Reply to thread

Message

<blockquote data-quote="thch" data-source="post: 2450633" data-attributes="member: 562032">tech notes:phase.&nbsp; phase is easiest to envision in the &quot;steady state&quot; sense.&nbsp; this means if you were to play a tone, and waited 100 cycles, then compared the input and outputs, you&#039;d notice a shift (phase) between the two.&nbsp; (plot Sin(x) and Sin(x + 120deg)&nbsp; and Sin(x - 120deg) for examples of phase offsets)for a single frequency, this will appear as a bounded phase offset (there is no way to see a difference between 10deg and 370deg).&nbsp; this phase offset can be translated into a time delay (for this frequency).&nbsp; this is where 10deg and 370deg would be different.&nbsp; a phase delay of 10deg corresponds to a short time delay, while 370deg would be a longer delay.typically, people only care about how much frequency X is delayed with respect to other frequencies &quot;near&quot; X.&nbsp; eg, you only care about the delay on frequencies from 20hz to 20khz.&nbsp; if there is 1 hour of delay at 1Mhz, you could care less!&nbsp; this is where group delay comes in.now is the interesting question -- why wait for 100 cycles to look at the steady state response?&nbsp; why can&#039;t you look at the first cycle?&nbsp; well as it turns out, the start of the first cycle isn&#039;t the same as the start of the 2nd or 3rd cycle.&nbsp; on the first cycle, the wave isn&#039;t changing, then suddenly increases from 0 to a peak. on the 2nd+ cycles, the wave IS changing as it crosses through 0 to a peak.a number of math techniques can be used to show and quantify &quot;how much&quot; things are affected by this &quot;transient&quot; response.&nbsp; the Laplace transform is among the best.&nbsp; the Fourier transform is also applicable.as for the staggered Question.&nbsp; its not uncommon to have a 1st-3rd combo.&nbsp; the phase of one speaker is usually inverted as there is more phase shift in the 3rd order filter.(among the more interesting topics for phase is the realization that ACOUSTIC effects like distance affect the phase of the signal to your ear, and as such some filters will work better in practice even when they look like crap on paper.&nbsp; this is also why the LR filters are not really a good choice for car audio -- the tweeter is probably no where near the mid, and the major LR benefit of phase matching is lost!)I have a wealth of information on the subject of filters.&nbsp; I should be able to answer any questions.</blockquote>

[QUOTE="thch, post: 2450633, member: 562032"] tech notes: phase. phase is easiest to envision in the "steady state" sense. this means if you were to play a tone, and waited 100 cycles, then compared the input and outputs, you'd notice a shift (phase) between the two. (plot Sin(x) and Sin(x + 120deg) and Sin(x - 120deg) for examples of phase offsets) for a single frequency, this will appear as a bounded phase offset (there is no way to see a difference between 10deg and 370deg). this phase offset can be translated into a time delay (for this frequency). this is where 10deg and 370deg would be different. a phase delay of 10deg corresponds to a short time delay, while 370deg would be a longer delay. typically, people only care about how much frequency X is delayed with respect to other frequencies "near" X. eg, you only care about the delay on frequencies from 20hz to 20khz. if there is 1 hour of delay at 1Mhz, you could care less! this is where group delay comes in. now is the interesting question -- why wait for 100 cycles to look at the steady state response? why can't you look at the first cycle? well as it turns out, the start of the first cycle isn't the same as the start of the 2nd or 3rd cycle. on the first cycle, the wave isn't changing, then suddenly increases from 0 to a peak. on the 2nd+ cycles, the wave IS changing as it crosses through 0 to a peak. a number of math techniques can be used to show and quantify "how much" things are affected by this "transient" response. the Laplace transform is among the best. the Fourier transform is also applicable. as for the staggered Question. its not uncommon to have a 1st-3rd combo. the phase of one speaker is usually inverted as there is more phase shift in the 3rd order filter. (among the more interesting topics for phase is the realization that ACOUSTIC effects like distance affect the phase of the signal to your ear, and as such some filters will work better in practice even when they look like crap on paper. this is also why the LR filters are not really a good choice for car audio -- the tweeter is probably no where near the mid, and the major LR benefit of phase matching is lost!) I have a wealth of information on the subject of filters. I should be able to answer any questions. [/QUOTE]

Verification