Random offset image sequence based on particle ID

[olivier jeannel]

Hi gang,

I know how to set image-texture based on particle ID (image01 set to 
particle ID 01, image02 set to particle ID 02,etc.)


I'd like to have one image sequence (a video), with time offseted based 
on Particle ID. Example :
Particle 01 gets sequence from 0 to 100, particle 02 gets sequence from 
50 to 150, etc.


How do I achieve that ? I'd really like to avoid having 100 various 
videos filling my hard drive...


Thanks !

Re: Python on Linux

[Xavier Lapointe]

Hmm, but this can have nasty side effects (recently experienced it).
Someone pointed out to me that you can actually pass the *env *keyword
argument in the subprocess call and pass a *copy *of the current
environment instead (with necessary modifications).


On Fri, Sep 14, 2012 at 12:33 AM, Alan Fregtman wrote:

> Hey X,
>
> About the 3rd item, you just need to make sure the system libraries are
> loaded before XSI's own by putting them first in the 
> *LD_LIBRARY_PATH*environment variable.
>
> Here's a workaround sample code to start subprocesses in Linux that worked
> for me last time:
>
> import osimport subprocess
>
> inLinux = Application.Platform.startswith("Linux")if inLinux:
> exe = XSIUtils.BuildPath( pDir, executable)
> ldEnv = "LD_LIBRARY_PATH"
> sysLibDir = r"/usr/lib64"
> ld_oldVal = os.environ[ldEnv]
> os.environ[ldEnv] = sysLibDir+":"+os.environ[ldEnv]
>
> command = r'/path/to/awesome/tool blablabla arguments here'
> proc = subprocess.Popen( command.split(), stdout=subprocess.PIPE, 
> stderr=subprocess.PIPE )
> out, err = proc.communicate()print "stdout: %s" % outprint "return code: %s" 
> % err
> if inLinux:
> # Reset to old values like the good samaritan coder we are. :p
> os.environ[ldEnv] = ld_oldVal
>
>
> As a matter of fact, it was thanks to you that I figured this one out at
> the time. :p
>
> Cheers,
>
>-- Alan
>
>
> On Tue, Sep 11, 2012 at 8:19 PM, Xavier Lapointe  > wrote:
>
>>
>>- You might have to import __*future*__ in order to have access to
>>some statement like *with*
>>- The @property decorator does not work
>>- Keep in mind that if you use a newer version of linux that the one
>>supported, you might have trouble starting subprocesses, since the libc
>>version coming with soft will override the one on the system (but it's
>>still possible).
>>- For plugins it should be too bad, but for libs, I would consider
>>running some unittest if I were you
>>
>>
>> Cheers
>>
>>
>


-- 
Xavier

Re: Pixel 7 Vienna

[Felix Geremus]

I'll be there for the first time. Definitely interessted in meeting some 
people from the list in home sweet home vienna


Cheers,

Felix

Am 17.09.2012 21:58, schrieb Stefan Kubicek:
Friends of mine are organizing the Vienna Siggraph Chapter aka "Pixel 
Vienna" again from Oct 05 - 07.2012.


The schedule has just been posted, some interesting speakers again 
this year:

http://www.pixelvienna.com/7/program/schedule

Last years video can be seen here (including Eric Mootz demoing his 
ICE tools at ~1:40):

https://vimeo.com/49339044

Would love to know if ne1 from this list intends to go and wants to 
meet there.


Cheers,

Stefan

Re: Pixel 7 Vienna

[carl callewaert]

It is a very good event and great parties! Last year I was one of the
speaker.

c

On 2012-09-17 4:58 PM, "Stefan Kubicek"  wrote:

>Friends of mine are organizing the Vienna Siggraph Chapter aka "Pixel
>Vienna" again from Oct 05 - 07.2012.
>
>The schedule has just been posted, some interesting speakers again this
>year:
>http://www.pixelvienna.com/7/program/schedule
>
>Last years video can be seen here (including Eric Mootz demoing his ICE
>tools at ~1:40):
>https://vimeo.com/49339044
>
>Would love to know if ne1 from this list intends to go and wants to meet
>there.
>
>Cheers,
>
>   Stefan
>
>
>
>
>
>-- 
>---
>Stefan Kubicek   Co-founder
>---
>   keyvis digital imagery
>  Wehrgasse 9 - Grüner Hof
>  1050 Vienna  Austria
> Phone:+43/699/12614231
>--- www.keyvis.at  ste...@keyvis.at ---
>--  This email and its attachments are
>--confidential and for the recipient only--
>

Pixel 7 Vienna

[Stefan Kubicek]

Friends of mine are organizing the Vienna Siggraph Chapter aka "Pixel Vienna" 
again from Oct 05 - 07.2012.

The schedule has just been posted, some interesting speakers again this year:
http://www.pixelvienna.com/7/program/schedule

Last years video can be seen here (including Eric Mootz demoing his ICE tools 
at ~1:40):
https://vimeo.com/49339044

Would love to know if ne1 from this list intends to go and wants to meet there.

Cheers,

Stefan





--
---
Stefan Kubicek   Co-founder
---
  keyvis digital imagery
 Wehrgasse 9 - Grüner Hof
   1050 Vienna  Austria
Phone:+43/699/12614231
--- www.keyvis.at  ste...@keyvis.at ---
--  This email and its attachments are
--confidential and for the recipient only--

RE: CLSID error and XSI base

[Matt Lind]

Yes, I tried it, but it's not as easy as that.

In the case of shaders, not only do you need the GUID of the shader, but also 
the GUID of each and every parameter of that shader as it's the parameter GUIDs 
that Softimage stores to remember a shader's connections within the rendertree. 
 You also need a functioning .dll so when Softimage parses the .spdl on startup 
it'll try to execute/load the .dll that it points to.  The .dll  can be a dummy 
that simply returns true or false, but it has to function within the API 
framework and return the data types declared by the shader.

Now that Softimage (as of 2011) has moved to shader defs using parameter names 
instead of GUIDs, I'm not sure how relevant this is anymore.  Or at least, for 
new scenes created from scratch.

Matt




From: softimage-boun...@listproc.autodesk.com 
[mailto:softimage-boun...@listproc.autodesk.com] On Behalf Of Amaan Akram
Sent: Saturday, September 15, 2012 12:31 AM
To: softimage@listproc.autodesk.com
Subject: Re: CLSID error and XSI base

I wonder what would happen if you were to create a dummy SPDL in a workgroup 
and give it the GUID of the missing one that is making XSI fail to load a 
scene. Never tried, but often wondered... did anyone ever try this?
On 15 September 2012 02:39, Matt Lind 
mailto:ml...@carbinestudios.com>> wrote:
Pure guess:

Not sure if still the case, but early versions of XSI used GUIDs to identify 
objects, plugins, etc...  Kind of like using an ID to look up a record in a 
database table.  Problem is when the item being looked up is not installed, 
there's no available information to determine what it was or how it fit into 
the larger scope of the data.  I'm also guessing much of the data was stored in 
data structures like graphs to represent the scene data which would imply you 
need to have a valid object before you can get the next object.  Example: you 
would need a valid shader to know which other shaders were connected to it's 
inputs as each shader parameter has it's own GUID.  If the shader is not 
installed, then the GUID would point to a missing record which in turn means 
other locally connected items couldn't be retrieved either.  If relationships 
are complicated enough, it might not be possible to clean up the scene.

I think that is some of the reason why the self installing plugin architecture 
was written as well as converting from GUIDs to parameter names for ICE 
attributes and shaders in recent releases.

Matt




From: 
softimage-boun...@listproc.autodesk.com
 
[mailto:softimage-boun...@listproc.autodesk.com]
 On Behalf Of Andreas Bystrom
Sent: Friday, September 14, 2012 6:20 PM
To: softimage@listproc.autodesk.com
Subject: Re: CLSID error and XSI base

while on the subject, what's the reason that xsi has always handled this case 
so poorly? a missing plugin would always render the scene non-functional but 
having xsi clean up and basically destroy the scene just because a plugin is 
missing or even a different version was installed isn't terrific.



On Sat, Sep 15, 2012 at 3:49 AM, Ben Rogall 
mailto:xsi_l...@shaders.moederogall.com>> 
wrote:
The original is here:
http://shaders.moederogall.com/

Haven't touched that page in a long time.

Ben


On 9/14/2012 10:27 AM, Luc-Eric Rousseau wrote:
it sounds like it; the clsid probably changed in the process.  I could
find zbumb for xsi with google search.. it shouldn't be too hard

On Fri, Sep 14, 2012 at 10:44 AM, Eric Thivierge 
mailto:ethivie...@gmail.com>> wrote:
If I remember correctly zbump was integrated as the default bump shader a
while back. Am I correct Luc-Eric?


Eric Thivierge
http://www.ethivierge.com




--
Andreas Byström
Lighting TD - Weta Digital



--
3D Artist/TD @ The Mill, London
http://www.amaanakram.com

Re: [BUG] Softimage crashing on workgroup unload containing a custom renderer and at least on .xsirtcompound file

[Martin Chatterjee]

Hey Philipp, long time no see! :-)


> *- when using events and python to Add/Remove workgroups the only ones
> that _never_ crash for me are workgroups containing only python plugins.
> Everything that has a DLL/SO in it crashes eventually in my tests. *


To be fair it's been a while since my tests - but I'm pretty sure that our
event-triggered workgroup switching only started crashing after having a
custom renderer installed.

We had quite a few dll's in there without any problems (compiled commands,
operators, shaders)

That's on Windows, by the way - might make a difference as well...

Cheers, Martin


--
   Martin Chatterjee

[ Freelance Technical Director ]
[   http://www.chatterjee.de   ]
[ https://vimeo.com/chatterjee ]

Re: [BUG] Softimage crashing on workgroup unload containing a custom renderer and at least on .xsirtcompound file

[Martin Chatterjee]

Hi Sajjad,

back from my vacation - hence the late reply. ;-)

What I did not mention in my initial mail(s) is that we stopped trying to
switch workgroups on the fly *via events* a while ago because this also
reliably crashed on our side.

We've been using a custom loader/saver dialog for a number of years now and
in there I can act upon the button onClicked() handler to call my workgroup
switching code.

My gut feeling is that these might be two different bugs/problems.

Cheers, Martin

--
   Martin Chatterjee

[ Freelance Technical Director ]
[   http://www.chatterjee.de   ]
[ https://vimeo.com/chatterjee ]



On Wed, Sep 5, 2012 at 1:01 PM, Sajjad Amjad  wrote:

> Apologies for bringing this slightly old thread back to life. I have been
> testing workgroup switching without having to restart Softimage.
>
> Specifically, I have two workgroups, where each workgroup has just one
> add-on installed in it. No .xsirtcompound files exist in any workgroup.
>
> The addon in each of the two workgroups is a unique version of VRay.
>
> I want to ideally trigger the workgroup switch using the onEndSceneOpen
> event (as every scene defines it's renderer version in a custom property)
>
> If I use the siOnEndSceneOpen event for switching, Softimage always
> crashes. Unlike Martins case, the rate of success is 0. However, if I
> attach the logic inside the event callback to a buttons OnClicked callback,
> the switching works flawlessly after the button is clicked.
>
> Martins original post required an xsirtcompound file to be present in the
> workgroup for Softimage to reliably crash, but I can reproduce this crash
> every time without the file. It seems to me that this problem is a bit
> deeper than it seems. Seng-Siong Lim wrote saying that it was only
> reproduced when using a script for switching and manual switching worked
> fine. In my case, the switching in either case is triggered through a
> script (OnClicked or OnEvent) but only fails when fired through an OnEvent
> callback.
>
> If anyone else has notes to share on this, please share.
>
>
> Cheers,
> Sajjad
>
>
> On 16 August 2012 16:31, Nic Groot Bluemink wrote:
>
>> Oh aye, I wouldn't argue with that. It's that at the time live
>> loading/unloading of renderers was a (poorly) 'known limitation' of sorts -
>> the kind of annoyance you have to live with and work around. If changes
>> have been made to the system since and it's now a bug, then +1 to fixing
>> that :)
>>
>>
>> On Thu, Aug 16, 2012 at 4:17 PM, Steffen Dünner <
>> steffen.duen...@gmail.com> wrote:
>>
>>> I would rather vote for fixing it where it's broken, i.e. on the
>>> Softimage side. :)
>>>
>>> Cheers
>>>
>>> 2012/8/15 Nic Groot Bluemink 
>>>
 Hey Martin,

 Though it's little help, I recall running across a warning in the SI
 docs that unloading custom renderers in-session could introduce
 instabilities back when I was investigating the same thing. I can't find
 the exact entry for you, it was weeks before I accidentally stumbled across
 it last time, but that's pretty much where I threw my hands in the air and
 wrote a workgroup switcher that would restart sessions instead here at
 Kettle.

 Nic
>>>
>>>
>>
>>
>> --
>> Technical Pretty Picture Making Person
>> Kettle 
>>
>
>

Re: Animation/keyframe setting that's driving me crazy.

[Stefan Kubicek]

Restarting XSI doesn not help? Have you tried deleting your preferences?
Do you have any plugins with custom events installed that get triggered 
onTimeChanged that could casue this?
Disable Plugins systematically to find out.



Hi,

I think I've accidentally changed some setting somewhere and it's driving
me nuts. The upshot of it is that if I save a keyframe on a parameter (can
be shader, ICE node, anything in a PPG) and then skip to another frame and
try and adjust that parameter to a new value it just snaps back to the
value of the first keyframe. I can get around it by just clicking the
keyframe button again and then adjusting it in the animation fcurve editor
but that is not how I want to roll! I've got a suspicion that it's
something to do with a SetValue "PlayControl.Key" that now happens whenever
I shuffle through the timeline to a different frame. I could well be wrong
though.

Please help! Please! I'm begging you all for help.

*wimper* *cry*

Anthony

P.s. I'm including a screengrab of that sort of shows both the SetValue
"PlayControl.Key" thing and also how whatever current frame I'm on in the
timeline is also accompanied by a little lime green marker.




--
---
Stefan Kubicek   Co-founder
---
  keyvis digital imagery
 Wehrgasse 9 - Grüner Hof
   1050 Vienna  Austria
Phone:+43/699/12614231
--- www.keyvis.at  ste...@keyvis.at ---
--  This email and its attachments are
--confidential and for the recipient only--