MUFOOF

[Xtaz]

Hi Spliners :

 

ill tell about MUFOOF

Exist countless uses for the passes, one of them, is the capacity to work with Multiple Frames in Only One Frame ( MUFOOF ). Here i will show you, how to set parameters..

I use it to simulate a fast movement, GI, fire ..........

The general concept is to produce an action (it doesn't import how many frames it has) that will be executed in one frame ....

The secret for that is to alter the properties of the action inside CHO.

 

First project -> Create a fast spinning helix

 

-> model a simple helix and setup bones ( or download it here )

 

 

 

 

-> Create a new action , select the bone and select :

frame -------- set Z rotate to

25_______________90

50_______________180

75______________-90

100______________0

 

->of this you sort things out created a turn of 360 degrees in 100 frames

 

-> import model and action to a new Cho

in the action properties under CHO reset :

lenght = 1

repeat = 100 ( number of action frames )

 

 

and in rendering options set :

Passes = 100 (10x10)

motion blur to 100%

 

 

now render the scene and VOILA, you have all your action in one frame

 

 

Next step will be how to fake radiosity using one light...

here is the result ...

 

[heyvern]

You are... the lord of...

 

MUFOOF...

 

--------------

 

Can you tell us how that is pronounced?

 

Like a sneeze? (mä-füf)

 

or like a... cow? (mü-füf)

 

Vernon "!" Zehr

[Xtaz]

Hey Vern !!!!

 

lets call it ...hmmm ... hmmm .... I as Lord of the Mufoof declare its name Müfüf

 

I rendered a video to show how its works ... if the final scene will be 30 frames then you will use repeat = 30*100( # action´s frames ) or repeat=3000

 

Click here to download AVI 160Kb

Click here to download MOV 25Kb

[sbk]

That's cool, though the name seemed mildly rude at first glance.

 

Then again, this from the woman who misspelled fluorescent in the title of the post that gave rise to this intriguing lighting method.

 

I claim insufficient coffee at the time of my original post.

 

But MUFOOF is neat. I applaud your luminary genius.

 

S.

[oakchas]

very cool!

 

Now a question, oh Lord Of MUFOOF.

 

Sometimes a still of something like a propeller, as you show it, has slightly more "density" in the blur at certain "stations" of the clock.

 

The photo captures the blur of the propeller, as you have , but it is also possible to detemine that it is a two or three bladed prop, because there is a concentration of the blur at, say (for example of a 2 bladed prop) 10 and 5 o'clock...

 

Is it possible to recreate this with your method... or would that have to be done in post?

[heyvern]

I think you could figure this out using some math... I can't give you specifics... but...

 

The action can be divided by the number of passes related to frame rate possibly, not sure exactly, to force those propeller "positions" to render "more often" than the others.

 

Or you could have those positions in the action "hold" for more than a few frames... maybe add a jitter at those positions to blur it some more.

 

I had this problem with a very fast pattern moving along... the frame rate was not in synch with the rotation and it appeared to go backwards. I had to figure out how many repeats of the rotation would line up with my frame rate.

 

p.s. I have used the brain cells that contained my math education for Bewitched reruns.

 

Vernon "!" Zehr

[oakchas]

I had this problem with a very fast pattern moving along... the frame rate was not in synch with the rotation and it appeared to go backwards. I had to figure out how many repeats of the rotation would line up with my frame rate.

 

In some cases, IRL (in real life) this actually occurs.

 

Does anyone remember the "space ray guns" (I don't really remeber what they were actually called) that shot a ringed propeller off the end of the gun?

 

If you shot it straight up, the propeller disc would sail in the air... but it sometimes appeared to be reversing it's rotation direction in mid flight. By the laws of physics it should not have done that, and continued to go upward, but it certainly seemed to.

 

My father told me that was impossible... (I was, like, 6 or so)...

 

I "proved" it was possible by shooting the disc at the top of the molding above a door frame. The disc spun along the top of the frame, and then reversed direction, but when it reversed, it did not fly away from the wall off of the molding ledge, it merely continued along the ledge in the opposite direction until it fell, still spinning, of the end of the ledge.

 

Anyway, it was an interesting phenomenon, and at ~6, I had no way to explain it... it just did it... and the old man was amazed, too.

 

Pardon the digression.. but I guess I'm saying sometimes things really are as they appear. (Maybe A:M follows these odd laws of physics, too? )

[Xtaz]

Is it possible to recreate this with your method..

 

no problem about that .... I just recreate a new helix whit 2 blades 0º and 180º ( from the main helix ).. I didnt touch in bone position ... and in CHO I rotate Z axis in 60º ....

 

 

Was this the expected result?

[oakchas]

whoa!

 

Very close! Great!

 

Thanks... I'm sure I'll find a use for this!

 

you are indeed the Lord of MUFOOF!

[Xtaz]

thanks Oaks ... lets continue

 

Second Project -> Simulating radiosity

 

First step is create a room whit color walls.. in this case red in left and green in right side ... i added a texture in the ground just to enrich the scene ...

 

 

then I create a new light in room model . You can position it anywhere ...

 

 

create a new action and then you must position light in correct place, in this case in the red wall edge, set light color to red... then in frame #100 drag light until green wall edge and set light color to green...

 

 

You can download a zipped file which contains model and action here

 

Create a new Cho, delete all lights and ground import room model and put some props inside it .... relocate camera ....

Import room model and action ....

 

 

in action properties under CHO set lenght = 1 and repeat = 100

in rendering options set passes = 100 (10x10) and motion blur = 100%

 

Voila !!!!!

 

[trajcedrv]

Live and learn, they say!

 

Way too cool Xtaz! Thanks for sharing, I want more!

 

This is one of the uses of A:M that I would never think of...

 

drvarceto

[pixmite]

XTAZ:

 

The propller spining beyond the Nyquist rate concept is fantastic and its been something I've looking to simulate for a long time. I have one additional challenge though... Is it possible to use a value drawn from a pose to set the number of cycles through an expresion? This would allow one to control the PRM of the propeller like a throttle throughout an action.

 

Pixmite

[Xtaz]

DrVarceto and Pixmite ... thank you for your comments ...

Pix ... Ill try something later, i believe it possible... i need to learn more about expressions.. but i promisse to try ....

 

Third Project -> Flame

DOWNLOAD PROJECT HERE

 

- Create a simple flame model and its groups ( adjust colors )

 

 

- give it a better flame model

 

 

- Create a new action and follow steps :

 

 

- create a new Cho

-Import flame model and action

-click in flame´s action under CHO and set LENGHT=1 and REPEAT=100

-Adjust rendering options MOTION BLUR=100% and PASSES = 100 (10x10)

 

- model a incandescent prop adjust lights ...

 

... RENDER AND ....

 

 

and if you are thinking about animate it here is a test

http://xtaz.com.br/anima/match2.mov

 

for now its all folks ... next step will be Global Ilumination !!

[aaver]

[...]The propller spining beyond the Nyquist rate concept is fantastic and its been something I've looking to simulate for a long time.

 

An example of this :-)

[attachmentid=11008][attachmentid=11009]

 

I have one additional challenge though... Is it possible to use a value drawn from a pose to set the number of cycles through an expresion? This would allow one to control the PRM of the propeller like a throttle throughout an action.

I've thought about that, but since the absolute rotation is a function of all previous RPM settings, i see no simple way to achieve this in A:M. You would need a Sum() or rather a Quadrature() function for this.

 

It would be simple to have a plug-in calculate the quadrature of one channel and put the result in an other, but it would be much more awkward for the user. I'm not sure the average user will use a Quadrature() function enough for Hash to take the trouble. I'm almost sure it would be a waste... ;-)

M01.mov

Propeller01.zip

[pixmite]

[...]The propller spining beyond the Nyquist rate concept is fantastic and its been something I've looking to simulate for a long time.

 

An example of this :-)

[attachmentid=11008][attachmentid=11009]

 

I have one additional challenge though... Is it possible to use a value drawn from a pose to set the number of cycles through an expresion? This would allow one to control the PRM of the propeller like a throttle throughout an action.

I've thought about that, but since the absolute rotation is a function of all previous RPM settings, i see no simple way to achieve this in A:M. You would need a Sum() or rather a Quadrature() function for this.

 

It would be simple to have a plug-in calculate the quadrature of one channel and put the result in an other, but it would be much more awkward for the user. I'm not sure the average user will use a Quadrature() function enough for Hash to take the trouble. I'm almost sure it would be a waste... ;-)

 

I guess I'm too stubborn to give up on this yet. Here is my latest idea which is currently giving me mixed results; I take the concept that XTAZ introduces with this thread one step further by instead of assigning the cycle length of the prop rotation action to 1 in the choreography, I assign it the value of a pose via an expression.

 

The pose doesn't do anything except generate a percentage from 1 to 100. I then create an action that changes the pose value over time. By applying both of these actions to the model in the choreography, the cycle length (therefore the RPM) changes over time.

 

So far I gotten mixed results, sometimes it looks like the prop never moves when I render an image, sometimes it does, and still sometimes I get a syntax error with the expression. In short, I need to learn more about expression assignments, especially since I'm trying to assign a variable to a parameter instead of a function.

 

Pixmite

[aaver]

[...]I guess I'm too stubborn to give up on this yet. Here is my latest idea which is currently giving me mixed results; I take the concept that XTAZ introduces with this thread one step further by instead of assigning the cycle length of the prop rotation action to 1 in the choreography, I assign it the value of a pose via an expression.

 

The pose doesn't do anything except generate a percentage from 1 to 100. I then create an action that changes the pose value over time. By applying both of these actions to the model in the choreography, the cycle length (therefore the RPM) changes over time.

Yes of course it does and if that was your challenge, I misunderstood it. What I was trying to do was to control the actual angular velocity for the propeller with a pose slider and that I still think is very hard to do in A:M, but I might be wrong. Actually, I hope I'm wrong :-)

[pixmite]

[...]I guess I'm too stubborn to give up on this yet. Here is my latest idea which is currently giving me mixed results; I take the concept that XTAZ introduces with this thread one step further by instead of assigning the cycle length of the prop rotation action to 1 in the choreography, I assign it the value of a pose via an expression.

 

The pose doesn't do anything except generate a percentage from 1 to 100. I then create an action that changes the pose value over time. By applying both of these actions to the model in the choreography, the cycle length (therefore the RPM) changes over time.

Yes of course it does and if that was your challenge, I misunderstood it. What I was trying to do was to control the actual angular velocity for the propeller with a pose slider and that I still think is very hard to do in A:M, but I might be wrong. Actually, I hope I'm wrong :-)

 

I think we're trying to accomplsh the same thing (control the velocity of the spin with a pose slider) If I suceed with my model I'll post it.

[pixmite]

 

I think we're trying to accomplsh the same thing (control the velocity of the spin with a pose slider) If I suceed with my model I'll post it.

 

Well I think I did it, but I’m not quite sure. The reason being is that I’m not quite sure I fully understand what AM is doing, but I have been able to get the effect I was looking for so I thought I’d share.

 

The first attempt, based on XTAZ’s prop model concept from this thread:

 

I created a pose that did nothing and named it throttle, and then in an action I changed the pose value over time. I added this action to the chorography and set the cycle length for the prop to be an expression whose value is set to the pose value. As the animation plays out, the cycle length was changing in accordance with my action, however it didn’t seem to have any effect on the model. The velocity wasn’t changing. I’m guessing that AM may only look at the cycle length parameter of the chorography when it starts, so despite the fact that I was able to dynamically change it, my model only seemed to respond to its initial value.

 

The Second attempt (the one that works, I think):

 

I added a second bone to the prop model (calling it throttle) to which no control points are assigned. I then created a throttle pose, which scaled this new bone in the Z direction. I set the Z rotation value of my prop model bone to the Z scale value of the throttle bone. Multiplying this expression by ChorTime() and a constant I was able to control the rate of rotation with the pose slider. I set various pose values in an action to simulate the non-linear rotation nature of an aircraft engine going through a start up sequence and I got the effect I was looking for!!!

 

Note: Some strange behavior that I don’t really understand: If I reduce my throttle value over time too quickly the prop starts to spin backwards? This occurs before the Nyquist rate is reached. Any ideas what I’m missing/not understanding?

 

If anyone is interested in the modle file I can posted later today.

 

Pixmite

[oakchas]

please post a .mov if not a model... both would be nice... to see the effect.

Thanks for working on this

[aaver]

[...]Note: Some strange behavior that I don't really understand: If I reduce my throttle value over time too quickly the prop starts to spin backwards? This occurs before the Nyquist rate is reached. Any ideas what I'm missing/not understanding?

Since you are multiplying the Z scale "function" with the ChorTime() you should expect this behavior. I also had this idea before I realized that setting the RPM with a slider is probably not possible in A:M (for reasons I have explained in an earlier post).

 

What is happening is that when the sum of the derivative of your "function" and the derivate of ChorTime() is negative, the propeller will spin back wards. If you have a propeller spinning at different RPM:s but always in the same direction and plot the absolute angle over time you will never have mixed sign derivatives in your plot. Always positive or negative.

[pixmite]

As requested here is the ver 12.0 model file. I tried to upload a 1MB AVI but the server grew impacient an deleted the process (56Kb line), so you'll have to render it yourself. Let me know what you think of it?

 

Pixmite

Prop_Throttle_Exp_3.zip

[Rodney]

I just wanna say... MUFOOF is awesome stuff!

Keep it up guys as this is MORE THAN USEFUL information.

 

Thanks to all who are exploring and posting their work here.

This is the kind of stuff that this Tutorials Forum is made for.

 

Rodney

[Xtaz]

Hi everybody ... thx for all comments about MUFOOF.... it have a big potencial in your hands

I did some attempts trying to recreate a rpm change ... now i dont have time to explain it but ill upload a MOV to you see ...

the prop begins at rpm X , in 7sec it changes to 2*X then in sec 12 it changes to X/2

http://xtaz.com.br/projetos/proprpm.mov

25 sec / 2.31Mb

 

tonite i write how i did it !!! no expressions

[aaver]

[...]tonite i write how i did it !!! no expressions

Yes I agree, it's much easier to do this without expressions. The trick is to realize that the rpm is proportional to the slope of the absolute rotation angle. With that in mind, it shouldn't be too hard to control the rpm in your animations even if it would be even more convenient to have a pose slider doing it, was it possible.

[oakchas]

tonite i write how i did it !!! no expressions

 

My guess is that you projected a still onto a plane and rotated that. (plane was attched to back of nose cone) The effect "almost" works...

 

It would be interesting to see in front of a fuselage, in a daylight scene... This might create a problem with how to make the blades appear where there is a black spot now.

 

The "shadow" (dark reflection) on top right of the nose cone (I think) is what gives it away.

 

But I'm just guessing! And it's still a pretty cool look! Something that could be added to a still to make it look like a photo.

 

I wanna know if my guess is good! What'd I win?

[Xtaz]

Hey charlie .. the proccess is very, very basic ... the MUFOOF base itself in the model´s position, then if you change it the MUFOOF will follow it... what I did was create a secondary action and I added it to model setting Blend Method to ADD in this action i didnt change any others parameters

 

SECONDARY ACTION :

 

select prop bone and set new Z rotation positions :

frame-----Z

15 _____ 90

30 _____ 180

45 _____ -90

60 _____ 0

75 _____ 90

90 _____ 180

105____ -90

120 ____ 0

 

I had to set 2 turns to complete the curve correctly

 

select all frames and shrink them

 

 

 

select frames copy , select the frame that you want to paste it .... paste ....

the amount of copy depends on the duration of the animation

 

 

lets change rpm in frame 120 .. select all frames before frame 120

 

 

shrink them

 

 

action2 animation:

http://xtaz.com.br/projetos/propaction2.mov

6 sec / 555 Kb

 

It would be interesting to see in front of a fuselage, in a daylight scene... This might create a problem with how to make the blades appear where there is a black spot now.

 

I did a test using this tutorial, and i used it whit a daylight ... you can check results....

http://xtaz.com.br/projetos/109propeler2.mov

11 sec / 1.02 Mb

 

I hope you help me to perfect MUFOOF

 

thanks all..

 

comments ??!?! Sugestions ?!?!?

[Dagooos]

I don't have any suggestions but it's a nice effect looks very realistic

[Xtaz]

Hi spliners lets continue .......

 

Project IV - Simulating Global Ilumination

 

now ill show how to fake GI using MUFOOF ....

 

- create a simple spiral model like the example and set a new NULL bone in the center.. or download it HERE

 

 

- create model whit sun light named sunlight

 

- create a new sunlight action then import spiral, right click sunlight bone and choose:

NEW CONSTRAINT - PATH ..... point to spiral..... go to frame #100 and change ease to 100% ( constarint properties )

NEW CONSTRAINT - AIM AT .... point to spiral´s NULL bone

 

- create a cho, delete all lights then import sunlight and it action setting CYCLE LENGHT=1 and REPEAT=100

put some objects below the spiral to test light.

 

 

- in TOOLS -> OPTION -> RENDERING adjust MULTI-PASS=100 and MOTION-BLUR=100% ....

 

RENDER IT !!!!

 

 

Thats it .....

[Stuart Rogers]

now ill show how to fake GI using MUFOOF ....

There are definite signs of multiple (virtual) lights there - I think you might need more lights (maybe on their own spirals, each with a rotation offset around the y axis). But it's definitely a promising approach!

Thats it .....

Is that all you have to say?! My project was going to take weeks to render as it is with a simple three light set up, and now you tempt me to use your MUFOOF - I'll be rendering until doomsday, damn it!

[amarillospider]

Fabulous Tricks XTAZ! Thanks! Keep them coming!

 

-Alonso

[serg2]

Thank.

Good idea, Xtaz - do you experiments with Penumbra parametr?

[heyvern]

Can I ask a stupid question?

 

Okay thanks...

 

How is the spiral light trick... different from other "random" single light actions or "multiple" light sky light rigs?

 

The trick is getting a light in sort of a random spot for each pass... so... is this just another way to do it?

 

Vernon "!" Zehr

[Stuart Rogers]

How is the spiral light trick... different from other "random" single light actions or "multiple" light sky light rigs?

I think the idea is to get the light to pass uniformly over the entire sky dome. It doesn't have to be a spiral, but with a spiral it should be relatively easy to see that you do get uniform coverage. This could be taken a step further - by modifying the spiral you could hint at non-uniform global illumination. For example, if you had a scene with half your sky clear, and the other full of brooding clouds, you could tilt the spiral over, and maybe distort it towards the brighter parts of the sky...

EDIT: By distorting the spiral, I have in mind making the coils closer together at the point where the sky is to be brighter. This could also be done by not dstorting the spiral but the play with the 'ease', so that the light lingers longer at the place where the sky should be brighter.

EDIT EDIT: I think my cold's starting to make me babble ... time for an early night, I think.

[Xtaz]

 

How is the spiral light trick... different from other "random" single light actions or "multiple" light sky light rigs?

 

 

Hi Vernon ..... I believe that the main difference among the methods, is the render time ... i did a small test using Yves´skylight method against MUFOOF

I used 144 passes to simulate the number of existent lights in the Yves´ dome...that increased the acuracy of the MUFOOF image.. check it ...

 

 

using yves´s dome whit:

multi-pass OFF render=2'52"

2 passes render= 2'55"

9 passes render=7'12"

[heyvern]

Uh...

 

Now I am confused... I have been using multipass with Yves's skylight rigs...

 

I downloaded that big pile of skylights... and I usually use the "8" version. I love the results.

 

Vernon "!" Zehr

[aaver]

I believe that the main difference among the methods, is the render time

 

I disagree. With all due respect I think the main difference is in quality. With your spiral method, you will never get rid of the banded look unless you have a very large light together with multiple rays from each light, but probably not even then. Your method with a moving light within a single frame was first thought of some years ago and John Henderson addressed the banding problem in his application Scycast. Jeff Lee has also done this in his JSL Skylight Rig Utility

 

Although the methods are slightly different, a static rig like Yves' is capable of producing as good results as a dynamic rig.

 

Yves', John's and Jeff's rigs are designed to cast uniform light from the whole dome, while your rig casts much more light from above than from the sides. If you like to compare the methods, you have to remove most of the lower lights in Yves' rig and you really don't need all the 148 lights to light this scene. My suggestion for a quick comparison would be to use the rig with 8 lights and lift the whole rig so the top light is maybe 150% higher. Set lights to cast 2 shadow rays and render with 9 passes. For even better quality, use the 20 lights rig and 16 passes. My guess is that with the same quality, you will find Yves' rig to be much faster.

[DrRIEGER]

Hello everybody,

 

I'm testing a lot the differents options for a professionnal work with a lot of animations (surgical techniques).

 

I follow this topic with attention, and I have 2 very important questions :

Q1 :

Under the Multipass subTab on the Rendering Settings Panel, what is the action of SOFTEN ?

(the online manual is not clear for me)

I made some tests and set soften ON : it increase A LOT the rendering time with no real difference (to my eyes) vs soften set OFF.

 

Here are my settings :

 

Export EXR in D1-PAL

Buffers : alpha ON

lights ON each light on its layer

Render FINAL

Multipass : 25 , Soften OFF ... then ON to get a comparison.

Motion Blur : 100 %

 

Q2 : what is the incidence of the Motion Blur percentage in this Dynamic Light Rig ?

Q2a : Considering I have a STATIC scene with a flying camera.

Q2b : Does the reply be the same if some elements in the scene have movements ?

Q2c : what is the consequence on the rendering as MotionBlur percentage decrease ? (I just notice that if No Motion is set, this Dynamic Light Rig does not work).

 

Many thanks for the replies.

 

Alain

[oakchas]

I did a test using this tutorial, and i used it whit a daylight ... you can check results....

http://xtaz.com.br/projetos/109propeler2.mov

11 sec / 1.02 Mb

 

I hope you help me to perfect MUFOOF

 

thanks all..

 

comments ??!?! Sugestions ?!?!?

 

Marcos!

 

A-W-E-S-O-M-E!!! I don't think I can help you perfect it, the propeller looks "dead on" to me in the movie.

 

This is really neat Marcos... I'm copying this whole topic to my PC! It has been very useful! I don't understand the differences/ benefits of the lighting situation (sklight rig vs spiral light)... but I'm not ready to understand that yet.

[Odog2020]

I have to say that I love messing with this Mufoof stuff. But why I am here is this. I did the example about the room, and came away with favorable results, except for having my floor to reflective.[attachmentid=12646]

 

Then I went ahead and tried the 'Spiral' sun light thing, and ended up with a dreadfull picture.

 

[attachmentid=12647]

 

Can you tell me if it is because I have a top on the bus stop, or did I not follow instructions correctly??

 

Thanks

[Xtaz]

Hi Odog ....

what type of light did you use ? in my example i used sun light. Check intensity and width values and increase them ... I believe the problem will be solved ...

[8up]

wow

[John Bigboote]

6 years later... still good stuff!