I hope your right because one of the focuses of this thread is to gain greater understanding in how to identify what good, great, and fantastic look like in the measurement domain.
Last edited:
Well ok. I should remeasure, but here is from a file I found from when I was tuning (I'm using XSim to plot the files and calculate time domain shots, but the XSim model is just the EQ'd speaker). Looks like the highest octave phase was needing some more work at that stage. I didn't get too aggressive in the bass since that will get position dependent (you can see I have a floor bounce issue at 200Hz that it seemed best to not push too hard). The sub was a little hot in level then too, I have it lowered now. The thing at 500Hz seems position dependent, too, I think that's related to other nearby chairs.
Here's what a 1kHz square wave looks like
And the step response (this moves around with position, too, of course, since low frequencies dominate step response).
All that said, I didn't find that having the higher frequencies with 'waveform coherent phase' made a striking difference, plain response flatness was the most important there (and setting up room surfaces to control reflections). It probably depends on how well things were done in the recordings a lot. Synergy horns are already quite coherent, and minimize room reflections - in some recordings the applied reverb can be so obviously audible it can get distracting with Synergies. In the bass, linear phase made a big diff, though.
Here's what a 1kHz square wave looks like
And the step response (this moves around with position, too, of course, since low frequencies dominate step response).
All that said, I didn't find that having the higher frequencies with 'waveform coherent phase' made a striking difference, plain response flatness was the most important there (and setting up room surfaces to control reflections). It probably depends on how well things were done in the recordings a lot. Synergy horns are already quite coherent, and minimize room reflections - in some recordings the applied reverb can be so obviously audible it can get distracting with Synergies. In the bass, linear phase made a big diff, though.
I'm glad you're in for the test Jim,
Don't know if it will help but here's my left and right channel at the listening position. Still fighting some reflections (the best I can without upsetting the girlfriend) and I only have 3 hidden damping panels in the room, otherwise it's a normal (small) family living room.
I know what's causing the disturbance in right channel, looking for a nice clean way to cure it and keep the girlfriend happy.
Contrary to you I do need some smoothing, my room and comb filtering from my arrays are both very real phenomena's. The downward slope is even more pronounced than yours, my metal ceiling
might have something to do with that. Yes, I will remove that ceiling in time. It is un-damped(!) and I hear it ringing during sweeps.
Despite all that it actually sounds pretty good already. Still a lot more work to do, while keeping the room as undisturbed as possible.
Here's the damping panel behind my listening position:
Got to find your ways right?
Don't know if it will help but here's my left and right channel at the listening position. Still fighting some reflections (the best I can without upsetting the girlfriend) and I only have 3 hidden damping panels in the room, otherwise it's a normal (small) family living room.
I know what's causing the disturbance in right channel, looking for a nice clean way to cure it and keep the girlfriend happy.
Contrary to you I do need some smoothing, my room and comb filtering from my arrays are both very real phenomena's. The downward slope is even more pronounced than yours, my metal ceiling
might have something to do with that. Yes, I will remove that ceiling in time. It is un-damped(!) and I hear it ringing during sweeps.Despite all that it actually sounds pretty good already. Still a lot more work to do, while keeping the room as undisturbed as possible.
Here's the damping panel behind my listening position:
Got to find your ways right?
Well ok. I should remeasure, but here is from a file I found from when I was tuning (I'm using XSim to plot the files and calculate time domain shots, but the XSim model is just the EQ'd speaker). Looks like the highest octave phase was needing some more work at that stage. I didn't get too aggressive in the bass since that will get position dependent (you can see I have a floor bounce issue at 200Hz that it seemed best to not push too hard). The sub was a little hot in level then too, I have it lowered now. The thing at 500Hz seems position dependent, too, I think that's related to other nearby chairs.
Here's what a 1kHz square wave looks like
And the step response (this moves around with position, too, of course, since low frequencies dominate step response).
All that said, I didn't find that having the higher frequencies with 'waveform coherent phase' made a striking difference, plain response flatness was the most important there (and setting up room surfaces to control reflections). It probably depends on how well things were done in the recordings a lot. Synergy horns are already quite coherent, and minimize room reflections - in some recordings the applied reverb can be so obviously audible it can get distracting with Synergies. In the bass, linear phase made a big diff, though.
How do you get your SR to go flat to the right like its a DC servo? All SR's Inhave seen before start falling down in a triangular shape - even for reportedly excellent speakers.
It's supposed to look somewhat like that X, the more you are right in the time domain, the flatter it gets. The example you saw from Stereophile was a Quad if I remember it right. Look at this post from mitchba: http://www.diyaudio.com/forums/multi-way/259384-why-i-think-actives-better-10.html#post3999587
It can never be totally flat though, but it can be better than the examples in Stereophile.
So the more "horizontal" the STEP goes, the closer to actual timing you are. Not that easy to get, mine does not do it yet due to a time delay around box/speaker resonance.
And early reflections in the measurement mess it up too.
It will tighten up the bass connecting the harmonics of the low rumble to the impact at higher frequencies. I believe it also makes higher harmonics smoother.
To quote one of my guru's:
John Dunlavy was one of few that actually explained what he was after.
I believe in his vision, but he was no marketing man and grossly ignored by many who believed he used too cheap components so it just couldn't be right.
It can never be totally flat though, but it can be better than the examples in Stereophile.
So the more "horizontal" the STEP goes, the closer to actual timing you are. Not that easy to get, mine does not do it yet due to a time delay around box/speaker resonance.
And early reflections in the measurement mess it up too.
It will tighten up the bass connecting the harmonics of the low rumble to the impact at higher frequencies. I believe it also makes higher harmonics smoother.
To quote one of my guru's:
John Dunlavy was one of few that actually explained what he was after.
I believe in his vision, but he was no marketing man and grossly ignored by many who believed he used too cheap components so it just couldn't be right.
Attachments
Last edited:
In the case of Dunlavy's plot the impulse is at 9 ms It crosses zero at what, 14 ms? 5 ms is one cycle of what frequency? The one Mr. Waslo shows is actually more impressive, but corrected with Rephase I believe?
If my interpretations are right this will give you the "snap" most are looking for. That Synergy is impressive...
I should try pre correction with Rephase and final with DRC sometime...
And we should all have a room like Jim's to play with...
If my interpretations are right this will give you the "snap" most are looking for. That Synergy is impressive...
I should try pre correction with Rephase and final with DRC sometime...
And we should all have a room like Jim's to play with...
Last edited:
So what you and some others are saying is bypassing passive XO's and switching to active filters for the XO is the way to go?
Could you elaborate on the difference between time smoothness and time coherence?
Totally wrong terms from a non native English speaking individual
.But what I'm trying to say is the speakers move fluently from the tweeter to the mid and on to the woofer and even the port. Much thought went in there to get it like that. But it gradually takes more time from the top to the bottom. Or maybe the mid is even a bit sooner than the top as you concluded from the phase plot. But it's still all very gradual, thanks to second order filters, correcting networks on the speakers themselves and the stepped baffle design.
But true time coherence would mean instant sound from all speakers at once. That would relate to the STEP shown by Mr. Waslo. Almost impossible to get passively, but with modern technology we can correct it.
So that's my rambling explained.
To me, lets first see what a correction of the complete speaker looks like. But a correction suite like Acourate is designed to do just that. Correct every single speaker separately and apply crossover slopes and correct the resulting multi way after that.
Let's take it one step at the time though. You could read up on Mitch's setup with active speakers to confirm what I just said: Computer Audiophile - Advanced Acourate Digital XO Time Alignment Driver Linearization Walkthrough
Fun stuff?
Well essentially we are still trying to solve that Group Delay problem .
But most probably not like Jim envisioned it.
The way I worked it out for myself is with the analogy of watching a band play life.
If they all are on key and most important on time, locked in the groove, watch how the band is enjoying themselves and just have to move along to the music.
If one of them is off... no such "band is on Fire!" miracle is going to happen. It will still sound decent though, provided they are any good to begin with.
.But most probably not like Jim envisioned it
.The way I worked it out for myself is with the analogy of watching a band play life.
If they all are on key and most important on time, locked in the groove, watch how the band is enjoying themselves and just have to move along to the music.
If one of them is off... no such "band is on Fire!" miracle is going to happen. It will still sound decent though, provided they are any good to begin with.
Last edited:
Predicted results of DRC correction over 4 cycles
Here is the improvement that can be expected with DRC correction with a 4 cycle window length. This is the window size that I use myself and it is considered to be a very small amount in the world of impulse response correction. I notice no ill effects outside of the sweetspot with this correction strength.
Here is the improvement that can be expected with DRC correction with a 4 cycle window length. This is the window size that I use myself and it is considered to be a very small amount in the world of impulse response correction. I notice no ill effects outside of the sweetspot with this correction strength.
Attachments
Last edited:
Well essentially we are still trying to solve that Group Delay problem
But most probably not like Jim envisioned it
Yes. But as X pointed out, the topic has expanded into related issues beyond GD that apply to the perception of tight bass response, which is to my liking.
There are few areas that we haven't explored too much and/or I am still a bit fuzzy about.
I few thoughts...
Generally speaking, what we are discussing eventually comes back to preserving the sound of the original recording. So how it is reproduced can be broken down into into component parts or the chain.
Source -> Amplification/Processing -> Speaker -> Room
(by amplification/processing, I basically mean everything between the source file or play medium and the power amplification stage)
Each has its role to play. Each has its limitations including the things it does well and easily, and the things that it struggles with.
IMO, if we were to rank each component in terms of its effect on degradation of said preservation, I think it would go like this (worst to best)
1) Room
2) Speaker
3) Source
4) Amplification/Processing
So what we are doing with DSP, EQ, Audyssey, FIR and so on is to try to compensate for shortcomings later in the chain through manipulation early in the chain.
Through another lens, the part of the chain that does its job the easiest (least source degradation) is trying to make up for the part that does its job the worst (most source degradation).
It reminds me (a bit) of a track relay where you have 4 team members. A very slow one, a slow one, a fast one and a very fast one. And in order to decrease the total time, we give the very fast one steroids and amphetamines while we do little to help the rest of the team.
I get where my analogy falls short. But it is interesting to me that our approach to fixing the larger issues happens where technology makes it easiest, rather than where the real problems are. Again, I get why this is the case. I am just calling a spade a spade.
Can I look at a revised mdat that includes the DRC corrections so I can look at all the data?
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.