My WAG. That because the individual burning pieces are perceived as “getting larger” as they move, in that they remain bright while leaving a diminishing tail. We then perceive them as coming closer.
If you look at fireworks with more intensely burning elements which leave longer trails, you can actually see the difference in size between those which come toward you and those going away. When the difference in size is apparent, the illusion disappears.
Does this have anything in common with the “spinning dancer” optical illusion, that your mind can switch directions back and forth?
wow
It’s related in some ways to the Spinning Dancer in that there’s ambiguity about the direction of motion. I think that TokyoBayer is on to something. If the size of each of the glowing pieces from the explosion tend to expand or swell as they burn, then the illusion could be generated that they’re getting closer. Things getting closer get bigger. Things getting bigger tend to look like they’re getting closer. Maybe all the pieces expand as they burn, and that could look like they’re* all *getting closer. Best theory so far.
I don’t think that’s it.
When I watch fireworks on youtube I don’t see much difference in size or brightness until pretty late between the ones moving towards vs away, I think the brain simply doesn’t know and assumes towards for all of them.
Here’s a video showing the drone’s perspective of flying thru a fireworks display.
I can see several different burst effects. Sometimes the burst spreads out in a pancake-style burst. Very flat when seen from ~above.
When the drone is heading straight into the bursts, the OP’s effect seems to generally hold up.
But when the drone is turning, the perspective is changing just enough that it’s clear that bursts are flying in all directions. So, it seems to be an effect of human perception.
BTW: I strongly condemn the dangerous and idiotic practice of flying drones into fireworks displays no matter how amazingly cool looking it is.
USCDiver, thanks for that video, it was stunning.
On a related subject: there are some fireworks effects that form specific shapes upon explosion - a heart, a smiley face, really basic stuff. It is in this case that I really do ask myself the following: a two-dimensional shape should only look “right” from a very narrow angle. However, I can only assume that the people standing somewhere else, turned 90°, are seeing the same shape. How is that possible, I ask you?
I’ve been observing closely, and it is apparent to me that the fireworks are angled to project slightly towards the crowd. The majority of the blast is directed towards towards the audience making the “illusion” mostly based on reality, barring a few “balls” launched in all directions.
One such directed blast can be viewed from the side at the 1:44 mark of the drone video (Fireworks filmed with a drone - YouTube)
At the 2:33 mark, the burning streaks are clearly all heading towards the camera, relative the background (Fireworks filmed with a drone - YouTube)
The smiley faces are similarly made by targeting a pattern towards the audience.
At the two minute mark this can be seen. (Fireworks filmed with a drone - YouTube)
I saw several “missaimed” smiles and other shapes at a live show today - those in particular do not look the same from all directions!
[quote=“USCDiver, post:3, topic:692384”]
Here ya go, this is what it looks like to be in the middle of the explosions
[/QUOTE]Thankyou, thankyou, thankyou, thankyou, thankyou 
Which crowd? The crowd can be all around the display and they can’t control the orientation of the shell when it explodes. Here is some info:.
http://www.paramountpyrofx.com/launching.asp
Some fireworks are not designed to be symmetrical, that’s one of them.
You can tell they are not all moving towards the drone by watching the change in speed and arc.
“When igniting patterned shells, pyrotechnicians usually ignite several at once as they cannot control the orientation that any one shell will have at the moment that it bursts. If the cardboard insert happens to be angled along the line of sight of the crowd, for example, they’ll only see a line of stars when it bursts open rather than seeing the stars’ arrangement head on. If several shells are used at a time, at least one of them is likely to be oriented correctly.”
I stand corrected on the fireworks at the 2:33 mark.
Here is a video from an airplane, which show the symmetrical “ball” fireworks.
There are many kinds of fireworks. Some shoot symmetrical balls. Some shoot patterns in a ring. Some shoot fans in one direction or two directions. Some create a fountain. Some have secondaries so the flaming balls change direction in flight.
For symmetrical balls, it is difficult to see that the ones flying away are flying away, they are interpreted as either being in the plane of view or flying towards you. Probably combination of the challenge of perspective for small objects and the lighting.
The fountain may be falling towards the audience, because if the fountain is aimed upwards at an angle toward the audience, it will fall over forwards.
Interesting to hear they shoot multiple patterns at one time so one may be facing the crowd. Another explanation, you have to be significantly off aligned from the perpendicular before the distortion becomes enough to make the pattern unintelligible. Looking at a heart from an angle will show a heart that has one lobe bigger than the other, for example.
This is the crux of my question. What, exactly, is the challenge? What does the lighting have to do with it? With a spherical explosion, why do all the parts appear to be coming toward the viewer, regardless of where he sits?
At that distance, your depth perception is practically nil. What’s more, is the effect is against a flawless night sky (usually), offering your brain no context for depth.
So, it’s easy for your eyes/brain to interpret any trails coming toward you when they’re really going away from you.
So then why don’t some or all explosions look as if they’re going away from you? This doesn’t seem to get at the core of the phenomenon.
Such as non-spherical ones, like rings or flares? Most likely because you’re getting some sense of parallax from those due to your viewing angle, which acts as a depth cue.
Also, if your brain is interpreting a spherical explosion like a hemispherical explosion where all trails are coming toward you, then it’s most likely because that’s just how you “saw” it as a kid and the illusion persists this way due to nothing more than that’s just how the illusion ingrained itself.
With some effort, you can probably visualize all the trails moving away from you.
When there are two possible interpretations of how something could be, sometimes your brain latches onto one. It is why some pictures of craters on the moon look like domes, for instance - the lighting cues trick your brain to seeing them as convex instead of concave.
Did you see that optical illusion of the spinning dancer above? So I looked at it, and ran through the entire sequence, and she was spinning clockwise. I tried but could not make the direction of perception change. It was only clockwise. I backed up, and then looked at her feet, specifically the reflection of her feet. Suddenly, the direction swapped, and she was spinning counterclockwise. Then staring at her image I couldn’t get it to go back clockwise.
The brain does weird things. The key for the fireworks display is the distance from you and the dark plain background gives you little reference for depth perception. The bright flares against black sky are small. Some of them get bigger as they burn, which is perceived as getting closer, even if they are receding. It is difficult to tell the difference between the front of the burning flare or the back of the burning flare, so you see the back and perceive it as the front, so instead of going away from you, it is coming toward you.
This seems to be the best explanation, akin to TokyoBayer’s theory. The now-famous drone in a fireworks display video shows a bunch of spherical explosions (starting around the 1:45 mark) that all seem to be coming out toward the camera. The best explanation is that it appears that way because the swelling of each point of light is interpreted as its getting closer to the viewer. If it’s spherical, a large number of them would be going in the opposite direction, and many would at least be going NOT directly toward the viewer. Yet, that’s how it all looks. I think that is the best theory.