Soundhack Objects

Sep 1, 2010 at 8:38am

Soundhack Objects

I know what I’m doing today :)

Sep 1, 2010 at 9:16am

i wonder when the windoze versions will be ready? :(

Sep 1, 2010 at 9:47am

More fun than a barrel of monkeys! Always loved SH,

Sep 1, 2010 at 10:17am

super cool! thanks

Sep 1, 2010 at 3:29pm

good to have people you can always count on.

Sep 1, 2010 at 3:54pm

wow, great to see the algorithms from the spectral shapers plugins are there too -for free

Sep 1, 2010 at 3:58pm

I’m finding some of these objects kind of disappointing. Binaural~ doesn’t take audio-rate degree input, and seems to have no internal smoothing filter, so every adjustment crackles, severely limiting its uses. That’s simply bad implementation imo. Bubbler~ seems to alias quite badly on many settings, something I don’t remember from using the plug-in version a while ago. I’ll need to test them all now, just for piece of mind. I don’t want to use things that degrade audio quality unless I’m doing it knowingly!

I’m excited by some of these objects, but also wary of using them, as some of them seem kind of badly put together :( I’ll wait for a newer version I think.

Sep 1, 2010 at 4:02pm

No good deed goes unpunished

Sep 1, 2010 at 4:37pm


I’m not trying to be inflammatory…………….I’ve emailed soundhack support as well as posting here (don’t worry I was very polite). Some of the externals are just not ready for use yet in my opinion. I don’t think badly of them for this, nor am I meaning to seem grumpy and ungrateful.

I advise people to be wary when using +.delay~. I had an ugly noise burst when I loaded audio into the help file; luckily I had my headphones off. Just a friendly warning is all :)

Sep 1, 2010 at 6:54pm

if binaural is not interpolating and if it does not allow signal input for the degree
parameter it simply reproduces the plug-in version.

setting up a binaural filter with both, many points AND interpolation is quite CPU
expensive when based on fft or fir.

Sep 8, 2010 at 9:28pm

Yes – the windows versions are done…

A bit of a response. The angle is interpolated, and the output is overlap-add between FFT frames. High frequency hash can still occur if you move too fast. it is not a bad implementation, just the nature of dynamic convolution filters.

I have a newer implementation of binaural that doesn’t have the shortcomings of the convolution approach. I still need to clear some permissions though.

All of the delay externals have fully interpolated parameters. Bubbler (the granular delay) does exhibit aliasing on certain settings, as it has several sources of randomness modulating delay time and grain windows down to .5 ms. I chose to leave the possibility of this kind of sound in the external, rather than polish it out. It also has saturation at the sum of the feedback and input. This will always cause harmonic distortion when driven by signals over -6dB. Again, my design choice, I like the possibility of over %100 feedback and over %100 grain density without hard clipping.

I’ll try to find why the delay gave a noise burst. That is something that does need fixing!

Any other bugs – please email the patch to



You must be logged in to reply to this topic.