There is an overlap in my understanding as you've said before the bass region has perceptual properties such as requiring a few cycles to be identified. For example, if a small vehicle cabin sized room were producing simple modes at 1kHz, would fixing it spatially be the dominant fix despite whatever the source location?
On a tangent, unless the answer to the above question is no, do the modes in the region of concern need to be clean, 6 wall modes to fit into your above description?
Group delay very much enters into the picture. Think of phase change with frequency as being a continuum from negative delay to positive (positive slope to negative slope.) For one and only one very unique situation where there is zero change in phase with frequency (zero slope) the delay will be zero. (This could be any phase value however. Take phase reversal in a speaker. Zero delay but 180 degrees of phase.) This is virtually never going to happen in practice (except the one situation that I mentioned,) but is theoretically possible. It is certainly NOT the situation in our discussion here.
Last edited:
I can't quite figure out the question. If there is a single mode, then no "spatial fix" would be possible no matter where the source was located. The mode shape remains the mode shape regardless of the source position.So I do not know what a "dominate fix" means in this situation.
In virtually no situation are the modes "clean" in the sense that only singular modes exist. Basically this cold only ever occur for the very lowest mode or two, typically, 20-30 Hz in a normal room. But even then if the room is damped then the modes spread and they begin to interact with one another. In my room there are no modes that stand alone even down at 25 Hz. So what you are asking virtually never exists hence my description would apply especially when they don't. But then I am still not clear on the question.
We're probably thinking/talking from different perspectives I guess.
I don't really see phase rotation and time delay as being the same thing.
FIR, and programs like rePhase, have made it where we can have about any shape phase trace we want, show up at the measuring mic.....all with the same delay.
I mean, it's not called "rePhase" for nothing
The way I see it a phase shift is a time delay but only for an infinitely long sine wave. It amounts to the same thing but only upto 360 degrees.
I'm not familiar with FIR filters and digital delays, but I'm guessing they can shift phase and delay any shape waveform, ie, it's not frequency dependant and doesn't cause group delay?
I'm not familiar with FIR filters and digital delays, but I'm guessing they can shift phase and delay any shape waveform, ie, it's not frequency dependant and doesn't cause group delay?
Last edited:
Is this because they are in the modal region, or in spite of them being in the modal region and just because of our perception at these frequencies?
I don't think that It's simply a difference of perspective, because your second paragraph is incorrect. The "shape of the phase" will dictate the delay. Different shapes have to have different delays. We can have the same magnitude responses with different delays, that's true, but that's not what you are claiming.
Yes, I am definitely claiming the shape of the phase does not have to dictate delay.
It doesn't have to with FIR, given enough sample time to alter phase to will.
The necessary sample time shows up only as pure time delay, no different than distance. Phase can be moved all around within a defined sample time, with a constant delay.
you might want to download rePhase and play around with it...it's a hell of a learning tool....or has been for me anyway
I don't know what to say, except that your understanding is incorrect. Phase change with frequency is time delay - period - and this is true in either the digital or analog domains. I'd try and explain the error in your example except that I can't follow it.
Maybe I'm misusing or misunderstanding definitions....
So,... my understanding is "time delay" means just that, and nothing past that. And that time delay is a constant, and frequency independent.
Examples being things like distance, fixed dsp latency, a digital delay, or FIR filters.
My understanding is "phase change with frequency, or phase rotation over frequency", is a relative relationship of frequencies' phase versus the phase of a given reference frequency at some point in time (usually nyquist freq, with all "time delay" removed).
I can see that "phase rotation" requires that there is a varying time shift vs frequency,
but I think of this time shift more as a "time rotation" vs frequency...without a clear term to describe it. Or without me having found one yet
So I see time delay, and phase rotation, as describing two very different time properties....
Please advise and thank you for hanging in with me!
If you don't understand the difference between delay and phase, maybe you should avoid commenting and giving advice on these matters.
//
Consider the standard physics representation of a signal as a complex exponential V(w) = A(w) * exp^(i * w * t) where w is angular frequency and t is time. This describes a vector that rotates around the origin with an amplitude of A and it makes w full rotations every second. At any instance in time there is a "phase" of this vector, but as time progresses this phase continuously advances. This means that phase and time are simply two different ways of describing the exact same thing - progression of time or progression of phase. They are the same thing not different things.
This is easier to see in the analog domain as things get a little foggier in the digital domain.
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.