What is the terminal velocity in Doom?

If you make an object fall in Doom (even vanilla) for a large enough amount of time, will it stop at a certain speed limit, or will its movement accelerate indefinitely? I tried this test map in vanilla Doom, and the view continued changing, until I actually DID hit the floor (due to hitting the exact spot I presume). The bug I was exploiting was how floors as low as -32768 cause the Z coordinate to wrap around when landing on the floor (even in GZDoom, sadly). Test WAD (vanilla Doom 2): http://www.mediafire.com/?5u0gj4otrwacb3b (WARNING: NOT FOR EPILEPTICS due to the increasingly fast floor-ceiling loop!)

So, is there a maximum falling speed in Doom?

Secondly, if I may ask so, how much is the g acceleration in Doom?

Terminal velocity occurs in real life because there's air resistance to balance the acceleration due to gravity. Doom doesn't have any air resistance (you only get friction when standing on the floor), so you'll just accelerate forever.

Not even a speed cap (to avoid glitches)? But okay, original and most of modern Doom doesn't need a falling speed cap because you almost never encounter such situations, not even in Eternity.

fraggle said:

Terminal velocity occurs in real life because there's air resistance to balance the acceleration due to gravity. Doom doesn't have any air resistance (you only get friction when standing on the floor), so you'll just accelerate forever.

Ya, but as you get closer and closer to the speed of light, you will start to get slower and slower

Nope, no vertical speed cap when falling, but there are some other caps in place when floating or moving up steps:

• Moving up steps is instantaneous, so you could consider this as "infinite" acceleration (z pos is simply modified to the destination's height)
  
• Floating up/down is done at a maximum rate of +/- 4 mu/tic (map units/tic). It's not even a proper acceleration, as floating is done at a constant speed and the z pos of a monster is simply altered directly.
  
• There is indeed a GRAVITY constant, defined to be 1 mu/tic per tic, or 1 mu/tic^2. Unlike floating, this is actually applied to momz (to the momentum/speed), so it causes a proper acceleration. However, it's kind of anisotropic: if you had zero vertical speed, you get a -2*GRAVITY acceleration, otherwise -GRAVITY.
  

If you want to translate mu/tics to "real" speeds, it will depend a lot on which interpretation of Doom's map unit values you buy into.

Commonly accepted/debated ones are:

• 10 map units are a foot.
  
• 16 map units are a foot.
  
• 8 map units are a foot.
  

Personally, I consider 16 horizontal map units to be a foot, and 10 vertical map units to be a foot (this is also kind of confirmed in heretic, where there is a FOOTHEIGHT constant defined to be 10 map units). 8 mu/foot IMHO doesn't fall neatly into either: it's a way too large scale for horizontal dimensions, and also for vertical ones. So for height, 10 mu/foot it is.

A tic is 1/35th of a second, so:

GRAVITY = 1 mu/tic^2 = 1225 mu/sec^2 = 122.5 ft/sec^2 = 40.425 m/sec^2 which is about 4 times stronger than normal gravity. For some reason it's hard to realize this in first person view, but it's easy to see how abnormally fast other objects do fall down (gibs, other players, dead monsters etc.)

@Mr. T (speed of light): that's not what happens: it simply gets harder and harder to accelerate towards it, in a way that is not explained by classical mechanics alone, but there are no speed reversals or slowdown effects simply because you're travelling near that particular speed.

See here.

Maes said:

GRAVITY = 1 mu/tic^2 = 1225 mu/sec^2 = 122.5 ft/sec^2 = 40.425 m/sec^2 which is about 4 times stronger than normal gravity. For some reason it's hard to realize this in first person view, but it's easy to see how abnormally fast other objects do fall down (gibs, other players, dead monsters etc.)

This is getting weirder and weirder. Not only that the default Doom value actually seems quite natural to the human, but ZDoom uses "800" instead of "1225" when talking about gravity (probably because of Quake).

Maes said:

• 10 map units are a foot.
  
• 16 map units are a foot.
  
• 8 map units are a foot.
  

I always assumed that 64 map units = 2 meter. That'd match best with Doomguy's size. And btw, Feet as a measurement have no place in any scientific discussions!

printz said:

This is getting weirder and weirder. Not only that the default Doom value actually seems quite natural to the human, but ZDoom uses "800" instead of "1225" when talking about gravity (probably because of Quake).

That "1225" number is a normalizing factor for converting squared tic times to squared seconds.

E.g. an acceleration of 1 [mu/tic^2] for 35 tics should give a final speed of v = at = 1*35= 35 [mu/tic], which is equal to 1225 [mu/sec].

If I want to compute this directly in seconds I will use a=1225 [mu/sec^2] and t=1 [sec]. Dunno in what context that 800 number appears in ZDoom or quake though.

Graf Zahl said:

And btw, Feet as a measurement have no place in any scientific discussions!

Tell that to the bulk of Americans and British that populate here. But I know what you're referring to by scientific. In any case, at least the foot has a precise size these days: 30.48cm.

EDIT: I tested a low-gravity map with gravity set to 800/1225 (assuming that 800 might be Quake's number of tics/second^2). I found the falling to be pretty acceptable. Has anyone jumped off a large height in real life to know for sure? :P

American scientists use the metric system all the time.

Well, not always.

The problem with imperial units is that they they have non-decimal multiples and subunits.

You have 12 inches that are 1 foot, 3 feet that are a yard and a furlong that is 220 yards, etc.

Of course, I supposed it's possible to use decimal prefixed with inches, too: e.g. kinches, minches, Minches, cinches etc. and simplify at least engineering-level calculations. I'm not sure which approach engineers are encouraged to use, but you can always do calculations in the most convenient unit and only convert to whatever other at the end, no matter how complex the relationship.

E.g. if I'm given a bunch of data in imperial units, even though I worked all my life with metric, I will prefer doing intermediate calculations in imperial, since the data will be dimensionally compatible and I will have less roundoff errors than converting all to metric beforehand, plus it's OK to write something such as "5.5 feet" in an intermediate calculation rather than "5 feet 6 inches". I will convert only the very final results to metric.

Feet and inches would make a lot of sense - if we used a duodecimal counting system instead of a decimal one. There are actually very good arguments for having one, though it's certainly unlikely to become popular any time soon.

Graf Zahl said:

I always assumed that 64 map units = 2 meter. That'd match best with Doomguy's size. And btw, Feet as a measurement have no place in any scientific discussions!

Height-wise that does actually equate to Maes' 10 map units = 1 foot theory, which I believe is also the one I accept.

Surely terminal velocity in Doom would be the point at which the floating point number for velocity wraps around to positive again?

What bugs me is that inches are already divided into 16 subunits when on the ruler, but they're the only imperial unit divided like this. 16 subunits would have been great compatibility with binary computing.

The old-fashioned measures were never unified. They were ad-hoc units corresponding to whatever felt practical and available at the time for this task. Each trade had its own units; each thing had its own definition, a bushel of apples and a bushel of wheat were neither the same volume nor the same mass.

Gradually, things like "twelve inches to a foot" were introduced to simplify things a bit, but the foot hadn't been created to be as much as twelve inches, and the inch hadn't been created to be one-twelfth of a foot. Furthermore, go to a different country, or even a different region, and such things were no longer necessarily true, as the local units called foot and inches had different lengths.

The metric system was designed from the top-down to be a unified and universal measurement system. A mass of apple would use the same unit as a mass of wheat, or a mass of water. Likewise for volume. Distance would use the same units whether you're measuring a person's size or a road's length.

I particularly like how a 1x1x1cm cube container will hold 1ml of water, which weighs ~1g at normal room temperature / pressure.

The one I tend to remember is that a cubic metre of water weighs a metric tonne. Can save a lot of embarrassment trying in vein to move/lift a moderate sized vessel full of water.

printz said:

What bugs me is that inches are already divided into 16 subunits when on the ruler, but they're the only imperial unit divided like this. 16 subunits would have been great compatibility with binary computing.

There are 16 ounces in a pound.

Yeah, if it weren't for the fact that we have 10 fingers and 10 toes, 8 or 16 would have been a very natural choice of base for a number system.

Grazza said:

There are 16 ounces in a pound.

Not if you're measuring powder charges for a muzzle loading cannon. ;-)

fraggle said:

Feet and inches would make a lot of sense - if we used a duodecimal counting system instead of a decimal one. There are actually very good arguments for having one, though it's certainly unlikely to become popular any time soon.

Most of the imperial units can trace their lineage back to Babylonian units, and Babylon used a base-60 number system. 8 and 12 are both factors of 60, and 16 is the square of 4, which also goes into 60. Base-60 was a very powerful system in an age where little in the way of advanced mathematics had been invented. The units don't link into each other nicely (like Fraggle's example) because they never needed to. Science didn't exist, engineering was extremely rudimentary, and nobody saw the need to integrate the various forms of measurement. Also, they arose out of local tradition in a very organic and not at all systematic process while the SI units were carefully and consciously designed for their purpose from scratch.

Base 60 is very cool because it allows you to divide by 3, in addition to whatever base 10 allows.

printz said:

Base 60 is very cool because it allows you to divide by 3, in addition to whatever base 10 allows.

Base 10 lets you divide by 2, 5, and 10. Base 60 lets you divide by 2, 3, 4, 5, 6, 10, 12, 15, 20, 30, and 60.

You'll never get US civilians to switch to metric. It will never happen. Ever. Science already uses SI, so that's all that matters. It's cute that people have tried and tried.

What was this thread about again?

As an American, I really wish we would switch over to metric. And everyone should use a 24-hour clock. I still can't comprehend when I meet people who claim they don't know how to do "army time" as if the army is the only entity that utilizes it, or as if it's any harder than a 1st grade math problem.

I'm pretty sure you meant that both 10 (not 8) and 12 are divisors of 60.

But I'm not posting to bash people for imperfect vocabulary. I think it's interesting that base 60 was likely created as a common base between 10 and 12. Base 12 actually was very popular, that's why there are words like 'dozen'. Both 10 and 12 originate from human hand. 12 is the highest number you can count to using your hand. Use your thumb, and count each part of your remaining fingers. You should have 4 fingers left 3 parts each, so 3 * 4 = 12.

10 divides by 1, 2, 5, and 10.
12 divides by 1, 2, 3, 4, 6 and 12. Traders without the knowledge of fractions might prefer base 12, because it's easier to divide stuff in equal amounts.