Hm, I tested tp partio with this small project first, it seems to work fine, like in tutorials, then copied the tp-partio into large project, set the input and output and that's it.
Not sure what it all means. What file sizes do you get? Are they usable?
I cleaned the project of the geometry and simplified it as much as I could, and now I am seeing 3 MB BGEO snapshot appearing, and much lesser memory footprint. However, the simulation stops at frame 217 and Cinema hangs, computer has to be restarted. I then start simulation at frame 217, and Cinema freezes at 433. It seems it can process frames in chunks of 217. Here is this project, perhaps it will help your debugging.
@Xavier: What do you mean? There's the filename section right on top. Just enter the path and filename (any file of your sequence) and it will recognize it, e.g. c:\data\myparticles_0050.bgeo will be recognized as a bgeo sequence with 4 digits.
In your tutorial there is a tab where you can select what type of file to make and where the data path goes. In my version that tab does not exist. So I can of course open files with that Input file section but if the plugin doesn't give me the option to export I have nothing to open.
I see now! I didn't quite get that I had to manually enter the file path. I guess a feature request would be a dropdown menu that lets you select your filetype with a field that allows you to select the directory without having to manually input. (just like it was before) Am I missing something or were those features removed?
Is fixing the issue with the simullations running in 215 frame chunks and freezing C4d after that on your plate?
I can confirm that this indeed is an issue, while running a simulation of 1600 frames, and having to restart the machine after each 215 frame run. The project is attached to my message below.
brought the Plugin I am using it for C4d r15 i simply can not cache out to disk as that parameter is not there and what format do i cache out as since i dont have any options
brought the Plugin I am using it for C4d r15 i simply can not cache out to disk as that parameter is not there and what format do i cache out as since i dont have any options
Input and Output are using the same filename now. The filetype is being recognized based on the extension. c:\output\particles###.geo for example will cache out to c:\output\particles0000.geo, c:\output\particles0001.geo etc. It'll also read from that file.
Thanks Michael i just now have the problem of caching one frame and not all frames as one would expect, even when i press write all i am just getting one frame
i bake TPs that are driven by xparticles. when i turn on "auto read from disk" i get a different simulation (its similar but bigger). once i hit read current frame though the particles jump to their correct position. i already tried to load the particles into a completely fresh scene without anything but tp partio to make sure nothing is affecting the particles but unfortunately the same thing happens.
@peter: i'm going to record a new tutorial soon. until then, please note that the read and write now use the same file, and you have to input it like this:
/Users/michael/Temp/particles/p_0049.hclassic where "p_" can be anything, "0049" is interpreted as a 4 digit sequence which will be used to enumerate the particle sequence (so, frame 1 will be stored as 0001 and so on) and the extension selects the file format (.hclassic being the old houdini geo format)
@Florian: I worked out what the problem was, it had to do with Cinema's priority system. I have fixed this as well as the global scale parameter now, it's already compiled for OSX and I'll do the same for Windows on Monday. It'll then go up in aescripts...
thank you for the fix with the scale!!! it works now. unfortunately i experienced a problem with using the "baked" particle group in a cloner object. the scale of the particles flickers like crazy?
can you confirm that? or is it something on my side!
thank you for the fix with the scale!!! it works now. unfortunately i experienced a problem with using the "baked" particle group in a cloner object. the scale of the particles flickers like crazy?
can you confirm that? or is it something on my side!
thanks a lot!
Florian - do you have more than one object under your cloner? If so make sure it's Object mode is set to "iterate" - anything else it will cycle through when generating particles - at least that is what I experienced when I used the cloner for particle geometry.
Hey, one thing I forgot to mention, if the TP Partio particles are the only particles in your scene, you could try and check the "Free all particles" checkbox under Read Particle Data. That will delete all particles in each frame and recreate them the next, in the same order. This could potentially help against jumping in cloners.
I have just purchased the TP Partio v0.16b2 version, I am trying to follow the tutorial of basic using, however, i don't have the Filename settings Option as Michael mentioned in this video tutorial... So i cannot really saving any caches of the TP...
Input and Output are using the same filename now. The filetype is being recognized based on the extension. c:\output\particles###.geo for example will cache out to c:\output\particles0000.geo, c:\output\particles0001.geo etc. It'll also read from that file.
Hi Michael, I'm on r15 with version16b2. I'm having the problem mentioned below where I'm only getting a single frame cached when I play back into a new tp particle group. Am I missing something about compatibility? I see my .geo files saved and growing as the simulation runs, but when I play them back I either get a single frame, or some weird explosion behavior every once in a while. Any clue what's happening?
As mentioned on Ting Ding's April 30th Post "i don't have the Filename settings Option ... So i cannot really save any caches of the TP..." This is my exact problem.
I've been hitting "write all" and watching and my simulation over and over and not able to get a grib on how to save to a specific location, specify a name, a file type etc....I attached a picture of what I "should" be seeing from the v15 Oa3. I am not understanding the response you made about naming convention or why these options aren't available. I hope you can help!
the "Input File" is a combined input AND output file. "Read Current Frame" and "Auto-read from disk" will use this file, and so will all the "Write …" buttons below!
This concept was introduced in a way similar to the Houdini file nodes that also use can switch between Read and Write mode.
It says r13 and up in your descriptions. Does that include r16?
can I use TP Partio for caching in order to support Team Render (Net Render)? If so, how?
Thanks in advance,
Philippe
Hi Philippe,
I have just tested it with R16.027 and it works. I read in 40 frames with ca. 1.6Mio Particles and wrote them out again. No issue that I can see at first glance.
It should also work in Team Render, but the plugin needs to be installed on every client and the cached files need to be accessible from each one under the same path (a network share for example)
thanks for your answer and for the test. Unfortunately I am experiencing freezes in the caching. Is there a way we can talk directly? I would not like to post my scenes publicly. Lloyd has my details.
Allright, I think I got it to work, Partially. Let me update you:
Here is what I did:
I turned OFF "[development] Track Particles", I do not know if this was the cause, and I do not know what it is supposed to do.
I turned off auto-reading from disk in the TP Partio node whilst I was writing the caches.
After this my cached started filling up without freezes. I did the furst four particle groups like this, turned off the TP expresso, turned auto-reading on. It worked, I could now random scrub through my animation and the cached particles showed up nicely. When I started doing the last two groups, the freezing returned.
Now I turned off the auto-reading on the existing first four caches, then the writing of the new groups worked again. Basically you cannot seem to do reading and writing at the same time, even if it is between mutliple separate chache elements. Something seems to go wrong with the I/O and it freezes C4D if you do that.
So far so good.
Couple of issues I still have:
When naming the cache file XXX2_0000.geo, it ignores the underscore and the name ends up XXX2#####.geo, turning the underscore ito an extra incremental digit. This does not happen when the part before the underscore ends on a letter (and not on a number as in the example).
Render to view (or render viewport region) seems to do nothing, or it is interrupted immediately and stops. Ineractive render region does not work either. No viewpoet rendering is possible. This is very annoying when you are building and testing a scene. I tried doing it with auto-read off - doesn't help. Render to picture viewer works, however,
After playing around for a while with my animation, the auto-reading stopped. No more cached particles in my viewport. It worked again after I restarted C4D. I figure this is a memory issue - can that be?
Main problem remaining
Cached particles started to dissappear from the picture viewer render, even if they do show up in the viewport. Restart doesn't help, even if the render is the first thing I do after I boot up. Picture viewer renders used to work earlier in my testing, From some camera angles I can get a frame to render cached particles in the viewport, but only if it is the first thing I do after a reboot. It all seems very random. That is of course a killer problem.
So this is where I am stuck now, I would appreciate some help. I still need an email address if you want my scenes.
The error probably comes from having multiple versions of it installed in different folders (maybe you have one in the Cinema plugins folder and one in your profile plugins folder?)
If auto-reading is turned on, what it does is it loads in the particles on every frame, which is what you usually want to read in a cache. If you are writing out a cache you don't want that, because it will try to read and write at the same time - and it will probably try to read first, getting an empty file and writing an empty file out, or something similarly weird. So yes, you need to turn that off. Ideally you use one TP Partio object for reading and another for writing and turn off the reading object during writing. That's how I do it.
I noticed a memory issue after a while and it is related to the partio library I think, but I have yet to work out exactly why that is.
The naming issue is a known problem. I'll look into it but I can't give a time frame on it.
Please note that while TPPartio seems to be working in R16, it has not been tested thoroughly yet. Again, I can't give a time frame.
the picture viewer rendering and the dispaly of cached particles in the viewport seems to work fine now that I removed the extra plugins.The viewport rendering is still not working.
I will keep replying here on the stuff I already wrote here, anything new I find I will reply to your PM.
i am missing something basic here. the settings from the plugin and the demo video differ greatly and i am kinda confused how to work the plugin. i can't get it to bake a sequence. i just end up with one file. how do i do that?
Hi, because of the confusion caused by the changed Input File parameter, I've uploaded a new video showing the setting. Please either have a look at the plugin page or directly at vimeo:
Anyway to get PShape to work with this? I'm using a PPass node connected to Pshape. I'm referencing an .Hclassic houdini file and I can see the tp "ticks" in the view port, but can't get geometry to stick to the particles. Any pointers?
Sweet plugin! How should i do to import a BGEO sequence exported from Houdini? When i choose the sequence as input file I cant get the particles to show up in C4D.
Sweet plugin! How should i do to import a BGEO sequence exported from Houdini? When i choose the sequence as input file I cant get the particles to show up in C4D.
I am on R16 OSX.
Thanks in advance!
Houdini changed their BGEO/GEO format a while ago. You need to export the particles as .bhclassic / .hclassic instead. That's using the old, supported format.
the "Input File" is a combined input AND output file. "Read Current Frame" and "Auto-read
from disk" will use this file, and so will all the "Write
…" buttons below!
This concept was
introduced in a way similar to the Houdini file nodes that also
use can switch between Read and Write mode.