Bunny Greenhouse, February 10, 2009 8:18:54 PM CET
Chairlift - Evident Utensil (dir: Ray Tintori)
buffering is suffering
Time to Pretend: The Sequel "More Lo-Fi Action, Same Great Price"
huh? how is this anything like 'time to pretend'? spence, if you're gonna irrationally hate on videos, at least "unlink" your screen name so you don't suffer the same fate.
I wasnt saying it as a negative really, its like Time to Pretend in the way that both use a Lo-Fi device as the videos structural frame. Ray Tintori does it best, its why Time To Pretend is great and why this video is decent. The same can't be said for the numerous ripoffs flooding the scene lately. I dont hate on videos, mainly for the fact tat I actually know the blood, sweat and tears that goes into every type of video, from low budget - mega budget, and i'm not gonna hide behind an anonymous avatar when i do :-)
yea, there are striking similarities to the time to pretend video. after that awful killers video, i would much rather see tintori stick with his low-tech devices
this is really awesome
i've been waiting years for someone to take digiglitch to its proper level. and now i finally understand why i suspected it would be this beautiful: it's a direct depiction of the fourth dimension. (hi flv here.)
"direct depiction of the 4th dimension."
murder, she wrote.
been reading flatland lately, prgsk? or whence this kind of verdicts?
as for the glitch/ bad encode effect, tintori's not the first guy to have thought about it, (i think all the directors who at some point or another have downloaded a bootleg off the internet did) but kudos to him for being the first one to abuse it in such a manner.
Yeah, really interesting visuals here. The footage has what looks like some uh ... LARPing in it. Hmmm.
no, leg - i mean that quite literally. it's a simple way to capture a shape of time.
ponder that, if you wish, as you rest on this plush digiglitch rug:
1. i am a christian orthodox so whatever you show me - the wild ass'skin, sandpaintings or this rug, it works no magic on me.
2. a "shape" of time it's not some glitches, no matter how tempting that might sound. time looks like this, if you will:
(a physicist's hourglass)
Can somebody technically explain how this glitching is done? Is this a plugin...?
fuck i feel like my drugs just kicked in.
you can do it in a couple of ways, i guess:
1. fireware capture whilst testing poorly encoded mpeg's. (or vhs tapes, if you want vhs glitches)
2. ae boris filters in the 'witness protection' category and some tweaking?
3. but this is a wild guess, you play with the settings when importing your video?
If I had to guess I'd say it could be the time displacement filter in AE, similar to what (I assume) Bruntel did in Eraser, but instead of using liquid footage to time-displace it you use a digitally glitched out version of the proceeding clip. Or something like that. It does look like a bad firewire cable effect but it seems to controlled/storyboarded to be done just by messing with equipment. Looks nice though. How'd Moyes do it?
this effect is very reminiscent of the one in takeshi murata's work
really shitty youtube of his 2005 video "monster movie"
i think it is largely some kind of codec-level manipulation of the footage
from Chairlift's myspace:
"Evident Utensil is the first music video ever shot for datamosh compression. "Datamoshing" occurs when the set of compression instructions for moving pixels around a set of frames gets applied to another set of frames. In more technical terms, when the Key Frames (described frames) are removed and the Delta Frames (modifications) persist, leading to smooth transitions between clips instead of a cut. The term "datamosh" was coined by the art collective Paperrad, but the technique had been previously used by video artists Takeshi Murata and aPpRoPiRaTe."
Nice. A definite style emerging for Tintori.
paperrad's cranberry umbrella mosh;
takeshi murata's pink dot;
sven koenig’s aPpRoPiRaTe.
Gotta say, I kinda dig the embrace of the imperfection of codecs, etc., that's been going around. It's kind of like the artful use of film scratches and flashes back in the day (though we all know that's played out at THIS point for the most part). I'm sure this style of video will become cliche soon enough but for now, I think it works. This is sort of a colorful cousin to the NIN video for "The Hand That Feeds."
I'm still unsure of exactly how to do this. aPpRoPiRaTe says:
For lossy video compression several techniques to save space are applied. The one I exploit is that of delta frames. An encoded movie doesn't contain all full frames but only a few of them (key frames) – the rest of the frames are just saved as the difference between different frames. Any frame that is not a key frame is calculated out of the last key frame and all following delta frames. By simply manipulating / deleting the key frames of a moviefile it's easily possibly to transmogrify that file with minimal effort. Such a manipulation is technically just copying of data just a little different from how it usually would be done - but it reveals the nature of found movie files as collaborative works.
Youtube members say:
According to Nate Boyce who is a friend of Takeshi Murata he is manipulating codec divx. Which I don't really know what that means, but I think most of it is done by creating patches in jitter.
Most modern video codecs are made up of keyframes and in between frames. The keyframes are full images, but the in between frames only record what has changed from the previous frame. When it's working correctly, it creates the illusion of continuous sequential images, but with much smaller file sizes. Takeshi is manipulating this process in a way where he is arbitrarily inserting and removing keyframes, so that the inbetween frames (which represent motion) become the focus...
The secrets of Takeshi Murata - he turns the video files into to text. when the videos are in text form he manipulates them by deleting code etc. apparently he did a lot of commercial work so he was around tech savy guys and this is how he learned.
oh my god
the world has gone mad
20 or so comments of praise and adoration for this tripe......is ray tantori an actual film-maker.....i think not until i see evidence of genuine directing....using actors, showing us emotion, showing us a true understanding of the language of film.......i know its only a music video but you all cream yourselves so squirmishly.
There are real film-makers out there who post on antville and the films get left behind. Think ace norton, WIZ, karni and saul, de-turah, etc etc.
attack attack.....dont be rough
Got to disagree, a-morphic. The main purpose of a music video is to illustrate the music in such a way that you add to it. This video achieves that perfectly. The music and image together is quite beautiful imo. Don't get me wrong, I love seeing the techniques you've listed above put to good effect but there's also nothing wrong with experimentation. Watch Tintori's short films and you'll see he has an understanding of the "language" of film. It's good to see him cut and paste the words a bit to come up with something different.
that was rad, very impressive.
a-morphic: Two of his films have premiered at Sundance. What were you saying?
the thing i find interesting is that his videos do not reflect his narrative short films at all. range?
Download Finished is a site by Sven Koenig that applies template glitching effects to any movie that you can select off of P2P networks. For specifics on delta keyframes, you might have to ask on one of those geeky compression forums.
from the download-finished faq:
How exactly is the found footage input transmogrified?
For lossy video compression several techniques to save space are applied. The one DOWNLOAD FINISHED exploits is that of delta frames. Unlike the original film, an encoded film does not consist of 25 full images per second: The amount of full images is reduced to save space, and the parts in between full images (key frames) are calculated on the fly. The frames generated from the differences between key frames are called delta frames. By simply deleting the key frames of a film the file data is transmogrified to reveal the nature of the found footage files as a collaborative work with a very complex data structure.
Are the processed films mine?
Yes. And for that matter: they belong to everyone else, as well. After processing, you are free to save your film (or anyone elses, for that matter) to a DVD, write a nice exposé; and send it off to a film festival of your liking. Or cut the movie up and insert parts into other film works you compose. Make a music video. Extract stills. Etc. etc.
and allow others to wax lyrically about the direct depiction of the 4th dimension.
Even better on the 2nd viewing. It's could stand to be a little less loose, but I'm sure this was plenty of effort for the money.
The most prominent previous example of this sort of thing that comes to mind is Kris Moyes' 1st Presets vid.
Though the earliest usage I can recall was this peice by David OReilly.
Cool to hear more about how this is done. I wondered before if it was actually live capture of some odd scrubbing technique.
I do find it odd that there is no tool to do a cleaner version of this, applying the motion of footage to some other image.
Overrated. This is okay. Perspective.
Was going to post the David O'reilly vid. But najork beat me to it. Yeah, David really pioneered this.
Also, big congrats to him for winning the golden palm for PSS! :)
Golden Bear award you mean. That's rad.
Yes of course, sorry :) Can't wait to see his new episodes!
Am I the only one who is still completely in the dark as to how to do this technique?
How exactly does one "delete delta frames"? Is there an app that says, would you like to modify "delta" keyframes during compression?
Exactly what tools are used to do this and how? In all my years in post, I've never once heard of the term, "delta keyframes".
All of this info as extracted is very nebulous. It's like as if this is what I were to say when explaining how to composite a 3D scene:
"Well after the files get transferred from your chosen NLE, you can take the sequence and use this as a base while combining your respective passes at will. One finds the use of normals as de facto reasoning for proper depth and blur rendering of chosen scene. The trick of course being the approach, either layer or node based."
- - -
It's like, wtf....
btw, there's def more at work than just the keyframe business - eg., just to cite one instance, the very end sequence, and other more murata-like effects.
From najork (found via google):
So for anybody who isn't sure how this works: video codecs like MPEG-4 use motion compensation to cut down on the bit rate. Only a few keyframes of the video are encoded in full, about one every few seconds; the rest ("predicted" frames) store a rough estimate of how much each block of pixels has shifted since the previous frame, along with just enough actual pixels to make up the difference between the estimate and the real picture. So if there's a single moving object on a static background, all that needs to be stored is the area of the background that's been uncovered since the previous frame.In this case, what I suspect they did is encode their raw video clips with no keyframes (except the very first one), then spliced them together, so the decoder applies the motion vectors to the wrong original image. It looks like they also duplicated the same frame several times in some places, to get those swirls of color.
an improvised low tech way would be to screen-cap some playback manipulation. Have both clips encoded together with minimal keyframes. Play it back and hit pause where i want the freak-out to start. Then jump ahead to where i want the ease in to start. Hit play and it should start melting until you hit another keyframe.
Moreover (maybe the closest we've gotten yet):
9.3. Encoding two pass MPEG-4 ("DivX")
The name comes from the fact that this method encodes the file twice. The first encoding (dubbed pass) creates some temporary files (*.log) with a size of few megabytes, do not delete them yet (you can delete the AVI or rather just not create any video by redirecting it into /dev/null or on Windows into NUL). In the second pass, the two pass output file is created, using the bitrate data from the temporary files. The resulting file will have much better image quality. If this is the first time you heard about this, you should consult some guides available on the net.
- - -
So basically is it possible that one file is the log, and that log file gets converted with the second file to create one movie file....?
From the DivxLab page: "During the first pass the encoder analyzes the source video complexity and stores this information in a statistics file. No video is output during the first pass. Later, the encoder is started again to perform the actual encoding. During this second pass the statistics file that was created on the first pass is loaded to inform the rate control algorithm of the complexity of every frame of video that it should expect to encode, and this enables the rate control to efficiently distribute the available bits throughout the video in a manner that improves quality consistency."
So yeah you could hypothetically pause the process after the first pass, edit the ".log" file, then perform the second pass and you'd have incorrect compression. I'm not sure exactly how you do this, but it might be accomplishable using AVIsynth and VirtualDub. Apparently VirtualDub has a frame counter that will tell you if a frame is a keyframe or delta frame.
You should read the rest of that DivxLab page because it's packed with all kinds of nuggets. If you find more feel free to post it, I'm kind of curious myself. Btw what's the link for that screencap melt quote?
from Ray Tintori:
"You'll never figure out how I did this, you little Antville nerds. Get off my wiener."
Weird, Kanye's new video has the exact same concept.
Jesus this effect is not that difficult. Its not even a secret.
There is no mystery.
Maybe a certain somebodies wiener has relocated to his forehead and is possibly inflated by his ego?
what track is kanye's new video for?
not sure, i think it was directed by neon?
who cares how it was done, it's really striking. it does this strange thing where you're attempting to cohere the images you're glimpsing but before you can grasp them the world's off melting again. like it purposely tweaks your sense of pattern recognition.
oh yeah: moshing by pitarch, 2007.
Update: How-to courtesy of Bob Weisz.
Paul B. Davis/Beige