Friday, March 07, 2008

Perception


Consider this post an addendum to yesterday's expose of Kos' nonsensical "darkened Obama" musings.

I spoke to my old friend, the video editor. He confirmed pretty much everything said in my post below -- and then he pointed out one obvious fact which utterly demolishes the Kos theory: Everything in that controversial Clinton commercial is dark. The broadcast version seems to have been fine, but the online incarnation is murky. Hillary is dark. The shot of the White House is ridiculously dark. The kid in bed is way too dark. The shot of Obama is actually the brightest image in the whole damned show!

If Kos and TroutNut had bothered to look at their own materials, they would have understood that Obama was not singled out -- the same technical problems visible in that one shot affected every shot. The entire video is too wide, not just the shot of Obama. Blame the Pixel Aspect Ratio glitch, as outlined previously.

(Side note: For some reason, many people remain blind to aspect ratios. They buy widescreen TVs and then play 1.35:1 material stretched out to fit from side to side. The damnfool idjits think that the picture is supposed to look that way!)

Now let's zoom out for a wider view.

As we saw in our previous installment, most videos placed online are darkened by lossy codecs. Why did this phenomenon go unnoticed (except by professionals) until now?

Even though image degradation is all over the internet, it registered on the national consciousness only when a black candidate ran for president. If a codec darkens a shot of a white woman, nobody cares. If Obama looks dark, the sirens go off and fear-junkies screech that the darkening occurred by intention.

What does that fact say about our attitudes toward race?

Here are three images culled from the current Talking Points Memo page. The brightness is "off" in all three instances. The shot of Stephen Harper was probably over-exposed to begin with. The shot of McCain was taken with the light source in the background, which tends to throw off an exposure meter. The still of Hagee comes from a video, and I'll bet you dollars to donuts that the image went dark due to the codec. The original broadcast probably looked much better.

Did you notice these problems before I drew your attention to them? No. Why? Because all three men are white.

In the days and months and years to come, we're going to see lots and lots and lots of pictures of Barack Obama online. Many of them will be too dark. Please do not presume that conspiratorial forces are at play. Obama will "go dark" for the same reasons that Hagee and McCain went dark in the examples presented here.

Another point.

The believers in a Grand Plot Against Obama tend to mutter ominously that Hillary's commercial contains "subliminal" content. All such accusations are paranoid nonsense.

Get ready for a shock, conspiracy fans: Subliminals are impossible on television.

(To read the rest, click "Permalink" below)


The word "subliminal" means an image that the conscious mind cannot perceive. In the United States, television broadcasts at 30 frames per second. A single-frame image -- 1/30th of a second long -- is perceptible, and therefore not subliminal. My editor friend tells me that he can easily spot a single "flash" frame when video is running at twice normal speed.

Forgive a bit of self-quotation:
I can't tell you how many times I've argued with film-illiterate conspiracy buffs who knew -- just bloody knew -- that subliminal images appear in The Exorcist. Every time I encounter this claim, I explain that, while quick cuts do exist in that movie, subliminals are impossible. Film runs at 24 frames a second, and 1/24th of a second is not subliminal; that's why you can see scratches and splices in a bad print. 1/3000th of a second is subliminal, but to flash an image so briefly requires a specialized piece of equipment called a tachistiscope. (For the same reason, no subliminals can ever appear on television, which has a 30 fps refresh rate.)

The usual response: "There must have been a tachisti-whatzit in the theater!"

Actually, I've spoken with the projectionist who ran The Exorcist during its first run at the National in Los Angeles back in '73. (In those days, big films did not open wide.) There was no such device.

Point won? Nope. The ne'er-say-die conspiracy buff would invariably accuse me of not knowing anything about movie technology (!) or of being part of the conspiracy.

(Before you say it: Yes, I know all about Vicary's stunt in the 1950s. He used a tachistiscope, and his results had no scientific value due to the lack of controls. And yes, I know about the Silverman/Weinberger experiments of the 1980s, which have been questioned due to some unsuccessful attempts at replication.)
Actually, I wasn't hard enough on Vicary. He flat-out lied about his results. Yet to this day, conspiracy-crazed boobs continue to repeat the lie!

Some controversial scientific studies indicate that carefully-chosen subliminals can have a modest effect. The most (in)famous experiment of the 1980s incorporated the phrase "Mommy and me are one." But the effect is subtle -- and without a tachistiscope, you just can't do it.

What about Wilson Key and his "embeds"? For those of you who have never read his books (which were wildly popular in the 1970s), Key thought that Ritz crackers were imprinted with the word "sex." He also said that advertisers paid airbrush artists to place penis imagery in photographs of icecubes.

Bottom line: Key was a blowhard pseudo-scientist and his books were garbage. (See here.) Back in the 1980s -- a time when I earned decent money -- I was a damned fine freehand airbrush artist. Nobody ever asked me to "embed" any hidden words or pictures into a photograph. I've never even heard a rumor of another artist receiving such a gig. Advertising pros laugh at Key's assertions.

Mind you, artists do sneak in hidden images as private jokes. I've done it. In the early 1980s, I illustrated a series of books designed to teach BASIC to kids. In one book, I hid the outline of a naked lady in a drawing of a castle. You know who gave me the idea? Wilson Key!

There's an old story (which may or may not be true) about an illustrator in the early 20th century who got a commission from Delmonico's restaurant. They asked him to create a pen-and-ink drawing for the Christmas menu. He provided an elaborate, exquisitely-detailed cityscape titled "Christmas in New York." After the menu was printed and in use, someone noticed a startling detail: In one tiny window of one small building, the artist had drawn an ultra-teensy image of Santa Claus performing an unnatural act with a reindeer.

I don't think that this image unconsciously sparked mass acts of animal abuse.

And I don't think that anyone's vote ever has or ever will be affected by a "subliminally" manipulated image of Barack Obama.

Subliminal advertising is a paranoid fantasy. If any blogger says otherwise, ask him or her to provide experimental data.

22 comments:

Anonymous said...

Sorry, but this seems to be an other aisle (?)of the same operation : -> http://www.google.com/search?q=%22marc+rich%22+%22viktor+bout%22&num=100&hl=en&newwindow=1&safe=off&start=0&sa=N

Anonymous said...

They could have (if they were the kind of people who listen to FLACs instead of MP3s) or if they really wanted to-- they could have used Huffyuv, Lagarith, LOCO or any of the other dozen losless codecs available now. They didn't have to use Sorensen's MAMMY-- might as well have added a blackface filter...

Anonymous said...

Visual literacy in this case requires the viewer to observe accurately, and then to verbally describe what they are seeing accurately. In so many case in the internet where visual literacy is needed, in order to understand an image you also have to have an understanding of the processes involved.

I'm glad you took the time to go through this. It's not an easy task to educate people on something like this.

You have, however, chosen a format for the blog, with these anorexically skinny columns of text requiring endless scrolling, that makes it almost impossible to connect your explanatory text with the images that illustrate it. Putting it up as a full page article might give people the chance to make that connection more easily.

Joseph Cannon said...

Are you out of your freaking MIND?

First, YouTube doesn't allow lossless video, for reasons that would be apparent to you if you weren't insane.

Second, do you really mean to say that everyone who uses a lossy codec is a racist? So, like, EVERY video available on YouTube is racist? That "Sweet Charity" clip online was put there by a racist?

Are you intentionally trying to set me off by tweaking my nose, or are you simply a nut?

By the way -- have you people noticed that Jon Stewart and Steve Colbert look darker in their online clips (the official ones on the Comedy Central site) than they do when broadcast?

Are you saying that some insidious conspiracy is trying to blacken Stewart and Colbert?

Joseph Cannon said...

FF: First, understand that my "Are you out of your mind" comment was written in response to scalarparty, not to you.

Second, the formatting on your screen depends on the settings on your monitor. I've tested Cannonfire on a variety of screen resolutions, and it usually looks pretty good. But things may start to go funky if your resolution is low or if you use very large text.

Joseph Cannon said...

One last thing, scalarparty. FLAC and MP3 are audio codecs, not video codecs.

Personally, I transfer all my FLAC and CD files to m4a, 200 kbps VBR. If you can hear a difference between that and the original, your ears must be so pure they bleed on Easter Sunday.

Audio survives compression much more readily than does video, in my estimation. Even when I have a well-made DivX file, I always have to brighten the image in playback to make it look acceptable.

That's why the gods of the internet gave us VLC player...

Anonymous said...

Colbert Rocks!
http://action.credomobile.com/2008/03/colbert_attreason.html

AitchD said...

You mean Wilson Bryan Key.

In this 'subliminal messages' context, you play a little too fast and loose with the notion of what 'subliminal' means. It means something like 'unnoticed' much more than it means something like "an image that the conscious mind cannot perceive", as you want my lying eyes to believe. At this moment I am sublimating the annoying machine vibrations of my frost-free refrigerator, along with trillions of other potential stimuli. If I didn't know your background and your grasp of the matter, I'd say you were confusing ideas about 'sublimation' with 'persistence of vision' and the necessary illusion of motion it produces.

The 1/30th (or 30 fps) second in (pre-digital) video isn't mechanical like the 24 frames/sec movie frame rate (which is the arbitrary rate that conforms to our sense of true, i.e., real-life motion). The video frame rate is dependent on AC (on-off) electricity, 60 Hz in North America, 50 Hz in Europe. I think the image resolves 30 times a sec (on), alternating with nothing (off) 30 times a sec.

When I was in France I could see the TV flicker-effect, the horizontal 'bar' crawling vertically (very much like when a movie camera records a TV set that's on) because my brain was programmed at 60 Hz for the video illusion. After an hour or so, my lying brain 'learned' to perceive the images at their 50 Hz resolves. I guess you (when you're in a good mood) could say that electrical 'bar' I saw was subliminal.

Do you wonder if infants develop neurons, via TV watching, which adults and older children never developed if they weren't exposed to TV as infants?

By the way, if you look closely at The Rolling Stones' "Sticky Fingers" album sleeve, you can see a red tongue sticking out from parted red lips!

Joseph Cannon said...

H, I am sorry I misrecalled the man's name. Thanks for bringing it to my attention. I've corrected.

Joseph Cannon said...

H, I should have responded to this point earlier...

"In this 'subliminal messages' context, you play a little too fast and loose with the notion of what 'subliminal' means. It means something like 'unnoticed' much more than it means something like "an image that the conscious mind cannot perceive", as you want my lying eyes to believe."

WRONG.

Dead wrong. Look up the scientific literature. You have it all ass-backwards.

An image is subliminal only when it cannot be perceived consciously -- even if a trained observer is told ahead of what to look for.

If we broaden the definition of the term to mean "unnoticed," then the term becomes subjective and meaningless. You'd be surprised how many morons out there do not notice things that are thunderingly obvious.

Let me give here a story that I privately told a correspondent in email...

Toward the end of the run of "Titanic," I went to the megaplex to see another show. Since I had about twenty minutes to kill before my movie started, I popped in to catch a bit of "Titanic" (which I had already seen), in the hopes of, er, um, seeing Kate Winslet naked.

Guess what? The movie was being run flat. It was supposed to be anamorphic.

(If you don't know what those terms mean, spend a day on this site...

http://www.widescreenmuseum.com/

Education and fun!)

Back to our tale...

About twenty people were in the theater, sitting in sheep-like contentment as they watched a movie where everything was thin and the heads were cut off.

I nudged one of the other patrons. "Hey," I whispered. "I'm not supposed to be here, so I can't complain to the manager. But shouldn't YOU complain?"

"Complain about what?" he asked.

"The picture's not right!"

"Not right how?"

"It's projected wrong. Everything looks thin. The heads are gone. Kate Winslet looks like she has lost a LOT of weight..."

Long pause. Then the guy says: "Oh. Yeah."

Now I ask you: Is this a case of subliminal imagery?

By your definition, it was. The jackass did not notice what was right in front of his eyes. Neither did anyone else, apparently.

Or is this a case of people being just plain fucking morons?

Sorry, but we have restrict our definition of the term. "Subliminal" means imperceptible, not ignored.

Subliminal also does not refer to techniques which might be described as psychologically manipulative. All art and entertainment is psychologically manipulative -- every brush stroke of every painting, every note of every song.

Anonymous said...

Do you wonder if infants develop neurons, via TV watching, which adults and older children never developed if they weren't exposed to TV as infants?

Based on the famous Hubel/Weiss experiment with kittens and vision, this could be true, if there were some television watching specific neuronal receptors, comparable to edge receptors in cats' visual cortex.

I don't know of any reason to think that the if part of the above is true, however.

...sofla

AitchD said...

I think you're out to brunch, but too rigid to argue with. Once upon a time, we learned in Psych 101 that the normal person can't recall verbatim a string of digits if it exceeds seven digits. That was before area codes and the 'new' pattern recognition they've fostered.

Yeah, our school showed Woodstock without an anamorphic lens. (They also didn't have a projector that masked the frame, so boom mics sometimes interfered with poetic faith, say, in Saturday Night Fever.) Before that, we ordered 2001: ASO and had to rent a lens, plus we had to rent an academy's auditorium, plus we realized its screen was tiny, so we had to put up 3 white bedsheets. Hey, I'm on record for being the first (and maybe still the only one) to point out that when the bone is tossed, becoming an orbiting bomb, and Floyd's pen is floating, Kubrick's cinematic cut suspends three objects simultaneously: the first tool/weapon, maybe the last, and the one that got us from Moonwatcher to Floyd - the device that extrapolated back and then ahead during the era which is not depicted in the movie.

Everything I read about the meaning of 'subliminal' refers to a 'threshold' of 'consciousness'. Don't those terms suggest subjectivity (or idiosyncracies even) to you?

Have you ever READ Hawthorne's "The Birthmark"?

Your anecdote and point about "Titanic" - "Now I ask you: Is this a case of subliminal imagery? By your definition, it was" - is so insulting that now I regret not deserving it.

sofla, I had that very kitten experiment in mind.

Hyperman said...

I would make it an electoral campaign rule of not having the right to use the other candidate image and logo in your own communication material.

The Hillary supporters would react the same way if they saw an Obama ad on YouTube where Hillary looks like a 70 years old grandmother, even if it's because of an encoding or codec issues (which I believe is really the case here).

I think the source of this sensitivity is that other Hillary's campaign tactics have a similar racial smear "flavour". Among some examples that pop into my mind: the "Turbanobama" picture leak or the "I don't have any information that would let me believe he is Muslim" when asked about the viral email "accusing" Obama of being a hidden Muslim (instead of replying "He is christian like me").

The "I don't have any information that would let me believe he is Muslim" is what I consider to be the real subliminal message
here, sounds like NLP.

Anonymous said...

Hey-- don't bust on NLP-- it's been helping boring guys get laid for 30 years now... it would be great to have an all-NLP debate! That would be fun!! And while we're at it, let Hillary use all the Sambo pics she can gather and let Obama find all those creepy-laugh clips, and all the goofy bad-side-of-Hillary pics, her Medvee, mudvee, medeve-- whatevers, and really sling out the mud (heck maybe Obama won't have any information that would let him believe in Mena/Parking Meters/blow (not the Monica kind)....

Anonymous said...

Cannon sez:

"Get ready for a shock, conspiracy fans: Subliminals are impossible on television"

Incorrect. They are indeed possible. You are forgetting about the audio signal also. Audio can carry subliminal signals quite nicely.

Roughly one minute of searching uncovers the following patents, which deal with the tranmission of subliminal signals, video and audio, via real time television:

(WO/1990/012470) SUBLIMINAL AUDIO AND VIDEO PROCESSING FOR REAL TIME TELEVISION

http://www.wipo.int/pctdb/en/wo.jsp?IA=WO1990012470&DISPLAY=DESC

System for implementing the synchronized superimposition of subliminal signals.

Also these US Patents:

4230990 October, 1980 Lert, Jr. et al. 358/84 Broadcast program identification method and system

4616261 October, 1986 Crawford et al. 358/142 Method and apparatus for generating subliminal visual messages

4777529 October, 1988 Schultz et al. 358/143 Auditory subliminal programming system

4807031 February, 1989 Broughton et al. 358/146 Interactive video method and apparatus

http://www.freepatentsonline.com/5128765.html

Anonymous said...

Speaking of perception, it looks like Hillary's gotten an assist to consolidate her complete domination of the Moron Contingent on the Left:
http://bp2.blogger.com/_sVfYYHr3cdo/R9Qx64q5ATI/AAAAAAAAAgU/T--TQXiaCv4/s1600-h/globeenquirer2.jpg

Joseph Cannon said...

Anon: genuinely fascinating. I had restricted my points to the whole notion of visual subliminals, of course.

That said, you can find patents for all sorts of nonsense. What we need are studies -- published in peer-reviewed journals, using control groups and proper scientific methodology, and perfectly replicable -- that this shit actually works.

Hyperman said...

I fully agree that patents are not a garantee of something that work !!!

From what I studied about it in my marketing major, it doesn't work on TV or in print (like hidding naked ladies in the ice cubes of a liquor ad picture).

But I believe NLP works in politics / propaganda, a careful wording and juxtaposition of words can have powerfull effect. Example: 9/11 and Saddam.

Anonymous said...

As I recall, one reason interest in subliminal messaging has died off in the last couple of decades is that subliminals can't make somebody do or feel something he doesn't already want to, and even then they're only modestly successful.

I mean, I can't tell you how many political commercials I've seen where a candidate was shown in black and white and his speech was inaudible, but I don't think Washington is full of silent-movie stars.

Anonymous said...

sofia, although the hubel/wiesel experiments exposed a fundamental set of visual cortex development in those kittens, one that generalizes to (probably) all species' visual systems, the notion that kids growing up watching a lot of TV will have 'different' cortical systems than those of us who didn't is actually worth entertaining (so to speak).

but i would posit that those kids are actually losing many connections that we had and have, not gaining something we don't.

the very act of watching TV narrows the visual and auditory fields, restricting perceptual scope for locating and identifying not just objects and actions themselves, but their 'meaning' (import, significance, relevance) for the moment when they emerge in experience.

i think what this means is it makes it harder for kids to grow up and feel comfortable in the REAL real world without straight lines and sharp angles and predictable outcomes, i.e., ya know, nature.

as for the 'magical seven (plus or minus two)' datum, aitchd, miller observed that across a large variety of data, not just numbers. it's a capacity thing, sort of like bandwidth. and area codes didn't change that factoid; 10 digit numbers are still harder to memorize. what people do is remember the area codes for locations, and they know where their friends live, and most often they live in the same area. this is just the step beyond the old system that used neighborhood names for the first three digits, and folks remembered them that way.

no new pattern recognition going on there; just more and more reliance on numbers as identifiers compared to using names 'back then'. we're still using the same old pattern recognition apparatus we've always had for many millennia, in fact the same one the kittens exposed in those experiments.

AitchD said...

You mean when you see on TV a steaming, greasy, cheese-stretchy pizza slice being lifted, a large number of viewers don't want something to eat even if they're not really hungry? Or the reverse question: if 99.99% of the viewers don't buy the Lexus, it proves the Lexus ads don't work?

The way TV works on its primitive or fundamental level, especially TV advertising, is by putting images into your mind that you don't choose. Who puts those images there probably doesn't have your best interests at heart. Good luck getting rid of them.

You have to fix your gaze at a constant distance to watch TV (and a movie). For some people, this induces a trance-like state if their brain can't tell if its body/mind is 'conscious' or 'unconscious'. I've heard that people who are in a trance are susceptible to what I've heard called 'post-hypnotic suggestion'.

The issues about 'embeds' of pornographic images, or any kind of image, are distracting and irrelevant.

AitchD said...

What a pleasant surprise, dr! Will you follow me down?

"sofia, although the hubel/wiesel experiments exposed a fundamental set of visual cortex development in those kittens, one that generalizes to (probably) all species' visual systems, the notion that kids growing up watching a lot of TV will have 'different' cortical systems than those of us who didn't is actually worth entertaining (so to speak)."

That wasn't what I've wondered about ever since I read about the kittens. I wondered about the inherent nature of 'video' information on (a) TV and how we perceive it, thus: standard definition (CRT) TV beams its light through some 3 million phosphors while refreshing the info 30 times a second, while at any instant the eye can only grasp an area comprising some 70 (seventy) phosphors (or bits of information). We know how the eye 'scans' the whole territory very rapidly to make visual sense out of the 3 million bits. Further, since a TV beams out light, we have to fix our gaze about one inch in front of the screen to bring it into focus - if we focus on the screen we see the phosphors or dots. (Whatever we make of entire generations who have been gazing at nothing but light for hours on end is another topic!) So, watching TV requires that we process 3 million bits of information per second, doesn't it? I don't know if the human brain has done anything like that before TV. What I wonder is if infants develop specialized (?) neurons when they begin to watch TV, and (gulp) if since the 1950s, there's a new kind of human brain capacity.

"but i would posit that those kids are actually losing many connections that we had and have, not gaining something we don't.

the very act of watching TV narrows the visual and auditory fields, restricting perceptual scope for locating and identifying not just objects and actions themselves, but their 'meaning' (import, significance, relevance) for the moment when they emerge in experience."

I read that before TV, boys/men's rate of dyslexia was like 85%-90% greater than girls/women's rate, but since TV it's about the same.

I'd say there's three distinct kinds of TV that kids watch: the kind that 'brings' stuff into your living room, the kind you program with DVDs or 'On-Demand' and the like, and the kind you aren't taking into account up there, which extends your eyes and ears when the camera is at the ball park or on the Moon. That last kind, which extends our central nervous system, letting it be in two places at the same time, was a new thing on Earth. A book you wouldn't want to fall on your toes could be written about how our sense of 'live' TV has evolved.

"i think what this means is it makes it harder for kids to grow up and feel comfortable in the REAL real world without straight lines and sharp angles and predictable outcomes, i.e., ya know, nature."

The straight lines and angles also exist in the external world, they've been around longer than history, and they're uniquely and exclusively human, which I suppose infants come to realize long before they're verbal, i.e., wherever there are straight lines and sharp angles, there's human fingerprints. And conversely, as you say.

"as for the 'magical seven (plus or minus two)' datum, aitchd, miller observed that across a large variety of data, not just numbers. it's a capacity thing, sort of like bandwidth. and area codes didn't change that factoid; 10 digit numbers are still harder to memorize. what people do is remember the area codes for locations, and they know where their friends live, and most often they live in the same area. this is just the step beyond the old system that used neighborhood names for the first three digits, and folks remembered them that way.

no new pattern recognition going on there; just more and more reliance on numbers as identifiers compared to using names 'back then'. we're still using the same old pattern recognition apparatus we've always had for many millennia, in fact the same one the kittens exposed in those experiments."

I've been improving owing to remembering confirmation numbers when I pay bills online After logging off); I don't print the message (Socrates predicted that writing will produce forgetfulness). I mentioned area codes, though I had this article in mind:

http://www.newyorker.com/reporting/2008/03/03/080303fa_fact_holt

"... the average Chinese four-year-old can count up to forty, whereas American children of the same age struggle to get to fifteen. And the advantages extend to adults. Because Chinese number words are so brief—they take less than a quarter of a second to say, on average, compared with a third of a second for English—the average Chinese speaker has a memory span of nine digits, versus seven digits for English speakers. (Speakers of the marvellously efficient Cantonese dialect, common in Hong Kong, can juggle ten digits in active memory.)"