I used to be able to press the R key and it would let me record a sequence of frames as PPM files. Pressing R no longer does that. Is there a new shortcut now? Or is that feature now completely disabled?

QuanTech:

@Videogamer555 (View Post)
enter tpt.record(true) in the console. The confirmation box still shows up.
The usual shortcut was removed i think because people would press the R button by accident and use up lots of space on their hard drive (though i dunno how they would 'accidentally' click Confirm when the box pops up)

Thanks, but please tell the developers to put back in the R shortcut, and also to add a way to reset the frame count. Otherwise the frame count remains the same, even after the saved files are deleted. You currently have to turn off TPT and restart it to reset the frame count.

jacob2:

@Videogamer555 (View Post)
There is no reason for the 'r' shortcut to stay. If you really need it though, you could write a short Lua script to add it back. I would write one, but on a phone right now ...

If you test tpt.record, you will see that the frame count issue was resolved when I did the recording rewrites. Each recording goes into a separate dated folder. tpt.record will return the name of the folder.

Cool. Thanks. The shortcut was great though. I frequently used it to record things, and then analyze the frame sequence later for scientific analysis using other software that I wrote myself.

jacob2:

@Videogamer555 (View Post)
Very interesting. What kind of analysis? Are you detecting the elements by pixel?

Basically the reason I removed it is because I figured the output was unusable without a script to fix it up for you. Especially because of that frame bug where it wouldn't reset the counter. Seems like you were also using some software you wrote to use the files.

Wrote a program in VB6 that loaded the PPM frames and then scanned one line over time. This generated a new single picture, in which your dimensions were x and t instead of x and y. The vertical dimension of the output image is now frame number, instead of the vertical dimension of the input images. This way you can see all the particles on a single line of input over time. An explosion, when viewed like this looks like a cone, as the particles move outward over time. If you think of the input frame set as 3d, with x,y,t being the available dimensions, then what I'm doing is basically turning that "space time cube" on its side and taking a slice from it after rotating it, and saving that slice as the output image.