Stephen King's IT (1990) - Original two-part version! [RELEASED] — Page 2

Jetrell Fo said:

SpookyDollhouse said:

Jetrell Fo said:

TheHutt said:

Why do a composite of LD and 2 tape VHS? The LD is the uncut version (192 min). No need to compose it with VHS.

It was merely a suggestion, not a direction.

The VHS is uncut as far as I can tell too. At least the 2 tape version.

There’s literally no need to do that though since the Laserdisc is uncut. Ideally, editing the cut footage from LD into the Bluray transfer would be best.

I thought there was no need for a composite yet here we are discussing a different option for a composite. I’ll never understand why people want to dismiss suggestions so easily but it’s all good, as long as the project is done in a way that everyone enjoys, that is what counts.

It’s not about dismissing anything so easily, there’s literally no point to use a lower quality source for alternate footage if higher is available, which in this case is the laserdisc and 576i broadcast. Someone already pointed this out. It’d be extra work (and once again no point) in mixing VHS footage with laserdisc footage when it all exists on laserdisc anyway.

My point was not about VHS but a composite project designed to maximize the presentation … that is what compositing can help with.

Of course it does.
For example, if there will be identical footage on BluRay, then I will try and re-use it. For example, if the end credits of Part 1 are identical to the BluRay end credits, I could use these instead. Same for the opening of Part 2 (if it is the same as Part 1, that is).

I am curious if the 576i source will be of better quality than the (very good) NTSC LD by RidgeShark. 😃

My next step will be gettin the BluRay (the steelbook I preordered is scheduled for Oct 12th) and match the colors, the noise level and the framing of the LD source to the Blu.

TheHutt said:

Of course it does.
For example, if there will be identical footage on BluRay, then I will try and re-use it. For example, if the end credits of Part 1 are identical to the BluRay end credits, I could use these instead. Same for the opening of Part 2 (if it is the same as Part 1, that is).

I am curious if the 576i source will be of better quality than the (very good) NTSC LD by RidgeShark. 😃

My next step will be gettin the BluRay (the steelbook I preordered is scheduled for Oct 12th) and match the colors, the noise level and the framing of the LD source to the Blu.

the end credits for both parts are different.

Thanks for the info! Then original credits they are.

Wow, looks great! Definitely the best so far. Maybe the laserdisc can be used to help patch up the “True Entertainment - Richard Thomas Day” text.

Looks great!
The logo can be probably patched up by LD footage. The squeezed part of the credits could probably be lifted from BluRay.

Hello @JayArgonaut,

Any news on Part 2?

I am glad to see that digital and analog footage can be used to make one great project for this.

I’m excited to see the end results.

@JayArgonaut, could you send them to me as well? I’d like to give it a shot, too, in case RidgeShark is busy.

JayArgonaut said:

Here’s part 1, if you like what you see, I’ll share the grabs for part 2.

TheHutt said:

@JayArgonaut, could you send them to me as well? I’d like to give it a shot, too, in case RidgeShark is busy.

@JayArgonaut, +1

I’m still waiting for my steelbook of IT.
The front part of the intro credits looks the same, so that HD intro credits could be used for beginning of part 2 (as far as I see, only the last photo in the credits is different).

JayArgonaut said:

Sorry Hutt, I clean forgot! I went straight to the next stage by sending RidgeShark copies of the recordings so that he could get working on them.

I think others want this, can you or RidgeShark share?

My French steelbook is scheduled to arrive on Saturday. Looking forward to delve into it. 😃

Hello,

I am having some trouble with color matching the LD footage to the BluRay footage. In Part 1, two shots need to be inserted from LD to BluRay footage.

1. The full shot of the wife yelling “No” twice (this shot is faded out on BluRay).
2. The “IT” writing on bathroom wall shot with “To be continued” title card.

I used AviSynth and the script ColourLike to create two histograms:

1. from the beginning of the wife yelling scene
2. from the previous shot of the “IT” writing on the wall.

The first shot works quite well.
BluRay footage: http://www.thehutt.de/it/02_blu.png
LD footage (upscaled + added noise, no color correction): http://www.thehutt.de/it/02_ld_nocorr.png
LD footage (upscaled + added noise, color correction): http://www.thehutt.de/it/02_ld_corr.png

However, on the final “IT” writing, I have some trouble. The histogram I lifted from the previous “IT” writing (a couple of shots before) and it looked quite okay.

BluRay footage: http://www.thehutt.de/it/01_blu.png
LD footage (upscaled + added noise, no color correction): http://www.thehutt.de/it/01_ld_nocorr.png
LD footage (upscaled + added noise, color correction): http://www.thehutt.de/it/01_ld_corr.png

However, if I apply this histogram to the final shot missing on the LD), there are some unwanted artifacts. Also, as this shot ends in a fadeout, applying color correction here results in some really weird artifacts (especially where the black isn’t actually black).

LD footage (upscaled + added noise, no color correction):
Frame 001: http://www.thehutt.de/it/03_nocorr_frame001.png
Frame 200: http://www.thehutt.de/it/03_nocorr_frame200.png

LD footage (upscaled + added noise, with color correction):
Frame 001: http://www.thehutt.de/it/03_colorcorr_frame001.png
Frame 200: http://www.thehutt.de/it/03_colorcorr_frame200.png

Here is the AVS script I am using (for the final shot which has these problems):

Import(“d:\Anwendungen\Multimedia\AviSynth\plugins\QTGMC\QTGMC-3.32.avsi”)
AviSource(“d:\video\It\EndofPart1DV.avi”, fourCC=“CDVC”)
AssumeBFF()
ConvertToYV12(interlaced=true)
QTGMC( Preset=“Slower” )

Pulldown(0,3)
Trim(2237, 2484)

Crop(12, 0, -4, -0)

eedi3_rpow2(rfactor=2, cshift=“spline36resize”, fwidth=1440, fheight=1080, threads=1)
AddGrain(10)
AddBorders(240, 0, 240, 0, $000000)

ColourLike(“d:\video\It\IT_1a_hist_s.txt”, “d:\video\It\IT_1b_hist_s.txt”)

So my question is: is there a way to make ColourLike work even with the fadeout? Or are there any better alternatives?

Also, the end credits of the BluRay are not identical with the end credits on Part 2 of the TV series (in fact, the BluRay credits look like they’ve been redone for the BR from scratch, as they feature an entirely different font). These will have to be replaced as well.

TV end credits: http://www.thehutt.de/it/tv_endcredits.jpg
BluRay end credits: http://www.thehutt.de/it/blu_endcredits.jpg

Perhaps DrDre’s color matching tool could be useful here? http://originaltrilogy.com/topic/Color-matching-and-prediction-color-correction-tool-v13-released/id/18128/page/1

I would try color-matching the laserdisc to JayArgonaut’s broadcast copy, which looks closer to the blu-ray than the laserdisc does and also since it has the missing scenes. Then I would tweak the color temperature and skin tones in that result to better match the blu-ray, which has a warmer, yellowish look compared to the more blue-ish broadcast copy.

As for AviSynth processing, while I love QTGMC for processing native 50i/60i footage and also for some problematic telecined films, it’s not the best choice in this instance. QTGMC has a way of smoothing small/fine details away and creating a subtle water-color, waxy look.

See here for a comparison between QTGMC and TFM on part 2 - http://screenshotcomparison.com/comparison/188849 That roof is a real stress test for any of these filters, but also notice all the other little fine details that get smudged due to QTGMC. Check out that “Derry Inn” sign, the various bushes and trees, and the roof of that yellow car.

Since Warner did a really good job with the 3:2 pulldown on the laserdisc, you really don’t need to use QTGMC at all. A plain, accurate inverse telecine will give you the maximum resolution afforded by this source.

These are basically my previous settings on my earlier upscaled example, using TFM and TDecimate to perform the inverse telecine:

AviSource(“d:\video\It\EndofPart1DV.avi”)
ConverttoYV12(interlaced=true)
fields=last.AssumeBFF.SeparateFields()#.DFMDeRainbow()#remove # to turn on, beware of unwanted desaturation
backward_vec2 = fields.MVAnalyse(isb = true, delta = 2, pel = 2, overlap=4, idx = 1)
forward_vec2 = fields.MVAnalyse(isb = false, delta = 2, pel = 2, overlap=4, idx = 1)
fields.MVDegrain1(backward_vec2,forward_vec2,thSAD=400,idx=1,plane=3)#reduces color noise with no negative effect
Weave()
TFM(PP=0,mChroma=false)
TDecimate(mode=1)
Crop(12, 0, -4, -0)
Spline64Resize(1440,1080)
Super = MSuper(sharp=2)
bvec3 = MAnalyse(Super, isb=true, delta=3, blksize=16, overlap=8, search=3, truemotion=true, dct=5)
bvec2 = MAnalyse(Super, isb=true, delta=2, blksize=16, overlap=8, search=3, truemotion=true, dct=5)
bvec1 = MAnalyse(Super, isb=true, delta=1, blksize=16, overlap=8, search=3, truemotion=true, dct=5)
fvec1 = MAnalyse(Super, isb=false, delta=1, blksize=16, overlap=8, search=3, truemotion=true, dct=5)
fvec2 = MAnalyse(Super, isb=false, delta=2, blksize=16, overlap=8, search=3, truemotion=true, dct=5)
fvec3 = MAnalyse(Super, isb=false, delta=3, blksize=16, overlap=8, search=3, truemotion=true, dct=5)
MDegrain3(Super, bvec1, fvec1, bvec2, fvec2, bvec3, fvec3, thsad=500)
UnsharpMask(strength=30,radius=3,threshold=0) #fairly subtle- brings details out a tad
GrainFactory3(g3str=3,g2str=6,g1str=4,g1size=3,g2size=3,g3size=3,g1shrp=8,g2shrp=8,g3shrp=1)#still needs tweaking
Trim(2237, 2484)
ColourLike(“d:\video\It\IT_1a_hist_s.txt”, “d:\video\It\IT_1b_hist_s.txt”)

Filters:
http://avisynth.org.ru/mvtools/mvtools2.html or http://avisynth.nl/index.php/MVTools (for MDegrain1 and 3)
http://avisynth.nl/index.php/WarpSharp (contains unsharpmask)
http://avisynth.nl/index.php/TIVTC (contains TFM and TDecimate)
http://avisynth.nl/index.php/GrainFactory3 - lots of options for grain
http://avisynth.nl/index.php/DFMDeRainbow - I usually turn this off because it can desaturate reds and some other colors, but it is very useful for the rainbowing on the opening credits of part 2. Perhaps apply this first, then run the footage through DrDre’s program to compensate and regain more accurate colors.

Hopefully some of this will be useful. Also, I will get you part 2’s credits soon!

Hello RidgeShark,

Muchas gracias for the precious information! I’ll check it out. 😃

Small update:

DrDre’s color matching tool works like a charm! Too bad one cannot use it with AviSynth (or can one?) and thus I’ll need to render a color-corrected intermediate.

The following piece of code doesn’t work, unfortunately:

fields=last.AssumeBFF.SeparateFields()#.DFMDeRainbow()#remove # to turn on, beware of unwanted desaturation
backward_vec2 = fields.MVAnalyse(isb = true, delta = 2, pel = 2, overlap=4, idx = 1)
forward_vec2 = fields.MVAnalyse(isb = false, delta = 2, pel = 2, overlap=4, idx = 1)
fields.MVDegrain1(backward_vec2,forward_vec2,thSAD=400,idx=1,plane=3)#reduces color noise with no negative effect

AviSynth reports an error that MVAnalyse cannot be found (although it works fine if called standalone, but not as a fields method). Strange.

Small status update;
Part 1 has been color-matched and upscaled. Of course it isn’t the same quality as HD, but it looks quite okay when color-matched.
I experimented with several options (color-matching prior to upscaling, adding noise before/after color-matching) and I seem to have found the result I like best.

So right now it’s this:
-Upscaling and adding grain first,
-Color-matching in AfterEffects after that.

Next week I’ll try to get Part 2 done. Though I am still missing the end credits for Part 2.