Vertical speed when jumping

Started by Dower
24/05/2010 20:43
Page 1 of 1
Dower24/05/2010 20:43

So I saw in a topic posted a long time ago that when you've not recently jumped, your vertical speed at jump off is 268.25 units/second. My question is: What is your vertical speed when jumping X seconds (or frames) after your last +jump.

Example:
I +jump (my vertical speed is 268.25 at jump off), then wait 1 second and +jump again. What is my vertical speed at jump off the 2nd time?

Help appreciated :)

TlofT24/05/2010 21:20

jump, run slow, jump 1s later or jump 1s after the first jump (basically a bhop)?

get some ljstats, make min distance 0 and then do some physics

Dower24/05/2010 21:27(edited 24/05/2010 21:28)

Originally posted by TlofT
jump, run slow, jump 1s later or jump 1s after the first jump (basically a bhop)?

get some ljstats, make min distance 0 and then do some physics


That's a good tip, i'll try this when i have some spare time and post my results. Still hoping for someone who knows the engine to post an exact formula.

Edit: To be able to do that test I'd have to know the size of the cs model, since ljstats measures from the back of it and jump off, and at the front of it at landing.

ova24/05/2010 21:39(edited 24/05/2010 21:39)

Originally posted by Dower
Originally posted by TlofT
jump, run slow, jump 1s later or jump 1s after the first jump (basically a bhop)?

get some ljstats, make min distance 0 and then do some physics


That's a good tip, i'll try this when i have some spare time and post my results. Still hoping for someone who knows the engine to post an exact formula.

Edit: To be able to do that test I'd have to know the size of the cs model, since ljstats measures from the back of it and jump off, and at the front of it at landing.


The answer u are seeking is not 2.

Pulsar!25/05/2010 03:54

You can calculate vertical speed in CS with the same formula as you calculate it in real world :) With a=800 u/s*s.

V=a*t, t - time.
Time can be calculated as:

t=sqrt(2*H/g), H - height.

Dower25/05/2010 13:27(edited 25/05/2010 13:27)

Originally posted by Pulsar!
You can calculate vertical speed in CS with the same formula as you calculate it in real world :) With a=800 u/s*s.

V=a*t, t - time.
Time can be calculated as:

t=sqrt(2*H/g), H - height.


Yes, I know this and I have the "normal" vertical speed which is 268.25 units/second

But what I'd like to know is what this speed is when you've recently jumped. We all know you jump slower and less high if you bunnyhop or just jump twice in a row.

SUBWAYZ25/05/2010 14:06

Only the height will decrease while bunnyhoping. First jumps is 63.xx high. Following bhops height is something from 52.xx units to 32.xx units, depending on if you do standups or not. Also the standups timing changes height. Reason to decreasing height is lowering frame amount between jumps.

correct me if i'm wrong

DiDo-Grecord holder25/05/2010 14:37

geek time

Lt.RAT25/05/2010 18:22

Originally posted by Dower
Example:
I +jump (my vertical speed is 268.25 at jump off), then wait 1 second and +jump again. What is my vertical speed at jump off the 2nd time?

Your second vertical jump would be the same as first.

But, when u press + jump, your Vertical velocity is sqrt( 2 * 800 * 45.0 ) = 268,328157. And jump height is 45.0 (as in formula). There exists variable called "fuser2", in old cs version (p47 and earlier) fuser2 at jump off equals to 1500, in new version it`s [1000*(2*sqrt(2 * 800 * 45.0)/(800*1.04) + 2*sqrt(2 * 800 * 45.0)/800) ] = 1315.7894 (inside engine its calculated with less precision)

Each frame after jump (0.01 sec for 100fps) fuser2 decreases: [fuser2]

When we on ground (walk code) our horizontal velocity decreases (if fuser2 > 0) : [velocity.xy] after it all calculations of ground friction and speed gain from strafes.

When we try to bunnyhop (there is no walk code) our vertical velocity sets to 268,328157 and decreases (if fuser2 > 0): [velocity.z]

Really our second jump (etc) vertical speed depends on elapsed time from previous jump (airtime). Coz non-constant time sampling its hard to say exact value. But it could be calculated for some airtime variation.

If we in air more than 1315.7894 msec our fuser2 is zero, and we havent velocity penalty. Also coz of change 1500->1315.+ all old demos not comparable to new ones.

PS yes, it`s 268,328+ and not 268.25. eDark took 268.25 value from demo, and in demo we have values transmitted from server to client with loose of precision (though it listen server demo). Also it`s easy to proof is someone needs comprehensive explanation.

Dower25/05/2010 18:39

Originally posted by Lt.RAT
Originally posted by Dower
Example:
I +jump (my vertical speed is 268.25 at jump off), then wait 1 second and +jump again. What is my vertical speed at jump off the 2nd time?

Your second vertical jump would be the same as first.

But, when u press + jump, your Vertical velocity is sqrt( 2 * 800 * 45.0 ) = 268,328157. And jump height is 45.0 (as in formula). There exists variable called "fuser2", in old cs version (p47 and earlier) fuser2 at jump off equals to 1500, in new version it`s [1000*(2*sqrt(2 * 800 * 45.0)/(800*1.04) + 2*sqrt(2 * 800 * 45.0)/800) ] = 1315.7894 (inside engine its calculated with less precision)

Each frame after jump (0.01 sec for 100fps) fuser2 decreases: [fuser2]

When we on ground (walk code) our horizontal velocity decreases (if fuser2 > 0) : [velocity.xy] after it all calculations of ground friction and speed gain from strafes.

When we try to bunnyhop (there is no walk code) our vertical velocity sets to 268,328157 and decreases (if fuser2 > 0): [velocity.z]

Really our second jump (etc) vertical speed depends on elapsed time from previous jump (airtime). Coz non-constant time sampling its hard to say exact value. But it could be calculated for some airtime variation.

If we in air more than 1315.7894 msec our fuser2 is zero, and we havent velocity penalty. Also coz of change 1500->1315.+ all old demos not comparable to new ones.

PS yes, it`s 268,328+ and not 268.25. eDark took 268.25 value from demo, and in demo we have values transmitted from server to client with loose of precision (though it listen server demo). Also it`s easy to proof is someone needs comprehensive explanation.



Thanks a lot! Exactly what I was looking for :)

kREWAN25/05/2010 19:00

Originally posted by Lt.RAT
Originally posted by Dower
Example:
I +jump (my vertical speed is 268.25 at jump off), then wait 1 second and +jump again. What is my vertical speed at jump off the 2nd time?

Your second vertical jump would be the same as first.

But, when u press + jump, your Vertical velocity is sqrt( 2 * 800 * 45.0 ) = 268,328157. And jump height is 45.0 (as in formula). There exists variable called "fuser2", in old cs version (p47 and earlier) fuser2 at jump off equals to 1500, in new version it`s [1000*(2*sqrt(2 * 800 * 45.0)/(800*1.04) + 2*sqrt(2 * 800 * 45.0)/800) ] = 1315.7894 (inside engine its calculated with less precision)

Each frame after jump (0.01 sec for 100fps) fuser2 decreases: [fuser2]

When we on ground (walk code) our horizontal velocity decreases (if fuser2 > 0) : [velocity.xy] after it all calculations of ground friction and speed gain from strafes.

When we try to bunnyhop (there is no walk code) our vertical velocity sets to 268,328157 and decreases (if fuser2 > 0): [velocity.z]

Really our second jump (etc) vertical speed depends on elapsed time from previous jump (airtime). Coz non-constant time sampling its hard to say exact value. But it could be calculated for some airtime variation.

If we in air more than 1315.7894 msec our fuser2 is zero, and we havent velocity penalty. Also coz of change 1500->1315.+ all old demos not comparable to new ones.

PS yes, it`s 268,328+ and not 268.25. eDark took 268.25 value from demo, and in demo we have values transmitted from server to client with loose of precision (though it listen server demo). Also it`s easy to proof is someone needs comprehensive explanation.

geek

PoshitO25/05/2010 20:51

so bassically 1605.328 - 268.328 equals 1337 ???

TlofT26/05/2010 03:35

interesting lt rat.

Grrr28/05/2010 12:21

hm, so that explains why we jump higher with standup bhop(or I'm wrong and theres smth more?) and why Ryderc couldn't finnish his dd/bhop combo on his avatar (missing the 256 block)

Lt.RAT28/05/2010 22:22(edited 28/05/2010 22:24)

Dunno about Ryderc, but with stand-up effect you are right. Also there exists technique "Drop bhop" which really like LJ/WJ after 1.3+ sec, or it can be called "no slow down bhop" as compromise between prevalent terminology and engine work. And standard "Drop Bhop" should be for jumps with airtime lesser than 1.3158 seconds.

Also some ppl try to do standup for WJ, but in most cases it`s useless. In some cases for WJ + CJ there is no "standup effect" but with duck after dd we can stay in air longer => more time for prestrafe => more pre speed.

And yeah, if we talk about bhops, we need to mention another "no-slow down bhop", after flight from ladder. Because "fuser2" sets to standard value, only after +jump, after flight off from ladder fuser2 equals to zero, and there is no slow down on ground, and for following jump.

nucLeaR28/05/2010 23:41

So a Drop Ladder Stand-up bhop can break the maximum human distance with normal settings ?

Also you said that dd+duck (the "russian walk") keeps you more time in air than a normal CJ (that's the explanation for aa 100 fast CJ faster speed?)

Lt.RAT29/05/2010 01:15

Originally posted by nucLeaR
So a Drop Ladder Stand-up bhop can break the maximum human distance with normal settings ?

No, such jumps would be like WJ for engine physics. And maximal distance like in WJ.

Originally posted by nucLeaR
Also you said that dd+duck (the "russian walk") keeps you more time in air than a normal CJ (that's the explanation for aa 100 fast CJ faster speed?)

Can`t get your logic between 100aa and airtime :)

More airtime in cj prestrafe -> more time to strafe and gain speed (in duck we gain 1/3 speed from normal strafes)

Really max gain speed per frame in 10aa and in 100aa - equal. But with 100aa much easier to do strafes more close to Ideal strafes.

Hard to explain on words, better on graph, but i`ll try :D
With 10aa we move mouse:
Slow - we gain speed, but not too much.
When we try to move our mouse more faster, there is a time when our gain speed per frame starts decrease. Faster we mooving our mouse less gain. And between slow and fast phase there is ideal mouse movement for speed gain.

For 100aa, Slow and Ideal phases as in 10aa, but when we try to move our mouse faster we still gain enough speed, and much more than for 10aa. Less decrease in gain for 100aa, more mouse movement range for ~Ideal strafes.

Grrr29/05/2010 13:28

Originally posted by Lt.RAT
Dunno about Ryderc, but with stand-up effect you are right. Also there exists technique "Drop bhop" which really like LJ/WJ after 1.3+ sec, or it can be called "no slow down bhop" as compromise between prevalent terminology and engine work. And standard "Drop Bhop" should be for jumps with airtime lesser than 1.3158 seconds.

Also some ppl try to do standup for WJ, but in most cases it`s useless. In some cases for WJ + CJ there is no "standup effect" but with duck after dd we can stay in air longer => more time for prestrafe => more pre speed.

And yeah, if we talk about bhops, we need to mention another "no-slow down bhop", after flight from ladder. Because "fuser2" sets to standard value, only after +jump, after flight off from ladder fuser2 equals to zero, and there is no slow down on ground, and for following jump.


Yeah, I noticed that.. but I didn't knew that the height varies by every 0.xxx units (from the fuser2), I thought that its drops constantly (like 13 units), my all knowledge is based on this plugin http://xtreme-jumps.eu/e107_plugins/forum/forum_viewtopic.php?119520.post :P It helped me to check out where we got standup effect and I was rofl-ing on the ppl who invent those new "techniques" like u said standup-WLJ, standup-CJ. Its true that we got more airtime using CJ with full duck, but its seems more useful on non-legal settings (like 100aa where u can strafe ducked succesfully, and u gain lot easier speed), anyway standup-CJ is useful about doing some trickjumps (like that on fof's video, and 1 tj which I scripted) but as for getting better pre, its almost imposs to get over 290 prestrafe with it on legal settings, while u can get over 300 on 100aa. I recommend for pepole to use that plugin.. so the questions and the "new technique threads" to be reduced.. and some1 should upgrade it :)

Originally posted by nucLeaR
Also you said that dd+duck (the "russian walk") keeps you more time in air than a normal CJ (that's the explanation for aa 100 fast CJ faster speed?)


yeah, u got more airtime after the unduck, so u get bigger speed for your prestrafe (but since u gain speed harder while ducked on 10aa its not a big plus there) :P

Page 1 of 1

Please log in to post replies.