I am creating ambience loops punctuated with intermittent oneshot sounds for variety and detail. I’ve come across several niggles:
1) To create intermittently sounding oneshots, I am using the spawn feature. This is fine, but I really want to be able to set the spawn time to represent the time between the end of one sound and the start of the next, rather than the start of one sound and the start of the next. This is because a) my oneshot sounds can be very different in length, and b) I am a control freak. Being able to do this would be handy for a few other reasons as well, particularly when creating sentenced loops with sounds of different lengths.
2) 3D position randomisation. Now, assume my event always plays from the listener position. If I set 3D position randomisation on the event, I would expect the event to be offset by that amount in a random direction each time the event triggers. My minimum radius is 100, so I set my 3d position randomisation setting to 100. This should mean that the event will play from somewhere on the minimum radius. It currently always plays from the event origin. It’s not clear whether the event should always play 100 units from its origin (and therefore somewhere on the surface of a sphere described by that radius), or somewhere within a radius defined as 100 units from the origin (i.e. somewhere within the sphere’s volume). Your documentation (page 87) refers to a ‘cube surrounding the event instance’. Do you mean a sphere? Not sure how you can define a cube with a radius
3) When I select 3D position randomisation on the sound definition rather than the event, it works as I would expect – it plays the sound within a volume defined by the 3D randomisation radius. I would like to be able to EITHER play the sound somewhere within the volume, OR play it always at the same distance from the event but with a random position.
4) Volume issues. In order to achieve a good 3D spread for these random sounds, I need to define a wide 3d position randomisation value. However, for the volume to remain consistent, this value MUST be less than the inner radius value, otherwise the volume will start to drop off if the sound gets played too far away. I should not have to compromise my inner or outer radii to achieve a particular 3D random position result.
So, I think it would be ace if you could do either of these things:
1) Create a random 3D pan control that simply randomises the 3D panning of a sound when it gets triggered. Allow the user to set this on sound definitions or events (setting on an event will apply the same random setting to all sound definitions triggered by that event at a particular time). This will allow the user to alter pan positions without worrying about volumes at the same time.
2) Allow the user to switch between playing a 3d randomised event (or sound definition) within the volume described by the 3d randomisation parameter, OR on the radius described by it.
- PeteWard asked 10 years ago
Thanks for the reply
1) I could do that, but then the timing would no longer be random. I guess if the loop was long enough nobody would notice. In any case, this still gives me a time difference between the start of one sound and the start of the next – what I need is a difference between the END of one sound and the start of the next. As well as making it easier to temporally spread oneshot sounds of different lengths, this would allow me to set up loops made of segments of sound of different lengths that crossfade with each other seamlessly. This is a very neat way of making non-repeating loops out of very little actual sound data, and works well for constant sounds like fire, running water, machinery, etc. The current system allows such segmented loops to be set up, but the segments need to all be exactly the same length. This can easily result in faint, regular pulsing unless the crossfades and times are set up exactly correctly, and it’s easy to notice because our brains are so good at pattern recognition.
2) That would make sense.
3) The problem with setting it as a 2D sound stems from our implementation of ambient sounds. We have a 3D sound track the listener position while the listener remains within a defined volume. If the listener crosses the boundaries of that volume, the 3D sound is made to stick on the boundary. The result is that within the volume, the sound appears to be 2D, but outside the volume, the same sound appears to be 3D. If ambient oneshot sounds were set to 2D, they would not behave correctly with this system.
I think all that’s missing is a random pan option. At the moment, I am forced to use a random position to derive a random pan. The difference is that position implies pan and volume, whereas pan simply implies pan. Alternatively, if the distance from the event origin could be fixed for the 3D position randomisation, it would remove the volume variability and just give a pan change.
1) To create intermittently sounding oneshots, I am using the spawn feature. This is fine, but I really want to be able to set the spawn time to represent the time between the end of one sound and the start of the next, rather than the start of one sound and the start of the next.
Would this be better implemented with a variable delay of the start time in conjunction with the current spawn functionality?
2) 3D position randomisation. Now, assume my event always plays from the listener position. If I set 3D position randomisation on the event,
Your documentation (page 87) refers to a ‘cube surrounding the event instance’. Do you mean a sphere? Not sure how you can define a cube with a radius
Haha, got me. I originally wrote ‘sphere’ – however it was explained to me that because the variation from the origin position occurs in 3 planes, making the resulting plot (of maximum variations) look like a cube, not sphere.
As for 3) your asking for a ‘A variable positioning with fixed distance’ right?
- templar answered 10 years ago
Introducing a variable delay of the start time for a spawning sound definition would be very useful, but it still doesn’t measure the time between the end of one sound and the start of the next. Still useful though.
re. random positioning – yes, variable position with fixed distance is exactly what I need, because I want to be able to randomise pan position without affecting volume.
Pete – Here’s some suggestions I have:
1) Put your one-shot sounds on a layer that has a parameter with a velocity of 1. Then your min/max parameter value is equivalent to seconds, and you can add one-shots to what essentially becomes a ‘timeline’. Might give you a bit more control than using spawn time alone…
2) I’m pretty sure that 3D position randomization means that the sound can instance anywhere within a spherical volume of X units.
3) You can’t specify a fixed distance with a random position. Position and distance are effectively the same thing. You could instead specify that the sound is 2D and randomize panning and/or volume and achieve a similar effect.
- Symbiotic answered 10 years ago
Please login first to submit.