spoRv

G,

I didn't want to be sarcastic, really! After looking *carefully*, at a lot of frames processed with my script, I must admit that there are good things, but bad ones too...

So, I used the plugins you suggested, and it was better; yes, my script retain more details, but also more noise... please download the new clip, resized, and then LimitedSharpenFaster and MCTemporalDenoise:

SENDSPACE

and let me know if could be a good choice for OUT ruLes 2.0 -> by the way, the processing time was incredibly high! 0.05 to 0.09FPS!!! Many hundred (up to one thousand) hour to process the whole movie? NO WAY!

So, I must find a way to accelerate the dub time - my PC will be fried after 40 days! - I thought to save the unprocessed upscaled version in lagarith, then let the script use this last file, instead of doing all in real time (load captures, median/average, upscale the result, then sharpen/denoise)...

OR I should find another script.

What about "Avisynth best script to enhance Star Wars restoration project that uses only laserdisc captures" contest? (^^,)

g-force said:

Just watched the OUT sample you posted. Large halos, trails of grain.

I thought you were giving up on the "enhancement". You are using a horrible denoiser, and an even worse sharpener.

Would you like to give me a gift, in form of denoiser/sharpener/upscaler avisynth script?

I made two test clips - one for OUT, one for SET; enhanced, upscaled and merged with unprocessed to smooth a bit the image.

SENDSPACE

results are not comparable to BD, but not bad, considering the source is laserdisc!

SET lacks some details, because I used three different players (two of them with less resolution than the third) and only one edition; next time will be (almost surely) better, using the best player with three different editions.

Feedbacks will be appreciated.

OK, let's dump enhance script #2...

Can you point me to the best (in your opinion) avisynth enhancement script (sharpener, denoiser etc.), so I could build up a comparison, and eventually find a solution or some inspiration?

just uploaded here: Sendspace - it was really hard to find... it's an avisynth plugin treasure!

You_Too said:

While the untouched capture has rainbowing, noise and stuff, for some reason the enhanced versions have halos around edges, looking a bit like edge enhancement. For example in the first shot, the edges are sharp and clean in the first pic and in the enhanced there's dark lines running along the edges of the moon and the star destroyer.

Maybe something went wrong when you did the median? Were the captures perfectly aligned?

You_Too, here you are another image - this time PNG, 1,2MB, maybe the JPEG compression made the problem worst last time... there are four versions: top left, 925 enhanced; top right, average NOT enhanced; bottom, two slightly different enhanced versions: could you tell me where the problem lies, and if the enhanced is a little better than before?

If there are dark lines along edges on the first pic, the fault lies in the enhancement script; if there are on the second pic then the fault is on the capture alignment - that for this test was offset by 1 pixel on chroma on one capture; if there are in the one (or both) of the bottom pics, the problem lies in enhancement AND average...

That is why sometimes is better to use only one laserdisc player (or at least the same model, different units) with different edtions, because the captures will be perfectly aligned.

On the other hand, using different players, with different qualities and defects, result in an interesting mix when medianed and/or averaged, but more difficult to align perfectly. What do you think?

And, thanks for the constructive feedback!

You_Too said:

It's quite complicated actually. Not like our GOUT project but still.

I did it all in avisynth. Since there was a lot of blue and cyan in the 97 SE I first used selectivecolour to lower those colors a bit. Then I made histogram files for colourlike, using some shots from the 97 SE and some from theatrical prints like the Senator screening.

I then made a 50/50 of the histogram conversion result and the cyan/blue fix so I got the best of both and finally, I made an additional RGB curve to correct the pink tint in the Mos Eisley scenes.

I admire your hard work! But it's too much for me... my restoration experience is limited, and my time too, as I have family, a dayjob, and... cats! ,,,^..^,,, and I'd like to release this project ASAP.

But some color correction is needed, so I tested a lot, and find a simple, but effective, method; merge the original laserdisc image (that is less pinkish than GKAR. by the way) with the same image, auto color corrected. I know it's not the perfect, definitive solution, just only a compromise - but a working compromise is better than pinkish images! (^^,)

Well, let me explains some of my later tests: I captured the french laserdisc (side 1 for the moment) with three different players, directly without passing through any dvd recorder; the players used are all Pioneer, CLD-D925 via S-Video, LD-V4300D and CLD-600 via composite; then I averaged/median the three captures, to obtain less noise; then I averaged/median the resulted image with the auto color corrected one. For comparison, I used GKAR, unprocessed, only converted from DVD to AVI (Lagarith YV12) using DVD2AVI.

Here you are some screenshots: top left image is raw capture from CLD-600; top right is the same capture, enhanced; bottom left is the average/median result of the three captures, enhanced, color corrected, resized to match GKAR; bottom right is GKAR, unprocessed. As I wrote before, the GKAR picture is always pinkish, while the laserdisc is pinkish only sometimes - at least in this side... obviously, GKAR has a better definition.

Starship is white on SET, pink on GKAR; planet is brown on SET, pink on GKAR; engine booster is right in SET.

Wall is white on SET, pinkish on GKAR; C3 is golden on SET, golden-ish on GKAR

Stormtrooper is white on SET, pink on GKAR

Darth Vader is black on SET, dark violet on GKAR

Leia's face color is more natural on SET, and whites are "whiter"

Sand is yellow in SET, orange-pink in GKAR

R2 is white in SET, pinkish in GKAR; the rocks color in SET is more similar to OT

I don't know if, in GKAR desert, they are sand waves, or MPEG waves... really, I'm not joking!

Maybe Luke is a bit pale on SET, but better pale than sunburnt as in GKAR!

The ground is brown in SET, violet in GKAR

Note RD's dome; still silver on SET, but what color is on GKAR?

C3 is golden on SET, pinkish on GKAR; in GKAR Luke's clothes are pink instead of white

Still working to improve the quality of the laserdisc captures; I'm almost sure that using three different editions (french, english, german) the noise will be lower; as the picture of 4300 and 600 is noisier and less detailed, I think I'll use only the 925 for all the captures - but I could also test the DVL-909, and maybe give a try on the Philips LDP-600WS...

OR, I could try the "extreme capture experiment (version 1)":

1. capture three times each edition; TOOT them to avoid drop-outs
2. repeat for three laserdisc editions
3. results are three TOOTed captures, one for each edition
4. average the three TOOTed captures
5. repeat for at least three laserdisc players
6. results are three averaged captures, one for each laserdisc player
7. median the three averaged captures

                     |capture 1|
          |->english-|capture 2|->tooted-|
          |          |capture 3|         |
          |                              |
          |          |capture 1|         |
cld-600---|->french--|capture 2|->tooted-|->average-|
          |          |capture 3|         |          |
          |                              |          |
          |          |capture 1|         |          |
          |->german--|capture 2|->tooted-|          |
                     |capture 3|                    |
                                                    |
                     |capture 1|                    |
          |->english-|capture 2|->tooted-|          |
          |          |capture 3|         |          |
          |                              |          |
          |          |capture 1|         |          |
cld-d925--|->french--|capture 2|->tooted-|->average-|->median
          |          |capture 3|         |          |
          |                              |          |
          |          |capture 1|         |          |
          |->german--|capture 2|->tooted-|          |
                     |capture 3|                    |
                                                    |
                     |capture 1|                    |
          |->english-|capture 2|->tooted-|          |
          |          |capture 3|         |          |
          |                              |          |
          |          |capture 1|         |          |
ld-v4300d-|->french--|capture 2|->tooted-|->average-|
          |          |capture 3|         |
          |                              |
          |          |capture 1|         |
          |->german--|capture 2|->tooted-|
                     |capture 3|

OR, I could try the "extreme capture experiment (version 2)":

1. capture three times each edition; TOOT them to avoid drop-outs
2. repeat for at least three laserdisc players
3. results are three TOOTed captures, one for each laserdisc player
4. average the three TOOTed captures
5. repeat for three laserdisc editions
6. results are three averaged captures, one for each edition
7. median the three averaged captures

                     |capture 1|
          |->cld600--|capture 2|->tooted-|
          |          |capture 3|         |
          |                              |
          |          |capture 1|         |
english---|->cldd925-|capture 2|->tooted-|->average-|
          |          |capture 3|         |          |
          |                              |          |
          |          |capture 1|         |          |
          |->ldv4300d|capture 2|->tooted-|          |
                     |capture 3|                    |
                                                    |
                     |capture 1|                    |
          |->cld600--|capture 2|->tooted-|          |
          |          |capture 3|         |          |
          |                              |          |
          |          |capture 1|         |          |
french----|->cldd925-|capture 2|->tooted-|->average-|->median
          |          |capture 3|         |          |
          |                              |          |
          |          |capture 1|         |          |
          |->ldv4300d|capture 2|->tooted-|          |
                     |capture 3|                    |
                                                    |
                     |capture 1|                    |
          |->cld600--|capture 2|->tooted-|          |
          |          |capture 3|         |          |
          |                              |          |
          |          |capture 1|         |          |
german----|->cldd925-|capture 2|->tooted-|->average-|
          |          |capture 3|         |
          |                              |
          |          |capture 1|         |
          |->ldv4300d|capture 2|->tooted-|
                     |capture 3|

totalling 27 captures (each 3 sides) = at least 54 hours of captures, and about 1.5TB of hard drive!

Any idea or experience to share?

ilovewaterslides said:

Finally! I have it, i noticed two problems with the French audio track.

...

And the audio goes out of sync at the end.

I investigated, and found that side 4 of the french THX laserdisc was out of sync... I put both english and french soundtracks on Audacity, to find the exact delay; I sync'd the french track with the english one, when no dialogs are present, and I saved the french track... that was out of sync for dialogs... then I tried to sync "by eye" the dialogs, but the effect was too much out of sync...

At the end, I reached a compromise, so both the dialogs and sound effects are out of sync (one positive, one negative), but the result is better than before (where both were off sync of about 700ms...)

Here you are a rough test clip, just to hear if the sync could be tolerable enough:

SENDSPACE (10MB)

You_Too said:

Don't know if this is any helpful but here's how those two shots look in my and DJ's 97 SE project, sourced from the TB broadcast.

Is it possible to know which method is used for color correction?

And, did you color correct the whole movie, only selected scenes with the same settings, or selected scenes with different settings?

pittrek said:

Too green. WAY too green

yes, I know, it was made this morning (was half asleep)... I forgot to add further tests... here they are:

YUV, chroma R and B changed - still too pinkish

HSV, saturation and value changed - colors quite good

RGB, green and blu changed - still too green

YUV, chroma R and B changed - too blue

Now, what is the image with the least uncorrect colors?

Test 2 has the same settings than test 1, apart saturation, so you are right... when saturation is applied on test 2, it is like test 1 - pinkish...

So, I stacked test 2 and test 2 with saturation, and applied the VDub gradation curves filter, and here you are the results:

This is the only setting I touched on the filter:

Must be tested further, but maybe this is the way to go...

About CLD-600: its image is definitely too soft... is it its own fault, or is it the Pioneer DVD recorder's comb filter - which, by the way, has the detail setting at max?

Here you are fast test I made right now; GKAR is converted from VOB to AVI (with DVD2AVI) then reduced to 704x340 to match captures resolution.

Captures are rough, taken with Pioneer CLD-600#1 composite to Pioneer DVR-320s, from there to PC via Svideo; source French THX laserdisc.

Test2 color is pretty good.

Further testing on monday, "second capture test day" of SET ruLes!

pittrek said:

Thanks man. Or thanks lady ? Anyway thanks for archiving the web :-)

Sorry, I always forget that Andrea is a female name in US and UK... I'm a man, and I changed my signature to state that - in Italy is a male name, and there are at least half million of Andrea here! (°L°,)

ilovewaterslides said:

Finally! I have it, i noticed two problems with the French audio track. The voices are sometimes too low it's hard to understand what they are saying. And the audio goes out of sync at the end. Beside these two problems, your preservation is really well done and looks very good. Nice job! ;)

Thanks for the feedback; about the volume problem, it may be due to the fact I converted the 2.0 PCM Dolby Surround to 5.1 AC3 - the plugin didn't work to well; for the sync problems, I checked it and you are right, so I tested video with original track, and it's still out of sync... I checked the original capture, and it's out of sync, too! So, either the capture has some sync problem, or the original french soundtrack is out of sync on the laserdisc itself... I must investigate further!

g-force said:

 It's hard to tell with the .gif yo posted. I'm probably not the person to ask anyway, as I've never seen a denoiser that I like.

I don't like denoiser a lot, too, but it's a "necessary evil" when processing a laserdisc capture, do you agree?

Anyway, I'm still testing denoisers (and related settings)...

I remember I found this thread (and thankfully the whole forum) when I was looking for info about the Pioneer HLD-X0 and X9, the latter bought some time ago (as I cannot find an X0 at a reasonable price -> less than a used car).

This thread, with all its technical info, restoring ideas, surprising images... gave me the "restoration spirit", and I'm sure many of you have started your own project thanks to this thread, too, isn't true?

So, thank you, XØ Project, and thanks to its members!

Happy to spread the x0project website... is it time to build an x9project website involving the use of Technidisc laserdiscs as source? I'll put the HLD-X9... (^^,)

I'm happy to see your interest into this project: hope to start soon to capture - after further tests, really needed to discover the best PAL laserdisc in the world my home...

You_Too said:

Oversharpening artefacts in the source + digital broadcast compression artefacts. At least that's what I think.

Anyway, nice idea and looking forward to see the results of this project!

I know that "gibberish" were digital compression artifacts... I wanted only to test your attention (^^,)

Infact, I started to watch GKAR ("know your enemy...") but after some minutes I stopped; it's true that the resolution is there, but also many distracting artifacts, and wrong colors...

I'm curious to see the results of my project, too!

Jetrell Fo said:

Sweet, I'm looking forward to this immensely.  I am also excited to see where this project goes.

Rockin!!!!!!

Stay tuned! If it takes the same time of [OUT ruLes], it will be ready a month after the captures will be done...

CapableMetal said:

I think that's your reason right there. The broadcast versions are all anamorphic, so have much more vertical resolution than the letterboxed laserdiscs. I'd much rather watch with a 'clean' picture missing some very fine detail than one that has it but is covered with digital artifacts!

Wait! The capture tests are not over... maybe I will find a laserdisc player with more resolution - and I have some suspect it could be the 925...

Are you going to remove the awful pink tint from certain scenes on Tatooine? I've never seen a '97 release that doesn't have it.

Let's test! Tell me which scene is the worst, in your opinion, and I'll capture it this very night - until midnight, of course, I have /still) a day job!

I'm thinking about TPM [ruLes] using the (second) best laserdisc ever made, and two different copies of the laserdisc.

That would be excellent, the laserdisc has the theatrical cut, I believe, and we have the DTS audio for that also, but only a digital broadcast to sync it to.

A [TPM ruLes] with a DTS soundtrack? I'm in!

BTW, my DTS syncs are almost prepared for you, I'll PM you when they're available.

Perfect, thanks a lot! Your work is precious for my project, really!

My pleasure, of course. Seeing as you're using PAL sources it should be easy to make your SET ruLes project conform to the 'standards' we have set in the other thread, which is basically just reel change frame numbers, and will mean any audio you create can be used for other projects (my audio has been to conformed to the numbers there), and other audio could be applied to your video (such as other language tracks) in future.

I'm also happy to lend any other assistance should you need it on the audio front ;)

Thanks thanks thanks! This will be the written proof, when I will ask for your help! (*-*,)

Yesterday I finished my first "test capture day" of SET ruLes!

I tested the Pioneer LD-V4300D (unit #2, as in my first laserdisc capture comparison test, done more than a year ago, someone told me that the unit I used that time, #1, may have some problems); with and without both the DVD recorders (used simply as comb filters), and different connections (S-Video, RCA, BNC); I used the french and the german laserdiscs, captured small portion (2 minutes) of the same scene, then aligned temporally and spatially, then took the GKAR DVD and converted to AVI, for comparison - no other manipulation was made on the GKAR.

These are some of my conclusions:

• the 1997 SE has all the four sides cropped, compared to OT ... indeed, its picture is zoomed, too - as the image is 340pixels high, vs 324pixels of the OT THX laserdiscs...
• the LD-V4300D seems to have less details than the CLD-D925 (unit #1) I used for OUT, as I made direct comparisons between the same scenes captured with both the players (with 4300 yesterday, with 925 a year ago)...
• the BNC output seems slightly better than the RCA - maybe due to the better connections
• the colors are wrong - as many here have noted, white is often either blueish or pinkish, but wait... maybe I found where the problem lies, and a solution... follow me!

Well, my capture card internal comb filter is good, but not the best; the direct connection of the 4300 (as it was for the 925) is worst than the S-Video; as the 4300 has not a S-Video out, but only the composite out (RCA and BNC) and RGB (yes, maybe I have to test it too...) the only solution is to feed the composite laserdisc player output to an external comb filter, and here the two DVD recorders come in help!

The Panasonic DMR-ES15 has a better comb filter than my capture card, but no settings at all, so I take my Pioneer DVR-320S out of a dusty closet where it was sleeping for a long time, and, surprise! It has a lot of comb filter settings, that I forgot he had!

Between them, there is an interesting one: white AGC (stands for automatic gain control). I capture with the white AGC on, and the colors are wrong - as through the Panasonic and direct composite to capture card, BUT... if I turn it off, and set the white and black to the max... BAM, the colors are right - or, at least, better than with AGC on, and similar to OT. So the problem could lies in the white AGC, that maybe must be included in a lot of comb filters, like capture cards, TV sets etc.. or, because the Pioneer DVD recorder comb filter has so many settings, that set the white and black level at the max, it helps to recover the original colors (or most part of it)... I don't know, I'll leave this problem to the experts!

I managed to do the following comparison with avisynth and virtualdub: I took two different captures - french through the 320 S-video, AGC off, no color correction, and german through ES15, with saturation=0.8 - then overlaid the two (the alignment IS NOT perfect, because taken from two different A/D converter, so the two captures have different widths, but hey, this is just a test), then enhanched with the script I used for [OUT ruLes] - the one with the new denoiser; at the end, resized to match the GKAR height. GKAR, as written before, was only converted from MPEG2 to AVI using HuffYUV - as the laserdisc captures.

The wall on the left is white on SET, blue on GKAR; the light on right is white on SET, pink on GKAR.

Wall is white on SET, pink on GKAR, but Vader's buttons are less detailed, and...

what is that gibberish on GKAR's picture, bottom right?!?

OK, I admit SET lacks details also if its picture is enhanced, but, as explained before, it *could* be because the laserdisc player used for the captures is not the best OR the very unit I used is faulty; because the spatial alignment is not that perfect; because there are only two captures instead of three... I'm trying to find some other excuses, but maybe the missing details are due only to the fact that those laserdiscs could resolve about 544x340=184960pixels, where GKAR has 704x416=292864pixels... so GKAR DVD has about 50% more resolution than the SE laserdiscs... I'll find out the truth!

But, as you could see, no blockiness, no compression artifacts, better colors... hope to find a better capture method, to improve resolution.

Soon other capture tests...

CapableMetal said:

Does this mean you aren't going to finish capturing the OUT?

Absolutely not! It means that, as I worked A LOT with ANH for the [OUT ruLes] project, now I know its reel changes, missing frames etc. and my mind is "fresh" enough to work on another ANH, this time the 1997 SE... if I finish all the OUT and then begin the SET, I'll surely forget many many things...

After I've done the ANH SET, I'll go for TESB OUT, then TESB SET, then ROTJ OUT, then ROTJ SET... at the end, if I will be still alive (xx_), I'm thinking about TPM [ruLes] using the (second) best laserdisc ever made, and two different copies of the laserdisc.

Talking about it (the Pioneer HLD-X9), if someone would be so kind to provide me the OT Technidisc laserdiscs, I could capture them with it...

BTW, my DTS syncs are almost prepared for you, I'll PM you when they're available.

Perfect, thanks a lot! Your work is precious for my project, really!

_,,,^..^,,,_ presents the [SET ruLes] project

Special Edition Trilogy restored using Laserdisc editions

Project mission: restore the 1997 Special Edition Trilogy using audio and video from various Laserdisc editions.

Video sources: analog video captured directly from Laserdiscs; other analog sources are admitted (VHS, CED, VHD, Beta etc.) only if the material is not present on the Laserdisc format.

Audio sources: uncompressed PCM and AC3 soundtracks captured directly from Laserdiscs; DTS Digital Surround theatrical soundtrack from CD-ROMs, other analog sources are admitted (like VHS, vynil etc.) only if the material is not present on the Laserdisc format.

Subtitles sources: Project Threepio (for english and german), literal subs for italian.

Project prohibition: to use DVD, BD, DVB, HDTV or any other audio or video from any digital sources; exception is DTS original theatrical soundtrack, because it was originally released on theaters and never commercially availble after that, and because it is of the same kind (bitrate, bit depth, frequency) of any other laserdisc DTS soundtrack.

Supplemental material: will be released in a future [ruLes] project.

Episode IV - Star Wars: A New Hope [SET ruLes]

Source Material

Video: THX PAL laserdiscs

• Star Wars Trilogy: Special Edition - CLV
• Star Wars: Coffret Trilogie: Edition Spéciale - CLV
• Krieg Der Sterne Trilogie: Special Edition - CLV

Video: PAL VHS HiFi

• Star Wars: Guerre Stellari - Edizione Speciale (only for italian crawl and alignment)

Audio: THX PAL laserdiscs

• Star Wars Trilogy: Special Edition - English DS
• Star Wars: Coffret Trilogie: Edition Spéciale - English DS (with French subtitles OOP)
• Krieg Der Sterne Trilogie: Special Edition - German DS (with German crawl)

Audio: Theatrical DTS soundtrack

• Star Wars: A New Hope - Special Edition - English DTS CD-ROMs

Audio: THX NTSC laserdisc

• Star Wars Trilogy: Special Edition - English AC3 5.1, CLV/CAV

Audio: PAL VHS HiFi

• Star Wars: Guerre Stellari - Edizione Speciale - Italian DS

Subtitles: english, german, italian (full and forced); french and spanish will follow.

Capture setup & processing info

Hardware:

• Pioneer LD-V4300 PAL/NTSC laserdisc player (6x) *
  
• Pioneer CLD-D925 PAL/NTSC laserdisc player (3x)
  
• Pioneer CLD-600 PAL laserdisc player (2x) *

• Panasonic DMR-ES15 DVD recorder (as comb-filter) *
• Pioneer DVR-320S DVD recorder (as comb-filter) *
  
• Pioneer SP-D07 Dolby Digital decoder (as AC3 RF demodulator)

• Panasonic NV-FS88B S-VHS recorder
• Sony MDS-J520E MiniDisc recorder (as analog/digital converter)

• Sony Vaio VGX-XL201 PC Core 2 Duo E6300 1.86GHz, 3GB RAM, 250GB HDD
• AverMedia BDA A16C internal video card (with Philips SAA713x chipset)
• Western Digital Elements 1TB external USB HD (only for Star Wars)

• Monster Cable S-Video 5pin cable (very high quality)
• Monster Cable S-Video 5pin cable (high quality)
• LINN composite RCA video cable (very high quality)
• Audison stereo RCA audio cable (very high quality)
• SONY optical S/PDIF cable (very high quality)
• SONY optical S/PDIF cable (high quality)
• unbranded BNC/RCA & RCA/BNC adapters (medium quality) *
  

* this tools will not be used anymore for this project

Software:

• VirtualDub as capture software
• VirtualDub and AviSynth for video editing
• AviSynth + SoundOut for audio conversion

Settings:

• video captured at card native resolution 720x576 @25fps, YUV, HuffYUV compression
• PCM digital audio from Laserdisc to PC via S/PDIF bit-perfect 16bit 44,1KHz stereo
• AC3 digital audio from Laserdisc to DD decoder via S/PDIF, then to PC via S/PDIF bit-to-bit perfect 16bit 48KHz 5.1 surround
• DTS theatrical digital soundtrack extracted from CD-ROMs
• analog audio from VHS recorded through the MiniDisc A/D converter 20bit Sigma-Delta modulation and then to PC via S/PDIF bit-perfect PCM 16bit 44,1KHz stereo

Project status: RELEASED

Captures:

testing various combination of laserdisc players (different models and units), DVD recorders (different models) and comb filter settings, connection methods (S-Video and composite, RCA and BNC) and brands.

After many tests, I decided to use only the Pioneer CLD-D925 as its video quality (resolution, noise, drop outs) is better than the others; the connection used is S-Video, as the laserdisc player comb filter is better than the capture card's.

Video treatment:

all laserdisc video captures will be time and space aligned, then with AviSynth median and average scripts will be overlayed to eliminate some noise and drop-outs, and put out unwanted video parts (as burned subtitles); color correction, at the contrary of OUT, was necessary... after many tests, I found the best compromise to retain the original colors when correct, and to correct them when wrong - in particular during the Tantive corridors initial scenes and Mos Eisley; little noise reduction and sharpen; upconversion from PAL 720x576 4:3 to 1280x720 16:9.

Video tracks:

• the original video format is 720x576@25fps 2.25:1 letterbox PAL, with a video resolution of visible image around 700x340, captured using the HuffYUV lossless codec; video is upconverted to 1280x568@23.976fps 2.25:1 as source, inside a 1280x720pixel 16:9 frame, using x264 codec and compressed at 8000kbps.

Audio treatment:

the original soundtracks are: laserdisc PCM 16bit 44.1KHz 1441kbps stereo, Dolby Surround encoded; laserdisc AC3 16bit 48KHz 384kbps 5.1 surround; DTS CD-ROMs 16bit 44.1KHz 1441kbps 5.1 surround; VHS HiFi recorded in PCM 16bit 44.1KHz 1441kbps stereo, Dolby Surround encoded.

All the soundtracks will be converted only when necessary; the conversions will be: from PAL 25fps to 23.976fps (pitch corrected when needed); from 44.1KHz to 48KHz; from PCM 1441kbps to AC3 384kbps; from DTS 16bit 44.1KHz 1441kbps to DTS 16bit 48KH 1536kbps and 768kbps. Only minimal treatments will be done, right for the conversions (NO audio enhancement).

AVCHD Audio tracks:

• english soundtrack from NTSC THX laserdisc, AC3 5.1 Dolby Digital 384kbps 48KHz
• english soundtrack from english PAL THX laserdisc, AC3 2.0 Dolby Surround Encoded 384kbps 48KHz

BD Audio tracks:

• english theatrical soundtrack from CD-ROMs, DTS 5.1 Digital Surround 1536kbps 48KHz
• english soundtrack from NTSC THX laserdisc, AC3 5.1 Dolby Digital 384kbps 48KHz
• english soundtrack from english PAL THX laserdisc, AC3 2.0 Dolby Surround Encoded 384kbps 48KHz
• german soundtrack from german PAL THX laserdisc, AC3 2.0 Dolby Surround Encoded 384kbps 48KHz
• italian soundtrack from italian PAL VHS HiFi Stereo, AC3 2.0 Dolby Surround Encoded 384kbps 48KHz

Subtitles: to be refined; SRT format, both full and forced, English, German, Italian; French and Spanish will follow.

Subtitles tracks:

• english full 100% literal - from threepio project
• english forced 100% literal - from threepio project
• german full
• german forced
• italian full 100% literal - adapted by me from the italian script
• italian forced 100% literal - adapted by me from the italian script

Conversion from 25p to 23.976p: done with TsMuxeR

Final formats:

• AVCHD folder, 720p/23.976fps with two english audio tracks and six subtitles (three forced and three full)
• BD folder, 720p/23.976fps with five audio tracks (including DTS theatrical one) and six subtitles (three forced and three full)

• other formats will maybe available in the future

Actually AVCHD and BD are both available on myspleen.

ATTENTION: to view the AVCHD version of SW:ANH [SET ruLes 1.0] correctly, it should be burned onto a DVD9 or BD-R, and played on a stand-alone BD player or a Playstation 3! - To view the BD version of SW:ANH [SET ruLes 1.0] correctly, it should be burned onto a BD-R, and played on a stand-alone BD player or a Playstation 3!

Jobs to do for the next version:

• french and german literal subtitles synchronization
• adding eventual subtitles in other languages
• adding the eventual french and spanish soundtrack, if could be found
• correct eventual video, audio, subs errors

To do later:

• author AVCHD, DVD, Blu-Ray discs properly (menus, audio and sub tracks etc.)
• provide opening crawl in german and italian (A LOT later...)

CREDITS:

• Capable Metal for the synchronization of the DTS and laserdisc AC3 soundtrack, and for color correction tips
• Jetrell Fo, Borisanddoris and Eldonante for their work about the 1997 SE Trilogy DTS theatrical soundtrack
• marvins and spielt mit for their french and german OUT literal subtitles, that will be used also for SET soon
• AntcuuFaalb, ilovewaterslides, Capable Metal, msycamore, You_Too, g-force for their support, comments, hints and technical help about my [OUT ruLes] project - really helpful for this new project, too
• Harmy, Moth3r, Zion & laserman & MeBeJedi @ X0project, and all the other restoration projects founders and members, to have inspired me to start my own Star Wars restoration projects!

Self comments: There is no official DVD or Blu-Ray disc of the 1997 Special Edition trilogy (and never will be, I'm afraid to add); thankfully there are many DVB broadcasting around, released on DVD (like GKAR, TB, Reivax, Flunk)... but all of them have some of the following problems: MPEG2 "blockiness", compression artifacts, motion smear, image too sharpened or blurred, wrong colors, missing frames, burnt-in subtitles, PAL wrong audio speed, low bitrate, only one or two audio languages, not the best audio quality, lack of subtitles (check my own post here about these issues).

What I'd like to achieve is: eliminate compression artifacts, minimize motion smear (if possible), have the "right" image (not too sharp, not too blurred), correct colors, keep all the possible frames intact (as there are three different edition involved, it's feasible), retain only soft subtitles and not burnt-in ones (especially inside the picture), set proper audio/video speed @23.976fps, set the best bitrate available for any given format (minimum 7500kbps video bitrate for AVCHD and BD with 1280x720 resolution), add some interesting audio soundtracks (DTS theatrical, AC3 5.1, PCM Dolby Surround) in different languages, plus add several subtitles, including proper italian one.

g-force, AntcuFaalb, what do you think about the new denoiser?

I'd like to release a new version, but as my PC will do the conversion in FIVE days (OO_)I want to be sure it is good enough... Thanks to AntcuFaalb; he found that strange glitch - caused by the denoise filter - probably never noticed by anyone else, including me...

Maybe a little clip will be more explanatory! Please let me know which scene do you like to see, to show the new denoiser potential - around 30 seconds, if possible, as it will be dubbed in more than an hour...

Thanks, Zion!

So, here you are the link to a .zip file containing an untouched mirror of www.x0project.com website, taken in the year 2006, including all the images and videos:

http://www.sendspace.com/file/blvro9

There are interesting articles about the Pioneer HLD-X0 and restoration techniques. Note that the images and video included theer are NOT available on the archive.org... so, what are you waiting for? Download it, right now! (^^,)

Zion, I found into my laserdisc folder an offline copy of the www.x0project.com website, complete with images and videos!

I ask here your permission to let it publicy available, as there are interesting facts there that many would be happy to know.

And Happy New Year to everyone!

AntcuFaalb said:

_,,,^..^,,,_ said:

I'm testing new denoise filter - soon I'll post some screenshot comparisons.

Denoisers often do more harm than good, especially if they're temporal.

GOSH! Where are my green lasers?!? (°°_)

Thanks to have pointed to this error - this IS a really constructive critic, indeed!

I agree with you, but as laserdisc picture is analog and prone to noise, denoiser is needed... here you are a comparison between unprocessed, OUT ruLes 1.0, and the new denoiser (plus a bit more saturation) - the green laser is back, yooppeeh! (^^,)

Hope this time the overall result is better. (by the way, no FluxSmooth was used...)

I'm testing new denoise filter - soon I'll post some screenshot comparisons.

Next version will include original 2.0 Dolby Surround encoded tracks, converted to AC3, because they will sound better than the current 5.1 encoded with FreeSurround.

Meanwhile...

Happy New Year!

I forgot to mention... when calculating the video bitrate, I used 384kbps for each audio track, but at the end I changed my mind and used 448kbps... so, the resulting file it too big for a dual layer DVD!

I think the next release will use the same video bitrate, and the soundtracks lower bitrates (from 448 to 384kbps) to stay inside DL DVD size limit. OR I'll include DD 2.0 Dolby Surround encoded. Or should I lower the video bitrate?

g-force said:

The combination of the residual grain being both large and slow-moving is not very appealing. There appears to be some other slight temporal issues that I can't tell are from the original capture, the denoising, or the median/average of the captures.

I think I can release a SD file, with only average/median applied, and without any denoising, @25fps, and another like the former, with the new denoising, @25fps; if the problems come with the first, than probably they are in the original capture (I don't think median and average could cause such issues); if the problems arise in the second, than the denoiser/sharpener is the guilty; if don't appear in both, maybe it's TsMuxeR which did a poor job of converting from 25fps to 23.976fps.

The audio looses sync slightly over the duration of the movie.

Could you please be more precise? I presume you are talking about the english track; I just checked few minutes at the end, and it seems the audio is in sync with the video... Don't know, the audio tracks were left into the original captures, and the same cuts were done both for audio and video, so they MUST (theoretically) stay in sync... again, maybe TsMuxeR messed up all, or I did something wrong during the conversion from 25fps to 23.976fps! You'll may check one of the file I'll release, cited in the previous answer.

Any chance you could upload the best of the raw captures before combining them and post-processing? Might allow better feedback as to how to improve your final product.

May the two files I'm talking about in the first answer be enough? As I'm really, really sure the four captures are aligned with a pixel perfection...

If you think this is my first attempt at restoring a movie, and it took a month (about 150/200 working hours) to do it, I'm proud of the work I've done!

And, G., "this isn't too bad" to me means "quite good"! Thanks!