December 19, 2010 at 8:30 am #3835
Hello again Peter,
I’m sorry, I wasn’t home yesterday, so only answering to you now… So, the bubble test you sent me is at least showing something… Let me try to describe it to you.
I see just a blue dot at start, growing slowly and slowly fading away. But it do not go up… It’s circling (to be more precise I could say it’s spiraling). Huh, it’s even more complicated than that… On its origin point, it’s not moving at all… It’s only growing and fading away. But when I double-click in the renderer it’s changing… Let’s imagine the renderer window is divided in its X and Y axis in the middle of the screen. The meeting point of those 2 axis is the center of the screen. It’s also the place where the bubble is born. If I click below the horizontal split, on the right hand side of the screen, the bubble starts spiraling anti-clockwise. The farther it is from the split, the bigger the spiral. If I click up the middle horizontal split, still right hand side from the vertical split, it’s also spiraling anti-clockwise. Everything on the left hand side of the vertical split makes the rotation clockwise.
For the Update mode, I was already in Precise mode. I’ve switched to Fast, and … well, there is a difference, but only in the spiral speed.
As well, after various tests, it seems I also get anti-clockwise spirals on the left hand side, so, it could be random seed…
But in all the tests, I couldn’t see any bubble up effect. It’s very much more a spiraling effect to me.December 19, 2010 at 12:40 pm #3836
Great, thanks for the description. It makes it a bit easier to isolate the problem on a single particle effect. It just sounds like it’s ignoring the fact that it has weight. The sinning of the particle sounds ok, as it’s set to spin randomly overtime, and the handle of the particle is offset so that it spirals round like that.
If you open up the bubble effect and click on the bubble emiiter, take a look at the weight and weight overtime graphs. The nodes on there should be on -167.5 and 1 respectively. Try moving them about and seeing what happens. You might have to double click on the preview window to see any changes.December 19, 2010 at 5:07 pm #3837
I checked the parameters you mentioned, but it seems one parameter is not alright. The weight overtime was set to zero. By changing the value, I now can see the bubble working as intended (a bubble slowly growing, and going up the screen until it has faded away.
It seems the particle totally ignored that weight overtime parameter. I call that really strange. Just to be sure, I downloaded again the same particle, opened it without a single change, and the weight overtime was still at zero. The weight parameter is set just below -175%. Looks like very close to -167,5 to me.December 19, 2010 at 8:47 pm #3838
hmm interesting! So it must be a problem when it loads the emitter, that some graphs aren’t being loaded properly for some reason. I think we’re getting closer!December 21, 2010 at 6:26 pm #3839
Hello, could you test this attachment. It’s exactly the same effect, except the weight overtime is now slightly different. The previous version of the effect, the weight overtime was something like 0.9999, but this one it is 1. I want to test the theory that it isn’t converting the number into a floating point number properly for some reason, so 0.999 would be converted to 0.
Thanks!December 21, 2010 at 6:38 pm #3840
I call that a good guess ! It’s all working perfectly as it should be. The weight overtime is now at 1 without a single problem. So that could be a reason there…
Good guess ! We are going forward :pDecember 21, 2010 at 7:05 pm #3841
Thanks for testing 🙂
Ok, so now I have to figure out why, but at least I feel like I’m slowly getting to the bottom of it! Strange that the vaders game is working fine as that essentially uses the same loading functions, must be a difference somewhere… And of course it’s also strange that it only affects some Macs.December 21, 2010 at 7:16 pm #3842
Yes, I agree there with you. I don’t understand why it works on some macs and not on others… Could my main language be some answer ? My main language being french, we do have some specials characters, like our fabulous é, à, è, ô, û or ö
Could the coding be related to a coding ? Just guessing there… No idea at all… I don’t know where did Italian Job come from, but its name suggest italy, which is the same coding as mine, but is not the same as english speaking countries. My call there (and I know this from developing website), is that some program parts are reading a file which on my computer, due to different language, takes other values. It’s just like typing a text with iso-8891 encryption for the characters, and read it on UTF-8 format. All the accentuated character, quotes, or points will be translated into something that cause an error.
Once again, this is just guessing, but it seems that if no english native ever reported this problem, but latin people did, could be linked with the locals of the mac.December 21, 2010 at 8:39 pm #3843
That’s a good theory! I’m not sure if it’s the reason, as it’s only converting numbers. But then again maybe it’s a decimal point thing, as I know in UK we use . to separate decimal point, but in other countries they use a comma. Also, I’ve had quite a few asian customers and their character set can be even more complex, but no complaints as yet, but maybe they all use windows? Anyway, I made a new version which changes the c command that converts string to floats.
It was using atof which apparently has been deprecated (but really should still work), so now I’m trying strtod, but whether that will make any difference I’m not sure. new version can be found here: http://www.rigzsoft.co.uk/files/TestFiles/TimelineFXEditor.zip
Thanks again for helping to test!December 22, 2010 at 7:07 am #3844
So, I just tested the new version, but it seems it do not solve it. I tested the first bubble effect, and the weight overtime is still at zero. I tried the packs I had been testing before, but once again there, I have the same files working and not working…December 22, 2010 at 10:21 am #3845
comankhParticipantDecember 22, 2010 at 10:27 am #3846
Youhou ! I’m not alone 🙂
I’ve got the very same particle animation as you. So it’s reasonable enough to think that both our files are not corrupted in some way, and as it is the very same, well, I guess that’s not our Macs…
Just to verify my assumption it might be something to do with the coding, could you confirm what language is installed and running on your mac ?December 22, 2010 at 1:04 pm #3847
checked on both Polish & English systems, no difference.December 22, 2010 at 2:25 pm #3848
Thanks for the posts, it’s slowly helping me get to the bottom of it! I hope to have something for you to try this evening. It definitely seems to be due to node values being less than 1, being set to 0. If I take all the values of the cloud 7 explosion that are 0.xxx and set them to zero then it begins to look like the effect you’ve posted there, so I just need a more robust way of converting the values that are in the xml file (which stores all the effect data) into floating point numbers.December 22, 2010 at 5:46 pm #3851
Could you run this version and just try loading the first bubble effect I attached? It will show a pop up message with 2 values on it, weight overtime and float. If you could let me know what it says that’d be great. Only try with the bubble otherwise if you load in an effect library with lots of effects it will show the pop up for every effect in the library.
I’m interested to see what the values are that it returns, both should be the same, but I’m assuming they won’t be on your Mac because it isn’t converting the weight overtime to a float properly. If they are the same then I guess the error comes after the loading procedure.
You must be logged in to reply to this topic.