designing different delays - implementation questions
i am trying to design lossy delays, e.g. delays in which each successive echo of the sound has been affected somehow - in particular, delays in which successive echos sound less present or more spatially diffuse. i understand that a weak implementation of this would be simply to drop an eq or reverb after the delay and use some fancy routing to only affect the wet output - but this, as a single effect applied once to all output, is not what i'm looking for.
i understand that what i need to do is to turn off the delay unit's internal feedback and implement the feedback myself by routing the delay's output back into itself. my problem of course is that i cannot seem to avoid a catastrophic increase in volume.
a certain amount of knob tweaking and volume attenuation here and there seems to solve the problem for some particular input at a given amplitude, but i'm looking to solve the problem in general - e.g. no matter what you feed into this filter, there will be no exponential volume increase.
[ s p h 3 x ]
You could use multiple DDL's
This would get really device intensive if you are trying to accomplish super long delay effects. If however, you are only trying to get, say 2-10 delays, this is what I would do.
1) Run your source noise maker into a SpiderAudio.
2) Run each of the 4 audio outputs to its own DDL.
3) Set each DDL to MS mode and set each DDL's time to a slightly higher time value than the before it (i.e. 100ms, 200ms, etc.).
4) Set each DDL's Feedback so that only one or two repeats occurs.
5) Route the DDL's into whatever affect you want to use. Like maybe each into its own RV-7 and have the RV-7's set to increasingly larger room sizes, feedback levels etc.
6) Route the RV-7's into a mixer and use the mixer channel level sliders to shape the delay response.
7) Use another DDL or other effect(s) as a send effect(s) from the mixer. Then set each send knob to increasingly higher values for each channel on the mixer.
This set-up will only give you a few delay repeats, but you can add as many SpiderAudios, DDL's, RV-x's etc. until your computer begs for mercy. Plus you get pretty precise control over your repeat levels using the mixer's channel levels.
This might not be exactly what you are lookng for, but it was the first thing that popped into my head. I thought it would make some interesting noises so I wanted to write it down before I forgot it.
Hope this helped,
How to fix that feedback problem
You are routing a DDL back to itself?
I'm picturing that you are using the left input channel on the DDL as the dry/source signal. Then you take the right output channel and route it to the right input channel. Is this correct?
If so, you can insert a 6:2 mixer in the right channel feedback loop. Then use a Malstrom Mod in 1-shot mode to auto attenuate the signal using an appropriate waveform.
Or as a different take on the above, route the DDL's left output to the 6:2 and use the same Malstrom set-up to attenuate the channel level.
Re: designing different delays - implementation questions
you can seperate your sound with spider, letsa say for 6 channels ,and you put effected delays on each channels. then you put each channel in the mixer and randomise the volumes im mixer with lfo, for that youll need to make a combinator.. . so youll get each sound......ash damn i cant speak....... welll.. doesnt matter. awell you will get each delay effected. and they will sound different. you can ... well. you can do everything. out.
|All times are GMT +2. The time now is 08:04.|