Moth3r

msycamore said:

Any ideas to what's causing this.

It looks like some sort of chroma sub-sampling issue.

Thread re-opened after PM from the OP.

TheWool said:

...

At the moment I've de-muxed the mkv container into a .h264 file and .dts file...

Sorry, but we can't offer any help with editing illegal downloads. Confirm (by PM) that you own and are working from the official BDs to get your thread re-opened.

I proposed a template message to say the project is deprecated, but the details are retained for reference only (see Howard the Duck, for example).

Warb: Thanks, that's much better.

Frink: TL;DR.

I object to the title change.

It should be "Who is Ignoring You, and Who are You Ignoring? (was: Who is Ignoring You? (was: Hello all, I'm back!))"

Chewtobacca said:

TV's Frink said:

Haha, J T34 is ignoring me now.

Really?  That surprises me.  Perhaps the cat picture you posted upset him.

Think he is easily upset. Check earlier ITT.

Orinoco_Womble said:

... Some people have mentioned that the 2-ch audio sample rate for PAL is 44.1KHz (same as CD), but NTSC is 44056Hz.  I'm pretty sure at best 'cheapie' audio cards with optical SPDIF in will only manage 44.1KHz.  Any idea if this is true? bob23 mentions it here and Buster D mentions it here.  It's also mentioned in the LD FAQ here, but others have said it's a load of poo.  Dnewhous thinks so according to this post here.  Who's right?

The 44056Hz rate was also mentioned here; I'm leaning towards the "load of old poo" answer.

Also, you mentioned in this thread about having to up-sample the 2-ch 44.1KHz audio to 48KHz for DVD player compatibility.  If this is the case, then why not just capture at 48KHz and have the audio h/w do the re-sampling for you? Or is this one of those cases where the cheap HW will give audio artefacts and it's more reliable (pure?) to use s/w for the up-sample?

I was under the impression that some audio cards could not resample a digital input in realtime. I could be wrong, I haven't captured many PCM streams to my PC. If the card can do it then fine - but I would still compare the results against a software upsample.

Lastly, Darth Editious mentions having to strip out padding from the captured AC3 raw audio stream here.  Has anyone else come across this and can suggest an easy method to remove the padding please?

I've also read this in several places - if you zoom in on the waveform of a captured AC3 in an audio editor, you can apparently see "packets" of data with "silence" in between. Something like AC3Fix or BeSliced should be able to extract the real data.

PS: Not that I really need to capture LD - DTS digital audio streams, but has anyone figured that out?  I can't seem to find anything much on that at all.

I've heard that DTS is actually easier - no demodulator required, for a start.

ray_afraid said:

Could someone please shoot me a PM and tell em where I can d/l RTJ and BE? Just watched em on youtube and I must have a dvd!!

Torrents:
The NTSC conversions of the original BE and RTJ are still seeded on Myspleen; the Miami version may be on Demonoid but I can't check at the moment.

Rapidshare:
Links for various versions can be found at fanedit.info

Usenet:
Various versions of BE, and the Miami version of RTJ are still available if you use a provider with a long enough retention.

I would recommend that you try and get the Miami versions if possible, as these are the newest and have several updates and corrections from the original version.

Don't know if you checked Google's archive of alt.video.laserdisc from around 2003, but Karyudo's success in doing this was discussed around this time.

Here's one discussion:

Donald Haas said:

What Audio card allows the saving AC-3 right from the TOSLink datastream without decoding?

Karyudo said:

Ah, now there's a question! This is the hardest part. Because many
cards are TOSLink/coax capable, but then mangle the bitstream. Even
some cards with the same chipset do different things: one will be
fine, and another will butcher. You're gonna have to try some, I
guess. I've got an ISA card that works fine, but it's ISA. Some cards
-- even quite inexpensive ones -- with the CMI 83xx (<-- forget the
exact number) chipset are OK; others don't work.

This is one arena where throwing more money at the problem isn't the
fastest way to a solution. A cheap-o CMI-based card could well
out-perform (in this aspect, anyway) the most expensive Creative
Audigy something-or-other.

Best of luck -- and please post your results (good or bad), because
that's what everyone's looking for!

D. Carroll said

Sound cards that have codecs (coder/decoder) that conform to the AC97
spec will resample all audio to 48kHz.

Most "consumer" (read cheap) soundcards conform to the AC97 spec.
AFAIK, the only cheap soundcard that doesn't conform to the AC97 spec is
those based on the Cmedia 8738 chipset (thanks to its built in
proprietary codec). 

However, "Prosumer" (read expensive) cards based on VIA's (formerly
ICEnsemble's) Envy24 don't resample audio.  M-Audio makes some nice
cards.  Too bad their relatively cheap new consumer card based on the
Envy24, the Revolution, doesn't have any digital inputs.

I heard that Microsoft is working on an update to AC97 but knowing
Microsoft there will undoubtedly be some catch.

Karyudo has not been on here in years, but I know that Darth Editous and adywan have also managed to do this.

LeeThorogood said:

...

When you say "There is no need to deinterlace" whats the reasoning behind this?

PAL video runs at 25 frames per second; each frame is made up two fields. Out of the 576 horizontal lines in a frame, the odd lines are from the first field and the even from the next. In your captured video, you will see that each frame contains fields from the same instant in time - this is known as progressive.

Interlaced video would have every field from a different instant in time, thus the odd and even lines in a frame would not exactly line up at the edges of moving objects, and you would be able to see combing artefacts. Deinterlacing filters are designed to remove these combing artefacts and give you progressive frames.

Simply put, if the source originates from film the frames are progressive, but if it was shot for TV (on a video camera) then they would be interlaced.

(I'm not going to talk about IVTC now because that's something you only need to understand if you're dealing with NTSC video). 

To do the colour correcting I cropped off the black borders (letterboxing and overscan area) leaving a clean black background generated by my video editing software. I then used a space scene to carefully adjust the blacks until the space background matched the black borders.

If you are doing this by eye, then your black level setting is dependent on the calibration of your display. I suggest you run the video through a luminance graph or waveform monitor, if such a thing exists for your software, and watch the levels while you make your adjustments.

I once saw something similar on my laptop, again using MPC-HC.

I can't remember exactly how I fixed it - I think I had turned off DXVA for some reason, and turning it back on stopped the artefacts. I've never managed to get my head round the settings you find deep inside MPC-HC.

Oh and ditch the codec packs, they're unnecessary and usually end up causing problems.

I don't have an iphone myself, but there was another member who emailed me recently because he couldn't post to the forum from his iphone.

I'd say there's interest in this.

LeeThorogood said:

... just to check I did test the output from the Composite against the S-Video and found the S-video to give a slightly less noisy and slightly sharper image, so I went with this.

Do you still have those test images? Would be interesting to see.

Here are some frames from the ROTJ Master which is the original capture, edited, filtered (de-interlace, colour correction, noise reduction) and cropped using the Apple ProRes 422 (HQ) codec. this file weighs in at about 50GB.

There is no need to deinterlace.

LeeThorogood said:

I have moved the snapshots to an external server so they should load as quickly as possible now, if they are still slow its because they are between .25 and .5 of a MB in size.

Ahem. Does your school know you are abusing their bandwidth during the summer holidays? ;-)

LeeThorogood said:

I just wanted to mention that the colour correction filter I am using is only doing some very basic things, restoring blacks, fixing washed out colour, nothing to fundamentally alter the '97 SE look.

I think the black may have been "restored" a little too much - the shots look pretty dark. Otherwise (apart from the timecode and logo of course) the images look promising. 

Maybe I'm a little slow to catch on, but can someone explain to me where the "actual movie sequence" in-joke is from?

Anyone want to mock up Luke missing his first grappling hook throw with the title "Pics or it Didn't Happen"?

Jawa77 said:

Nope "Close the blast doors, close the blast doors! .... Open the blast doors, open the blast doors!" was there in the original 70mm Dolby 6 channel on opening day.

"Close the blast doors" was not in the 70mm Dolby 6-track mix.

There are in-theater recordings of the 6-track mix from 1977 that corroborate this fact.

Differing frame counts between versions is fairly common, and it is true that reel changeovers are more often than not the places where the differences occur. 

These are not, however, "encoding errors".

It's a sample of the original HDTV source.

Could be from Planet of the Daleks? I remember some Daleks getting pushed into a pool of "liquid" ice. (From reading the novelisation, I'm too young to actually remember watching it)

TV's Frink said:

How do faneditors do custom crawls for their SW edits?

Here's another option:

http://originaltrilogy.com/forum/topic.cfm/UPDATED-v0-4-Star-Wars-Opening-Crawl-After-Effects-Template-SEE-FIRST-POST/topic/9001/

zombie84 said:

There was a really good page explaining about the evils of crushed blacks with reference to photography and the SE but I've forgotten it.

That might have been on Mike Verta's site, link here:
http://www.starwarslegacy.com/site/pages/Editorials.html

So did Erikstormtrooper in post 1234. Darth Mallwalker identified (post 1235) that this version (from Demonoid) was a crappy IFO-hack-job. I don't know if the download on fanedit.info is the same hack or not.

Now I'm confised - this is the RtJ thread, JediTray was talking about RtJ, and I was talking about about RtJ.

So why in Post 1250 does Frink list the versions of BE on fanedit.info?

You know, the fan project wiki would be a great place to list all the version information, much easier than having to trawl the thread to find the details.

To answer JediTray's original question - if you want a disc without PAL speedup you have to look for the dual layer NTSC conversion of the original (pre-Miami) version.

Regarding NTSC conversions, I've just had a look through this thread and I think the following summary is accurate:

Original Version
Single Layer disc by caligulathegod used original audio
Dual layer disc by caligulathegod used slowed down audio

Miami Film Festival Version
caligulathegod never finished his conversion
some fuckwit uploaded an IFO-hack-job conversion to demonoid

I take a different approach, where possible.

Edit your HD video so it runs frame-for-frame in sync with the video from the DVD. (Hopefully this won't be much work, cutting out the occasional frame is not a problem, the only time it might get tricky is when the HD version is missing frames.)

Then just remux the edited HD video with a direct stream rip from the DVD, and it will be in perfect sync.