Why would you specifically use ww12345's to match colors to?
I believe he modified them during the conversion to NTSC.
Mentor's untouched release is probably a more accurate source.
Btw, saw the screenshots and hope you find some color in there besides red.
Funny thing is, I predict any BD release to be open matte since Cameron has a tradition of reformatting to fill TV screens.
HDBits has several versions. I see they have a 1080i DVB/mpg2 version from some newsgroup (not many details, not even AR) @ 15gb and a 1080p x264 version (possibly DTheater sourced with a downmixed audio track).
They also have a standalone rip of the original full bitrate DTS/DTheater audio track.
The rest are 720p (including an open matte one). Funny thing is I don't see the version that I originally got there which was the open matte 720p with the DTheater 1510kbps track.
The point being, I'm not sure there is a 1080i/p version that's open matte.
I've watched the open matte version and it felt a bit less theatrical. (I've got it in 720p with the DTheater DTS track.)
OAR is still important even where 16x9 TVs are involved. Cinematography is an art.
http://www.thedigitalbits.com/columns/the-rumor-mill/true-lies-abyss-2014-x-files-twin-peaks
Just sayin'...
It is my understand that Disney is fantastic with preservation... we just don't get to see THOSE transfers.
They archive high resolution scans and then modify them for home video release depending on what the flavor of the days is.
As far as Beauty and the Beast, haven't all versions been altered since the IMAX release in order to make room for Human Again? (Even when you watch the version without the song.)
I thought I was the only one hoping for a better version of the theatrical cut.
The whole devil worship plotline is awful.
It actually looks like you can see a bit of Herman's arm through her leg. Weird.
Yes, yes, all these years later and I finally watched this.
For all it's flaws (as a cut, not as a preservation), it's actually really good. I enjoyed it more than the last time I watched the theatrical version.
Anyway, I wanted to throw a thanks FanFiltration's way and confirm that this is still floating around before I delete it from my drive.
It doesn't look like a spot on the film... but it really doesn't look much like anything else either.
Sat through most of the full version today.
Definitely no transporter accident. It doesn't even make sense in context of the technology (receiver pad isn't needed if there is a sending pad (is it?)). Seems pointless (but DOES explain the reason for no science officer).
Cut all the times anyone says "External view," to see what's going on outside of the ship they are talking to on subspace. What the? Where is the camera guy standing? Awful awful plot device.
Just show the external view, don't pretend that people see it on their view screen. At the briefing I recommend that the screen be replaced by static when they cut back to the scene, not the empty space where the ship was.
From what you say, it sounds like ColourLike is just buggy. Has anyone informed the author that it doesn't actually perform like his instructions say?
Although re-reading the details thinking about what you've said, it could be intentional. It could be assumed to average together every single frame (with every=1) or every 3rd frame (with every=3). That blows...
Although it is not recommended, it might be possible to force it to do what we want by incorporating WriteHistorgram and ColourLike into the same script.
Some sort of For...do loop that counts off each frame, trims one frame at a time, generates a histogram for source and destination clips, applies colourlike to that frame, and then increments.
Theoretically possible, probably crazy slow to make work.
About ColourLike:
WriteHistogram(clip, string outputFile, int every = 1)
'Every' lets you set every how many frames it samples the colors. 1 means every frame. So yes, it balances the color on a frame by frame basis.
I see no mention of a setting to find an average historgram for the whole film... and why would you want to?
The problem is, in my tests, ColourLike just doesn't work particularly well.
Most recently I attempted to use it to recolor the Matrix BD with the original DVD as the source color palette. It did next to nothing. I got better results mixing the BD luma with the DVD chroma (obviously not ideal).
Has anyone actually had good results with Colourlike? (Never heard of RGBmatch before.)
TheDigitalBits had an interesting article a little while back regarding color changes made for the 3D version of Top Gun which has a shift towards orange. ( http://www.thedigitalbits.com/columns/the-3rd-dimension/update-on-top-gun-3d )
The important bit is this:
Many colorists, especially when mastering for video have been trained to hate the color red because of the bleeding issues with NTSC televisions. As such, they instinctively avoided it, which is one of the reasons why “the film has never looked this way before.” The last master for Top Gun was run at least 7 years ago, and in that time we’ve jumped several generations of technology in terms of digital color timing tools. In 1986, everything was chemical – you did the best you could to get the colors you wanted. Today, you can turn everything any color of the rainbow you desire.
It makes me wonder how many of the changes we perceive is more because of our familiarity with old home video masters and not the theatrical colors.
I can agree with that logic.
Most sources say a 4k scan is 100% of the detail of a full size print. I'd think a 16mm has half that.
Well the first post in the thread will be 4 years old come June...
My computer is carved out of stone, and while I can't play back 1080p smoothly, I can edit and encode it... as long as I'm patient.
I don't think there is a minimum for working with it, just real time playback.
Yeah, stupid confusing, but when the use a 'k' number (thousands of lines) they are talking horizontal lines, but when they use p or i (progressive/interlaced) they are referring to vertical lines.
It's why UHD/4k TVs are going to be confusing to the public. It almost sounds like it's 4 times the resolution of 1080p, but it's more like double. (Besides the fact that seeing the difference will be nearly impossible.)
My mom still gets confused trying to record one show on her VCR while watching another. I can't even attempt to explain HD.
Oh, and good script, ntcuFaal, I like that. It takes the eyeball out of the equation.
If you are familiar with avisynth, something like this:
avisource("SotS.avi") #I'm assuming some lossless format.
spline36resize(1920,1080) #or whatever the appropriate AR conversion is.
a=subtitle("1080p")
b=spline36resize(1280,720).spline36resize(1920,1080).subtitle("720p")interleave(a,b)
Then use something like VirtualDubMod to view the avisynth file.
Everything will be 1920x1080, but you can step single frame through it with even frames being 1080p and odd frames being 720p upscaled back to 1080.
If you can't see any detail difference between them, then 720p should be fine.
Anyone with a better method?
I can answer ONE of those questions.
A 2k scan provides 2000 pixels horizontally. 1080p is 1920 pixels wide, so yeah, a 2k scan is pretty much 1080p.
Now if the 16mm reduction has 2k worth of information is another question all to itself that I cannot answer.
If no one knows, the best way to go is to take some frames and compare a 720p resize to the source and see if there is any less detail.
If it's there, it would be worth preserving.
ww12345 - Sorry if it has been asked and answered, but at what resolution are you scanning it at? Are you going for Blu-ray?
Space Kaijuu said:
The Rigid Tools poster wouldn't be too hard to reinsert into the original since it's only a few frames in the animation and it's part of the background painting rather than some sort of cell or animation element. Is the image above the best that's out there, or are there any higher resolution frame captures?
I'm about halfway done restoring the altered frames from the Jessica Rabbit panty shot scene. The latest edit to the scene actually removes the panties from her butt and makes it look even more like she's not wearing panties than the original did. Seriously, Disney?
Love to see the screenshot comparison there.
No, GetFLV, TubeDigger and a few other similar programs actually save the incoming data from the websites. They are true rips, not screen captures.
Flash is frequently tries to fix their code to stop them, and they have constant updates to get around it. It's a fun little arms race.
Oh, one more thing. Several of these streaming services can be ripped right from your PC with inexpensive software like GetFLV.
Check out their list: http://www.getflv.net/site.php
drngr said:
They must have added that exclusion to the list just recently.
Here is the US account for the same seller. I posted the Aussie one because it was slightly cheaper before, with the exchange rate.
http://www.ebay.com/itm/4-Port-1-x-4-Audio-Video-1080P-HD-HDMI-Splitter-Amplifier-Multiplier-Box-/220864833576
Amazon link: http://www.amazon.com/VeeCome-portable-Powered-Splitter-Support/dp/B0094OEQNU/
Thanks. Just noticed these are listed as 1.3 compliant. Hopefully they'll update to 1.4. (It's interesting they call it a 3D splitter even though 3D was added in the 1.4 specs.)
Btw, the ebay.com has the same exclusions, but the Amazon link seems good.