Does the auto mic setup distinguish between direct radiated sound and reflections?
Would it be useful to split set-up in a part with gated measurement, to get the direct radiated sound to be 'flat' by EQ, and a second, non-gated measurement in which it is the objective to get the reflections 'flat' too, but accomplish this part with accoustic treatment in stead of input signal EQ?
I’m assuming it accounts for (or at least measures) reflections as the mic is @ lp.
The Yamaha system (ypao) is a bit mysterious and not exactly precise as it changes several parameters with one adjustment (even manually) my plan is to upgrade to a better dsp setup (rme-adi2 dac,dbx driverack,or minidsp shd are on the shortlist) in order to have more control and measured response.
What I find so interesting is once you get a minimalistic passive xo in place the dsp is like frosting on the cake.
I’m fairly sure this is all going to lead to some interesting experiences........I’m also sure many have already figured all this out and I’m late to the party, but for those on the fence or think their too old for new tricks all I can say is ......technology has advanced to a point where I believe digital/dsp has surpassed analog. And affordable at that!
The Yamaha system (ypao) is a bit mysterious and not exactly precise as it changes several parameters with one adjustment (even manually) my plan is to upgrade to a better dsp setup (rme-adi2 dac,dbx driverack,or minidsp shd are on the shortlist) in order to have more control and measured response.
What I find so interesting is once you get a minimalistic passive xo in place the dsp is like frosting on the cake.
I’m fairly sure this is all going to lead to some interesting experiences........I’m also sure many have already figured all this out and I’m late to the party, but for those on the fence or think their too old for new tricks all I can say is ......technology has advanced to a point where I believe digital/dsp has surpassed analog. And affordable at that!
Don't know if it has been mentionned, but if you built lower distortion cabinet than you are used to ear, this should be something to consider : H2 harmonic...and about this, Nelson Pass did a little nice H2 generator. He explain well how this can change de depth field, depending it phase :
"So why is the phase important? Well, it's a subtle thing. I don't suppose everyone can hear it, and fewer particularly care, but from listening tests we learn that there is a tendency to interpret negative phase 2nd as giving a deeper soundstage and improved localization than otherwise. Positive phase seems to put the instruments and vocals closer and a little more in-your-face with enhanced detail."
This is from this, and the dedicated H2 generator thread being here. User Antoinel said that depending H2 generated phase, the resultant is same directional movement as ear drum or not.That should explain the effect.
"So why is the phase important? Well, it's a subtle thing. I don't suppose everyone can hear it, and fewer particularly care, but from listening tests we learn that there is a tendency to interpret negative phase 2nd as giving a deeper soundstage and improved localization than otherwise. Positive phase seems to put the instruments and vocals closer and a little more in-your-face with enhanced detail."
This is from this, and the dedicated H2 generator thread being here. User Antoinel said that depending H2 generated phase, the resultant is same directional movement as ear drum or not.That should explain the effect.
Thanks. Missing out on the back wave is no fun.
I wonder how his design handles intermodulated signal.
If the system just measures total amplitude, this might explain your experience with the system.
Yes... [in Jeremy Clarkson's voice]
You must be more lucky with your driver choice than me, I can't get away with passive crossovers with most things.
Welcome to the party!
Last edited:
yah.....it would have to be one heck of a driver! I like the idea of coax with a sub....is that close enough?
thanks robbin, it's refreshing to talk about actual sound!
It seems talking about sound here aggravates some more than others.
Thanks, it's nice talking with you.
I don't know about aggravating people, it is not my intention and what I say is never meant to be personal.
I do like to come in strong though, but that's just because I like feedback as much as my amplifier does. And if my signal isn't strong, the feedback isn't going to be either. My goal is the same as everybody here I would think, our common enemy is distortion of the perceived soundscape.
About crossovers.
I have a little trick up my sleeve that I use in my crossover.
To a certain extent it solves two problems of system frequency response dips/peaks als result of crossovers in more-than-2-way systems, especially if the crossover points are close.
The end goal is a minimum phase system, with flat frequency response.
It can be of use for your coax-and-sub combo, as it is more than 2-way.
Are you interested?
I don't know about aggravating people, it is not my intention and what I say is never meant to be personal.
I do like to come in strong though, but that's just because I like feedback as much as my amplifier does. And if my signal isn't strong, the feedback isn't going to be either. My goal is the same as everybody here I would think, our common enemy is distortion of the perceived soundscape.
About crossovers.
I have a little trick up my sleeve that I use in my crossover.
To a certain extent it solves two problems of system frequency response dips/peaks als result of crossovers in more-than-2-way systems, especially if the crossover points are close.
The end goal is a minimum phase system, with flat frequency response.
It can be of use for your coax-and-sub combo, as it is more than 2-way.
Are you interested?
For high volume, yes. That's three way, you have the same phase/timing issues.
Yes I’m interested.......you can’t know too much!
The only phase/timing issues I have is figuring out how I solved the issue!
Many say there’s no issue to solve, but after you solve it.....it’s quite obvious.
I can relate to this.
Time, and phase alignment improved my perception of depth too.
Do yo mean with prioritizing phase: inter-driver/crossover phase alignment?
And with flat frequency response: that of the total system?
Agreed, I will show it soon. I need some help with further improvement.
What was the 'issue' to begin with for you?The only phase/timing issues I have is figuring out how I solved the issue!
Many say there’s no issue to solve, but after you solve it.....it’s quite obvious.
Hybrid......minimalistic 1st order hp/2nd order lp xo....no resistors!, pseudo biamping stereo a/b have seperate dsp settings, full dsp for sub. It’s a bit convoluted!
Cool looking fwd to it......there was no issue just stumbled across this “optimum phase/timing” thing strictly by accident while dialing (aha moment) and then took me months to figure it out and actually be able to repeat ......now that I have a little better grasp I can move it in/out of focus with the ypao, just have no idea exactly how the ypao operates. There’s no explanations anywhere, it definately plays with phase,delay,and eq but the changes are not definable.
Hence the need for better dsp with real time measurements.
Edit.....Hopefully this subject is helpful or at least interesting to the OP?
Bob
Last edited:
I hope so, otherwise we'll hear from OP I suppose...
Talk about luck!
I ran into a similar problem with the DCX2496. For the longest time it just.. didn't 'click' into place. I never managed to figure it out manually, scratching my head for weeks.
When trying to input filter blocks on it, the frequency number on the display didn't match the frequency response I measured on the microphone... Just blaming my measurements, the precision of the calculated xo+driver response etc etc...
After a long time a good friend of mine, and also owner of one of those devices, had the idea to rule out speakers,amps and microphone and directly looping back into the record channel for Arta... Then I could finally see what I was doing.
We figured out the calculations for HP/LP filter frequency settings/expected response and accidentally discovered a huge distortion problem in the device whenever a +15dB filter setting was used. Distortion measurement of the devices was furthermore usefull and lead to scrapping basically the whole output boards for various reasons.
Talk about luck!
I ran into a similar problem with the DCX2496. For the longest time it just.. didn't 'click' into place. I never managed to figure it out manually, scratching my head for weeks.
When trying to input filter blocks on it, the frequency number on the display didn't match the frequency response I measured on the microphone... Just blaming my measurements, the precision of the calculated xo+driver response etc etc...
After a long time a good friend of mine, and also owner of one of those devices, had the idea to rule out speakers,amps and microphone and directly looping back into the record channel for Arta... Then I could finally see what I was doing.
We figured out the calculations for HP/LP filter frequency settings/expected response and accidentally discovered a huge distortion problem in the device whenever a +15dB filter setting was used. Distortion measurement of the devices was furthermore usefull and lead to scrapping basically the whole output boards for various reasons.
One chart is FR of my elderly DCX2496, razor flat in the audible range.
The other chart is distortion. If you don't know how many zeros are needed to convert "-90dB" (right across the band) to percentage, you can look it up on the web*.
Don't forget, when I measure the DCX2496, a lot of really old gear is also in the measurement loop adding to FR error and distortion.
B.
*quite a few
Attachments
Last edited:
Nope, no clipping.
This was before the ouput mod, with the original +22dBu at the outputs. To say I would have noticed clipping would be understating it lol.
Distortion is relative to signal, present at all amplitudes but only when the value in software is +15dB. At +14.9, it is gone. -15dB is no problem. It's not something you might immediately notice when listening but Arta showed it. It's not a big deal, I just put the filters at +14.9 if I really have to. Seems like a bug.
Your graphs look examplary.
Try this: dial in a 1khz LR12 LP EQ and overlay the captured response curve.
Now try a 1khz LR12 HP EQ and again overlay the response.
Look at the overlays, is the crossover indeed at 1kHz?
Because with both mine, and a third one of a friend, it does not.
From the top of my head the correction factor is the same across the device (1.474). Depending of either low pass or high pass, f must be either devided or multiplied. The only exception being inversed parametric filters, where you will have to use the square root of this number, iirc.
Last edited:
I've checked my notes for the DCX2496. Some corrections:
cascading the crossovers to try eliminate low level phase rotation between
the drivers. It's the filter trick I was talking about earlier.
About that, it works with 4-way too, and xo-points aren't fixed like Dueland's.
Unfortunately the DCX2496 by design isn't capable of setting up multiple
crossovers per output channel. A workaround compromised of parametric EQ
HP/LP combined with a BP filter to transform the slopes to L-R Q does work to
a certain level.
From Linkwitz' site:
So here you can see my 'problem'...
Linkwitz' solution consists of cascading the filters, and adjusting the output
levels of the drivers. Again from his site:
I liked this cascading style so I tried working it out a little more with the same
philosophy. Then this happened:
Excuse me for the huge image.
When I started experimenting with implementing this filter, it was subjectively
less distorting the perceived depth of field. Less lost energy from phase
abnormalies lowers perceived distortion IMO.
Further improvements were made, I integrated low level driver responses
beyond what is considered the normal crossover band, and low-level
cascaded those along the other channels too for example.
The DCX2496's design doesn't allow for cascading filters AFAIK, and even
with the workarounds, implementation quickly reached processing limits.
Going forward I'm working on an Equalizer APO based solution for my 'multi-
cascading' filter.
The DCX will remain a treasured piece, and I will continue using it for other
purposes.
The parametric EQ of DCX2496 gets used by me in this way for parallel
cascading the crossovers to try eliminate low level phase rotation between
the drivers. It's the filter trick I was talking about earlier.
About that, it works with 4-way too, and xo-points aren't fixed like Dueland's.
Unfortunately the DCX2496 by design isn't capable of setting up multiple
crossovers per output channel. A workaround compromised of parametric EQ
HP/LP combined with a BP filter to transform the slopes to L-R Q does work to
a certain level.
From Linkwitz' site:
So here you can see my 'problem'...
Linkwitz' solution consists of cascading the filters, and adjusting the output
levels of the drivers. Again from his site:
I liked this cascading style so I tried working it out a little more with the same
philosophy. Then this happened:
Excuse me for the huge image.
When I started experimenting with implementing this filter, it was subjectively
less distorting the perceived depth of field. Less lost energy from phase
abnormalies lowers perceived distortion IMO.
Further improvements were made, I integrated low level driver responses
beyond what is considered the normal crossover band, and low-level
cascaded those along the other channels too for example.
The DCX2496's design doesn't allow for cascading filters AFAIK, and even
with the workarounds, implementation quickly reached processing limits.
Going forward I'm working on an Equalizer APO based solution for my 'multi-
cascading' filter.
The DCX will remain a treasured piece, and I will continue using it for other
purposes.
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.