posted 21 February 2014 15:32            

I'm trying to achieve this:

First we hear a natural atmo with wind and birds. Then slowly we start to hear something tonal "through" that atmo which turns into a musical piece.

Simple cross fade would reveal too much about that music. I don't want to hear the actual instruments, but the tonality of the piece. Visual analogy would be an image of a solid wall and a human face pushing through the wall so that it still has the texture and the color of the wall, but yet we could tell whose face it is.

I've used Reaper DAW and some plug-ins. I figured that a vocoder using the atmo as the carrier could be the answer but the results tend to be distorted and ugly. I've also made quite interesting mixes with a convolution reverb by using an atmo as an IR and then played music through that. The results are very blurry though.

posted 21 February 2014 23:36            

Have you tried the Kyma Vocoder with the SideChain set to Noise in a CrossFade with the music? You could start with noise as the sidechain and gradually introduce the music. And the bandwidth could start out wide and grow more narrow to make the filters start ringing.

posted 24 February 2014 10:39            

Tunethruatmotest.kym

Thank you very much for your suggestions. I'm not sure if I understood correctly what you meant. FilterBank test is not good at all, but vocoding has something of the character that I'm after. Here are my trials attached. Though I'm not very happy with these, maybe it helps you to guide me further. I'd be very glad if you could do that.

posted 24 February 2014 14:14            

You could also try using the Morph1dSampleCloud as the SideChain to your Vocoder example in place of the Crossfade, just to try another variation.

posted 24 February 2014 14:20            

(Noise XFade Music)-->HPF-->Gain

in other words, with the same side chain you were using for the Vocoder but with a Gain as the last stage.

Try playing that with the gain boosted to around 16 and do the fade. You could try increasing Interval to 12 or 24 for better results.

posted 24 February 2014 16:52            

May be the cross filter could have some mileage in this situation. If you captured some of your atoms in the response, then putting a random clicky sound in the live input (maybe like a scratched record) would be like playing random copies of the captured atoms at different times and at different levels, and my give you a sound a bit like the original atoms.

Then you could cross fade the clicks with your music, (before the crossfilters live input) but this should not sound like a cross fade as both clicks and music are being treated by the same response. The music will have a blurring effect, but the cross filter gives you live control over the start and end of the captured response, so moving these points closer together will give you control over the amount of blurring. Also you may find one small area of the atoms that gives a better sound to the music so you can make your start and end points hone into that area.

The crossfadeing and the start and end changes will make a big difference to the signal level so the real skill is always compensating for this vast level change.

The best place to control this level is the level control in the cross filter itself, as this will guarantee the best sound resolution at all times. Adding gain at the output is like closing the door after the horse has bolted, and will not make the most of the crossfilters optimised clarity.

I hope this makes sense