Topics

Cycle speed and realism

V Bass
 

Over the years, I've seen a lot of automata and videos of automata and a thought has been slowly forming about what makes a piece interesting and engaging. And it comes down to speed.

To my eye, an automaton engages the viewer when the movement seems natural, and a lot of the appearance of naturalness depends upon timing. Imagine we're looking at an automaton with a couple of kids on a seesaw. The cycle time of this piece is 10 seconds - it takes five seconds for one end to rise, five more to go back down. Visualize that by moving your hands up and down and counting five seconds up, five down. 

Yawn. Have you ever seen a couple of kids go at that pace? Picture it in your mind and figure how long you think real pair of kids would take - 3 seconds for calm kids, 1 for hyperactive kids?

How can we control this? Thinking out loud here, you'd have to work from the final movement backwards, looking at cam profiles, lever ratios, gear train ratios, input (crank/motor) speed, adjusting all accordingly. It's a complicated chain of actions, but if the design doesn't take the end movement speed into account, it can end up looking like the whole mechanism has been soaking in syrup when it operates.

Oddly, I can't recall seeing an automaton that seemed unrealistic because the movements were exageratedly too fast.

What are your processes for getting realistic movement speeds in your pieces? Do you work backwards as above, forwards from the desired crank or motor speed, or set the two end speeds and make the middle bits fit?

Paul E Giles Jr
 

I'm a both ends towards the middle guy. I exclusively use cranks because I love to watch the grandkids work the device. Kids being kids I immediately reduce the shaft speed because their excitement means lots of rpm's. I also try to use a couple of elements and that means different speeds. And yes, I often slow down the speed only to increase it later for an object. 

 

This is an interesting question that at least to me involves some pretty complicated stuff.
First, let me say "To each h/her own".  
Here is how I do it for my own automata:
  • I know that I want both a hand crank and a motor.
  • I consider a comfortable crank speed to be approx 50-60 rpm.  Note that I am defining the crank speed by operator comfort.
    • Thus I use gear motors that are 50 rpm.
  • The speed of the automata action is very important.  Prior to creating my drive components I visualize how fast I think things should go.  I focus pretty carefully on this.  It's important.  
  • I believe that 'slower' is desirable more often than 'faster'. Not too slow is the trick.
  • I then create the drivetrain to meet the requirements of both ends.
    • The automata that I am creating right now needs three speeds:
      • 50 rpm for crank/motor input
      • 150 rpm for pulsing
        • I accomplish this by having three nodes on the 50 rpm shaft.
      • 1 rpm for driving the CamPak.
        • I accomplish this with a reduction train of pulleys and poly belting.  It's very smooth and quiet (and powerful).
Bottom line...the speed of things is very critical and important.  Wrong speeds make an automata not credible.  Correct speeds make it delightful and a pleasure to operate/watch.

I look forward to reading other comments.  Good question.

Cheers
-Jim Coffee-
one of the Automata Group Moderators
I create automata near San Diego, California



On Wed, 1 Jan 2020 at 14:21, V Bass <vrbass@...> wrote:
Over the years, I've seen a lot of automata and videos of automata and a thought has been slowly forming about what makes a piece interesting and engaging. And it comes down to speed.

To my eye, an automaton engages the viewer when the movement seems natural, and a lot of the appearance of naturalness depends upon timing. Imagine we're looking at an automaton with a couple of kids on a seesaw. The cycle time of this piece is 10 seconds - it takes five seconds for one end to rise, five more to go back down. Visualize that by moving your hands up and down and counting five seconds up, five down. 

Yawn. Have you ever seen a couple of kids go at that pace? Picture it in your mind and figure how long you think real pair of kids would take - 3 seconds for calm kids, 1 for hyperactive kids?

How can we control this? Thinking out loud here, you'd have to work from the final movement backwards, looking at cam profiles, lever ratios, gear train ratios, input (crank/motor) speed, adjusting all accordingly. It's a complicated chain of actions, but if the design doesn't take the end movement speed into account, it can end up looking like the whole mechanism has been soaking in syrup when it operates.

Oddly, I can't recall seeing an automaton that seemed unrealistic because the movements were exageratedly too fast.

What are your processes for getting realistic movement speeds in your pieces? Do you work backwards as above, forwards from the desired crank or motor speed, or set the two end speeds and make the middle bits fit?

Gus
 

It depends on many factors, doesn't it ?
How good is the element being manipulated, eg a head turn mechanics high tolerance, or sloppy ? Does it jerk because the mechanics aren't well machined/sanded/lubricated ? Unless that is what is wanted of course. Irrespective whether a fast or slow turn is wanted, the execution is crucial, otherwise it fails. Are multiple movements taking place, each one a different speed, as one would expect, eg a waving arm can imply something simply by how fast or slow it moves, and the same figure turning from the waist in conjunction might execute that quite fast, presenting a complete emotion. 
My approach is focused around the story I want to tell, without using words. I don't decide at the beginning how fast or slow it needs to be, that develops as the work progresses.  Often reworking cams, gears etc if it seems the story has no flow, fails to engage, or, as you put it, “Yawn”.
Good expression that ! 
The element of surprise is what I seek. Be that  fast or slow, matters not much to me.

Gus