MatCap GLSL shader for Soft - easily doable?

[Dan Yargici]

I've been messing around with the new MatCap viewport shading in the latest testbuild of Blender and have extreme jealousy that it's not in Soft.  :)

The shader is really simple from what I understand as it's just looking up a texture based on the normals, nothing more.

It's a really nice way to view a mesh while modelling.

Is this something easily achievable in Softimage's realtime view I wonder?

DAN

[Dan Yargici]

You can see a painfully bad example of it in action here:

https://www.youtube.com/watch?v=rgPSQEAWsPE

Why he doesn't tick the smooth shading option I don't know...

[Gustavo Eggert Boehs]

Hi Dan,

The MentalRay equivelent of a MatCap would be a mirror ball (or a grayball for shooting FG rays), they are part of the production library so in SI you will see this shaders are named Mirror Ball (mip) and Gray Ball (mip). Since the latest release of SI (with the HQ viewport and all) nVidias's MetaSL library comes bundled in. In it you will find MetaSL versions (which, for the most part, can be displayed in realtime through HQV) of most MR shaders, including the mirror ball.

You will find the shader you want under MetaSL - Environment - Map_Ball. Unlike the MR version, you cant plug this shader directly to your surface, so you plug it in your enviroment and plug a completaly reflective material for your surface shader.

For some reason the enviroment texture was not being degamma-ed, so I have plug a colorcorrection in its output... anyways, the result is pretty close to the rendered versio, outside some pinching that I have no idea were it is coming from.

Best Regards
--
Gustavo E Boehs
http://www.gustavoeb.com.br/blog

[Dan Yargici]

That's a great find Gustavo - thanks!

DAN

[Ed Manning]

 http://andy.moonbase.net/archives/651

[Dan Yargici]

Also great, thanks!

[Andy Moorer]

:)

Although I wouldn't be surprised if that real time shader was too archaic to run properly in the 2013 viewport... If that's the case, it shouldn't be too difficult to make one even if you aren't a shader guy, using the newly-released and free mental mill 1.2 (why they just released an update to a supposedly deprecated tool I have no idea, but what the heck.)

[Andy Moorer]

I tried out Gustavio's clever approach and it works, but on my machine at least it doesn't appear to be behaving quite like a true MatCap, in that the results seem to be in object space as opposed to camera space (but I've got a weird graphics card so it might just be me?) 

Plus I didn't like having to use it as an environment map. So I went into Mental Mill and mucked about, only to realize that all the metaSL shaders mental mill is built on are now in softimage. 

In essence, mental mill is now (effectively) in the rendertree for all softimage users. 

Cool!

But I wanted a "fire and forget" solution for MatCap shading, so I hunt-and-pecked my way through writing a metaSL MatCap shader directly.

It seems to be working fine on my box, but I'm a pretty crap shader writer so I would appreciate any suggestions for improvement from any gurus out there.

You can find it here - http://andy.moonbase.net/archives/1037

Cheers - AM

[Dan Yargici]

Andy, that's great!  Thanks!

That's exactly what I had in mind!

DAN

[Steven Caron]

great job andy!

so how was the workflow of making the shader? did you use all softimage gui and the shader code editor? did you use an external editor and kept re instancing the node? or did you use mental mill text/graph editor and output .msl from there to use in softimage?

s

[Rob Chapman]

Im guessing that he got it working in Mental Mill2 and copy / pasted the GLSL codes over to the SI shader :)

did the code work straight away with no conversion?  nice one by the way Andy, wanted to try it but sadly only got as far as installing mental mill2 before being distracted by open beta of the MR Mila shaders instead! 

[Andy Moorer]

@Steven - I just used textpad and kept re-instancing the node, it was annoying. I fiddled with the shader code editor but this was such a simple little shader it wasn't really worth my time to mess with it. Outputting msl from MM was my first idea, but a)I never found a way to, it would write an xmsl which pointed to a .mill file and b) each of the MM nodes seemed to have a lot of junk in them I didn't need. 

@Rob - close, MM setup was a bit kludgy/inefficient, so I pulled up a few example scenes, looked thru the msl to get a feel for it and then hacked away at a couple of nodes until I understood what was going on. You'll see that this shader is basically the guts of generate normals, convert to camera space and texture 2 lookup nodes, shaken, not stirred.

There was bupkis for documentation on metaSL (and I'm a shameless hack), so there wasn't much I could see but to do it this way. But it worked. :D

Now that I've realized we have all the good stuff out of MM, it wouldn't be difficult to implement the exact same thing as 3-4 of the factory metaSL compounds strung together, but I wanted it all in a single easy shader... and I felt like doing some digging anyway. ^_^

MetaSL is the right idea, but I have a feeling it's not going to go anywhere and that in the long run all this stuff will be deprecated for something like allegorithmic substance files, or what have you.

[Steven Caron]

ah, ok cool.

actually its still a lot better with plugin shader/shader defs which parse the .msl for you.

[Andy Moorer]

I'll keep exploring. I'm a complete noob to this kind of stuff so I pretty much just waded in and started banging rocks together hoping to get fire. "Me wants matcap" level approach.

But now I've gotten my feet wet, and it was fun, so I might wander down the path of doing some glsl and see how I like it.

[Gustavo Eggert Boehs]

This is great stuff Andy, very inspiring!

 

“Now that I've realized we have all the good stuff out of MM, it wouldn't be difficult to implement the exact same thing as 3-4 of the factory metaSL compounds strung together, but I wanted it all in a single easy shader... and I felt like doing some digging anyway. ^_^”

 

Me to, so I tried to convert your code to a little shader network, attatched to this is what I came up with... Looks like it works alright 😊

 

Gustavo E Boehs
http://www.gustavoeb.com.br/blog

 

De: Andy Moorer <andym...@gmail.com>
Enviado: ‎12‎ de ‎fevereiro‎ de ‎2013 ‎20‎:‎08
Para: soft...@listproc.autodesk.com
Assunto: Re: MatCap GLSL shader for Soft - easily doable?

 

@Steven - I just used textpad and kept re-instancing the node, it was annoying. I fiddled with the shader code editor but this was such a simple little shader it wasn't really worth my time to mess with it. Outputting msl from MM was my first idea, but a)I never found a way to, it would write an xmsl which pointed to a .mill file and b) each of the MM nodes seemed to have a lot of junk in them I didn't need. 

[Andy Moorer]

Awesome Gustavo, yes you broke it out perfectly!

The big plus with this multi-node approach is that you can do stuff I haven't implemented, like rotate the texture, or more easily mix several matcaps together, add bump etc (though I guess you could do that with the single shader, too.)

The shader I made is the bare bones, I strongly urge anyone interested to dive in. I'd love it if someone better at this kind of thing blew my little shader out of the water with something more robust.

If we can get a really strong solution it will be a major boon, just look at the mileage Zbrush users get out of matcap shading - it's a very robust way to shade, in particular for modeling/sculpting and for stills.

[olivier jeannel]

Hi list,
Asking this question with no shame ^^;
This might be basic, but what is the difference between a force and a
velocity ?

Sorry for being half brained...

[César Sáez]

They are completely different concepts, in short:

Force = mass * acceleration

Acceleration = Velocity / time

Velocity = distance-traveled / time

[César Sáez]

Actually speed = distance-traveled/time, velocity is more like position / time (it's a vector).

[Daniel Harjanto]

Try to answer the question,

Force will be considering mass as F = m.a, where m is mass, and a is acceleration.
While velocity doesn't consider mass.

Cheers,

--
Daniel Harjanto
Infinite Frameworks Studios
TD
http://misterdi.cgpot.com

[olivier jeannel]

Thank's C�sar ! Now it's clear :)

Le 13/02/2013 11:13, C�sar S�ez a �crit :

[Raffaele Fragapane]

It's actually incorrect more so than clear.

Distance traveled divided by time is not velocity. It's speed, a scalar value.

Velocity has to do with the rate of change and is represented by a vector providing direction, and it's magnitude representing speed. So while speed is the simple speed an object travels at, velocity also specifies direction.

A force is represented similarly to velocity and has/affects (depending on the approach, generally correct if considered affecting) mass, it's effect results in the affected body's velocity. A force is basically something affecting another entity into change. It can be summarized in push or pull efforts.

Acceleration is the rate of change of speed.

This is for Newtonian physics, or vectorial physics, the most common kind in CG related models.

On Feb 13, 2013 9:52 PM, "olivier jeannel" <olivier...@noos.fr> wrote:

Thank's César ! Now it's clear :)

[Raffaele Fragapane]

Reading that after sending it (when else would one proof read?) I realize it might come across as a bit aggressive with the opening comment.
My apologies to César if that's the case, it sure wasn't intended to be or to discourage someone else away from contributing to the already rare physics discussions.

Oliver, Wikipedia has excellent resources about these subjects, just look up Newtonian physics and wiki hop around to the specific words you wanted the meaning of.

[Leonard Koch]

Velocity is the speed and direction at which an object moves.

Force is something with a strength and a direction that causes acceleration on the object and has an effect on its velocity.

Think about it like this:

You have some object, that is flying through space with a VELOCITY. 

Maybe it is flying upwards, so the DIRECTION it is flying in would be [0, 1, 0].

And let's say that it is flying with a SPEED of 6 units per second.

Those are the two components our velocity is made of: In what direction is it flying and how fast is it flying there?

In ICE these two values are one vector [0, 6, 0] which you get by multiplying the SPEED (5) with the DIRECTION ([0, 1, 0]).

When you multiply a vector(a row of numbers) with a scalar(a single number) you simply multiply each element of the vector with the scalar.

[6*0, 6*1, 6*0] = [0, 6, 0] = the VELOCITY of the object

Now there is also the FORCE and this could be for example gravity or a wind blowing, just anything that will move the object.

What properties does a force have?

It pushes things in a certain DIRECTION.

It pushes things with a certain STRENGTH.

Let's say our force pushes to the left, so in Softimage it's DIRECTION would be [1, 0, 0].

And let's say that the STRENGTH of the forces push is 8.

Once again these two properties of the force are represented in one vector [2, 0, 0] which you get by multiplying the STRENGTH(8) with the DIRECTION ([1, 0, 0]).

[8*1, 8*0, 8*0] = [8, 0, 0] = the FORCE.

So now that we have a good idea of what a force and a velocity are let's see what happens when the force acts upon our little object.

The object is moving through space with the VELOCITY [0, 6, 0] and suddenly there is the FORCE [8, 0, 0] acting on it. Maybe an interstellar wind started blowing...

The FORCE is acting on the object and wants to push it in another direction/speed, which would give it a new VELOCITY since velocity = speed*direction.

But the object also has MASS, that is a measure of how resistant to changing velocity an object is. Let's say it has the MASS 4.

Now the force is still going to change the velocity of the object, but the effect will be reduced by the mass.

The formula for how it is going to be reduced by the mass, comes from the famous formula F=m*a ( force = mass * acceleration )

If we divide both sides of that equation by mass we get:

F/m = a ( force / mass = acceleration)

This tells us that we must divide our FORCE (F) by the MASS (m) to get the acceleration.

Force: [8, 0, 0] 

Mass: 4

[8, 0, 0] / 4 = [8/4, 0/4, 0/4 = [2, 0, 0] = acceleration

Now all you have to do once you have computed your acceleration is to add it your velocity, to see what new velocity the force pushed it in.

So: PreForceVelocity + acceleration = PostForceVelocity

Velocity = [0, 6, 0]

Acceleration = [2, 0, 0]

[0, 6, 0] + [2, 0, 0] = [0+2, 6+0, 0+0] = [2, 6, 0] = PostForceVelocity

After our Object travelling with a mass of 4 and a velocity of [0, 6, 0] interacted with a force of [8, 0, 0] its new velocity is [2, 6, 0]

A bit elaborate, but I hope it helps you gain some intuition.

[Leonard Koch]

Apologies, there are a couple of number typos in the first one.

Here is the corrected version:

Velocity is the speed and direction at which an object moves.

Force is something with a strength and a direction that causes acceleration on the object and has an effect on its velocity.

Think about it like this:

You have some object, that is flying through space with a VELOCITY. 

Maybe it is flying upwards, so the DIRECTION it is flying in would be [0, 1, 0].

And let's say that it is flying with a SPEED of 6 units per second.

Those are the two components our velocity is made of: In what direction is it flying and how fast is it flying there?

In ICE these two values are one vector [0, 6, 0] which you get by multiplying the SPEED (6) with the DIRECTION ([0, 1, 0]).

When you multiply a vector(a row of numbers) with a scalar(a single number) you simply multiply each element of the vector with the scalar.

[6*0, 6*1, 6*0] = [0, 6, 0] = the VELOCITY of the object

Now there is also the FORCE and this could be for example gravity or a wind blowing, just anything that will move the object.

What properties does a force have?

It pushes things in a certain DIRECTION.

It pushes things with a certain STRENGTH.

Let's say our force pushes to the left, so in Softimage it's DIRECTION would be [1, 0, 0].

And let's say that the STRENGTH of the forces push is 8.

Once again these two properties of the force are represented in one vector [8, 0, 0] which you get by multiplying the STRENGTH(8) with the DIRECTION ([1, 0, 0]).

[8*1, 8*0, 8*0] = [8, 0, 0] = the FORCE.

So now that we have a good idea of what a force and a velocity are let's see what happens when the force acts upon our little object.

The object is moving through space with the VELOCITY [0, 6, 0] and suddenly there is the FORCE [8, 0, 0] acting on it. Maybe an interstellar wind started blowing...

The FORCE is acting on the object and wants to push it in another direction/speed, which would give it a new VELOCITY since velocity = speed*direction.

But the object also has MASS, that is a measure of how resistant to changing velocity an object is. Let's say it has the MASS 4.

Now the force is still going to change the velocity of the object, but the effect will be reduced by the mass.

The formula for how it is going to be reduced by the mass, comes from the famous formula F=m*a ( force = mass * acceleration )

If we divide both sides of that equation by mass we get:

F/m = a ( force / mass = acceleration)

This tells us that we must divide our FORCE (F) by the MASS (m) to get the acceleration.

Force: [8, 0, 0] 

Mass: 4

[8, 0, 0] / 4 = [8/4, 0/4, 0/4 = [2, 0, 0] = acceleration

Now all you have to do once you have computed your acceleration is to add it your velocity, to see what new velocity the force pushed it in.

So: PreForceVelocity + acceleration = PostForceVelocity

Velocity = [0, 6, 0]

Acceleration = [2, 0, 0]

[0, 6, 0] + [2, 0, 0] = [0+2, 6+0, 0+0] = [2, 6, 0] = PostForceVelocity

After our Object travelling with a mass of 4 and a velocity of [0, 6, 0] interacted with a force of [8, 0, 0] its new velocity is [2, 6, 0]

A bit elaborate, but I hope it helps you gain some intuition.

[Matt Lind]

Velocity = net displacement / time.

Force = Mass * Acceleration


NOTE: Speed and velocity are very different.

Speed is distance traveled over time (scalar)
Velocity is net displacement over time (vector)

Example: Running around a track in a stadium.

If you make a complete lap in 60 seconds, then your speed is 6.66 meters per second, while your velocity is 0 meters per second because you haven't been displaced from your starting position.



Matt

[Steven Caron]

is that distinctions helpful here?

i mean we aren't just going to render the last frame of our particle going around the track, we are going to render 1440 frames (24fps*60secs) and at each frame the instantaneous velocity is going to have some direction and magnitude.

[Matt Lind]

Yes it’s important because there are many equations which rely on velocity as a variable, not speed.

 

If you’re computing a speed, you need to accumulate the distance travelled over time.  If you’re computing a velocity, you’re working with deltas.

 

In more explicit terms, if you’re computing motion vectors, you’re comparing the current frame to the previous frame.  If you’re computing speed, you’re comparing the current frame to the first frame.  How you code for those scenarios is vastly different.

 

 

Matt

[Alok]

Even then, we are not computing the displacement between the first and the last frame only (which, of course, will lead to the wrongful observation that no displacement has taken place yielding zero velocity). In this case the displacement at the last frame will be calculated between the last and the second last frame.

I think what Steven is saying makes sense.

On 13/02/2013 3:03 PM, Matt Lind wrote:

[Oleg Bliznuk]

I think we get zero velocity only if we integrate it over elapsed time ( and it is not an instant vel which is used most of time I guess ), otherwise speed = scalar magnitude (length) of velocity, and the velocity = diff between the last and the prelast positions

[Steven Caron]

i know velocity is a delta and i have used it many a time as part of an equation which is variable.

while your distinction is accurate i was concerned it wasn't helping olivier with that example which illustrates no displacement. i just dont think it was an intuitive example of velocity. in your example it would mean you sat in the bleachers to watch the race but left right before it starts and returned the moment it ends. so you say to your friend, "man these guys haven't moved an inch!" and she looks at you weird...

[Matt Lind]

If you have a problem with the definitions, talk to mathematicians and physicists.

 

I only put out the information for clarity as confusing speed with velocity is very common and there are times when it does indeed matter.

[Raffaele Fragapane]

I'm a big fan of sticking to definitions, but they normally need to be presented in sets to the uninitiated, or they become confusing.

IE: if you decide to put forward an example sampled across 6.6 seconds resulting in the case of no velocity despite much distance covered, the concept of discrete units and sampling intervals might need mentioning.

It's probably easier to go Lagrangian, and assume the 1fps sampling unit, and examplify on that, if you don't want to touch on calculus (which I find tends to be more intimidating than most other fundamentals for new comers).

[Steven Caron]

i have no problem with the definitions and i just thought you might actually be causing more confusion with your example.

[Matt Lind]

I gave the same definition and example as was used when I learned it in physics class all those years ago.  I don’t recall anybody having a problem with it then.

 

Matt

 

 

 

From: softimag...@listproc.autodesk.com [mailto:softimag...@listproc.autodesk.com] On Behalf Of Steven Caron
Sent: Wednesday, February 13, 2013 1:46 PM
To: soft...@listproc.autodesk.com
Subject: Re: Difference between a force and a velocity ?

 

i have no problem with the definitions and i just thought you might actually be causing more confusion with your example.

[Steven Caron]

sorry if my responses have perturbed you, that wasn't my intention. i knew this rabbit hole would open up and we would all start falling down it arguing about this, but its my fault for responding in the first place... knowing you wont agree with my statements.

i just hope olivier understands how to use force and velocity in ICE.

[Ed Manning]

C'mon, Matt --

You're sounding a little like Chris Cox here... ;-)

You may be technically correct, but if we are going to split hairs, then we could also differentiate between net or average velocity, which you describe, and instantaneous velocity.  (If I could figure out how to print proper equations and math symbols in email, then I'd probably go ahead and make a fool of myself trying to demonstrate my memory of 35-year old pre-calculus courses.)  Anyway, change in position over change in time isn't really what I remember calling velocity in physics or math class -- velocity is the *limit* of change in position over change in time, as the amount of change in time nears zero.  Or we could just say it's the derivative of position with respect to time, but then we need more definitions, which might mean more disagreements...

Suffice to say that you can of course have any number of velocity vectors sum to zero over time.  What might be more helpful to the OP would be using your example to introduce the concept of temporal sampling.  Velocity isn't necessarily position now minus position a frame ago, even in an animation application -- it can be computed over any time interval, including subframe steps.  And of course your example also illustrates temporal aliasing -- it's easy to miss important changes when you only look once every frame. When we compute motion vectors, we usually try to temporally subsample for exactly this reason. 

[Alok]

Let me make somethings very clear here for the sake of discussion:
Here we have to define whether or not we are talking of instantaneous velocity or simply velocity. As Raff touched the topic of Calculus (and also in almost all cases of studying the Newtonian Model of Classical Mechanics), most of the time we are dealing with instantaneous velocity, which as the name suggest, is the velocity of the object at the instant of time. In case the object starts and ends at the same point, once it has come to rest, both it's velocity and speed are zero. But just before that (and here comes deltas) at a moment of time it was not at the same place (it couldn't be as it is moving in some direction) and thus it will have velocity.

Now consider a man walking on the tread mill, does he have velocity (or speed)? Neither because there is no displacement (huh ? the odometer is lying and I am sweating for nothin'). We always need to consider instantaneous stuff. Specially in our field it is of utmost importance as we are dealing with simulation and particles.

Off topic, just for fun - Go read Zeno's Achilles and tortoise paradox (if you haven't already), which is somewhat related. You'll have fun !

On 13/02/2013 4:32 PM, Matt Lind wrote:

[Matt Lind]

I’m not perturbed, I just don’t think your assessment is accurate based on past experience of the subject being taught.

[Matt Lind]

I think its other people who are blowing this out of proportion.  After sending my initial email on the subject, I encountered this one from Olivier which precedes mine:

 

>Thank's César ! Now it's clear :)

> 

>Le 13/02/2013 11:13, César Sáez a écrit :

> 

> They are completely different concepts, in short:

> 

> Force = mass * acceleration

> 

> Acceleration = Velocity / time

> 

> Velocity = distance-traveled / time

> 

 

 

 

Seeing how I didn’t do much differently than Caesar as far as explanation other than to give an example, I don’t think I’ve done anything to split hairs or mislead people.  I gave a very introductory explanation as is normally given on the 1^st or 2^nd day of a physics course.  If you want to dive deeper and differentiate between instantaneous velocities and whatever everybody else is throwing about, then you are the ones splitting the hairs, not me.

[Steven Caron]

yep, i was... your example was the confusing part, not the definition.

[olivier jeannel]

... It was clearer before ... :DDD

[Raffaele Fragapane]

[Steven Caron]

please add...

http://cgmemes.blogspot.com/

On Wed, Feb 13, 2013 at 3:01 PM, Raffaele Fragapane <raffsx...@googlemail.com> wrote:

[Raffaele Fragapane]

I don't have an account, please feel free to do so in my place. I googled it though, didn't do all the post myself ;)

--
Our users will know fear and cower before our software! Ship it! Ship it and let them flee like the dogs they are!

[Andy Jones]

The example of running around on a track is wrong unless you're on a stationary planet.  Way to confuse Olivier, guys :(

Here's my stab:  Definitions aside, velocity describes how something is moving at an instant in time.  Force, however, is more closely related to how an object will accelerate over time (in that the acceleration of the object is directly proportional to the forces acting upon it).  So if a force is acting on an object, that will cause it to start moving differently as time passes, like what happens when you drop something.  When you drop a ball, gravity (a force) causes its downward velocity to increase more and more as time passes.

A noteworthy exception to this simple explanation of forces and velocity are forces that are actually dependent on velocity.  In practice, one of the most common such forces is the force of friction with a surface, such as a table top (or even the surface of air against the object, which is a component of the "drag" force affecting an object traveling through a medium).  For example, as an object travels faster and faster through the air, the force of friction with the air will increase.  This is why when you first drop a ball it speeds up on its way down, but once it's falling fast enough, it will stop speeding up and continue at a constant speed (and constant velocity, if you're a stickler for terminology).  Once the ball is falling at a fixed speed, you know that the force of gravity and the force of drag/friction are equal and opposite, such that they cancel out.

This is of course ignoring secondary effects, such as the heat generated by the people below arguing over email forums about how physics works.  In this case, as the arguing increases, the amount of hot air released increases, and the air surrounding the ball becomes thinner, reducing the drag force on the ball, and causing it to fall even faster, thereby increasing the speed at which the ball hits its intended target, hopefully reminding that target that it's silly to get worked up over Newtonian physics, since it's a simplified approximation of reality anyway.

- Andy

[Raffaele Fragapane]

But is the ball propelled by an internal combustion engine, or is it victim of the gravitational pull?

Because if it's the former I feel it's remiss of you to forget that cold air is more compressible (or rather more easily pulled in larger quantities) than hot air, and therefore results in a more favorable stoichiometric ratio chambered and more power.

Seriously, man, get your examples right! You can't omit details like that.

[Matt Lind]

There’s nothing wrong with the running track example.  You are overthinking the problem.  It’s an introductory level illustration to clarify the difference between speed (scalar) and velocity (vector), and is used in physics text books believe it or not.  If you do a google search on “physics definition of velocity”, the top three results all explain velocity as the change in position divided by the change in time. (e.g. v = x1 – x0 / t1 – t0).  If the start position and end position are the same, then the velocity is indeed zero over any time interval because it has no magnitude, and proven mathematically.  If you have a problem with that, take it up with the physicists and mathematicians.

 

 

 

 

 

 

From: softimag...@listproc.autodesk.com [mailto:softimag...@listproc.autodesk.com] On Behalf Of Andy Jones
Sent: Wednesday, February 13, 2013 6:08 PM
To: soft...@listproc.autodesk.com
Subject: Re: Difference between a force and a velocity ?

 

The example of running around on a track is wrong unless you're on a stationary planet.  Way to confuse Olivier, guys :(

[Eric Thivierge]

Can someone explain this "Physics" thing to me?

--------------------------------------------
Eric Thivierge
http://www.ethivierge.com

[Andy Jones]

http://xkcd.com/1145/

[Xavier Lapointe]

http://what-if.xkcd.com/28/

Teaser: Physics, Steaks, and Cooking.

--
Xavier

[Alan Fregtman]

Since this whole convo is about velocity, I feel the 1st "What If" is more relevant:

 

http://what-if.xkcd.com/1/

Teaser: baseball, 90% speed of light and doom.

 

[Alan Fregtman]

Added. ;)

[Adam Sale]

wow... 

[Andreas Böinghoff]

Interesting thread - I never really thought about the technical aspects of that. 

On 2/13/2013 12:37 PM, Leonard Koch wrote:

Apologies, there are a couple of number typos in the first one.

Here is the corrected version:

Velocity is the speed and direction at which an object moves.

Force is something with a strength and a direction that causes acceleration on the object and has an effect on its velocity.

Think about it like this:

You have some object, that is flying through space with a VELOCITY. 

Maybe it is flying upwards, so the DIRECTION it is flying in would be [0, 1, 0].

And let's say that it is flying with a SPEED of 6 units per second.

Those are the two components our velocity is made of: In what direction is it flying and how fast is it flying there?

In ICE these two values are one vector [0, 6, 0] which you get by multiplying the SPEED (6) with the DIRECTION ([0, 1, 0]).

When you multiply a vector(a row of numbers) with a scalar(a single number) you simply multiply each element of the vector with the scalar.

[6*0, 6*1, 6*0] = [0, 6, 0] = the VELOCITY of the object

Now there is also the FORCE and this could be for example gravity or a wind blowing, just anything that will move the object.

What properties does a force have?

It pushes things in a certain DIRECTION.

It pushes things with a certain STRENGTH.

Let's say our force pushes to the left, so in Softimage it's DIRECTION would be [1, 0, 0].

And let's say that the STRENGTH of the forces push is 8.

Once again these two properties of the force are represented in one vector [8, 0, 0] which you get by multiplying the STRENGTH(8) with the DIRECTION ([1, 0, 0]).

[8*1, 8*0, 8*0] = [8, 0, 0] = the FORCE.

So now that we have a good idea of what a force and a velocity are let's see what happens when the force acts upon our little object.

The object is moving through space with the VELOCITY [0, 6, 0] and suddenly there is the FORCE [8, 0, 0] acting on it. Maybe an interstellar wind started blowing...

The FORCE is acting on the object and wants to push it in another direction/speed, which would give it a new VELOCITY since velocity = speed*direction.

But the object also has MASS, that is a measure of how resistant to changing velocity an object is. Let's say it has the MASS 4.

Now the force is still going to change the velocity of the object, but the effect will be reduced by the mass.

The formula for how it is going to be reduced by the mass, comes from the famous formula F=m*a ( force = mass * acceleration )

If we divide both sides of that equation by mass we get:

F/m = a ( force / mass = acceleration)

This tells us that we must divide our FORCE (F) by the MASS (m) to get the acceleration.

Force: [8, 0, 0] 

Mass: 4

[8, 0, 0] / 4 = [8/4, 0/4, 0/4 = [2, 0, 0] = acceleration

Now all you have to do once you have computed your acceleration is to add it your velocity, to see what new velocity the force pushed it in.

So: PreForceVelocity + acceleration = PostForceVelocity

Velocity = [0, 6, 0]

Acceleration = [2, 0, 0]

[0, 6, 0] + [2, 0, 0] = [0+2, 6+0, 0+0] = [2, 6, 0] = PostForceVelocity

After our Object travelling with a mass of 4 and a velocity of [0, 6, 0] interacted with a force of [8, 0, 0] its new velocity is [2, 6, 0]

A bit elaborate, but I hope it helps you gain some intuition.

On Wed, Feb 13, 2013 at 12:24 PM, Leonard Koch <leonard...@gmail.com> wrote:

Velocity is the speed and direction at which an object moves.

Force is something with a strength and a direction that causes acceleration on the object and has an effect on its velocity.

Think about it like this:

You have some object, that is flying through space with a VELOCITY. 

Maybe it is flying upwards, so the DIRECTION it is flying in would be [0, 1, 0].

And let's say that it is flying with a SPEED of 6 units per second.

Those are the two components our velocity is made of: In what direction is it flying and how fast is it flying there?

In ICE these two values are one vector [0, 6, 0] which you get by multiplying the SPEED (5) with the DIRECTION ([0, 1, 0]).

When you multiply a vector(a row of numbers) with a scalar(a single number) you simply multiply each element of the vector with the scalar.

[6*0, 6*1, 6*0] = [0, 6, 0] = the VELOCITY of the object

Now there is also the FORCE and this could be for example gravity or a wind blowing, just anything that will move the object.

What properties does a force have?

It pushes things in a certain DIRECTION.

It pushes things with a certain STRENGTH.

Let's say our force pushes to the left, so in Softimage it's DIRECTION would be [1, 0, 0].

And let's say that the STRENGTH of the forces push is 8.

Once again these two properties of the force are represented in one vector [2, 0, 0] which you get by multiplying the STRENGTH(8) with the DIRECTION ([1, 0, 0]).

[8*1, 8*0, 8*0] = [8, 0, 0] = the FORCE.

So now that we have a good idea of what a force and a velocity are let's see what happens when the force acts upon our little object.

The object is moving through space with the VELOCITY [0, 6, 0] and suddenly there is the FORCE [8, 0, 0] acting on it. Maybe an interstellar wind started blowing...

The FORCE is acting on the object and wants to push it in another direction/speed, which would give it a new VELOCITY since velocity = speed*direction.

But the object also has MASS, that is a measure of how resistant to changing velocity an object is. Let's say it has the MASS 4.

Now the force is still going to change the velocity of the object, but the effect will be reduced by the mass.

The formula for how it is going to be reduced by the mass, comes from the famous formula F=m*a ( force = mass * acceleration )

If we divide both sides of that equation by mass we get:

F/m = a ( force / mass = acceleration)

This tells us that we must divide our FORCE (F) by the MASS (m) to get the acceleration.

Force: [8, 0, 0] 

Mass: 4

[8, 0, 0] / 4 = [8/4, 0/4, 0/4 = [2, 0, 0] = acceleration

Now all you have to do once you have computed your acceleration is to add it your velocity, to see what new velocity the force pushed it in.

So: PreForceVelocity + acceleration = PostForceVelocity

Velocity = [0, 6, 0]

Acceleration = [2, 0, 0]

[0, 6, 0] + [2, 0, 0] = [0+2, 6+0, 0+0] = [2, 6, 0] = PostForceVelocity

After our Object travelling with a mass of 4 and a velocity of [0, 6, 0] interacted with a force of [8, 0, 0] its new velocity is [2, 6, 0]

A bit elaborate, but I hope it helps you gain some intuition.

On Wed, Feb 13, 2013 at 12:16 PM, Raffaele Fragapane <raffsx...@googlemail.com> wrote:

Reading that after sending it (when else would one proof read?) I realize it might come across as a bit aggressive with the opening comment.
My apologies to César if that's the case, it sure wasn't intended to be or to discourage someone else away from contributing to the already rare physics discussions.

Oliver, Wikipedia has excellent resources about these subjects, just look up Newtonian physics and wiki hop around to the specific words you wanted the meaning of.

On Feb 13, 2013 10:08 PM, "Raffaele Fragapane" <raffsx...@googlemail.com> wrote:

It's actually incorrect more so than clear.

Distance traveled divided by time is not velocity. It's speed, a scalar value.

Velocity has to do with the rate of change and is represented by a vector providing direction, and it's magnitude representing speed. So while speed is the simple speed an object travels at, velocity also specifies direction.

A force is represented similarly to velocity and has/affects (depending on the approach, generally correct if considered affecting) mass, it's effect results in the affected body's velocity. A force is basically something affecting another entity into change. It can be summarized in push or pull efforts.

Acceleration is the rate of change of speed.

This is for Newtonian physics, or vectorial physics, the most common kind in CG related models.

[Alok Gandhi]

"If you have a problem with that, take it up with the physicists and mathematicians."

Sorry Matt but I think you're wrong, and you can consider me a mathematician (I have a Masters in Mathematics and a Bachelor in Physics, Chemistry and Mathematics). 

Just taking a pure math approach now to set things right. 

Velocity is a vector as we know with a magnitude as Speed and a direction.

So we can write:

Where 's' is speed and 'v' is velocity and   is the magnitude of the velocity.

Now in the above equation, we cannot have 's' as non-zero and 'v' as zero. Because if 's' is zero , 'v' will be zero and if 's' is non-zero so will be 'v'.

Here is a reference:

http://bit.ly/XOAM50

Cheers !

Alok Gandhi

Lead TD

Modusfx

[Matt Lind]

Like I said multiple times already.  Take it up with the physicists and mathematicians.  The example given is from a physics text book.  You got a problem, take it up with the author.

 

Sheesh!

 

 

 

 

From: softimag...@listproc.autodesk.com [mailto:softimag...@listproc.autodesk.com] On Behalf Of Alok Gandhi

Sent: Thursday, February 14, 2013 5:15 AM
To: soft...@listproc.autodesk.com

[Grahame Fuller]

The formula
[Inline image 1]
is valid for instaneous velocity and speed, but not average velocity and speed over an interval. I think that maybe that was the point of the example in the textbook.

Suppose we take 4 samples around the track. The average velocity is given by:

Vavg = AVG([ 6.66, 0], [0, 6.66], [-6.66, 0], [0, -6.66]) = ([ 6.66, 0] + [0, 6.66] + [-6.66, 0] + [0, -6.66])/4 = [0, 0]/4 = [0, 0]

However because speed has no direction (you cannot travel at -10km/h for example), the average speed is given by:

Savg = AVG (6.66, 6.66, 6.66, 6.66) = 6.66

Of course, this distinction is secondary to the more important distinction that velocity is a vector (length and direction) while speed is a scalar (magnitude only).

gray

From: softimag...@listproc.autodesk.com [mailto:softimag...@listproc.autodesk.com] On Behalf Of Matt Lind
Sent: Thursday, February 14, 2013 01:42 PM
To: soft...@listproc.autodesk.com
Subject: RE: Difference between a force and a velocity ?

Like I said multiple times already. Take it up with the physicists and mathematicians. The example given is from a physics text book. You got a problem, take it up with the author.

Sheesh!

From: softimag...@listproc.autodesk.com<mailto:softimag...@listproc.autodesk.com> [mailto:softimag...@listproc.autodesk.com] On Behalf Of Alok Gandhi
Sent: Thursday, February 14, 2013 5:15 AM
To: soft...@listproc.autodesk.com<mailto:soft...@listproc.autodesk.com>
Subject: Re: Difference between a force and a velocity ?

"If you have a problem with that, take it up with the physicists and mathematicians."

Sorry Matt but I think you're wrong, and you can consider me a mathematician (I have a Masters in Mathematics and a Bachelor in Physics, Chemistry and Mathematics).

Just taking a pure math approach now to set things right.

Velocity is a vector as we know with a magnitude as Speed and a direction.

So we can write:

[Inline image 1]

Where 's' is speed and 'v' is velocity and [Inline image 2] is the magnitude of the velocity.

[Alok]

Totally agree with you Graham, and that is why is my earlier post I had clearly mentioned the distinction between average and instantaneous velocities. Moreover, in case of CG we never talk of average velocities, it is always instantaneous. So the example of average velocity was not relevant.

-----
No virus found in this message.
Checked by AVG - www.avg.com
Version: 2012.0.2238 / Virus Database: 2639/5601 - Release Date: 02/13/13

[Bradley Gabe]

In the past 2 years, I attended college courses in Physics 101 and 201 at UT Dallas, as well as a Princeton Review course for the MCAT's which include heavy study in basic physics, especially about the definitions of fundamental units and concepts.

As it turns out, Matt is correct in that the example of displacement returning to origin will result in an average velocity of 0. Both my physics professor as well as the Princeton Review instructor took special care to mention this specific example more than once because it is a common trick question on the MCATs.

HOWEVER, and this is important. It is absolutely NOT the first and only example provided to explain the difference between velocity and speed. Nor did it come up during the first lecture about velocity. It was brought up long after we had repeatedly reviewed the concept, in an attempt to trip us up in order to demonstrate the exception.

Defining velocity and then using zero displacement to explain it is similar to introducing the letter 'p' and using the word 'pneumonia' as your very first example. If you have a problem with that, take it up with the lexicographers and wordsmiths?

-B

[Andy Moorer]

This is one of the most intelligently absurd threads I've ever followed. :)

[Raffaele Fragapane]

It's fascinating the way a car crash between two vehicles driven by physicists arguing with each other as they are flung out of their seats and through their windscreens would be.
Can't take my eyes away, horror and awe together.

Ultimately, I'm with Spiderman on this.

[Grahame Fuller]

You waterski while twirling a web strand behind you?

Pictures or it never happened!

gray

From: softimag...@listproc.autodesk.com [mailto:softimag...@listproc.autodesk.com] On Behalf Of Raffaele Fragapane
Sent: Thursday, February 14, 2013 04:58 PM
To: soft...@listproc.autodesk.com
Subject: Re: Difference between a force and a velocity ?

It's fascinating the way a car crash between two vehicles driven by physicists arguing with each other as they are flung out of their seats and through their windscreens would be.
Can't take my eyes away, horror and awe together.

Ultimately, I'm with Spiderman on this.

[Alan Fregtman]

If a parrot's velocity is [0,0,0], is it dead?

:p

[Raffaele Fragapane]

Is it in a forest? And is anybody watching?

[Eric Thivierge]

Is the parrot sitting on a leopard?

--------------------------------------------
Eric Thivierge
http://www.ethivierge.com

[Ahmidou Lyazidi]

Schr�dinger parrots ?
Ahmidou Lyazidi
Director | TD | CG artist
http://vimeo.com/ahmidou/videos


2013/2/15 Eric Thivierge <ethiv...@gmail.com>:

[Ed Manning]

laden or un-laden?

On Thu, Feb 14, 2013 at 5:55 PM, Ahmidou Lyazidi <ahmid...@gmail.com> wrote:

Schrödinger parrots ?

[Andy Moorer]

African or European? 

Hey, Alan started the Monty python references, that made the question mandatory.

[Simon Anderson]

Bring me a shrubbery

--
-------------------
Simon Ben Anderson
blog: http://vinyldevelopment.wordpress.com/

[Ed Manning]

NI! NI! NI! NINININ!!!!

[Bradley Gabe]

Oh, what sad times are these when passing ruffians can say Ni at will to old ladies.

[Eric Lampi]

Aaaaalbatross!

--
Freelance 3D and VFX animator

http://vimeopro.com/user7979713/3d-work

[Steven Caron]

https://i.chzbgr.com/maxW500/7055969792/h07769A83/

[Matt Lind]

Thanks for the validation, Brad.

 

Each instructor has their own methods because this was indeed the first example presented to me when I took physics years ago to make the distinction between speed and velocity.  The professor said straight up it would be the very first question on the exam (it was) and anybody who got the question wrong would fail the exam.  Nobody failed, to my knowledge.  It wasn’t treated as a trick question, it was fundamental material.

 

He often reused examples across different topics so we could focus on the relevant new information and cut down the noise/relearning.  For example, the running track was reused (with slight modifications) as an introduction to ‘work’ illustrating no work is performed if there is no displacement.  Since he already proved the theory with velocity, he didn’t have to waste time re-explaining it for work.  While tough at first due to the abstract nature, I found his approach very refreshing as it allowed me to tackle tougher problems because I was exposed to the edge cases upfront and conditioned to not think of them as edge cases.  Edge cases are a product of our tools (math) to explain observed behavior.  The universe doesn’t have a concept of edge case.

 

I think my professor chose this approach because he once mentioned students had problems progressing from newton’s laws to quantum mechanics because as they entered more advanced levels, they tended to forget previously learned material still applied (or they didn’t see the connection).  He conditioned students to think more about the problem rather than merely choose a formula and plug in numbers.  Often his toughest problems required the least amount of math to solve.  Some of his lectures could be really dry, as he was a dry guy himself, but he was probably the best physics professor I had of many.

 

 

Matt

 

 

 

From: softimag...@listproc.autodesk.com [mailto:softimag...@listproc.autodesk.com] On Behalf Of Bradley Gabe
Sent: Thursday, February 14, 2013 1:21 PM
To: soft...@listproc.autodesk.com
Subject: Re: Difference between a force and a velocity ?

 

In the past 2 years, I attended college courses in Physics 101 and 201 at UT Dallas, as well as a Princeton Review course for the MCAT's which include heavy study in basic physics, especially about the definitions of fundamental units and concepts.

[Raffaele Fragapane]

[Sandy Sutherland]

Ah - this is where you can tell the difference between a force and a velocity - as if you took your parrot with a volocity of [0,0,0] and gave it a force of your foot in it's butt ----------------> [Len100] then not only would you find out if the velocity of [0,0,0] meant it was dead, but you would also change the velocity to [100,0,0]

S.

                                                              
Sandy Sutherland | Technical Supervisor

---

From: softimag...@listproc.autodesk.com [softimag...@listproc.autodesk.com] on behalf of Alan Fregtman [alan.f...@gmail.com]
Sent: 15 February 2013 00:22
To: XSI Mailing List

Subject: Re: Difference between a force and a velocity ?

[Greg Punchatz]

Please, just let me have a little bit of peril!

Sent from my iPhone

[Grahame Fuller]

http://www.youtube.com/watch?v=gENVB6tjq_M

gray

From: softimag...@listproc.autodesk.com [mailto:softimag...@listproc.autodesk.com] On Behalf Of Raffaele Fragapane
Sent: Thursday, February 14, 2013 11:15 PM
To: soft...@listproc.autodesk.com
Subject: Re: Difference between a force and a velocity ?

[Inline image 1]

[Dan Yargici]

Hi Andy, I was musing about this some more this morning and I have a suggestion. :)

How big an endeavor would it be to have this as a Display mode?  i.e setting a viewport Display mode to 'Matcap' or 'Litsphere' and having it override all materials.  Perhaps setting the litsphere image via custom preference?

A more immediate way to set it up would certainly be awesome.

Like I say, it's just a suggestion, kudos for making the shader in the first place! 

DAN

On Wed, Feb 13, 2013 at 5:40 AM, Andy Moorer <andym...@gmail.com> wrote:

Awesome Gustavo, yes you broke it out perfectly!

The big plus with this multi-node approach is that you can do stuff I haven't implemented, like rotate the texture, or more easily mix several matcaps together, add bump etc (though I guess you could do that with the single shader, too.)

The shader I made is the bare bones, I strongly urge anyone interested to dive in. I'd love it if someone better at this kind of thing blew my little shader out of the water with something more robust.

If we can get a really strong solution it will be a major boon, just look at the mileage Zbrush users get out of matcap shading - it's a very robust way to shade, in particular for modeling/sculpting and for stills.

[Andy Moorer]

Gosh Dan, I have no idea. It's an interesting thought.

One downer for me with the HQ viewport is the sometimes painfully lengthy delay when you switch to it. It may not seem like much time but it's enough to discourage people from using it routinely.

I have a love/hate thing going with the HQ viewport. When I make myself use it, and take the time to adjust some display settings, it can be a gorgeous way to work. When laying out a scene or doing some modeling having the greater aesthetic display environment has a real effect on my feeling and mood as I work. We are artists, and the "space" we work in makes a difference, one which our more rational/calculating minds tend to discount. But it's also painful when it pauses to think and breaks the very creative flow that is one of it's strengths.

I would urge the devs to keep working on the HQ view... its a first release effort, and a good start, but they haven't gotten it to a point where people love it yet. And they can get it there. To stop now would be to stop at a point where it adds no great value for the effort put into it.

But back to your thought about a high-speed zbrush-like creative viewport...

It's certainly possible to "fake" a display mode with a script which saves out all materials, switches them with a matcap shader while switching the viewport display to HQ or returns the materials. I considered this briefly but decided it was unwieldy and intrusive, at least at the scripting level I'm accustomed to.

But I LOVE the idea of a high-quality-high-speed view which uses something like a global matcap as a way to sculpt or simply display geo in a more aesthetic manner than the shaded view and/or get the benefits of fake complex lighting directly with a matcap.

(steps on soapbox aka - fungus time!)

I can only hope one of the guys out there who read this thread and know a thing or two has an idea and interest in the challenge. Softimage is a tool for artists. Softimage needs more artists to adopt it. Artists are sensitive to UIs and how their work is displayed. Zbrush has a painful UI but a beautiful environment for the actual art the artist is focusing on, and it immediately appeals. Softimage should tap into this effect. Engineers like displays like CAD systems, but artists want a sense of light and mass while they work.

(steps off soapbox) 

The Matcap shader and scale:

This brings up an issue with the way I've implemented this shader which becomes apparent sometimes in a scenario where you texture an entire scene across the whole camera field with a single matcap: the scale of the underlying texture of the matcap is affected by the size of the object. If you put a finely textured matcap that looks great on a small sphere and then put that same matcap on a huge backdrop object, this issue becomes apparent. I haven't looked into whether zbrush has a trick they use to deal with this - at the moment with my shader the only solutions are:

1) in a case like this, avoid matcaps which reveal much texture and just use them for light qualities (boo)

2) make several scale variants of the matcap where the texture and resolution of the source image compensates (now you've turned something simple and elegant into a lot of time/effort. boo again)

3) use the matcap in a rendertree setup in which objects recieve their "light" via the matcap but have flat, diffuse textures applied normally (better, but still you lose some of the benefits of the matcap. Worth trying out, though I haven't myself.)

[Gustavo Eggert Boehs]

A slightly lazier suggestion would be to setup a pass where this material overrides all other materials... just a thought

2013/2/18 Dan Yargici <danya...@gmail.com>

[Andy Moorer]

Good idea Gustavo, and it would be easy to script out a menu item to set up/switch to display this pass