likely due to some USB/filesystem/datarate/MKV-specific problem. It's a Blu-ray compliant encode, and I've confirmed it's played on every Blu-ray player I've thrown it at when it's in shiny disc form (not the 83, but quite a few super-old vendor-abandoned models).
Just to update, I'm having much better success now, but still running into snags. Using DGIndex and MPEG2Source per Moth3r's advice after more DirectShowSource problems, and now I'm seeing that in fact it's MOSTLY progressive with a few noticeably interlaced bits (ugh), so I'm now trying TIVTC and hoping for better-looking results. So far, so good, but slow...
EDIT: Did I mention I'm trying this on a Hong Kong film? Apparently crappy telecines are kindof a thing there. Who knew? Latest script below:
loadplugin("dgdecode.dll")
loadplugin("eedi2.dll")
loadplugin("tivtc.dll")
# d2v file created from vob file by dgindex
mpeg2source("input.d2v")
# inverse telecine to reconstruct progressive frames
tfm(d2v="input.d2v")
tdecimate()
# crop top and bottom from the 4:3 frame (60px top & bottom on a perfectly centered image)
crop(0,60,0,-60)
# scale image 2x vertically, interpolating missing fields (slow)
eedi2()
# resize to anamorphic ntsc frame
spline36resize(720,480)
That seems to be the issue. The VOB file shows a length of 1:31:55, the extracted m2v shows a length of 55:11, and that's exactly where it truncates. Rather than trying to get the script to see the correct length on the m2v (probably some framerate screwup in directshowsource(), I'll just use the VOB. It's easier in the long term anyway.
Thanks for the help so far... running into a snag. HCenc (GUI) is only encoding about 75% of the video, even though it says it's doing all frames and it completed successfully. I play the AVS through Windows Media Player and the whole thing plays all the way to the end. 75% is suspiciously close to 24/30, so I'm wondering if it's a framerate problem.
I've been trying to use an m2v file extracted from a VOB file as my source, but I'm wondering if it wouldn't be better to just use the VOB as the source. Perhaps by not using the VOB, I'm losing some framerate/other hints that would help out the encoder. Any other suggestions welcome.
Thanks for the catch on cropping! I think I got those 72 values by working with a PAL source, and forgot to change them for NTSC...
Yeah, I hate having to do the convertto's, but tempgaussmc_beta1() only works in yv12, and I really like its deinterlacing results. Deinterlacing/interpolated scale/reinterlacing seems to produce a less aliased image than the basic script, which IMO looks better, but yes, it certainly makes things a lot more complicated (and slower) than just scaling the image.
EDIT: Wait, I think I got what you were saying about having the full frame and not messing with the fields. I thought crappy old 4:3 letterboxed DVDs would be hard-telecined, but maybe they just have the pulldown flag just like modern DVDs... I'll try the whole process without deinterlacing/interlacing, but still using the interpolated scale, and report back.
I'm having some luck in testing with the following script. Feel free to critique; I'm new to all of this. I haven't actually converted a real video yet.
import("TempGaussMC_beta1.avs")
directshowsource("input.m2v")# crop top and bottom from the 4:3 frame
crop(0,72,0,-72)
# stabilized deinterlace (also doubles frames, so we undo that with selecteven)
converttoyv12(interlaced=true)
tempgaussmc_beta1()
selecteven()
converttoyuy2()
# scale image 2x vertically, interpolating missing fields (slow)
eedi2()
# stretch to anamorphic ntsc frame
spline36resize(720,480)
# compensate for the blurring effect from all of this resizing
sharpen(0.25)
# interlace the video
bob(0.0, 1.0)
# set NTSC video framerate
assumefps("ntsc_video")
EDIT: Script fixes.
I saw there was another upscaling thread, but that was focused on HD upscales and rather than derail that one I thought I'd start my own.
I'd like to convert a 4:3 letterboxed DVD to 16:9 anamorphic (all NTSC). That part is simple enough, but I'd like to retain as much quality as possible and I don't mind a slow script, i.e. use various deinterlacing/scaling tricks to avoid jaggies.
Does anyone have any generic scripts that would be handy to start with? I'm pretty new to avisynth, so verbose is good.
Harmy said:
OK, I'd like to revive this thread, since I want to start working on v2.0 next month.
If you guys have any suggestions about what could be improved over v1.0, now is the time to start discussing them :-)
I'd like to see if you can start your work using snicker's restored-clipped-detail version, if it's ready for the whole film.
Also I'd like to see restored starfields.
Such are the side-effects of using a serial from a keygen program. I'm sure if you contacted DTS-MAS support (or sales), you could work around that little issue...
Spaced Ranger said:
Harmy said:
No matter what I try, no matter what codec I use, I keep getting these rendering errors when rendering AVI.I hate the ghost in the machine. Hope you can excise yours without too much trouble.
This may be the oft-cited monitor calibration issue, but v.2.1 seems too strong orange-hued. I hadn't watched much through v.2.0 development, but it here seems weaker with some hue-steps off in the other direction. Like Goldilocks and the Three Bears' porridge, overlapping the two seems just about right:
Yeah I know this is a week late, but wow that's really nice. AFAICT it's mostly desaturating the orange and leaving the rest about the same. I wonder what this process would do to 3PO though.
timdiggerm said:
I think we just learned that Frink is a fellow Marylander.
Screw that. I think we just learned I might be older than Frink.
Going off to check out assisted living options now...
Harmy said:
OK, here's the promised video:
Absolutely nothing to complain about, Harmy. Very nice.
Also, is it my imagination or are you no longer in your cubist phase? If so, that's even better news!
This shot is from the Senator Tech IB print. Less-saturated blue walls, definitely black Vader.
Yes, admittedly this is from a camera with probable auto-white-balance issues, and who even knows about the projection, but still... I just remember this part of the film had some crazy blue problems in the 2.0 workprints and Harmy opted not to make it so blue based at least in part on the Senator reference.
chyron8472 said:
Also, Harmy, forgive me for saying so, but those 2.1 screenshots look really rather blue.
Yeah, Vader looks pretty blue (I get that he's reflective, but wow he's pretty blue). I don't mind the walls though. Isn't this scene pretty close to those "too crazy blue" shots from Verta's IB reference? Is it possible the Senator IB reference shots could help here?
Dunedain said:
Since this seems like it might be the best quality source available at the moment for the original Star Wars stereo 35mm soundtrack, has Hairy_Hen used this as the source for his restored/corrected 1977 Star Wars stereo soundtrack?
He's used it as a source for his 70mm mix, but for fairly small sections IIRC (an explosion on Tantive IV... and that may be it). H_H's 70mm project predates this one and I believe switching over to an entirely new source for most of the audio would have created more problems than it solved--keep in mind that while this recording is clearly the current best-of-breed for 1977 stereo, 1) it is not without its problems, and 2) its better dynamic range is not such a standout when compared to later audio revisions, from which h_h based much of his audio.
And yes, I'm also interested in hearing Empire and Jedi!
Harmy said:
I just decided to use the LPP print as a color reference here instead of the IB print, because it looks closer to what I also saw on some 8mm prints, so it was probably what it looked like on the more wide spread versions and also because I like it better. Though I made it all a little brighter than in the LPP transfer that I saw, as I believe it probably should be, plus it's really bright on the IB print, so with brightness I stuck to the IB as a reference, since that was scanned on professional equipment.
I suspected that's how you came up with the most recent look, and I think it's a good approach and produced good results.
