logo Sign In

Red5

User Group
Members
Join date
17-Nov-2003
Last activity
4-Sep-2015
Posts
94

Post History

Post
#300338
Topic
Help: looking for... 1997 SE TB Digital Broadcasts
Time
Originally posted by: skyjedi2005 now we get to the other subject aspect ratio the laserdiscs and widescreen vhs are letterbox. not sure if the dvb broadcasts were 2:35:1 anamorhpic or 4:3 letterbox . the set i have of gkar is anamorphic i think but i might have been stretched to anamorphic by the dvd creator, unless the person could be contacted to see whether or not it was oar. oh shit i just remembered that the gkar is slighly cropped might be 1:78:1 or some shit.

the gkar has a sharper picture than any laserdisc sourced dvd being pal and not ntsc but it plays back at pal speed so unless you enjoy pal speedup it is crap. plus their is some blockiness in the image in some areas from compression, maybe. i obviously cannot say whether it is in the raw capped broadcast or not.


I'm pretty sure the Gkar, TB, Reivax and the Flunk release are all genuine anamorphic transfers, although Gkar and TB were probably captured from the analog video output of a DVB receiver, but you're right Gkar, TB and I think Reivax appear to be slightly cropped especially at the bottom and maybe the AR also is slightly off, not much but maybe more equal to 2.25-2.30:1 or so.

I think the Flunk release should be correct AR though as it's claimed to be a digital transfer from a master tape, although as already mentioned the Flunk has several other quality issues.

Still using video from Gkar, TB, and Reivax and combine it with high quality audio capture from some of the SE LDs one should be able to put together very decent PAL and NTSC '97 SE version DVDs.

Post
#297482
Topic
George Lucas to host showing of Star Wars "1977" for AFI's 40th anniversary.
Time
Originally posted by: CO
"Come see the landmark film that started it all, although you won't see the mindblowing effects that revolutionized movies in 1977 cause you are going to see updated effects from 1997."

Frauds!

It may be even worse cause I think it's the 2004 (DVD) version they're showing, a three year old (DVD) movie, not much of a classic. At least the 1997 version was actually shown in theaters (if that means anything to a film institute).

Post
#296578
Topic
Info Wanted: Moth3r's PAL trilogy vs. anamorphically enhanced GOUT's
Time
Originally posted by: totsugeki
Originally posted by: Red5
I've measured the multiburst VITS (Vertical Interval Test Signal) on a few of my PAL LDs and at best the 5 Mhz burst was at -4 dB and quite noisy, so the 5.5 Mhz bandwidth of PAL LD is in practice more like just below 5.0 Mhz.

Does that mean that Laserdiscs have test signals embedded in them (during blanking)?

Yes, on all THX releases and infact it seems many of the rest of them as well, not always with a multiburst line though.


Also, I'm still wondering about Moth3r's statement about DVD's 6.75 MHz translating into 540 lines. Shouldn't it be 720 lines?

No, I believe Moth3r is talking about LINES OF RESOLUTION

For example, since the current DVD format has 720 horizontal pixels (on both NTSC and PAL discs), the true horizontal resolution can be calculated by dividing 720 by 1.33 (for a 4:3 aspect ratio) to get 540 lines. (On a 1.78 [16:9] display, you get 405 lines) In practice, most DVD players provide about 500 lines instead of 540 because of filtering and low-quality digital-to-analog converters.

Although todays DVD players with 108 Mhz DA converters should have no problem reproducing the full 6.75 Mhz bandwidth or 540 lines.

Post
#296535
Topic
Info Wanted: Moth3r's PAL trilogy vs. anamorphically enhanced GOUT's
Time
The active part of a PAL or NTSC scanline is 52 - 53 usec long and to fill that part with 704 alternating white-black pixels equals a period time of 148.148 nsec which is the same as a 6.75 Mhz signal.


And regarding the bandwidth of LD it's important to remember that the 6.75 Mhz bandwidth of DVD is at full signal strength (well, perhaps with a little drop from imperfect reconstructions filters), but analog LD bandwidth figures usually is somewhat arbitrary measured with a signal drop of 6 dB or even 20 dB.

I've measured the multiburst VITS (Vertical Interval Test Signal) on a few of my PAL LDs and at best the 5 Mhz burst was at -4 dB and quite noisy, so the 5.5 Mhz bandwidth of PAL LD is in practice more like just below 5.0 Mhz. But it's still very player dependant and I'm very sure a X0 quality PAL player would give both higher bandwidth and cleaner signal.

Post
#293414
Topic
Info: Are the GOUT DVDs blurred?
Time
The "PAL" version is also slightly more cropped, I think LFL resized the NTSC GOUT from 720x272 to something like 732x330 and then cropped it at 720x326 for the "PAL" version, I've no idea why they didn't just resize to 720x324.

The vertical detail is also slightly more intact on the NTSC and is clearly the best (less bad) choice if you want to resize to fake anamorphic as the "PAL" already been blurred and resized once. But for direct view without using a fancy scaler the "PAL" version might actually be the better (less awful) choice for PAL viewers.

I guess LFL used a rather primitive adaptive blur filter that mistook some (much) of the vertical detail for jaggies, but as LFL decided to use an 13 year old transfer to begin with why not also use a primitive blur filter on it, it seems to fit.

Post
#286605
Topic
.: The XØ Project - Laserdisc on Steroids :. (SEE FIRST POST FOR UPDATES) (* unfinished project *)
Time
Yes, NTSC laserdisc resolution is something like 560x480.

X0 was most likely captured in standard NTSC DVD resolution 720x480 or 704x480, but keep in mind that the actual picture area of the letterbox is only 704x272 The picture area of PAL letterbox is 704x324

So even if you captured a NTSC LD at 1920x1080 you would get 560x480 of real resolution and you just wasted tons of GB on your HD .

Post
#286577
Topic
.: The XØ Project - Laserdisc on Steroids :. (SEE FIRST POST FOR UPDATES) (* unfinished project *)
Time
I think the restoration of the laserdisc sources is like 99% of the work and upscaling it to 'fake' anamorphic only 1% work, so it seems reasonable to make both LB and anamorphic versions, when the time comes.

In any case the X0 will be much much better than the GOUT. Lucasfilm appear to have applied a vertical blur filter trying to hide jaggies and a rather poor deinterlacing job, so the GOUT does not even have the vertical resolution of standard letterboxed NTSC and this is very visible in some scenes where some of the vertical detail is completely gone. And in this respect the GOUT really looks like crap, upscaled or not (IMHO).

I trust the X0 team already done proper deinterlacing and even removed some of the jaggies 'by hand' and this is of course a much better base for a fake anamorphic upscale.

Next step would be to combine the horizontal resolution of the NTSC X0 with the vertical resolution of the PAL laserdiscs and gain some 19% in real vertical detail.

Post
#283978
Topic
.: The XØ Project - Laserdisc on Steroids :. (SEE FIRST POST FOR UPDATES) (* unfinished project *)
Time
Originally posted by: Zion
SC-DC Comparison

The '93 versions have a different fade out and fade in from the part where 3PO says, "Why I should stick my neck out for you..." to "Thank the Maker!". The fade out beginning at 01:19:28:18 (X0 TC Print time) happens about 30 frames sooner on the SC. Why does this happen? Your guess is as good as mine.

Additionally, in the shot of 3PO being lowered into the oil bath, the lights on the back wall actually change color on the '93 versions, but not on the SC. This is the exact same shot though, so there was obviously some sort of additional color timing done to this shot. It makes one wonder if the color changes to the canyon shots were done after the fact as well - as in these were not present on the original film print.


In the 1995 PAL release there's a few frames missing at every reel change including this point, but the fadein is not much different from the DC, but anyway, I think the release containing the most frames must be closest to the original.

The changing lights on the wall must be an optical glitch in the DC transfer, you can see that there's an abrupt change in brightness behind 3P0 just as the greenish lights changes to red. And in the 1995 PAL THX release which was done two years after the DC, the lights are the same as in the 1986 JSC.

If you look at Moth3r's 1977 WS bootleg which is overall brownish/yellowish in color, the greenish lights on the wall should actually be much more bluish, so maybe the blue colors on the print was fading to green aready in 1986.

In the bootleg the canyon scenes also appear darker, but this being a several generation video copy it's very difficult to judge.

In the 1995 PAL THX the canyon scenes appear slightly brighter than the DC but clearly not as bright as the JSC.

Post
#282219
Topic
.: The XØ Project - Laserdisc on Steroids :. (SEE FIRST POST FOR UPDATES) (* unfinished project *)
Time
I remembered that this thread already covered some of this.

6. IVTC errors - the master video will have had 3:2 pulldown applied to produce the NTSC framerate of 29.97fps. For the DVD, the video was inverse telecined (IVTCed) to convert to 23.976fps. Unfortunately there were some cadence errors in the master, resulting in some flickering artefacts on R2D2 in the scene with Ben and Luke after the Sandpeople attack. These artefacts are visible when viewing on a computer or a progressive display.


And more stuff here

And there's also the issue with the vertical filter that probably was applied to hide the poor de-interlacing job.

Post
#282156
Topic
.: The XØ Project - Laserdisc on Steroids :. (SEE FIRST POST FOR UPDATES) (* unfinished project *)
Time
Originally posted by: zombie84
It was originally surmised that the GOUT crawl was re-created based on the trailer that came out before sept 2006--inevitable, it ended up being that the EOD crawl that it was being compared to was simply such a poor transfer that not all detail was visible, thus inconsistencies appeared which actually weren't inconsistencies at all. Someone speculated that theres fake gate weave but to me this is pure heresay--its the real deal as far as i am concerned. Where LFL got this from is the real question--a telecine was already done for EOD but this is clearly different than that one, and based on the much better sharpness and fidelity it appears to be a newer one as well.

Yes the GOUT crawl looks genuine enough and so does the stardestroyer flyover scene.

You can clearly tell the difference between the 1993 master and this unknown mysterious GOUT source as it has a slight chromashift or colour-bleed. The GOUT segment seems to been spliced in right after the fox fanfare and to the cut where you see the stardestroyer coming toward you. What really irritates me is that why did they not use a segment only a couple seconds longer so it covered all the space scenes until the first interior view of the Tantive ship for instance.

Judging by the reasonable good quality of the whole GOUT segment, chances are that the mysterious source actually is a full SW transfer from a good quality print that GL is hiding under his bed. And maybe he's going to sell it to us one frame or one scene at a time, if his going to sell it at all that is.

Post
#282081
Topic
.: The XØ Project - Laserdisc on Steroids :. (SEE FIRST POST FOR UPDATES) (* unfinished project *)
Time
Originally posted by: Zion
Originally posted by: Red5
Could it not be that those scenes were simply brightened during the JSC LD tansfer process back in 1986, and it is the '93 version that is closer to the original?

As others have said, the countless other sources including location photos point to the fact that these scenes were shot in daylight. Comparing other shots between the JSC and the DC/GOUT, there is a definite color difference as far as hue and saturation are concerned. But these two scenes clearly have more differences than that. I'm 99.9% convinced that GL had this done deliberately for the THX LD release.


I did actually know that the scenes were shot in daylight but thanks for the rest of the info

I am surprised and really can't beleive GL would bother with deliberate changes in the THX release, considering how worn and noisy the print were at that time. Although maybe the worn print actually were the main reason for some of the 'deliberate creative' changes.

Post
#281949
Topic
.: The XØ Project - Laserdisc on Steroids :. (SEE FIRST POST FOR UPDATES) (* unfinished project *)
Time
Originally posted by: tweaker
Could it not be that those scenes were simply brightened during the JSC LD tansfer process back in 1986, and it is the '93 version that is closer to the original?


That's possible, but this is Zion we're talking about. Would he really get his shit backwards?


tweaker,

Please do not alter or add to my question if you do not know the answer anyway.

Zion,

I'd be surprised if there were any 'delibarate creative decision' to make those specific scenes darker in the 1993 release.

Post
#281935
Topic
.: The XØ Project - Laserdisc on Steroids :. (SEE FIRST POST FOR UPDATES) (* unfinished project *)
Time
Originally posted by: Zion I would have to go back and look at the 97SE, but there are clear differences between the '86 Japanese Special Collection LD and the '93 THX LDs - both of which are obviously different from the 2004 DVDs as well.

The most obvious changes are in the beginning of the film. The canyon scene with R2 and the Jawas was altered in '93 to make it appear later at night. Also, the wide shots of Luke and 3PO outside the homestead were altered to better match the closeups.

http://img407.imageshack.us/img407/8190/8693colorcompareee4.th.jpg

http://img407.imageshack.us/my.php?image=8693colorcompareee4.jpg


Could it not be that those scenes were simply brightened during the JSC LD tansfer process back in 1986, and it is the '93 version that is closer to the original?

Post
#253005
Topic
Star Wars Trilogy - Original Theatrical Versions: 2006 DVD Flaw List
Time
I think the noisy/grainy GOUT picture is mainly caused by the quite high horizontal resolution. The Laserdiscs that used the same mastertapes would not be able to reproduce much more than 500 pixels, whereas the GOUT really seem to have the full 720 pixels resolution of DVD (at least for the niose and the grain ). Also the GOUT picture seem to been sharpened without first doing a proper noisefiltering and this would emphasise the noise even more.

One other flaw that's been mentioned from time to time is the high-contrast over-bright 'video look' picture of the GOUT (this also applies to the NTSC and PAL THX LD releases).

Here's some examples compared to the THX WOW DVD:

GOUT (clipped white levels and missing stars, also poor vertical resolution)
http://img60.imageshack.us/img60/1614/gout1fd0.jpg

THX WOW (hmm, nice anamorphic picture)
http://img332.imageshack.us/img332/9732/wow1zx6.jpg


GOUT (clipped white levels and missing stars)
http://img347.imageshack.us/img347/16/gout2lh5.jpg

THX WOW (hmm, nice anamorphic picture)
http://img54.imageshack.us/img54/3720/wow2hm7.jpg


GOUT (clipped white levels, and hmm very good horizontal resolution)
http://img327.imageshack.us/img327/5342/gout3ni1.jpg

THX WOW (hmm, nice anamorphic picture, horizontal res. could be better though)
http://img339.imageshack.us/img339/4063/wow3er7.jpg

Post
#247751
Topic
Are the PAL GOUT DVDs upscaled from the NTSC masters?
Time
Originally posted by: Moth3r
Originally posted by: Karyudo
... anyone who has viewed a PAL copy of the same 1995 THX release can tell you that it is definitely different from the NTSC version. And it is better in most areas. You're right that it has its flaws -- most notably more dust and crap in the picture -- but it does have more resolution, more detail, and no ghosting.
Apart from the burn marks and the line dropouts, I'd say that the PAL laserdisc actually has less dirt and fewer scratches than the DC, but it does have the same level of DVNR artefacts (ghosting).

The fact that the burn marks 'slipped through' the DVNR on the PAL LDs but was filtered out on the DC, and that the starfields are more intact on the PAL, could be an indication that less DVNR was used. The PAL LDs does at least not suffer from stormtroopers with 4 eyes link.

I think the dropouts is a pressing problem and it varies from disc to disc.

But even if the PAL masters are not 'perfect' they would certainly be much better than any NTSC upscale. The PAL masters were not only used for VHS/LD releases, there were also a couple of D2 (or D3) broadcast masters made in 1995 (incidentally these also had the same burn marks).
And if the PAL masters were good enough for German, Swedish and Danish TV broadcasts they are most certainly also good enough for a bonus disc DVD transfer.

Post
#246710
Topic
Are the PAL GOUT DVDs upscaled from the NTSC masters?
Time
This is from the 1996 (Lucasfilm) THX site.


The Making of a THX Laser Disc

The THX Laser Disc Program began in 1993 to improve the quality and consistency of laser discs. The motivating force behind the program was the fact that on a properly calibrated Home THX Audio System a tremendous amount of variation in the quality of laser discs was noticed.

Tom Holman, along with Principal Video Engineer Dave Schnuelle, began investigating all of the many steps in the laser disc mastering and manufacturing process. They developed monitoring techniques and quality assurance principles to solve the many problems they found along the way. One technology that was developed was the THX Video Test Signal. The patented THX Video Test Signal is inserted into a video signal's vertical interval (the space leftover between each video field). This allows a computer to continuously monitor the black level, white level, chroma level, chroma noise, phase, frequency response, etc. of the video signal. This ensures that no video signal degradation occurs during duplication.

Here is a brief description of all of the steps that go into making a THX laser disc. Remember that a THX Laser Disc does not use any special surround sound format. THX Laser Discs have soundtracks recorded in Mono, Stereo, Dolby Surround, and/or Dolby Digital.

The first step is to calibrate the video monitor and the audio playback system. THX engineers use a Photo Research Spectra-Radiometer to carefully calibrate the black level, white level, greyscale, and color balance of the video monitor being used in the transfer. The Spectra-Radiometer measures the wavelengths of the light coming from the video screen and can display the actual spectrum on a computer. This gives the engineer valuable information on color and greyscale as well as on the actual purity of the phosphors being used in the monitor. The purpose of this calibration is to ensure that decisions made by the filmmaker on the color of the transfer are done under correct and repeatable conditions. The Photo Research device is calibrated regularly to National Bureau of Standards specifications.

The second step is to calibrate the dubbing machines. The frequency response, phase, azimuth, and level of each channel (2 channels for Dolby Stereo transfers, 6 channels for Dolby Digital) are checked to ensure accurate playback of the soundtrack master tape.

Once the system is calibrated, THX engineers monitor the transfer and answer any technical questions that arise. The transfer of picture and sound are done at the same time to ensure correct synchronization. All decisions about content, colorimetry, etc. are made by the film company, and many times the director, producer, or cinematographer is present. THX engineers are present in an advisory capacity only. Much of the look of a movie transfer is the responsibility of the telecine artist and the studio advisor.

Upon completion of the transfer (usually to a D-1 digital video tape), THX Laser Disc engineers step in to begin supervising the duplication process. The D-1 master tape is digitally transferred to a digital D-2 video tape. At this stage, the patented THX Laser Disc test signal is inserted into the video vertical interval.

On a parallel path to the video transfer, the 2 channel analog print master is transferred to digital audio. If the laser disc contains a Dolby Digital soundtrack, Dolby engineers work along with THX engineers to transfer the 6 track print master and convert it into the 5.1 channel Dolby Digital serial bitstream. One more Digital D-2 video copy is made and it is at this stage the soundtrack is married to the video print. The FM analog tracks are copied, the digital 2 track Dolby Stereo soundtrack is copied, and (if used) the Dolby Digital bitstream is transferred as well.

It is from this final digital D-2 master tape that the laser disc master is made. Since a laser disc contains an analog video signal, the digital video of the D-2 master tape must be converted to analog at this point. Once the laser disc master is made, THX engineers inspect a sample from each stamper to ensure that the quality of the original transfer is maintained. A test disc must go through a computerized test process using the patented THX Video Test Signal. It must also go through a review by a THX Laser Disc QC technician. To ensure the highest quality possible, every transfer and sample disc is looked at and listened to. THX Laser Disc QC technicians have the power to reject entire pressings of laser discs, and they have used that power on occasion.

The goal of the THX Laser Disc Program is to ensure that the look and sound of a movie is preserved through the complicated picture and sound transfer processes. You may notice some picture and sound differences between THX Laser Disc titles, but those differences were decided upon by the movie makers themselves.

The technologies developed by the THX Laser Disc Program become increasingly important when you look at the future of home video. THX engineers are currently working with many digital video formats, including DVD, to ensure that the filmmaker's intentions are preserved for the future.



Although there's no mention of DVNR, it is my impression that the DVNR took place during the Telecine and it was controlled by Telecine artist (or colourist). The DVNR was probably used like any other adjustment i.e. like the settings for colour or black level. I believe the DVNR was an 'automatic' process that had only one setting for Luma and one for Chroma with a scale 0-10, and I would suspect it was set to handle the worst parts of the print and then probably not changed during the transfer, and that's why some (many) parts were clearly over-filtered.

Anyway IMO also the D1 Telecine master would be DVNR filtered and after the D1 there were no more video tweaks applied.

The printing master would be the final D2 tape. The lack of dotcrawl in the GOUT suggest a non-composite source like D1.

Post
#245876
Topic
Info: 2006 OT DVD/Laserdisc - a comparison...
Time
Originally posted by: Moth3r
It certainly looks like a vertical blur has been applied to the official DVD; there is a noticeable loss of vertical detail when compared with the laserdiscs (both the DC and the PAL laserdiscs). I don't know why this might be the case, perhaps it was done to reduce the effect of IVTC artefacts (e.g. see here)?

I think the DVD could do with some processing to remove video noise/grain, and because it has superior horizontal detail/sharpness the results should look better than any laserdisc transfer, despite the vertical blurring.

Yes, the loss of vertical detail could be the effects of a not so successful anti-aliasing filter. My guess is that the GOUT was then also sharpened and that brought some of the aliasing or jaggies back, or rather it emphasized the jaggies the AA filter missed. If it was sharpened it would also explain some of the higher noise level.

Since the D1 master already was DVNR filtered, the remaining noise really is a bitch to filter as it's full of the DVNR artefacts, the same also applies to the LDs of course. But it is usually also best to noise filter *before* any sharpening is applied.

I've not yet seen that much of the NTSC GOUT but I think it's less sharpened than the 'PAL'.

Post
#243595
Topic
Are the PAL GOUT DVDs upscaled from the NTSC masters?
Time
Originally posted by: Moth3rHmmm. After spending time at Zion's site, there are some scenes that do look better on the PAL laserdiscs. Try shots 22 and 24.

Yes and also #36, it's a bit strange though cause they do not appear to even have the vertical resolution of upscaled NTSC.

I wonder if some sort of vertical filter was applied that messed up some of the scenes, and if this also affects the NTSC version.

Also the GOUT crawl and star-destroyer flyover scene did not come from any of the laserdisc masters, so here the 'PAL' could in fact be better than the NTSC.