Harmy's THE EMPIRE STRIKES BACK Despecialized Edition HD - V2.0 - MKV & AVCHD (Released) — Page 103

Just finished watching V2.0, a vast improvement over V1.0 I must say in terms of colour timing, sharpness and contrast, still not sold on the colour timing of the Hoth scenes though. Can't really explain but something just doesn't look right whenever I watch them, however I don't get that same feeling from watching the Hoth scenes from PSB, it just looks right, but that's just personal taste and I know I won't change your mind on this, Harmy.

Of all the GOUT upscales the ones that stick out like sore thumb for me are the unzoomed shots of the Falcon's cockpit, pretty sure those are priority number 1 as far as 35mm scans go for you.

One question: Why does the scene in which the Millenium Falcon enters the Asteroid Field being chased by the TIE Fighters look so murky now? IMO it was perfectly fine in the original version.

Anyway this is V2.0 and I'm pretty sure you will brush up on any imperfections when you're done with V2.5. Keep up the good work Harmy and may the Force be with you.

I notice a audio glitch in the 5.1 track. Lukes lightsaber ignites twice in the wampa cave. 

Finally watched this version today and all I can say is Thanks so much for all your efforts Harmy and contributors.

Once again it gives the feeling of watching the movie we all fell in in love with many years ago warts and all in a form and resolution that can be enjoyed on a big screen.

Thanks again Harmy keep it up.

Yeah they do look bad as far as i'm concerned, the rest of the movie looks amazing.

doubleofive said:

jacksparrow900 said:

I notice a audio glitch in the 5.1 track. Lukes lightsaber ignites twice in the wampa cave. 

Pretty sure that's supposed to be there in that track. As in, it was there originally.

 The double-ignition is present in every audio track of the movie, even other languages, so it's not a glitch.

For someone accustomed to watching the SE, which has only one saber ignition, it would certainly seem like a glitch.

Whether they did it on purpose, or if it actually is a mistake, is impossible to say for sure.  But every mix ever made of the original version of the film has the double ignition, so there's no way I'd ever try to remove it.

Actually, the funny part of it is that by removing the double ignition, the SE introduces a far worse glitch into the scene, which is that the subsequent sound effects become out of synch to the picture.  They now happen too early, by the same amount of time that the removed first ignition sound would have been.  Consequently you can hear the sound of Luke cutting himself free from the ice before he is actually seen to do so, as well as deactivating his saber while it is still obviously lit onscreen.

That's just plain sloppy work, and this mistake has been there since 1997.  Yet another example of why redoing something that is already complete can be a very slippery slope—you forget why you did things a certain way the first time, and in trying to correct perceived imperfections, greater errors may easily be introduced that were never present previously.

I'm sure this has been asked before but is there a link to a gallery of fixes for this new version? I can't seem to find it if it's been posted.

Great, thanks! I totally missed that in the first post.

I noticed earlier hairy_hen posted this:

"The '93 mix of ESB contains a few glitches, so I did a bit of restoration work on it since it was the primary source for the new 5.1.  This includes declipping, declicking, and noise removal, all with iZotope RX 3; as well as patching editing mistakes in the laserdisc source.  I shall therefore upload this corrected copy in AC3 form to be included in the MKV."

I'm assuming this corrected copy of the audio track is already muxed into the corrected v2 mkv on spleen? Is this correct?

Did hairy_hen end up uploading that track by any chance for others to download?

The mkv does have my fixed 1993 version as a 224 kbps AC3.

To clarify, there aren't many glitches I'm talking about.  The sound quality pretty much craps out for a second when the Falcon attacks the Avenger; there's a jump-cut in the music at the beginning of the carbon freeze scene; and a few instances of clicking and clipping scattered throughout.  All the laserdiscs I've heard contain these flaws, but the GOUT corrects the jump-cut and the Avenger flyby.  The '95 Faces vhs may have also had these fixes, if I remember correctly.  My correction of the Avenger scene sounds better than the GOUT, because I think they spliced it in from a different mix to cover it up, while I was able to use the Spectral Repair and the Decrackler tools in iZotope RX 3 to eliminate the distortion while keeping the rest of the sound intact.  I was pretty amazed it worked as well as it did, though it took several tries to get it right.

I did have to do some restoration on the 35mm stereo track also.  The main difficulty there was that the older laserdiscs were missing so many frames at reel changes compared to the GOUT, such that the cuts in the audio were quite noticeable.  The flow of the music would be totally interrupted in these spots, if a music cue was playing over a reel change.  To avoid loss of synch and to eliminate these annoyingly bad edits, I ended up patching segments of the 1993 mix into the 35mm to smooth over the gaps and allow it to play continuously throughout.  These patches were meticulously matched in level and EQ, as well as having their stereo width narrowed by 12%, in order to seamlessly match the rest of the 35mm.

I also used some denoising at the very beginning and very end of the track, since these parts contained excessive hiss.  Everything in between is largely untouched, except that mild limiting had to be employed to prevent the loudest parts of the track from clipping once the overall gain had been raised to match the 5.1 version.  These edits are all carefully integrated and I very much doubt if anyone will be able to hear them (I certainly can't).

In terms of content, the 1980 and 1993 versions only occasionally differ, having come largely from the same 4-track source.  The '93 is more dynamic overall (though the '80 isn't far behind, surprisingly), and has more bass, while the '80 has more emphasis in the midrange.  Each has a different vibe, but both sound great in their own way.

About that missing snowspeeder crash: I believe the explanation for its being missing in the 1993 mix is because it is not present in the 4-track master from which it was derived.  The theatrical printmasters had it, but this earlier generation copy did not, so when the '93 mix was being made for the Definitive Collection, they simply didn't notice its absence.  The track feels very quiet without it, which is probably why the 35mm version has it mixed very prominently (compensating for its prior absence).  Later copies, such as the 16mm mono and the SE, dialed back the level somewhat.

I shall make my tracks available for download together once I've finished them all.  It would be good to have them in a torrent or something, perhaps in various formats to be easily used with different kinds of video.

The better the fidelity of your equipment, the easier it is to hear the difference between lossy and lossless.  For most people, the main barrier to hearing the difference is not any deficiency in their hearing per se, but the simple fact that great quality analog gear tends to cost quite a lot more than they are willing to spend.  Once you have the opportunity to do some A/B comparisons on a real hifi system, you may well surprise yourself with what you can actually perceive.

Note that these kinds of differences are easier to hear with music than with sound effects.

For a really good explanation of the effects of different audio resolutions, watch the presentation called "Lost in Translation" by mix engineer Andrew Scheps: link

I had the opportunity to meet Mr. Scheps when he gave this talk again a few weeks ago, and he's a really cool and knowledgeable guy.  As part of the presentation he set up various listening comparisons between lossy and lossless formats on some ridiculously good speakers, and in every case the uncompressed versions sounded better.  His theory as to why this is true is that it's not even because the lossy was 'degraded' per se, but because removing too much information makes the sound more fatiguing and less engaging.  Lossy audio forces your brain to work harder to fill in the gaps, and over extended listening periods this will tire you out mentally, making the experience less enjoyable.  Higher quality audio, by contrast, can be listened to for a long time without any associated unpleasantness.

"should I be able to hear the difference from an average mp3 there?"

For a high-quality MP3, probably not. Some can.

There have been tons of tests on this, and the answer is that even with very good equipment most listeners can't tell the difference between a good mp3/AAC file and a higher-quality file in AB and ABX tests (where you listen to multiple clips blind and then have to identify which is which). This infuriates audiophiles who've spent fortunes on good equipment, but the science is there.

You want lossless sources of course so you can manipulate them without without introducing error, but for almost all listeners, compression is fine. (I know I'l have people chime in saying "I can hear it." Maybe you really can, but I'd wager if you took an ABX test under scientific conditions, you'd be surprised.)

128kbps mp3 vs FLAC is for me immediately distinguishable.

256kbps mp3 vs FLAC is for me just hardly distinguishable.

320kbps mp3 vs FLAC is for me absolutely indistinguishable.

I hope that words "distinguishable" and "indistinguishable" do exist :-)

Yes, of course, people wouldn't hear them. But people do hear that "something" is missing, throwing out frequencies causes "damage" also to the kept sound