4 ms·

A couple of questions that come to mind: • Is it possible to go backwards? Obviously not from the later iterations of this loop — all useful information has be

by svat 3y ago

A couple of questions that come to mind: • Is it possible to go backwards? Obviously not from the later iterations of this loop — all useful information has been lost — but from the first couple of iterations, say, is it possible to recognize "these are probably the room's resonant frequencies" and de-emphasize them? • What would the experiment be like if each round involved re-playing the tape in a different room (one with presumably different resonant frequencies)? Would it sound significantly different / would there be some possibility of regaining information after more iterations than in the former case?

squeaky-clean 3y ago
For the first question, kind of. In theory yes, in practice people like to move so it depends on how accurate it needs to be. It's easier to do with white noise than a sample. But ideally you would use an "impulse" (an extremely short pulse of sound) in a dead-silent room. You play the impulse and record that along with any tailing reverberation/resonance. This it typically used to "capture" the audio response of something like a room or speaker cabinet. (You can purchase Impulse Responses of Abbey Road Studios, or famous cathedrals. It can even capture a speaker response and is often used by guitarists with digital setups). But you could invert it to remove resonance. Keep in mind, even moving a millimeter will slightly change the phase of the resonances and thus the resulting response. So this would only work perfectly to remove a room response from a fixed speaker and a fixed microphone. But some mixing boards for live shows have an "anti feedback" system that works off this principle. It's not perfect, but it's good enough to prevent a microphone screech. For 2, That's an interesting thought. My guess is that it would take longer to become inaudible, and be much more "diffuse". Like being recorded in a cave or room with very complicated geometry. It's almost like how most digital reverb effects work. They tend to pass sound through multiple layers of delay (aka echo) while applying phase shifting. Sometimes they are designed to resonate, but usually they are designed so all the delay layers conflict and don't resonate too strongly in any given frequency. Part of why it resonates in such a pronounced way is the fact that it's a fixed room and almost certainly rectangular (Which you may notice many recording or mixing studios tend to avoid being a "box").
_kb 3y ago
Some form of equalisation is applied in most audio systems. Ideally this should correct the system and transducer response only. Attempting to EQ a room will make things worse as the response will differ across listening positions. For addressing room resonance, physical acoustic treatment (bass traps, diffractive and absorptive materials), or in the case of new construction optimising the structure of the room itself is the general approach, with DSP used to augment this. There's an interesting intersection of this once you start diving into the world of active acoustics too. In the venue space you have systems like Meyer Constellation or L-ISA that use a distributed set of speakers and microphones to be able to change the room acoustics in real-time, or even emulate a different space.

by svat 3y ago

squeaky-clean 3y ago

For the first question, kind of. In theory yes, in practice people like to move so it depends on how accurate it needs to be. It's easier to do with white noise than a sample. But ideally you would use an "impulse" (an extremely short pulse of sound) in a dead-silent room. You play the impulse and record that along with any tailing reverberation/resonance. This it typically used to "capture" the audio response of something like a room or speaker cabinet. (You can purchase Impulse Responses of Abbey Road Studios, or famous cathedrals. It can even capture a speaker response and is often used by guitarists with digital setups). But you could invert it to remove resonance. Keep in mind, even moving a millimeter will slightly change the phase of the resonances and thus the resulting response. So this would only work perfectly to remove a room response from a fixed speaker and a fixed microphone. But some mixing boards for live shows have an "anti feedback" system that works off this principle. It's not perfect, but it's good enough to prevent a microphone screech. For 2, That's an interesting thought. My guess is that it would take longer to become inaudible, and be much more "diffuse". Like being recorded in a cave or room with very complicated geometry. It's almost like how most digital reverb effects work. They tend to pass sound through multiple layers of delay (aka echo) while applying phase shifting. Sometimes they are designed to resonate, but usually they are designed so all the delay layers conflict and don't resonate too strongly in any given frequency. Part of why it resonates in such a pronounced way is the fact that it's a fixed room and almost certainly rectangular (Which you may notice many recording or mixing studios tend to avoid being a "box").

_kb 3y ago

Some form of equalisation is applied in most audio systems. Ideally this should correct the system and transducer response only. Attempting to EQ a room will make things worse as the response will differ across listening positions. For addressing room resonance, physical acoustic treatment (bass traps, diffractive and absorptive materials), or in the case of new construction optimising the structure of the room itself is the general approach, with DSP used to augment this. There's an interesting intersection of this once you start diving into the world of active acoustics too. In the venue space you have systems like Meyer Constellation or L-ISA that use a distributed set of speakers and microphones to be able to change the room acoustics in real-time, or even emulate a different space.