Found 41 posts.
Search results Show results as topic list.
Houdini Lounge » COPs in H10...
- houdiniuser42
- 42 posts
- Offline
Houdini Lounge » COPs in H10...
- houdiniuser42
- 42 posts
- Offline
twodWill this work under Linux? and if so, just for curiousity's sake, how? not via DirectX directly, right?
The major improvement to COPs in H10 is the ability to use the video card to assist some compositing operations
…
This works with DX10 capable cards …
and, although a little off topic, but are the newer ATI cards ( Radeon 3xxx & 4xxx series ) behaving better with Houdini?
Technical Discussion » Colored Shadows
- houdiniuser42
- 42 posts
- Offline
markfgillis
Am trying to color ray traced shadows… can only get shades of black.
I remember this being trivial in earlier releases… I think by using a shadow or light shader with shadow color parameter fields.
How do we do this now?
Set the opacity to something other than grey (i.e. the Of variable inVOPs).
If you apply a basic surface and set the opacity to something other than grey you should get colored shadows (though it might not be intuitive - if you set the opacity to 1,0,0 the surface will be opaque in red, passing the green/blue light through - a cyan shadow).
is this an improvement on Zerouni's examples? sounds similar to example one, but more simple?
============================
and specifically in regard to:
markcouldn't a Complement VOP be wired in to make the shadow color picking easier?
… (though it might not be intuitive - if you set the opacity to 1,0,0 the surface will be opaque in red, passing the green/blue light through - a cyan shadow).
============================
anyway, the 3 examples in a bit more detail than my earlier post:
taken from Zerouni's - Houdini on the Spot book. highly worth it. hope to see another one for H10. :wink:
pg 201 - 202
Example 1 uses 2 light sources, with one casting shadows and the other not… then you end up w/ the same “shadow color math” scenario that mark mentions.
Example 2 uses a custom shadow shader done in VEX. *shudders* i'll just recommend going to the book for this one. pg 202
Example 3 he states, "modify the surface shader so that the object becomes transparent only for shadows." again uses some VEX, but it's cleaner and leaner.
hope that helps some. good luck.
Technical Discussion » Colored Shadows
- houdiniuser42
- 42 posts
- Offline
i believe Zerouni gives 3 examples in the Houdini On the Spot book…
time being short, i don't have time to fetch the book and go over the examples given.
good luck.
time being short, i don't have time to fetch the book and go over the examples given.
good luck.
Technical Discussion » importing voodoo or icarus into houdini
- houdiniuser42
- 42 posts
- Offline
perfected??? uhhh errrr… can't speak for Voodoo, but Icarus does create a houdini .clip file…
in CHOPs put down a File chop with the .clip file
then attach an Export chop.
in the “node” param have it drive your camera (usually /obj/cam1)
in the path, i put: tx ty tz rx ry rz px py pz sx sy sz aperture focal
as far as i can tell, it appears that the names in the .clip file were ignored. at least in the export node… so the order is important and you can't skip params you don't need.
as for the .geo File sop… don't forget to turn on “display points” :wink:
back to Voodoo, given that a .clip file is a text file, and having output from Icarus to compare , it seems totally possible to write a python script to convert the .txt output from voodoo into a nice Houdini friendly (.clip) format… any takers?
good luck.
in CHOPs put down a File chop with the .clip file
then attach an Export chop.
in the “node” param have it drive your camera (usually /obj/cam1)
in the path, i put: tx ty tz rx ry rz px py pz sx sy sz aperture focal
as far as i can tell, it appears that the names in the .clip file were ignored. at least in the export node… so the order is important and you can't skip params you don't need.
as for the .geo File sop… don't forget to turn on “display points” :wink:
back to Voodoo, given that a .clip file is a text file, and having output from Icarus to compare , it seems totally possible to write a python script to convert the .txt output from voodoo into a nice Houdini friendly (.clip) format… any takers?
good luck.
Houdini Lounge » Intermediate Houdini Class
- houdiniuser42
- 42 posts
- Offline
A couple extra points:
The class meets “at the Academy of Entertainment and Technology (AET), a satellite campus of Santa Monica College located at 1660 Stewart St., Santa Monica, California.
At the AET campus, we provide FREE PARKING for currently enrolled students. ”
( from: http://www.smc.edu/apps/Comm.asp?Q=18 [smc.edu] )
also, if one wishes to save $30, there are optional fees that can be waived. Contact the Bursar's Office.
The class meets “at the Academy of Entertainment and Technology (AET), a satellite campus of Santa Monica College located at 1660 Stewart St., Santa Monica, California.
At the AET campus, we provide FREE PARKING for currently enrolled students. ”
( from: http://www.smc.edu/apps/Comm.asp?Q=18 [smc.edu] )
also, if one wishes to save $30, there are optional fees that can be waived. Contact the Bursar's Office.
Houdini Lounge » more new tutorials,
- houdiniuser42
- 42 posts
- Offline
zamousThere is a pretty good SHOP tutorial on 3DBuzz … it's from their older video tutorials and from what I can tell is strictly streaming now, but still useful.
Any chance of doing tutorials for SHOPS? I have a really hard time wrapping my head around shaders in Houdini–albeit I am still fairly new to 3d in general.
Look for them Under the “Effects (v6.1)” section.
- (1) Tornado: Basic Sprite Rendering
(2) Tornado: Basic Shaders
Good luck.
Technical Discussion » Intel Dual/Quad Core and Houdini
- houdiniuser42
- 42 posts
- Offline
somehow i wasn't informed about page 2…
thanks for the revisited discussion and the advice.
the Q6600 has been around for a year!!! this bothers me, but being financially handicapped doesn't help either. the weird thing w/ the Intel quads is the newer procs are a little faster, but have less cache… until you get up to the high end consumer stuff. all of this may be moot when Nehalem finally debuts… which offers those of us in my situation w/ something else to drool over and the hope it'll drive the older tech lower.
i've dreamed of an open system where being able to insert a card w/ another proc or 2 would be a snap… think mini rack beside your desk… grows with you. i suppose this is a Boxx or something. again, affordability plays a role in this.
more train of thought meandering: Rackable systems has received praise for being more energy efficient w/ a setup that takes 2 half depth motherboards and splits one very energy efficient psu amongst them.
when will laptops go quad??? has anybody ever seen someone bring their home pc into a library?
lastly, has anybody graphed how much wasted cycles go to those learning software? i'm sure this graph is growing exponentially over time as computers get faster and software gets bigger. :wink:
42
Edit: according to wikipedia… http://en.wikipedia.org/wiki/List_of_future_Intel_Core_2_microprocessors [en.wikipedia.org] intel will launch a mobile quad in Q4 2008.
game on!
thanks for the revisited discussion and the advice.
the Q6600 has been around for a year!!! this bothers me, but being financially handicapped doesn't help either. the weird thing w/ the Intel quads is the newer procs are a little faster, but have less cache… until you get up to the high end consumer stuff. all of this may be moot when Nehalem finally debuts… which offers those of us in my situation w/ something else to drool over and the hope it'll drive the older tech lower.
i've dreamed of an open system where being able to insert a card w/ another proc or 2 would be a snap… think mini rack beside your desk… grows with you. i suppose this is a Boxx or something. again, affordability plays a role in this.
more train of thought meandering: Rackable systems has received praise for being more energy efficient w/ a setup that takes 2 half depth motherboards and splits one very energy efficient psu amongst them.
when will laptops go quad??? has anybody ever seen someone bring their home pc into a library?
lastly, has anybody graphed how much wasted cycles go to those learning software? i'm sure this graph is growing exponentially over time as computers get faster and software gets bigger. :wink:
42
Edit: according to wikipedia… http://en.wikipedia.org/wiki/List_of_future_Intel_Core_2_microprocessors [en.wikipedia.org] intel will launch a mobile quad in Q4 2008.
game on!
Technical Discussion » Intel Dual/Quad Core and Houdini
- houdiniuser42
- 42 posts
- Offline
Late, but still pertinent!!
Barcelona came and well…
JColdrick, did you get your “new” computer? Have you been happy?
and now that it is more than 6 months old would you like to pass it along? :wink:
—————————————————
continuing from the thread: someone mentioned an upgrade is usually 2x faster. is this the norm? how about for a home pc?
trying to get the most bang for the buck, money is an issue. that said, i'm thinking that the 2nd upgrade (or addition / replacement) will happen much sooner as new technology will most likely make anything i get now very much obsolete.
so, consumer level single socket pc, 2-4gb ram, with possibility of going “starving artist” hd res: dual, tri, quad?
side q: how long should an average frame take to render on a home pc? reworded: how long can the average houdini artist wait for one frame to render (at home)?
E7200 vs. Phenom X3 8450 ( http://www.legionhardware.com/document.php?id=751&p=0 [legionhardware.com] )
but am also considering (for the extra cost savings while i sit scratching me head… aka “thinking”): the 45W AMD Athlon X2 4850e 2.5GHz
it's already known this would be rather slow compared to the others. does the 400mHz divisible thing apply here as well? would underclocking it speed it up? yes, i realize what i just said… but to 2.4gHz… lose 0.1 gHz, but potentially gain the lost memory speed.
on the E7200, ppl have said it has limited cache. but cache is transistors and those guys give off heat… heat costs money, but so does time. thoughts on cache?
or would a quad be more applicable now?
lastly, would probably get no less than 2gb of ram with (eventual) max being 4gb. how does this correlate when scaling cores?
thanks in adv.
Barcelona came and well…
JColdrick, did you get your “new” computer? Have you been happy?
and now that it is more than 6 months old would you like to pass it along? :wink:
—————————————————
continuing from the thread: someone mentioned an upgrade is usually 2x faster. is this the norm? how about for a home pc?
trying to get the most bang for the buck, money is an issue. that said, i'm thinking that the 2nd upgrade (or addition / replacement) will happen much sooner as new technology will most likely make anything i get now very much obsolete.
so, consumer level single socket pc, 2-4gb ram, with possibility of going “starving artist” hd res: dual, tri, quad?
side q: how long should an average frame take to render on a home pc? reworded: how long can the average houdini artist wait for one frame to render (at home)?
E7200 vs. Phenom X3 8450 ( http://www.legionhardware.com/document.php?id=751&p=0 [legionhardware.com] )
but am also considering (for the extra cost savings while i sit scratching me head… aka “thinking”): the 45W AMD Athlon X2 4850e 2.5GHz
it's already known this would be rather slow compared to the others. does the 400mHz divisible thing apply here as well? would underclocking it speed it up? yes, i realize what i just said… but to 2.4gHz… lose 0.1 gHz, but potentially gain the lost memory speed.
on the E7200, ppl have said it has limited cache. but cache is transistors and those guys give off heat… heat costs money, but so does time. thoughts on cache?
or would a quad be more applicable now?
lastly, would probably get no less than 2gb of ram with (eventual) max being 4gb. how does this correlate when scaling cores?
thanks in adv.
Houdini Lounge » HW req for 9.5, ATI / XGP, GPGPU
- houdiniuser42
- 42 posts
- Offline
twod: re: multi-GPU solutions
your knowledge and ability to be succinct far surpasses mine, but that won't stop me from blabbing further: I once saw a ppt online that talked about GPGPU, of which I cannot find (of course).
anyway, wikipedia (“BrookGPU”) goes along with everything you said and adds 2 “gotchas”:
1- it appears to be ATI only which unless the driver issues get resolved might be diminishing returns even on those that could supposedly use it.
2- currently, at least per wikipedia *ahem*, the DirectX implementation is more efficient than OpenGL.
on the other front, how many Tesla's are you tinkering with?
your knowledge and ability to be succinct far surpasses mine, but that won't stop me from blabbing further: I once saw a ppt online that talked about GPGPU, of which I cannot find (of course).
anyway, wikipedia (“BrookGPU”) goes along with everything you said and adds 2 “gotchas”:
1- it appears to be ATI only which unless the driver issues get resolved might be diminishing returns even on those that could supposedly use it.
2- currently, at least per wikipedia *ahem*, the DirectX implementation is more efficient than OpenGL.
on the other front, how many Tesla's are you tinkering with?
Houdini Lounge » HW req for 9.5, ATI / XGP, GPGPU
- houdiniuser42
- 42 posts
- Offline
i'm on a desktop… considering a laptop to *ahem* be able to “get away” and find a “quiet corner.” (background music = crying and being told my computer isn't mine even though it's older than this little, but cute 22mo old person)
of course that doesn't guarantee i won't get distracted and/or procrastinate somewhere else.
back to the topic, DDR is more expensive than DDR2. would make some sense to apply that money to a faster pc w/ cheaper (but faster) and therefore *more* ram.
ps - shortly after my first visit to Sidefx (LA), i found out i couldn't run H8… 512mb ram, but IGP… opted for a dedicated GPU to regain the whole 512mb rather than up the ram. i still stand by the decision, it is just taking me a long time to decide what to upgrade to and how to protect it from little hands. :wink:
pps - when starting this thread, i was considering starting a technical thread, as well, about how to speed up rendering - caching, saving geo out and loading back in, mantra tweaks, etc. but for now, I'm leaning more towards getting a qualifying pc. ops: :wink:
ppps - any donations and/or free babysitting *greatly* appreciated!
of course that doesn't guarantee i won't get distracted and/or procrastinate somewhere else.
back to the topic, DDR is more expensive than DDR2. would make some sense to apply that money to a faster pc w/ cheaper (but faster) and therefore *more* ram.
ps - shortly after my first visit to Sidefx (LA), i found out i couldn't run H8… 512mb ram, but IGP… opted for a dedicated GPU to regain the whole 512mb rather than up the ram. i still stand by the decision, it is just taking me a long time to decide what to upgrade to and how to protect it from little hands. :wink:
pps - when starting this thread, i was considering starting a technical thread, as well, about how to speed up rendering - caching, saving geo out and loading back in, mantra tweaks, etc. but for now, I'm leaning more towards getting a qualifying pc. ops: :wink:
ppps - any donations and/or free babysitting *greatly* appreciated!
Houdini Lounge » HW req for 9.5, ATI / XGP, GPGPU
- houdiniuser42
- 42 posts
- Offline
thank you both for such detailed replies.
Rather than commenting further and saying things like: After I'm elected president - my platform being unified GPGPU - i'll make certain to make it happen. or commenting about how hiring me would be the best investment ever, or even how sidefx should simply buy Intel, kidnap the inventor of the Connection Machine and come out w/ the “Houdini Box” (don't forget to use lots of LED's) …
“one processor for each particle.” http://www.karlsims.com/particle-dreams.html [karlsims.com]
Yes, rather than saying anything like *that* :wink: I'll simply state that it is possible to run H9 w/ 512mb of ram. (not that one wants to! but…)
Oh and for probbins, on the laptop(s) was the Nvidia graphics a consumer level GPU or was it a Quadro? with dedicated ram?
Rather than commenting further and saying things like: After I'm elected president - my platform being unified GPGPU - i'll make certain to make it happen. or commenting about how hiring me would be the best investment ever, or even how sidefx should simply buy Intel, kidnap the inventor of the Connection Machine and come out w/ the “Houdini Box” (don't forget to use lots of LED's) …
“one processor for each particle.” http://www.karlsims.com/particle-dreams.html [karlsims.com]
Yes, rather than saying anything like *that* :wink: I'll simply state that it is possible to run H9 w/ 512mb of ram. (not that one wants to! but…)
Oh and for probbins, on the laptop(s) was the Nvidia graphics a consumer level GPU or was it a Quadro? with dedicated ram?
Houdini Lounge » HW req for 9.5, ATI / XGP, GPGPU
- houdiniuser42
- 42 posts
- Offline
Three (or more) Q's:
1- are the HW req for 9.5 decided already? and they are?
2a- are current generation consumer level ATI cards and drivers (namely: 38xx and 48xx) working better w/ H?
2b- will the “new” XGP (AMD's external graphics for laptops) compatible w/ H? they claim 4Gb/s external PCIe w/ first generation only powering external screens.
3- will 9.5 and/or near term future generations use GPGPU (CUDA, Steam)? again, thinking more on a consumer level here - for those of us that can't afford the best machine, but still looking for a good boost in perf for the buck, but any info on what's happening at the apex ( pinnacle / acme ) of Houdini goodness appreciated as well.
4.2- what are the specs for the laptop used in recording the *very* helpful and appreciated sidefx video tutorials?
thanks in adv.
1- are the HW req for 9.5 decided already? and they are?
2a- are current generation consumer level ATI cards and drivers (namely: 38xx and 48xx) working better w/ H?
2b- will the “new” XGP (AMD's external graphics for laptops) compatible w/ H? they claim 4Gb/s external PCIe w/ first generation only powering external screens.
3- will 9.5 and/or near term future generations use GPGPU (CUDA, Steam)? again, thinking more on a consumer level here - for those of us that can't afford the best machine, but still looking for a good boost in perf for the buck, but any info on what's happening at the apex ( pinnacle / acme ) of Houdini goodness appreciated as well.
4.2- what are the specs for the laptop used in recording the *very* helpful and appreciated sidefx video tutorials?
thanks in adv.
Houdini Indie and Apprentice » VOPS -> vopsurface OGL Diffuse expression
- houdiniuser42
- 42 posts
- Offline
tstexyes, i got carried away w/ “proceduralism” and tried to fetch the output of the SHOP and put it into the color parameter of the OpenGL tab. (edit: I thought it was working as the color that was being output was the same as the default color of one of the ring node's color parameters. unfortunately, i didn't quite catch that the same is achieved by setting r = 0).
Not quite sure what you are trying to achieve.
I'm going to assume that you have created a SHOP surface definition using a VOPnet. After that I'm not quite sure. This: ch(“/shop/material1/vopsurface1/output1”) makes me concerned that you are trying to call a surface shaders output from another nodes parm - which is not possible.
What exactly are you trying to do?
(i left the message rather than delete it for future amusement and edification.)
the hope comes from the way the VEX Super Material appears to work. is there a tutorial on getting an OpenGL / viewport shader to approximate (preview) what will be rendered (in realtime)?
Houdini Indie and Apprentice » Fun with Lsystem
- houdiniuser42
- 42 posts
- Offline
I like it.
Could you give us some stats?
Could you give us some stats?
- How long did it take to render?
What are your system specs?
How many boxes?
Are you planning to do any animation w/ L-systems?
Houdini Indie and Apprentice » VOPS -> vopsurface OGL Diffuse expression
- houdiniuser42
- 42 posts
- Offline
I'd like to know how what expression will give me the rgb commponents of the color being output by my VOP network.
so far, i've used: ch(“/shop/material1/vopsurface1/output1”)
but, from what i can tell it only returns a float… and i need to be able to access the rgb of the color vector.
my vop network is rather simple:
*rings* piped into bias of *colormix*
piped into *lightingmodel*
piped into *output1*
thanks in adv.
so far, i've used: ch(“/shop/material1/vopsurface1/output1”)
but, from what i can tell it only returns a float… and i need to be able to access the rgb of the color vector.
my vop network is rather simple:
*rings* piped into bias of *colormix*
piped into *lightingmodel*
piped into *output1*
thanks in adv.
Houdini Lounge » Mac OSX Port?
- houdiniuser42
- 42 posts
- Offline
Wolfwood
(snip)
Also the expectations of computer programs is different depending on when you started using a computer. Just a short decade ago there was rarely good (if any) for software. You'd have to spend hours fighting it and learning by trial and error. Now a days, thankfully, software makers have started putting a lot of resources into documenations and video tutorials. Which is great, but the drawback is it doesn't force people to spend hours learning by trail and error. Instead of exploring the application they'll just ask for more tutorials. (Which is a completely valid request, but given finite resources of the software makers, don't let that hold you back.)
the fact that i've not posted (much?) before most likely won't help my comment, but i've been dying to say it somewhere.
Before I commence, I must state that i agree with you!
The (at least one) “problem” is that us newcomers do not have the history with the application. So, to look at an app that is on version 9.1 and have no experience with it, we see a gazillion options and a plethora of knowledge to digest. Furthermore, when a “new feature” is introduced in a new version of an app, it has usually been carefully thought out by those that have used the sw before. So, to those individuals the additional tool makes total logical sense, but to someone starting from scratch, it really is a, “huh? wha?” experience. (In some cases we might not even know there was a “problem” that needed fixing!)
This is even more apparent when a “pro” writes a book or a tutorial and skips steps. (Note: nothing off to the top of my head, just speaking generally here.) They don't do it to be cruel or even because they are rushed… they have just become so good at what they do that they do it naturally … like, breathing and therefore don't remember to slow it down and spoon feed it to us noobies! ;o)
Honestly, I've read a few “rant” threads these past few days and I think it is amazing that the Hard Core / Old Skool Posse maintain their composure and continue to attempt to help some of the denser of us. FWIW, this is appreciated… greatly.
Not only will this win over more ppl, it will instill this pattern of helping. This in turn will foster more growth and … well, I'm sure we can all agree that “growth” in and of itself is a good thing, yes? Best be a moving target or one will get caught or surpassed.
You have a fine product (albeit a bit different than most if not all the competition) and an excellent support group that I hope to be counted as one, someday… soon and for the rest of my life. (here's looking at you, kid.) ;o)
Lastly, if you hadn't already dispatched the hounds on me yet, would you give me a head start even if I told you I don't own or use a Mac?
Technical Discussion » Expression / Math question - animating the fireball
- houdiniuser42
- 42 posts
- Offline
Whoever came up w/ the fireball expression in the expression cookbook, i'm not worthy!
that said, as silly as this may sound, could someone point me in the right direction as to *where* to put $F in that formula to make it animate smoothly?
perhaps i was close … once, but my methods for learning are usually rather random.
related to this q is the speed of the playback. what's the workflow for messing around w/ algorithms on a slowish pc? and, are expressions cookable? … meaning is it necessary and/or wanted to do something like this to make the rendering as fast as possible.
thanks in adv. peace & 42
that said, as silly as this may sound, could someone point me in the right direction as to *where* to put $F in that formula to make it animate smoothly?
perhaps i was close … once, but my methods for learning are usually rather random.
related to this q is the speed of the playback. what's the workflow for messing around w/ algorithms on a slowish pc? and, are expressions cookable? … meaning is it necessary and/or wanted to do something like this to make the rendering as fast as possible.
thanks in adv. peace & 42
Technical Discussion » Animation question (the Stingray tutorial)
- houdiniuser42
- 42 posts
- Offline
circusmonkey
The simple way to do this would be to use a Sin and Cos function in the transform of your geometry object .
In the TX channel type sin($FF*6)*6 and in TY channel type cos($FF*6)*6
you will then have a looping circular motion for your sting ray
R
that crossed my mind too, but the math gets tricky. we need to be careful if we want it to loop correctly for “n” number of frames.
furthermore, if a circular and not elliptical orbit are required, the outside multiplier should be the same and therefore wouldn't creating our own variable and slider for the radius be helpful? i suppose 2 sliders could be used to allow elliptical orbits. hmmm … i've seen it in a video tutorial, but …. OK in help i found the “expression cookbook” worth a read.
RMB on the node, select Edit Paramater Interface and add a Float or Integer. rename it and remember the name! I chose Integer and named it “radius.” You can limit the range by checking “range”…
now in the translate params instead of sin($FF*6)*6 use sin($FF*6)*ch(“./radius”) make the same change for the “cos” expression.
think of houdini as an OS directory tree thingie. the expression cookbook helped me here. and, as memory slowly returned, you can drag and drop params onto other parameter textboxes. using my “radius” param, one can click and drag from the label (I used the same name “radius”) and drop it onto the param textbox you want it in. in this case translate X and translate Z. the path to the variable will be slightly different depending on whether you use Channel Ref or Relative Channel Ref.
note: you might want to stop the animation to get accurate values in your new param. sliding it while it was looping worked visually but the param didn't update. i'm on 9.0.725
now, who's up for getting the object to be tangental to the curve?
a possibly easier route and one that takes care of the tangental problem would be to translate the stingray off of the origin and then move the pivot in the opposite direction and then do a rotate on the Y axis. then the math would fade away and only $FF would need to be put in the rotate Y parameter.
peace & 42
Houdini Lounge » tutorials?
- houdiniuser42
- 42 posts
- Offline
digital tutors has 8 online vids to get ppl started.
3dbuzz has VTMs for an older version, but still very useful and they're coming out with new ones “any day now.”
and if it hasn't been hammered away enough, there are a ton of tutorials, including video tutorials, on sidefx's website. check the tutorial section, H9 blog, and old school blog.
peace & 42
3dbuzz has VTMs for an older version, but still very useful and they're coming out with new ones “any day now.”
and if it hasn't been hammered away enough, there are a ton of tutorials, including video tutorials, on sidefx's website. check the tutorial section, H9 blog, and old school blog.
peace & 42
-
- Quick Links