But how would I know if things can be even better 
.
I've done much of the same in my Car. fixing little things that seem unimportant otherwise, gradually improving the sound. Just to learn something.
My ears are as easy to fool as anyone else's, that's for sure. So I don't put 100% faith in that. If I could combine all the positive aspects I've heard and had so far in one single setup I would be very pleased. Some days I'm convinced I'm there... only to be disappointed the next day. We as humans have a multitude of filters active, every time we listen to something, being indoors or out. We can filter out a room without really thinking about it. Just record your current sound and play it back on headphones. You'd probably deny it sounds like that in your room.
Taking little steps to make it behave just a little better can go a long way to improve the sound. All this time I still have a functional setup, to play music or movies. But meanwhile I'm testing out different theories and look/listen to the results.
Sure I could say: there, I'm done, finished. But what would I have learned?
I think my rewards are in the (now countless) moments sitting down and listening and being overwhelmed with a piece of music. Taken by surprise basically.
This sort of thing can happen while seeing a movie etc. Even that experience is possible with a lesser system. But music is emotion to me. I'm just trying to maximise that connection to the emotion. As a fun hobby and learning experience.
(lol, all woolly talk from me trying hard to deny I'm actually obsessed with this stuff, it's an addiction, what can I say?)
.I've done much of the same in my Car. fixing little things that seem unimportant otherwise, gradually improving the sound. Just to learn something.
My ears are as easy to fool as anyone else's, that's for sure. So I don't put 100% faith in that. If I could combine all the positive aspects I've heard and had so far in one single setup I would be very pleased. Some days I'm convinced I'm there... only to be disappointed the next day. We as humans have a multitude of filters active, every time we listen to something, being indoors or out. We can filter out a room without really thinking about it. Just record your current sound and play it back on headphones. You'd probably deny it sounds like that in your room.
Taking little steps to make it behave just a little better can go a long way to improve the sound. All this time I still have a functional setup, to play music or movies. But meanwhile I'm testing out different theories and look/listen to the results.
Sure I could say: there, I'm done, finished. But what would I have learned?
I think my rewards are in the (now countless) moments sitting down and listening and being overwhelmed with a piece of music. Taken by surprise basically.
This sort of thing can happen while seeing a movie etc. Even that experience is possible with a lesser system. But music is emotion to me. I'm just trying to maximise that connection to the emotion. As a fun hobby and learning experience.
(lol, all woolly talk from me trying hard to deny I'm actually obsessed with this stuff, it's an addiction, what can I say?)
Last edited:
That was my point, which you summed up well.
It straightens electrical phase. I haven't read up on any of the background as you have, so for now the only advantage I can see is a linear impedance presented to the amplifiers, which I guess would bring some small THD benefit, from the amplifier.
But what impact does it have on the acoustic phase? Guessing again I would say little or none.
Being a natural sceptic with barley the knowledge I need to get out of bed in the morning, Id guess there isn't an audible change. Unless the amplifier THD is significantly variable with impedance, and not good in the beginning.
Perhaps.it is more complex than my simplistic musings, but id agree that the time alignment improvement would hugely outweigh the improvement (if any) from the impedance conjugate.
Would the conjugate network solve more than it causes, with so many elements?
Well who knows?
It straightens electrical phase. I haven't read up on any of the background as you have, so for now the only advantage I can see is a linear impedance presented to the amplifiers, which I guess would bring some small THD benefit, from the amplifier.
But what impact does it have on the acoustic phase? Guessing again I would say little or none.
Being a natural sceptic with barley the knowledge I need to get out of bed in the morning, Id guess there isn't an audible change. Unless the amplifier THD is significantly variable with impedance, and not good in the beginning.
Perhaps.it is more complex than my simplistic musings, but id agree that the time alignment improvement would hugely outweigh the improvement (if any) from the impedance conjugate.
Would the conjugate network solve more than it causes, with so many elements?
Well who knows?
Well who knows?
Not me, so it is a very good reason (for me anyway) to try
.But....
Here are things that indicate changes with an impedance correction network.
In REW I made impulse measurements. Based on that the components were chosen for the schematic I posted.
But if I look at the Group Delay of the impulse measurement I see this:
Group Delay plot of the original impedance measurement...
Group Delay plot of the corrected impulse...
Like I said, tiny differences. Don't look too closely at all the little bumps here, as they will vary from measurement to measurement,
mostly depending on sound (noise) level during the test. The peak at ~19 kHz is also always present with my impedance measurement laptop,
a very old Vista laptop from my son. But things like these differences indicate something is changing, as the sims that BYRTT showed indicated
as well, many pages back in this thread. Things like series and parallel resistors also make a change in these graphs. So to me it seems reasonable
to assume that sonic differences may occur. The quality of the components used may be a whole other chapter though. But I did use reasonable
quality components. Except for maybe the cored inductor. But it is operating well within it's comfort zone as far as I can figure this out.
Last edited:
Yes I can see improvement in group delay due to conjugate removing impedance bump at resonance, and the phase shift at the same point.
Whether it is audible or not largely depends on the level of group delay at resonance....WinISD or similar would give an indication.
Then there is the threshold of group delay at that given frequency. Given you have used sealed box array (assuming from the few pages I skimmed trying to read this monster thread) group delay will already be very low, and likley inaudible anyway.
Many discussions on this, but I'm not sure a concrete time was ever established.
Perfectionism is good, just seems a lot of effort for almost certainly no gain (in my mind at least)
Whether it is audible or not largely depends on the level of group delay at resonance....WinISD or similar would give an indication.
Then there is the threshold of group delay at that given frequency. Given you have used sealed box array (assuming from the few pages I skimmed trying to read this monster thread) group delay will already be very low, and likley inaudible anyway.
Many discussions on this, but I'm not sure a concrete time was ever established.
Perfectionism is good, just seems a lot of effort for almost certainly no gain (in my mind at least)
Thanks wesayso showing those GD plots with and without networks.
What about cone oscilations aren't cones visible being more stable sitting braked at point zero compared without the resonance network. For my A10.2 that point was quite clear and very visible, then knowing that the driver then when oscilating will be a generator and send current that go to amp including amps feedback network, think its better to be braked (motion compensation) out at the network.
What about cone oscilations aren't cones visible being more stable sitting braked at point zero compared without the resonance network. For my A10.2 that point was quite clear and very visible, then knowing that the driver then when oscilating will be a generator and send current that go to amp including amps feedback network, think its better to be braked (motion compensation) out at the network.
Last edited:
Its certainly an interesting math/theory meets real world experiment; I'm just not convinced I'd hear a difference.
To be honest I'm not sure I could hear the difference in a blind test of transient perfect setup such as yours, so much is lost in the recording capture I feel you have to be there to appreciate it - the steering of the array focus etc etc would be extremely difficult to capture in recording.
Heck I'm not sure id hear a difference on a multiway speaker between phase aligned and phase coherent speakers. But I'm sure it is a more worthwhile pursuit. Quadrature or inverted phase is most definitely audible. Transient perfection is something ill.mess with when I have time (when I'm old, retired and have enough time)
To be honest I'm not sure I could hear the difference in a blind test of transient perfect setup such as yours, so much is lost in the recording capture I feel you have to be there to appreciate it - the steering of the array focus etc etc would be extremely difficult to capture in recording.
Heck I'm not sure id hear a difference on a multiway speaker between phase aligned and phase coherent speakers. But I'm sure it is a more worthwhile pursuit. Quadrature or inverted phase is most definitely audible. Transient perfection is something ill.mess with when I have time (when I'm old, retired and have enough time)
I am not sure I follow.your point exactly.
A speaker is always a generator of back EMF with any movement of the cone.
At resonance, isn't the back EMF higher??? I.e. At peak value?
Assuming I'm not talking nonsense, then this EMF is absorbed within the conjugate network, which is acting as an impedance matching network (in a fashion) and as such I'm not sure that the damping of resonance is improved. Power transfer is improved.
It would be easy to measure if it is...a change of 'in box' Qes.
Of course if one is using a amplifier with a high impedance output in the range 4-8 ohms to 'match' the speaker, with conjugate network, then it absolutely would make a worthwhile difference (almost perfect power transfer characteristic and load matching)
Last edited:
When striving for time coherency, every little piece counts. The less digital correction needed to get the same (or better) results, the happier I'll be.
So far I've done all the things within this project to this level of .... (can't find a name for it, might call it "insanity")
But I do know the difference between good and better are little things. And I'd like to end up at: Great!
Lot's of baby steps to take still (lol).
So far I've done all the things within this project to this level of .... (can't find a name for it, might call it "insanity"
)But I do know the difference between good and better are little things. And I'd like to end up at: Great!
Lot's of baby steps to take still (lol).
Well I believe I do hear the difference. When it all aligns in perfect time it starts to sound more real, convincing, provided the recording is good enough.
Very hard to achieve though, in a normal living room and with some dynamics available. But it was my goal in this crazy endeavour. Still have a lot to learn though, I cannot put my finger on everything I hear and know why it sounds like it does. But it is an interesting puzzle.
Wesayo,
I agree, minimum tinkering with DSP would be my aim also.
Part of the reason I have avoided DSP is that Im not convinced its quite good enough. Now if it were 192/96 I would be more than happy (the difference between 48khz and 96khz I quite audible in my experience with cooledit and my soundcard 24 bit for home recording, beyond 192khz and including 32bit floating point, the difference is inaudible to me)
The other part is the low.in/out voltages.
The worst part is my first Serious speaker project and converting it to analogue active, well... That has taken 5 years. With DSP I don't think Id ever stop fiddling...
The next one is a 3 way, mid and HF I have already, bass driver is a little undecided. For me, that challenge is difficult in a passive or analogue crossover. DSP there...well that would make things easier.
I have to end my projects, or ill never get to the next in line!
I agree, minimum tinkering with DSP would be my aim also.
Part of the reason I have avoided DSP is that Im not convinced its quite good enough. Now if it were 192/96 I would be more than happy (the difference between 48khz and 96khz I quite audible in my experience with cooledit and my soundcard 24 bit for home recording, beyond 192khz and including 32bit floating point, the difference is inaudible to me)
The other part is the low.in/out voltages.
The worst part is my first Serious speaker project and converting it to analogue active, well... That has taken 5 years. With DSP I don't think Id ever stop fiddling...
The next one is a 3 way, mid and HF I have already, bass driver is a little undecided. For me, that challenge is difficult in a passive or analogue crossover. DSP there...well that would make things easier.
I have to end my projects, or ill never get to the next in line!
Last edited:
Well the idea for my project here originates from September 2011 so I do know how you feel. The DSP part in JRiver is quite acceptable with it's internal 64 bit processing. I could up sample everything, but leave the songs at it's native sample rate. I have experimented with it though.
My DAC does upsample everything though. I'm still not sure I hear the difference though I feel I do when listening to High Res material.
My DAC does upsample everything though. I'm still not sure I hear the difference though I feel I do when listening to High Res material.
My experience say its audioable, think its little ala be closer to the DC component. Example when i play acoustic guitar having ears so close to its body there is some what we can call little DC component when plectre attack strings and when by hand drum in guitars body. On a normal linear phase system such plectra and hand attack sounds more real than a one with a 360º phaseturn and if one also linearize system phase down to say 20Hz as Barleywater often show those attacks gets so real as it is as sitting with the guitar close to ones ears.
No matter low z amp my cones are very visible more calm, on a CD 2426H it was impossible to see
but audio able.Wesayso's line arrays have quite flat resonance peak compared other system and then maybe resonance network will be less audio able compared what could be for other system. Attach plot what two networks amp looks into and driver looks back at.
Attachments
I'll reserve judgement till I have both sides in line. I'm convinced there are (little) changes, but to hear that after FIR correction has had it's way to correct for identical issues (time correction etc.) it is a bit too hard for me to say anything just yet.
My impression is that it did "calm down" the corrected speaker, highly subjective here though.
Time will tell.
It would appear that the resonance peak in the impedance is effectively and almost totally flattened. I typically measure the Q of a speaker with an impedance sweep and in this case, there is a very broad peak and Q will be very low. Isn't a low Q speaker inherently able to respond quicker (more control authority - almost like it has a very large motor and a very light but stiff cone) to input as it doesn't have to overcome the intrinsic system resonance (there is none)? Kind of like why an AMT pleated tweeter sounds better than a voice coil tweeter - it has no Q resonance peak to overcome. It is "faster" and more dynamic with higher slew rates.
I started the day by listening to.... Led Zeppelin (the new High Res edition)
Wish I hadn't done that. I wanted to work on the other conjugation network.
Now way too much time had passed before going into my "workshop".
At least I got all the wood cut for the network. I'll get it done soon.
Only then my amp will be happy, now it's limping on one leg.
Wish I hadn't done that. I wanted to work on the other conjugation network
.Now way too much time had passed before going into my "workshop".
At least I got all the wood cut for the network. I'll get it done soon.
Only then my amp will be happy, now it's limping on one leg.
If your amp is limping... mine is crawling!
It's ok to enjoy the good stuff!
We sometimes get caught off on little details, we forget the big picture.
As for me, I screwed up something in the path yesterday, and I get echoes!
I need to nail down this thing as it is bugging me in the worst way!
It's ok to enjoy the good stuff!
We sometimes get caught off on little details, we forget the big picture.
As for me, I screwed up something in the path yesterday, and I get echoes!
I need to nail down this thing as it is bugging me in the worst way!
Here's a plot of the resulting impedance and electrical phase (left side):
And here's the right side:
Finally finished the second network...
I hope to run some test measurements soon.
Just to show it isn't a re-use of the left conjugation network here's a picture of the start of the second one
:
Last edited:
Did you hear any difference between the corrected and the non-corrected speaker?I started the day by listening to.... Led Zeppelin (the new High Res edition)
Wish I hadn't done that. I wanted to work on the other conjugation network
Now way too much time had passed before going into my "workshop".
At least I got all the wood cut for the network. I'll get it done soon.
Only then my amp will be happy, now it's limping on one leg.