How to Lie With Incompetent PhotoShopping

Via Matt Yglesias, the following pair of pictures purports to show that "an equal number of people can fit into a vastly smaller space if they're riding a bus than if they're in single passenger cars." See if you can spot the problem:

i-cc074262e520560343ea8dbc100912a5-busvcar.jpg

Answer below the fold:

The bus picture is on a vastly different scale than the car picture, greatly exaggerating the size of the effect.

I'm not denying the effect, mind-- just lamenting the deceptive photo editing used to make it look larger than it is. And the hell of it is, it's pointless stupid photo editing. The same point would be made by properly cropped and scaled photos. But whoever made it (Yglesias doesn't say where he got it) wasn't satisfied with the truth, and decided to jazz it up a little in a transparently obvious way.

This is what makes politics so depressing-- even people I basically agree with are so frickin' stupid...

(I will note that while I like public transit in principle, I'm not a big fan in practice, mostly because I'm significantly bigger than the average person. I'm basically never comfortable riding buses, particularly those with forward-facing seats. Trains tend to be a little better, but not much.)

Tags

More like this

(I think that's right-- we're drifting into phrasebook stuff, here...) Another day of tourism, another golf shirt with visible salt crystals from dried sweat. Did I mention that it's really frickin' hot? Today was a split day-- we spent the morning looking at temples near the Kyoto station, and…
Over at Dot Physics, Rhett is trying to learn his students' names: I have students sitting at tables (in this class and in labs). As they are working on something, I go around and write down who is sitting where. Yes, this means that you have to actually ask each student what their names is (I hate…
The old riddle goes: you were granted three wishes, and you have one wish left.  What do you wish for?  Everyone over the age of 6 knows: more wishes. When we figured out that there was oil under the ground, and figured out how to use it, it is as though we had been granted three wishes.  Now the…
As Kate and I set out to run errands the other day, the "Service Engine Soon" idiot light came on in my car (a 1999 Ford Taurus LX). This may or may not mean anything-- Kate got one of those in her Prius a while back and it was nothing-- but if it's actually an indicator of anything serious, I'll…

It is obviously true that the two photos are cropped on different scales, and maybe it makes enough difference to complain about. The place that I see this kind of thing done in a much worse way is when the media publish a bar chart and crop off everything but the very top so that small differences are greatly exaggerated.

It's not really photo editing or photoshopping though, is it? Everything is the right size, but the second picture is taken from further away/zoomed out. That's so obvious, just from the size of the people, that I have a hard time picturing it as a real attempt to deceive.

For me, despite (or because of) the change in scale, the groups of people do not look the same. From the distance that they stretch across the street, it looks like the group of people with the cars is larger. Maybe it isn't, but that's how it looks to me.

Yeah, it's hard to tell. tbh, I see this as a support for the 'cock-up not conspiracy' theory

I thought the problem was that some of the cars have been copied and pasted. Some of the cars near the front look pretty suspicious to me. If any of them have been cloned, though, the 'shopper has been competent enough to resize and smudge them slightly, so it's hard to tell for sure.

It's not really photo editing or photoshopping though, is it? Everything is the right size, but the second picture is taken from further away/zoomed out. That's so obvious, just from the size of the people, that I have a hard time picturing it as a real attempt to deceive.

I think they're taken from nearly the same place-- the angle of the street looks very slightly different, so the camera was probbaly moved side to side. The zoom is different, but that would've been trivial to fix with proper cropping and scaling. I could probably do it in GIMP, and nobody's going to mistake me for a photo editing master.

I don't really think this was a deliberate attempt to deceive. I think it's the result of incompetence and unscientific thinking-- whoever put that together doesn't know how to use their image editing software, and didn't recognize the importance of making a fair comparison. The two look superficially similar, and the desired effect is magnified by errors, so they ran with it.

Best I can tell, it was one of 3 things: 1) a Photoshop zoom in or 2) an in-camera zoom for the left-hand picture, or 3) an in-camera zoom out for right-hand.

Funny thing is, if they'd been kept at the same scale (i.e. both zoomed out), it would be even more impressive, as I suspect you'd see even more cars in the left-hand side.

By G Barnett (not verified) on 10 Jul 2008 #permalink

Uncle Al lived his first 17 years in Brooklyn, NY. YOU ride the IRT to its last stop in Flatbush. YOU ride the buses thereafter. Uncle Al got out.

If car sales reduced 30%/year public transportation would not need to publicly transport hairdressers, hotel workers, grocery clerks, and dishwashers. There would be no US economy, not even in politicians' lies.

Um, just to continue to play devil's advocate, perhaps the "amount of space" refers to the amount of space on the road they are taking up.

The masses of people are there to illustrate the number of bus passengers equals the number of individual drivers in the other photo. (So they should be taken to be about the same size). But the amount of space all those cars take up is incredible smaller than the space that the bus takes up.

Think about traffic jams and ponder how things might be moving more smoothly if half the cars were exchanged for a much smaller number of buses. Or, in other words, imagine you were tasked with driving your car though one of the streets pictured above.

I just got back from a trip to see the parents and used the King County bus system to get from Redmond to the airport both ways. It's fairly efficient, around 2 hours with one transfer between, but saves an amazing amount of money, and reduces some grief if a friend drops me off.

This is the first time I've done this and saw more than just one or two other passengers at the airport using the bus. Now it's becoming much more common.

By Carl Brannen (not verified) on 10 Jul 2008 #permalink

Um, just to continue to play devil's advocate, perhaps the "amount of space" refers to the amount of space on the road they are taking up.

That's exactly the point that they're making, and the effect is seriously exaggerated by the incompetent photo editing.

The bus in the right-hand photo appears to be a bit less that two of the car lengths in the left-hand picture, which makes it seem really impressive. Unrealistically so, as you know if you've ever seen a city bus in traffic.

A person in the right-hand picture appears to be just about half the size of a person in the left-hand picture, meaning that the bus in the right-hand picture is more like 3-4 car lengths from the left-hand picture. And assuming the scaling error is symmetric, that means that the bus looks a factor of 4 smaller (it's both half as long and half as wide as it should be) than it would if the scaling were done correctly.

That doesn't begin to compare to the legitimate space reduction that they're trying to demonstrate, but that's my point. By screwing up the scaling, whoever put that figure together has created something that's obviously deceptive when it didn't need to be deceptive.

I'm all for better public transit, and I'm glad that using the buses worked for getting you from Redmond to the airport, but I wouldn't call two hours "fairly efficient." That's a car trip that takes 30 minutes with no traffic, 40-45 minutes with typical traffic, and an hour at the most hideous Friday afternoon commute traffic.

So, "how to lie with incompetent photo cropping" rather? I spent a fem minutes staring at the photos without seeing any obvious 'shopping done, but figured it had to be something. Hmm, I wonder if he means that they're cropped differently? Nah, too obvious...

So stop "lying with incompetent post titling". :-)

It's a foto from a campaign of the Stadtwerke Münster (.de) from the early 90s promoting public transport.Photoshop might have been around then but I doubt it was used here to achieve the effect. Anyway it's an advert, it's supposed to be manipulative.

Münster has its own philosophy of indvidual transport and I liked it when I lived there. Visitors from abroad are frequently amazed by the huge abundancy of bicycles in the city.

This has little to do with using Photoshop. It is, rather, an example of framing. Seriously! I think it was mostly the photographer's fault and nobody tried to fix that error with Photoshop.

I'd say the entire exercise was poorly thought out and executed. Notice that some people are still walking from their cars in the left picture? And that the bus is next to instead of behind the group? They didn't plan to have the two situations compared, and the person taking the picture didn't even bother to hold the same zoom setting (which you can read off of the image metadata on any decent camera today) to keep the framing the same. To the contrary, the framing was chosen to make one seem more crowded than the other, probably at the time the picture was taken.

And it is mostly framing (or possibly cropping, which is the same thing). I checked that the aspect ratio of the area on the left matched the aspect ratio of the corresponding area on the right. (You can use cues on the roof at top left and on the wall to the right to set the width and top edge, and it all fits.) It hasn't been squeezed or stretched to mess with the perspective, as Chad noted.

And Photoshop does tell me that the pictures were not manipulated. I see no evidence of cars being cloned, for example, although the image quality is so poor that only truly incompetent work (such as someone detected in some scientific papers) would be spotted from this picture. It even tells me that, since the lighting is so different, that whoever did this didn't even try hard to correct for afternoon shadows in the second picture or the washout in the first one.

Oh, I missed Chad's second comment about scale. I estimated the scale change at about 1.73 or 1.74. If I rescale the bus up by that amount and put it in the left picture, it covers up about 6 cars in the right lane (some because of its height, but an ~80 passenger bus is pretty long). It takes up more than 1/3 of the vertical part of the picture.

I just posted it on my blog.
http://doctorpion.blogspot.com/2008/07/bus-v-car-rescaled.html