New Beta9 Features Thoughts/Explanations

hi

thank you alot for the very cool new beta9. There are a lot of new features but not all of them are documented allready (of course).

thougght i open a new thread where thew new features can be discussed/explained.

i make a start with a few questions.

-what is the type pin ind “dx9Texture” for? is it possible to render in a volume or cube texture?

-what exactly is that new filtered time mode??? does it mean that time gets calculated independent of the frame rate?

-can the new dft(value) node be used to build a frequency analyser like the fft (dshow9)
maybe with conjunction with scope (spread)? if yes would it be mopre accurate

some thoughts:

the new layering system is really cool much more clearer than the old renderpasses.
BUT unfortunatly it isnt possible anymore to load dx9 objects on the fly without connecting them to a group. working with send/receive nodes you have to know the sendstring before and you cant select a receive string before it exists…

mh just discovered the mainloop help patch which describes the thing with filtered time well…

-can the new dft(value) node be used to build a frequency analyser like the fft (dshow9)
maybe with conjunction with scope (spread)? if yes would it be mopre accurate

you can use it with the scope like fft but it would be more time consuming. DFT means discrete fourier transform, FFT fast fourier transform. only problem with FFT is that it works only for power of two spreadcounts.

better make some experiments with it … ;)

-what is the type pin ind “dx9Texture” for? is it possible to render in a volume or cube texture?

like joreg pointed out: this is not supported yet. sorry.