Optical Illusion in Racing Photos, Thoughts?

[Lunajammer]

There is an optical illusion in racing stills that's perplexed me for years... decades, where there appears to be a false perspective.

The car's far side appears wider than the near side, where you'd expect the opposite to be true. Like you can almost see the grill and the tail lights at the same time, as if the perspective is curving outward.

I did some perspective measuring in Photoshop which told me that I'm wrong on some pics and inconclusive on others. Any thoughts from photographers, psychologists or nobel laureates?

[BlackSheep214]

I don’t see any optical illusions you speak of. Perhaps the banks of the tracks is what got perplexed?

[Erik Smith]

I would guess compression from zoom lens and the bank of the track - although I have no way of proving anything or even why either of those would cause the effect. Maybe something to do with perspective. But, yeah, I'm no help.

[Lunajammer]

39 minutes ago, Erik Smith said:

I would guess compression from zoom lens and the bank of the track

I'm certain the zoom contributes to the effect, but bank of the track should just show you more of the 3/4 top angle and not alter perspective much. But at least I think you see what I see (more obvious is some pics than others).

[peteski]

I think it is all in your mind Mike.  Human eyes (well actually human  brain as the brain is actually interpreting the visual input form our eyes) can easily be tricked.  I believe that is the culprit here.

Just like

Though hard to believe, but all these long lines are parallel to each other. They look crooked because this illusion is tricking you in your perception of angles. Since the shorter lines keep changing their direction, our brain believes that the long lines too are uneven.

Edited March 4, 2022 by peteski

[Erik Smith]

59 minutes ago, Lunajammer said:

I'm certain the zoom contributes to the effect, but bank of the track should just show you more of the 3/4 top angle and not alter perspective much. But at least I think you see what I see (more obvious is some pics than others).

I do. Perspective should show the back of the car shorter than the side closest to the camera. 
 

Okay second theory - the camera man is actually moving the camera to blur the background whilst keeping the car in focus - thus, the actual exposure is catching more of the car as it moves across an arch, which the farthest away portion of the car needs to move faster to keep up - like the spokes of a wheel, the farther out from the hub, the faster they spin. This causes the rear, farther corner of the car to appear behind the closer rear corner, but, as the exposure continues, the farther, front corner appears to “pass” the closer front corner. If that makes sense. Which it may or may not, seeing as how I’m two hurricanes into the evening. Happy Mardi Gras!!

Edited March 4, 2022 by Erik Smith

[Erik Smith]

1 minute ago, Erik Smith said:

 

Edited March 4, 2022 by Erik Smith

[Lunajammer]

6 minutes ago, Erik Smith said:

Okay second theory

In summary I believe you're saying motion is contributing to the effect. I think so too. I just wonder how the illusion of distortion can happen without any apparent motion blur, which I'd think you can't have one without the other. (Bottoms up ?).

[Ace-Garageguy]

Having been a semi-pro photog for a brief time, and an amateur for decades, I know a little bit about this.

It's a particular type of distortion introduced by specifics of lens design.

For way more info than you probably want, read these:

https://www.imaginated.com/photography/photography-glossary/what-is-lens-distortion/

https://photographylife.com/what-is-distortion

Edited March 4, 2022 by Ace-Garageguy
ADDITIONAL INFORMATION

[Erik Smith]

17 minutes ago, Ace-Garageguy said:

Having been a semi-pro photog for a brief time, and an amateur for decades, I know a little bit about this.

It's a particular type of distortion introduced by specifics of lens design.

For way more info than you probably want, read these:

https://www.imaginated.com/photography/photography-glossary/what-is-lens-distortion/

https://photographylife.com/what-is-distortion

Interesting stuff. 

Assuming it’s pincushion distortion? - the photographers are using big zoom lenses. 

Edited March 4, 2022 by Erik Smith

[robdebie]

Two comments:

First: I think it cannot be a speed effect, since the photographer is panning his camera with the same 'speed' (angular velocity) as the car. For the camera, the car is standing still. The track is moving backwards though!

Second: in the 3D CAD software that I used in the past, you had several options how to display the object one was drawing. One choice was perspective or no perspective. If you work for a time with perspective 'on', and then switch to 'off', the object looks really akward, like the back side is bigger than the front side. I assume that applies to the racing shots too: we are so used to seeing cars with perspective, that a photo with barely any perspective looks strange. I'm guessing the photographers used very long telelenses, in the order of 400 mm or so. With such long lenses, you get very little perspective deformation. Concluding: it's the lack of perspective that plays tricks on you.

An unrelated example: if you know the Porsche 917K, it had a pretty wide and high rear deck. But since we hold our models at arms length or less, you don't see how big it actually is, in relation to the cockpit section. Only when you are confronted with race track photos, made from the front with (again) a very long lens, you see how ridiculously large it is. Like on this book cover. It's another example of the effect of lack of perspective.

I photographed the Heller model in almost the same attitude and with the longest lens I have, and it gives a similar effect.

Rob

Edited March 4, 2022 by robdebie

[Erik Smith]

In the photo below (taken from Mike’s example above), the red lines are parallel and at approximately the angle of the rear of the car. The blue lines are parallel and approximately the angle of the front of the car. If I were to continue to draw the lines from the front and back, they would eventually cross - thus there is perspective - it just doesn’t appear to be there without reference lines. So, optical illusion from the combined effects of compression, lens distortion, and most importantly, the human brain. 
 

In the bottom photo, all the cars are the same size. Darn brain. 

[robdebie]

Inspired by Erik's post, I did another analysis, to see whether there was a proper vanishing point. And indeed there was one, see right side of the picture below. I again conclude that the (relative) lack of perspective creates the illusion that the far side is larger than the near side.

Rob

[Lunajammer]

I love these answers because they are conclusive and confirm my findings as well but my brain keeps thinking I'm missing something. I did perspective lines too but I did not extend them to the vanishing point so they simply looked parallel. But that still confirms that they are at least not branching outward to the horizon.

2 hours ago, robdebie said:

Second: in the 3D CAD software that I used in the past, you had several options how to display the object one was drawing. One choice was perspective or no perspective.....   Concluding: it's the lack of perspective that plays tricks on you.

Excellent thought. In my Photoshop schooling we also did such an exercise and without perspective everything looks wonky.

[Lunajammer]

1 hour ago, Erik Smith said:

I was going to call you out on this one Erik because the cars are clearly not the same size and that the size difference was exaggerated, but guess what? They are indeed all the same size. My head's starting to hurt.

[DRIPTROIT 71]

And here I thought that the cars were actually shaped that way so that they could turn left better.?

[Oldcarfan27]

I think it may be action distortion from a single frame capture. Kind of like old time racing photography where the wheels looked oval, but clearly the car couldn't be driving on oval wheels, now could it? But it sure adds to the sense of motion!

 

[magicmustang]

Then there's Ed Roth's Orbitron. That's how he built it. Look at the slanted rear. 

Edited March 6, 2022 by magicmustang

[Ace-Garageguy]

7 hours ago, Oldcarfan27 said:

I think it may be action distortion from a single frame capture. Kind of like old time racing photography where the wheels looked oval, but clearly the car couldn't be driving on oval wheels, now could it? But it sure adds to the sense of motion!

Thing is, the SLR cameras the NASCAR images were undoubtedly made with function very differently from the much earlier cameras the slanted-at-speed images were made with.

Here's what makes the leaning "oval" tire effect (SOURCE: https://www.jalopyjournal.com/forum/threads/old-photos-why-do-cars-look-like-they-are-leaning-forward-at-speed.208476/ )

 

1)  "...you could do that by having a shutter which works like a window shade with a slit in it, that moves vertically as the photo is taken. Like a Speed Graphic camera, for example.

the bottom of the picture is taken first, then as the slit moves upward the rest of the picture is taken. So the bottom shows where the car was at the beginning of the shot, the top shows where it was at the end of the shot, and it moved forward during that time."

2)  "I'm thinking it's due to a vertical curtain shutter. Basically it's a horizontal slot in a plate that slides vertically past the lens to expose the film. If it moves from top to bottom, it'll 'scan' the car starting at the street and moving up the car as it drops. The latent image ends up inverted on the film in the camera. If the shutter travels slowly enough, the car's motion is registered a little further along as the light exposes the film."

3)  "Squirrel (first post) and ShortBus are exactly correct. I tried doing this with an old Speed Graphic press camera with a polaroid back, and it worked exactly as advertised. Put the camera on a tripod so the background is dead still, and shoot at the exact moment that the car is in the frame (yes, it takes some practice). Don't pan, as that will give you an upright car and a leaning background.

I used Polaroid B/W film and a very slow shutter speed with a small aperture (don't remember exact numbers, as this was about 30 years ago). "

[deuces wild]

Seems like the front fenders on the #12 Charger are "peaked up" a little more than the Petty Charger..... Or, I might be seeing things....??

[Lunajammer]

9 hours ago, Ace-Garageguy said:

the bottom of the picture is taken first, then as the slit moves upward the rest of the picture is taken. So the bottom shows where the car was at the beginning of the shot, the top shows where it was at the end of the shot, and it moved forward during that time."

And here I always thought it was one of those era specific tricks of the trade to exaggerate the excitement of speed. Very interesting.