logo Sign In

Dr. M's Reinventing The Wheel Edition (PAL to NTSC+) (Released)

Author
Time
 (Edited)

Major Update: 8/18/05

Current Status
Episode IV - New Hope: Finished - Distributed (v1.1 now being distributed to fix loss of second audio channel glitch)
Episode V - Empire Strikes Back: Finished - Distributed
Episode VI - Return of the Jedi: Finished - Distributing

Covers (by Erikstormtrooper):
http://www.erikstormtrooper.com/Rtw_ep4.jpg
http://www.erikstormtrooper.com/Rtw_ep5.jpg
http://www.erikstormtrooper.com/Rtw_ep6.jpg

(Full Tech Details at Bottom)

History:
I am am producing a new version of OT DVDs.
Out of fairness let me say up front I am not doing a new transfer. I really don’t have the setup to do anything better than what already exists, so I’m not.

I am creating a conversion to NTSC from Moth3r’s absolutely wonderful transfer which outstrips even the newest NTSC versions.
This is largely being done for my own benefit but release now seems required.

Before going any further, I will say, yes Moth3r knows what I am doing to his baby, and he seems to be largely ok with it.

But to be clear these will not be called Moth3r’s NTSC edition (his choice not mine). But in retrospect since this project has grown to be more than a simple format conversion, I absolutely respect that decision.
So since I am making hard work of something that is already extremely good, I have dubbed this “Reinventing The Wheel Edition” (RtW for short).

Why I am doing this:
Although I love PAL’s native higher vertical resolution, it’s 4% speedup (originally) was an issue for me. Also (and I don’t want to step on toes here) I don’t care for the washed out color in the original version. Foremost is that I DON’T OWN A PAL DVD PLAYER.

Now first the basics.
There are several ways to play a PAL DVD:
Obviously you can get a PAL or multistandard player.

Updated Tips for purists: If you want a true NTSC version of Moth3r’s discs you can use the Patch Method which uses IFOEdit to try and convince your DVD player that “this is not the PAL disc you are looking for”. It works on many units (but not all) and basically counts on manufacturers being cheap and using the same internal hardware internationally and just making tweaks for regions and formats. Compatibility with this varies.

Now if you augment this by demuxing the audio and video then using DGPulldown from 25 fps to 29.97 fps and remuxing/authoring you have a different beast altogether.
I saw this for the first time today and thought it was genius, what they are doing is using IFO edit to fool the DVD player into resizing the video to NTSC for you. The DGPulldown keeps the video at 25 fps on the disc but the DVD player performs telecining on the fly up to 29.97 fps for you (no I didn’t think that was possible either). The huge benefit is the audio does NOT have to be re-encoded. In fact nothing does. You’re just tweaking a bunch of flags and let the hardware do the rest for you.

This is perfect for Moth3r’s DVD’s since we know the audio to have been pitch correct to account for the 4% speed up.
Since this method would make a highly compatible version of Moth3r’s movie in NTSC, it’s almost a shame to not have done it originally… but I like overdoing projects (and it’s too late to go back now).

Finally you can re-encode (which is what I’ve done).
Since I do not have any raw avi files to work with, I am going from the public DVD version of these discs that everyone has access to.
Using VirtualDubMod or AVISynth, you change the framerate to 23.976 (NTSCfilm, just a bit over 1 fps slower than PAL), crop 72 lines from the top and bottom, resize to 720x360 and add 60 line borders above and below. Also the sound must be time expanded so it all syncs. That could be done in an afternoon to tell you the truth and was my original plan.

But why stop there?
As long as there’s filtering going on, why not improve the color, etc.
http://adventureclub.postrock.net/anh.jpg
http://adventureclub.postrock.net/anh2.jpg

Yes the above is from an actual NTSC frames (after re-encoding). Still not any more difficult to do really.

Actually what you see includes a 4% reduction in the white point (to brighten the whites) and 25% boost in the Yellow and Green saturation.

Edit: Return of the Jedi required slightly different handling. Unsharp Mask was used to sharpen, Yellow/Green was boosted only 10%, but Red, Blue and Cyan were desaturated by 15%.

At this point I could explain the increased lines of resolution of PAL laserdiscs vs NTSC, but I’ll save that for the PAL transfer threads: Here’s Moth3r’s.

What is important to understand is that an anamorphic NTSC DVD of Star Wars (which is in 2.35:1 aspect ratio) can use up to 360 vertical lines for the movie. This is about 50% more lines than even the PAL laserdisc source.

Now ideally I would have raw video so I could resize up instead of down for the disc, and also I wouldn’t have to recompress previously compressed video, but that is not the case unfortunately.

ORIGINALLY I had every intention of adding left/right black borders and reducing overall size to prevent further loss of picture in TV overscan regions. The reasoning being that there is excessive cropping at times in the original disc. Public opinion (and lots of explainations) convinced me to drop that idea in favor of preserving as many lines of resolution as possible.

BUT what about the audio? If you keep reading all the posts to follow you will see there has been much debate about how best to accomplish this.
I finally gave up on giving up and will be using an NTSC audio rip from the Definitive Collection. [Thanks to Arnie.d] A little jiggery-pokery with a shoe-horn and you have wonderful PCM digital audio from laserdisc source (bye-bye analog tape that’s been sped up, pitch shifted, analog recorded, ripped, encoded, decompressed, slowed down, and pitch-shifted again (no kidding that’s where this was going)).

Final thoughts:
I am not doing menus (not really a need).

For those concerned about further quality loss in the re-encode process, a Constant Quality first pass indicated that I’m using about 50% more bitrate than is need for a Q of 30, so relax.

Release:
Yes please. Not waiting for Episode VI, contact me if you want to help.

Feel free to talk amongst yourselves.
-Dr. M

Source Material
Video: Moth3r’s DVD, originally from the French 1995 THX PAL Laserdiscs
Audio: Uncompressed PCM “Definitive Collection” NTSC, Pioneer DVL-919 to Canopus ADVC55 (DV - analog capture) [Thanx arnie.d]

Original Hardware (taken from Moth3r’s details):
Pioneer CLD-D925 Laserdisc player
Toshiba VT-728B VCR
Leadtek WinFast VC100 XP video capture card
PC: Athlon XP 2700, 1GB RAM

Re-Mastering Hardware:
Athlon XP 3500+, 1 GB RAM

Original Software (Moth3r)
Capturing:
btwincap drivers
VirtualVCR
Huffyuv codec
Post-processing:
VirtualDubMod
AVISynth
Encoding:
CCE 2.70 (video)

Remastering Software:
Vobs files ripped/demuxed merged: (various)
VirtualDubMod to filter and frameserve
Womble Mpeg Video Wizard to edit and sync audio
Re-encoding:
CCE SP 2.67 - 6 (+1) pass VBR
CCE SP 2.70 - 5 (+1) pass VBR (RotJ)

DVD Authoring:
DVDMaestro

For quality preservation SFV files (disc checksums) were generated with Easy SFV Creator and are available.
The SFV files should appear on the DVD. If it isn’t there (smack the guy that sent it to you), then check your disc using these:
ANH’s SFV v1.1
ESB’s SFV
RotJ’s SFV

Dr. M

Author
Time
I was hoping to see a NTSC version of Moth3r's transfer. My dvd player can play PAL dvds no problem but my friends can't. The screenshots look great going to check out the clip later.

Ok just checked out the sample and it looks really good, at least on my computer. I would be willing to upload the sample to a.b.starwars if you dont mind Dr.M?
Author
Time
Your nuts Doc!

But this sounds like a neat project - It'd be nice to see a proper PAL to NTSC conversion your sample looked nice and clean and did not really look lie a re-encode I've not burned onto DVD or really been able to zoom into the moving video but that looks nice!

RiK

“My skill are no longer as Mad as the once were” RiK

Author
Time
What has not yet been decided is whether the audio track will be a time expanded version of the original (which was from a VHS source), or a native NTSC track replacement. I currently have access to a nice PCM recording from the Definitive Collection thanks to Arnie.d


I was hoping to host the 2nd version of this clip on my own since some people have problems with RapidShare, but it's currently not possible.
This clip is before I started thinking Picture Framing, and the audio is the original time expanded (with BeSweet) re-encoded to a low bitrate AC3 (for internet transfer ease).



Just a couple things, when you say "time expand" the audio, you do mean slow down by 4%, not actually "time strectching" which is what time expanding implies. You definitely want to pitch shift it down by 4%. Just checking, as it sounds like it was pitch shifted in your video clip, which would be correct.

Also I personally think that the colour/saturation is a bit too much, the faces of the rebels look rather a little too sunburnt and the white walls are starting to look a little pink, but of course this is subjective so its your call there. Maybe shifting the hue towards blue a little more would help, even though the official DVD is WAY to Blue, I think too blue is better than too red.

www.bardothodol.net

Author
Time
This is going to be a popular release I can imagine!

I agree that the saturation is just a bit too much. C3PO's metal "skin" shows it the best. I'd personally tone it down a notch.

What’s the internal temperature of a TaunTaun? Luke warm.

Author
Time
I've noticed a lot of the transfers have been way to red in the past (there's one rebel I tend to call "Flamboyant Guy" (official screenshot #4) because his face is often so red it looks like heavy makeup in most transfers). That's why when I was adjusting I based it on the reddest frame I could find (not this one though).
I did a simple saturation boost and put the brightest red in that clip to 75% of maximum (indicated by the boxes in the graphs below) and let the rest of the colors tag along. Now when I work with the full movie, there might be an even redder point (or other bright color) that I will set using instead. It's good not to exceed that 75% number, since beyond it colors might not be valid on all TV's.
Using ColorTools for VirtualDub set to vectorscope I show the following graphs for the R2/3PO screenshot:

http://home.earthlink.net/~originaldrm/officialvs.jpg
Official^
http://home.earthlink.net/~originaldrm/cc2vs.jpg
Cowclops v2^
http://home.earthlink.net/~originaldrm/moth3r35pvs.jpg
Moth3r's with 35% Saturation Boost^

Now in fairness my screenshot isn't the exact same same frame, but this should give you a good idea that I thought I used retraint.
This frame did NOT have the reddest object in this clip so you can imagine how far up the official one maxes-out at. Cowclops' new edition shows the reddest object to hit 80+% at some point during the same section of film.

As far as moving hues, I could "rotate" 5 degrees clockwise to match the Official DVD or 10+ degrees counter-clockwise for the CC's. That would probably be less desirable. And to move around individual colors isn't going to happen, I have a slight color blindness which prevents me from doing this by eye, and it stands to reason every scene may have different requirements. Though I will give it a look though.

Comments?

Dr. M

Author
Time
I think it should be somewhere between Moth3r's original and the clowclops release. But that's only my opinion based on my own capture.

A new release of Star Wars is like a box of chocolates, you never know what you're gonna get.

Author
Time
Originally posted by: Mavimao
This is going to be a popular release I can imagine!

Probably, except for a few points:


1. The sharpening process Moth3r used produces halos around dark edges, I used a different method which maintains clarity whilst avoiding halos:
http://www.haku.co.uk/pics/StarWarsDarthMoth3r.jpghttp://www.haku.co.uk/pics/StarWarsDarthCitizen.jpg



2. The scene where Obi-Wan and Luke lift up C-3PO after being attacked by the sand people is broken so there's no smooth transition between that shot and the outside shot of Obi-Wan's place.


3. Most importantly, if you merely resample Moth3r's audio from 25fps speed to 23.976fps speed the correct audio pitch will be lowered by 4.096% and everyone will have deeper voices.
This is because the PAL VHS audio source used for his release has been sped up from the original 24fps film but they (the VHS producers) didn't just resample it they also preserved the audio pitch, undoing that and keeping the audio as best as possible may be difficult because you'll have to resample whilst preserving pitch, I couldn't find any timestretching tool I was happy with, the ones I tried (proper packages) weren't competant enough to stretch and keep pitch without ruining certain pitches, it would ruin John Williams' score.

Use NTSC audio source if you can, but fitting it to the video source will be very difficult if the NTSC audio source doesn't have video to go with it as a guide to sync them up.
http://www.haku.co.uk/pics/LukeCruise.gif http://www.haku.co.uk/pics/dontcare.gif
***Citizen's NTSC DVD/PAL DVD/XviD Info and Feedback Thread***
Author
Time
This is because the PAL VHS audio source used for his release has been sped up from the original 24fps film but they (the VHS producers) didn't just resample it they also preserved the audio pitch


Really? I've never seen (heard) a PAL release with preserved pitch. I've only recently seen it used on UK broadcasts of Stargate, and you can tell it's had something done to it because the sound warbles and clicks.

Have you tried to align the audio track to the video yet? Because I'd imagine you'd be unlikely to get a perfect 25fps from your VHS player. I remember, for some reason, my friend and I pressing play at the same time on our VHS copies of Star Wars and after only a few minutes they were a few seconds out of sync.

I don't think you can ever time stretch/compress perfectly - if the tools you've tried haven't worked, have you considered that a combination of tape stretch and your player's mechanics could have conspired to lower the pitch to something similar to the true soundtrack?

DE
Author
Time
Originally posted by: Citizen

3. Most importantly, if you merely resample Moth3r's audio from 25fps speed to 23.976fps speed the correct audio pitch will be lowered by 4.096% and everyone will have deeper voices.
This is because the PAL VHS audio source used for his release has been sped up from the original 24fps film but they (the VHS producers) didn't just resample it they also preserved the audio pitch, undoing that and keeping the audio as best as possible may be difficult because you'll have to resample whilst preserving pitch, I couldn't find any timestretching tool I was happy with, the ones I tried (proper packages) weren't competant enough to stretch and keep pitch without ruining certain pitches, it would ruin John Williams' score.

Use NTSC audio source if you can, but fitting it to the video source will be very difficult if the NTSC audio source doesn't have video to go with it as a guide to sync them up.


Ahhhhh, right, this is where I was confused because comparing the Moth3r version and the Cowclops v1.0, I couldn't hear any difference in the audio pitch. I didn't realise that the original producers of the VHS audio source had preserved the proper pitch of the audio while stretching it by 4.096%. That makes sense now I thought C-3PO's voice sounded a bit deeper than I was used to in Doctor M's clip, but I just figured it was because I was so used to hearing it at PAL speed over the years.

Then, my suggestion for the sound would be to take the NTSC source sound you have and carefully fit that into the new NTSC'ed video you have. It will be fiddley but I definitely agree with Citizen about timestrecthing it, just don't bother! I've used many pro timestretch tools and they all make harmonically complex sound (eg music) sound horrible. Very few are capable of timstrecthing huge files by such precisely small increments as well, you would still have to mess around cutting and synching it up with the video, so just use the NTSC source sound and all will be great



www.bardothodol.net

Author
Time
Some kind soul has posted the Version 2 sample on alt.binaries.starwars which is very handy for those of us with an ISP (ntl) which has problems downloading from rapidshare.

A big thanks from me........

4 - 5 - 3 - 1 - 6 - 2

Discuss…

Author
Time
Hmmm, some interesting points to note in this thread:

Top and bottom edges have the least damage (and little can be done about this).
Left and right present the real problem. Because in addition to the lost regions, even more of the frame will be lost once this is view on a TV (because of overscan). The tighter framing in the PAL laserdiscs is well documented, but in addition, some of the image lost from the left and right could be due to the "capture window" effect.

Your nuts Doc! What about his nuts?

I did a simple saturation boost and put the brightest red in that clip to 75% of maximum I find this very strange - but that's exactly what I did when I captured the video. So in theory, you shouldn't be changing the saturation at using that method.

I've since found that the 75% thing (which equates to a maximum analogue voltage of 100 IRE for NTSC video) isn't a limitation for PAL, so I could have gone higher. Also, it's normally acceptable to go up to 110 IRE or even 120 IRE for the occasional isolated pixel.

:: Edit - check out frame 8098.

The sharpening process Moth3r used produces halos around dark edges,
No, the halos were not caused by the sharpening (see my post dated 18 April). It's actually an analogue oversharp problem, you can see that the halos are on the left hand side only (if it was digital sharpening it would be both sides). But my version is designed for analogue (CRT) displays, where such things are not as noticable, whereas I understand yours is intended for a digital projector?

I've never seen (heard) a PAL release with preserved pitch.
I hadn't either, so this is a bit surprising. Can anyone confirm this is definitely the case?

Have you tried to align the audio track to the video yet? Because I'd imagine you'd be unlikely to get a perfect 25fps from your VHS player.
No, but the audio on my DVD is a perfect 25fps, I spent some time matching it up frame-by-frame cutting out the occasional 40ms where necessary. Although it does maybe go out of sync for a little bit during some parts of the final death star battle.

Guidelines for post content and general behaviour: read announcement here

Max. allowable image sizes in signatures: reminder here

Author
Time
Hmmm, some interesting points to note in this thread:

The sharpening process Moth3r used produces halos around dark edges, No, the halos were not caused by the sharpening (see my post dated 18 April). It's actually an analogue oversharp problem, you can see that the halos are on the left hand side only (if it was digital sharpening it would be both sides). But my version is designed for analogue (CRT) displays, where such things are not as noticable, whereas I understand yours is intended for a digital projector?



Ahh so it's the capture card and/or the LD player that caused the halos? I captured mine on a Pioneer CLD-515 that doesn't have s-video so no combing filters, on a Canopus ADVC-100 which I find produces a very clear image.


I've never seen (heard) a PAL release with preserved pitch.
I hadn't either, so this is a bit surprising. Can anyone confirm this is definitely the case?


Before I posted that the PAL VHS audio source you're using has had pitch preservation I checked yours against the audio from the NTSC definitive LD:

StarWarsAudioSample-Moth3r-25fps.mp3 47kb
StarWarsAudioSample-Definitive.mp3 49kb
StarWarsAudioSample-Definitive-25fps.mp3 47kb

The Moth3r-25fps.mp3 and Definitive.mp3 sound identical, I sped up the Definitive clip to 25fps to show that Moth3r's audio is at the correct pitch whilst playing back faster, also note the filesizes.
http://www.haku.co.uk/pics/LukeCruise.gif http://www.haku.co.uk/pics/dontcare.gif
***Citizen's NTSC DVD/PAL DVD/XviD Info and Feedback Thread***
Author
Time
So much for a weekend project

A LOT to think about now.
As far as the colors, although I am dissatisfied with the redness of Cowclops, the dullness of the original Moth3r transfer, and the pinkness of the walls in my test version...
I still find any less color (eg, half-way between CC & Moth3r's which technically I am already) to be less visually pleasing, if not as accurate.

As far as the audio, ugh. Anyone have a good alternative? I have Sony SoundForge which has a decent Pitch Shift that could be applied after the time expansion, but I can't just say 4% and I although I'm good, I don't know how it'll sound.
I could cut, and expand silences between scenes/disc changes, etc with the NTSC to try to make it match so I don't have to alter the frames further.

Of course I could always wait for Citizen's release since he is using laserdisc sourced audio, the time may match better.

You guys are talking me out of this...

Dr. M

Author
Time
That is indeed strange regarding the original vhs audio being at the correct pitch. I guess (also considering that it was mastered to VHS probably about 10 years ago) that it was kept in sync with video when PAL'd from the original 24fps, causing the normal speed increase, then pitched down to the correct pitch, while still preserving the length fitted to the 25fps footage. ie:

24fps audio sountrack -> speed increased by 4.096%
then -> pitch shifted down by 4.096%

Pitch shifting gives much better sounding results than time stretching, they are achieved very differently. Time Stretching breaks the audio into very small incremented sections, and then timestretches them each by the same amount and then fits them back together. This is what can give the nasty warbling effect in very harmonically complex sound as it ruins the natural and non-uniform vibrations of the sound. This works by altering the time duration. Pitch Shifting is different by that it works by taking the frequency information and of the audio and reducing it by the desired amount.

Thats a seriously simple (and probably not very good) way of describing what happens, THIS goes into more detail, but as long as you understand the difference between the two mothods, you're sorted. Basically Time Stretching sounds shit unless you're only using it on simple Foley sounds or breakbeats or whatever if you are a music producer, it should NOT be used to alter the length of an entire audio track, especially when it contains big orchestral scores. The combination of Speeding up then pitch shifting down will give far far better results. If you're using Sound Forge, you've got the best tools to do either of these things.

But saying all that, you're still best using your NTSC PCM audio Doctor M, and fitting it to the video like Moth3r did with his VHS audio track. Thats gonna give the cleanest results overall.

www.bardothodol.net

Author
Time
Here's a comparison of the spectra of those two samples (the one from Moth3r's DVD and the one from the Definitive LD, scaled to match widths):

http://img14.imgspot.com/u/05/213/17/freq.gif

It looks to me like there's more going on than "simple" pitch shifting* - see how the spacing between sounds changes?

DE (*not an audio engineer)
Author
Time
Originally posted by: Darth Editous
Here's a comparison of the spectra of those two samples (the one from Moth3r's DVD and the one from the Definitive LD, scaled to match widths):

http://img14.imgspot.com/u/05/213/17/freq.gif

It looks to me like there's more going on than "simple" pitch shifting* - see how the spacing between sounds changes?

DE (*not an audio engineer)


ah, ok this what I mean when time stretching the audio information is taken and then sliced into much smaller incrememental pieces, indivdually stretched, then put together as a whole track again with the new duration. The time stretching algorithm analyses each segment working out how best and much to time stretch them in relation to the different frequency information of that slice and in relation to the other slices (see the link I posted above for a far better and techincal explanation). It doesnt just grab the front end and back end and pull it to the right length.

I have to hold my hand up and admit that I havent got the whole of Moth3rs ANH release, so can't comment if the whole soundtrack sounds like it had been time stretched when originally mastered to VHS, but I would wager that it had been looking at the graph Darth Editous has posted. This is also another good argument for people doing a PAL version to get the ORIGINAL PAL Audio from a PAL Laserdisc (maybe get it from the pan & scan UK LaserDisc release since these arent that hard to get hold of) I agree with the argument that the audio quality difference between VHS audio and LaserDisc audio is not very great, but problems like dodgy time/pitch shifting done back then, and problems like VHS tape warp/stretching caused by repeated play of the tape over time WILL effect the audio alot more.


pupil *

*nearly done my masters degree in sound design, but a little rusty on the theory as its been a while since I did any

www.bardothodol.net

Author
Time
Just make sure you calibrate your screen before messing with the colour (hue and saturation) - if your screen is off then the version will only look right on *your* screen and no one elses.
The halos on Moth3rs capture are caused because of the analogue bandwidth limitation, there is not enough bandwidth for the hard transitions from black to white, so you get the 'halos' - it is an analogue issue and can be solved either by filtering, or better still by remapping.
Author
Time
Hmm. First let me nit-pick and say that technically PAL audio isn't time stretched, it is time compressed since the runtime is less than the original. It runs faster, hence the increased frequency. Correcting it requires time expansion. For most PAL material, like it or not, that's really the best alternative out there if you're turning it back to NTSC.

Commenting on the rest (and I could be butt-ass wrong):
As far as pitchshifting. Typically it would not be needed as it only changes the frequency, not the actual speed/length of a clip. In this case (and I thought my ears were playing tricks on me too when I heard the butch C3PO) time expansion is STILL needed. Otherwise when the framerate is slowed to 24fps, the audio will run too short. We are (unfortunately for me) finding that evidentily it will need ADDITIONAL pitchshifting to reverse the whole process they put it through.

From what I'm seeing with soundforge though, pitchshifting doesn't seem to be done on a percentage basis making this much more wobbly a problem to deal with.

So the choices are looking to be: exhaustively edit in an NTSC audio track, give up, or wait for Citizen's release since he will probably not have to add/remove frames to make his audio track match (which will make for an easier match to the DC audio later). Anyone got an external drive they want to send to Citizen for me so I can get the avi files instead

Color: I am not not not color correcting, I played around a bit with it today, and I will say, categorically, it's not happening. Again with this, there is more going on then a few simple tweaks. I will however still use a saturation boost (something in the 25% to 35% region) (a trade-off yes).

I did finished d/l-ing the full ANH tonight, and am about finished working on my current project so I can take a serious swing at this Thursday.

Dr. M

Author
Time
Originally posted by: Doctor M
Hmm. First let me nit-pick and say that technically PAL audio isn't time stretched, it is time compressed since the runtime is less than the original. It runs faster, hence the increased frequency. Correcting it requires time expansion. For most PAL material, like it or not, that's really the best alternative out there if you're turning it back to NTSC.

Commenting on the rest (and I could be butt-ass wrong):
As far as pitchshifting. Typically it would not be needed as it only changes the frequency, not the actual speed/length of a clip. In this case (and I thought my ears were playing tricks on me too when I heard the butch C3PO) time expansion is STILL needed. Otherwise when the framerate is slowed to 24fps, the audio will run too short. We are (unfortunately for me) finding that evidentily it will need ADDITIONAL pitchshifting to reverse the whole process they put it through.


From what I gather it's not common practice for films on PAL VHS/DVD/LD to be pitch corrected after/during the 4.096% speedup to 25fps, which is why it surprised me somewhat to find it had been done to the SW VHSs Moth3r used for his DVDs, especially considering how old the recording is.

I was playing with timestreching whilst preserving pitch with CoolEdit earlier, I thought I'd found an optimal setting that didn't destroy the audio till I tried it on a much larger clip... it practically destroyed the intro music completely
I like the idea of resampling first and then adjusting pitch, may have to give that a try, otherwise if I'm not happy with that method I'll just stick to my original plan of resampling and having a 4% higher audio pitch.

So the choices are looking to be: exhaustively edit in an NTSC audio track, give up, or wait for Citizen's release since he will probably not have to add/remove frames to make his audio track match (which will make for an easier match to the DC audio later). Anyone got an external drive they want to send to Citizen for me so I can get the avi files instead


I have on my harddrive a 2.6gb avi of the definitive LD of ANH with no frames lost during the capture, manually removed the 3:2 pulldown then highly compressed it to mjpeg format because each frame is a keyframe, looks terrible but it doesn't matter because I only need it as a guide for syncing up to my PAL capture, the audio was captured & stored in the avi as uncompressed 48khz PCM...
http://www.haku.co.uk/pics/LukeCruise.gif http://www.haku.co.uk/pics/dontcare.gif
***Citizen's NTSC DVD/PAL DVD/XviD Info and Feedback Thread***
Author
Time
Originally posted by: Doctor M
Hmm. First let me nit-pick and say that technically PAL audio isn't time stretched, it is time compressed since the runtime is less than the original. It runs faster, hence the increased frequency. Correcting it requires time expansion. For most PAL material, like it or not, that's really the best alternative out there if you're turning it back to NTSC.

Still time stretching, its just whats its called. Also, this isnt what normally would be done to return PAL audo to NTSC audio speed, as the vast majority of PAL audio is just sped up by 4.096%, not pitch shifted or time stretched in any way. So you just need to slow it down (resample it) by 4.096% and its back to the normal speed and pitch again. Its just in this bizarre case we have here where the original pitch was preserved with pitch shifting/time stretching.

Commenting on the rest (and I could be butt-ass wrong):
As far as pitchshifting. Typically it would not be needed as it only changes the frequency, not the actual speed/length of a clip. In this case (and I thought my ears were playing tricks on me too when I heard the butch C3PO) time expansion is STILL needed. Otherwise when the framerate is slowed to 24fps, the audio will run too short. We are (unfortunately for me) finding that evidentily it will need ADDITIONAL pitchshifting to reverse the whole process they put it through.


I dont think you get what i mean. time stretching does it as one job, with pitch shifting you need to do it in two stages. To get it from PAL to NTSC speed, first slow down the track (resample it) to get it playing to the correct length (4.096% slower), then pitch shift it up by 4.096%.

From what I'm seeing with soundforge though, pitchshifting doesn't seem to be done on a percentage basis making this much more wobbly a problem to deal with.


Just playing in SoundForge, to pitch shift up by 4.096%, you need to increase the pitch by 69.5 cents, and to pitch shift it down by 4.096%, you need to decrease it by -72.4 cents. In my mind they should be plus and minus the same value, but Sound Forge doesnt seem to think so, I'm going by the transposition ratio at the bottom of the window, either up to 1.04096, or down to 0.95904.

www.bardothodol.net

Author
Time
Originally posted by: pupil
Just playing in SoundForge, to pitch shift up by 4.096%, you need to increase the pitch by 69.5 cents, and to pitch shift it down by 4.096%, you need to decrease it by -72.4 cents. In my mind they should be plus and minus the same value, but Sound Forge doesnt seem to think so, I'm going by the transposition ratio at the bottom of the window, either up to 1.04096, or down to 0.95904.


That is easily explained, and probably should be done before other people make the same mistake!

Take this example. Our film source is 100,000 seconds long, we make that faster by 4.096% which leaves us with a PAL version at 95,904 seconds. Now if we slow the 95,904 version down by 4.096% we end up with a copy that is 99,832 seconds long, not 100,000 seconds. The reason is that 4.096% of 95,904 is obviously less than 4.096% of 100,000.

So the problem was you were thinking in terms of the original source where SoundForge is correctly thinking in terms of the current source. Easy mistake and something I hadn't thought of until you mentioned it! So whenever you are adjusting PAL material to 24fps then if my maths is right then you need to slow it down by 4.2709376043%. Been way to long since I've done that sort of maths!

http://www.kineticpast.com/starwars/thecheatlaserdisc.gif
Ooh, a laserdisc. The Cheat's playin' something on a laserdisc.
Everything is better on a laserdisc. Whatever happened to the laserdisc? Laserdisc!

Author
Time
Originally posted by: Gillean
Originally posted by: pupil
Just playing in SoundForge, to pitch shift up by 4.096%, you need to increase the pitch by 69.5 cents, and to pitch shift it down by 4.096%, you need to decrease it by -72.4 cents. In my mind they should be plus and minus the same value, but Sound Forge doesnt seem to think so, I'm going by the transposition ratio at the bottom of the window, either up to 1.04096, or down to 0.95904.


That is easily explained, and probably should be done before other people make the same mistake!

Take this example. Our film source is 100,000 seconds long, we make that faster by 4.096% which leaves us with a PAL version at 95,904 seconds. Now if we slow the 95,904 version down by 4.096% we end up with a copy that is 99,832 seconds long, not 100,000 seconds. The reason is that 4.096% of 95,904 is obviously less than 4.096% of 100,000.

So the problem was you were thinking in terms of the original source where SoundForge is correctly thinking in terms of the current source. Easy mistake and something I hadn't thought of until you mentioned it! So whenever you are adjusting PAL material to 24fps then if my maths is right then you need to slow it down by 4.2709376043%. Been way to long since I've done that sort of maths!


eeeek, please dont tell me you worked that our in your head!

Right, I see where I went wrong now, simple mistake, thanks for clearing that up Gillean. So would I be correct in saying that you would change the pitch plus/minus 72.4 cents, depending which way (NTSC to PAL or PAL to NTSC) you are going. I think thats right now.

www.bardothodol.net

Author
Time
Originally posted by: pupil
eeeek, please dont tell me you worked that our in your head!


Haha, nah, not a chance! I've already got a head cold so it was hard enough figuring that out with the aid of a calculator! I'm sure that had I tried my head would have exploded!

http://www.kineticpast.com/starwars/thecheatlaserdisc.gif
Ooh, a laserdisc. The Cheat's playin' something on a laserdisc.
Everything is better on a laserdisc. Whatever happened to the laserdisc? Laserdisc!