Hi Mark
I am using the FAITAL PRO 5 PR 160 A.
Crosslite: yes, a fantastic toolset! And I am still just scratching the top of the possibilities 😉
You can send your measurments to Francisco… he is always happy, when he get‘s traces 😉 By the way: thy look pretty good!
What are you using for the FIR playout? Seems, that you are using very long filters?
At the moment, I am just using the FIR crossovers in the DSPs and for linearisation i am using IIR PEQs. In the past, I used also very long FIR filters and used the inverse oft the frequency response. That end up in very good step response and of course a very good frequency response. Ringing was also fine but…
It always sound a little bit boring. Now I am measuring the frequency response per chassis at ~10 positions in my listening area… taking the average and linearize the response of the used frequency responswithna bit overlap,for the crossover, the crossover comes on top of it.
This results in a not perfect frequency response, but to me this sound more natural.
My Theorie behind: to much correction, changes the sound, that it doesn‘t fit into the listening room. If I place a piano in my listening room, it will also not sound perfect… but this is, what the brain is expecting 😉
But, just my Theorie… 😉
Tomorrow i will build wheels for the second speaker, and than, later on i will do some measurments… maybe on Sunday.
I am using the FAITAL PRO 5 PR 160 A.
Crosslite: yes, a fantastic toolset! And I am still just scratching the top of the possibilities 😉
You can send your measurments to Francisco… he is always happy, when he get‘s traces 😉 By the way: thy look pretty good!
What are you using for the FIR playout? Seems, that you are using very long filters?
At the moment, I am just using the FIR crossovers in the DSPs and for linearisation i am using IIR PEQs. In the past, I used also very long FIR filters and used the inverse oft the frequency response. That end up in very good step response and of course a very good frequency response. Ringing was also fine but…
It always sound a little bit boring. Now I am measuring the frequency response per chassis at ~10 positions in my listening area… taking the average and linearize the response of the used frequency responswithna bit overlap,for the crossover, the crossover comes on top of it.
This results in a not perfect frequency response, but to me this sound more natural.
My Theorie behind: to much correction, changes the sound, that it doesn‘t fit into the listening room. If I place a piano in my listening room, it will also not sound perfect… but this is, what the brain is expecting 😉
But, just my Theorie… 😉
Tomorrow i will build wheels for the second speaker, and than, later on i will do some measurments… maybe on Sunday.
Hi Christian,
I'm looking forward to sending Francisco some examples of my work, to help ask about his FIR add on.
Still trying to get a better versed on Basic, to better what to ask about.
I'm using q-sys and 16,384 taps per channel @ 48k, lin phase. I use the same tap count on all channels because I don't want to fool with delays, and the q-sys Core can handle it.
I 1000% agree too much correction doesn't work. My normal method is to collect raw polars outdoors for each driver section, and average them to obtain a measurement to correct. That said, I've built enough of the conical unity horns to know 10-15 degree off axis makes a pretty good single axis reference.
Also, I never bother with room tuning...just quasi-anechoic. For room tuning, i just use real time EQ, like mixing live sound.
A bigger problem than polar averaging has been at what distance to measure, especially for the narrower pattern horns where the CD is farther from mouth.
Close to the mouth doesn't work ...inverse square laws are impacted, with the mic so close to outer ports. (this one is 75x60 degree)
The mouth tuning i showed above is very bass shy out in the room.
I tried the Vary-Time Transform noise reduction an all 5 sections.,...it has its own degree of smoothing in addition to impulse cleanup.
Have you tried it yet?
And It's not as neat as the 1/12th smoothing show before either, to avoid overcorrecting. 1/48th below.
It was more about learning the program., than trying for a real tuning.
But once i dialed up the sub and low sections, it did sound pretty damn good ...plus the syn11 sits on a double 18" sub cart.
I'm looking forward to sending Francisco some examples of my work, to help ask about his FIR add on.
Still trying to get a better versed on Basic, to better what to ask about.
I'm using q-sys and 16,384 taps per channel @ 48k, lin phase. I use the same tap count on all channels because I don't want to fool with delays, and the q-sys Core can handle it.
I 1000% agree too much correction doesn't work. My normal method is to collect raw polars outdoors for each driver section, and average them to obtain a measurement to correct. That said, I've built enough of the conical unity horns to know 10-15 degree off axis makes a pretty good single axis reference.
Also, I never bother with room tuning...just quasi-anechoic. For room tuning, i just use real time EQ, like mixing live sound.
A bigger problem than polar averaging has been at what distance to measure, especially for the narrower pattern horns where the CD is farther from mouth.
Close to the mouth doesn't work ...inverse square laws are impacted, with the mic so close to outer ports. (this one is 75x60 degree)
The mouth tuning i showed above is very bass shy out in the room.
I tried the Vary-Time Transform noise reduction an all 5 sections.,...it has its own degree of smoothing in addition to impulse cleanup.
Have you tried it yet?
And It's not as neat as the 1/12th smoothing show before either, to avoid overcorrecting. 1/48th below.
It was more about learning the program., than trying for a real tuning.
But once i dialed up the sub and low sections, it did sound pretty damn good ...plus the syn11 sits on a double 18" sub cart.
Hi Christian and Mark
I am a bit curious of how you two assess/compare Crosslite+ to Acourate. What is it that you can do in Crosslite+ that you can not do in Acourate? Is one easier to work with than the other?
I am asking, because I have been contemplating to purchase Acourate for some time now. I have the hardware to go 4-way FIR from a PC and have planed to upgrade to a more powerful fanless PC.
Have a nice weekend
Steffen
EDIT: I just discovered that there was a page 8 to this thread! 🙄 Maybe it is a bit early to ask? Hmm.
I am a bit curious of how you two assess/compare Crosslite+ to Acourate. What is it that you can do in Crosslite+ that you can not do in Acourate? Is one easier to work with than the other?
I am asking, because I have been contemplating to purchase Acourate for some time now. I have the hardware to go 4-way FIR from a PC and have planed to upgrade to a more powerful fanless PC.
Have a nice weekend
Steffen
EDIT: I just discovered that there was a page 8 to this thread! 🙄 Maybe it is a bit early to ask? Hmm.
Last edited:
Hi Mark
Q-Sys is able to use 16k Taps? Do you have any information or link to the Product?
Looking for a long time for Hardware based solution for large FIR filters.
Thanks!
Q-Sys is able to use 16k Taps? Do you have any information or link to the Product?
Looking for a long time for Hardware based solution for large FIR filters.
Thanks!
Comparting Acourate to Crosslite is not easy… depends on what you like to do.
Crosslite is very powerful in finding the right timings for a multi way setup.
Both are not easy. Acourate is easy if you just follow the macros… but this is for my ears in my room not the way to go.
Acourate and crosslite are both VERY powerful toolsets and I am using both.
Acourate is not easy to get help! Uli is very kind and will help where ever he can, but there is not a big user group.
Crosslite: if you buy the softwhere you get access to a Facebook user group where a lot of people try to help. Additional ther is once a week a zoom meeting, where you can answer your questions
Powerful PC: i don‘t think that you need a very powerful pc. I was doing 4 way two channel 88.2kHz with 65k Taps on an old 2012 Mac Mini which is not even getting warm.
Crosslite is very powerful in finding the right timings for a multi way setup.
Both are not easy. Acourate is easy if you just follow the macros… but this is for my ears in my room not the way to go.
Acourate and crosslite are both VERY powerful toolsets and I am using both.
Acourate is not easy to get help! Uli is very kind and will help where ever he can, but there is not a big user group.
Crosslite: if you buy the softwhere you get access to a Facebook user group where a lot of people try to help. Additional ther is once a week a zoom meeting, where you can answer your questions
Powerful PC: i don‘t think that you need a very powerful pc. I was doing 4 way two channel 88.2kHz with 65k Taps on an old 2012 Mac Mini which is not even getting warm.
Hi Mark
It seems, that the English website is better than the German 😉
But I cant find informations regarding FIR filters.
Measurements:
What do you like to see?
It seems, that the English website is better than the German 😉
But I cant find informations regarding FIR filters.
Measurements:
What do you like to see?
Hi Christian
Thanks for your explanation on the two softwares.
Very interesting that you could do a stereo 4-way system on an old Mac! As of now my hifi-PC is a 5. genaration intel i3-5010U NUC from 2015. Hmm, so I could actually get started wit my 3-way setup that I have now. Good news.
I am pretty shure that Mark uses the
Q-SYS Core 110f.
That is the hardware he always refers to. You can find them on Ebay. Before purchasing one you should speak to Mark about what to take into account/be aware of, as he knows of some issues.
Best regards
Steffen
Thanks for your explanation on the two softwares.
Very interesting that you could do a stereo 4-way system on an old Mac! As of now my hifi-PC is a 5. genaration intel i3-5010U NUC from 2015. Hmm, so I could actually get started wit my 3-way setup that I have now. Good news.
I am pretty shure that Mark uses the
Q-SYS Core 110f.
That is the hardware he always refers to. You can find them on Ebay. Before purchasing one you should speak to Mark about what to take into account/be aware of, as he knows of some issues.
Best regards
Steffen
Hi Christian,
The hardware I use are Q-Sys "Cores", which are basically Unix PCs dedicatedly solely to running Q-Sys processing schematics.
The open architecture schematics are made using Q-Sys Designer software on Windows PCs, which compiles the schematics into the Core.
So the Core ends up being a standalone hardware processor, without further requirement of a PC connection.
The schematics' can be anything....from next to nothing to running a Disney theme park. Just depends on the power of the Core.
Most folks in home audio go with a Core110f. Silent, USB, lots of analog balanced I/O.
My experience is that it's FIR capability is limited 6144 taps per ch, for maybe up to 8 channels. 6 channels for sure. 4k taps works for 8ch.
(All Q-sys is 48kHz only)
Discontinued Cores that use I/O cards like the Core250i, Core500i, and Core510i...can all run 16k taps per channel, which is the current limit allowed in schematic designs. I've used a 500i and a 510i, to run 15 channels of 16k taps per ch. Dunno about a 250i, but I feel pretty confident it can run at least 8 ch of 16K taps.
I have zero reservation with the only recently discontinued card Cores. QSC has made the strategic decision to enable all peripherals, amps included, with network capability, by passing local I/O.
Must admit, it's easier routing audio to my two 4 channel QSC network amps via ethernet, that running XLR lines from Core cards to other regular amps.
I've bought all my qsys gear used on ebay, Cores and amps.
With patience and constant monitoring, incredible buys can be had.
250i and 500i Cores are inexpensive right now. The trick is finding one with the right I/O cards one needs, already in it.
For under well under $1000, a processor can be had that totally smokes anything else I can find regardless of price (other than other newer Qys stuff)
Hi Steffen, I don't have any experience with Acourate other than having read the manual and Mitch's book a few times. I look to Christian here.
Frankly though, Acourate seem overly complicated to me. Maybe it's great for room correction, but I can't see using it for speaker design and measurement.
I think measurements and FIR generators are really two different things....at least from the perspective of DIY speaker building.
The hardware I use are Q-Sys "Cores", which are basically Unix PCs dedicatedly solely to running Q-Sys processing schematics.
The open architecture schematics are made using Q-Sys Designer software on Windows PCs, which compiles the schematics into the Core.
So the Core ends up being a standalone hardware processor, without further requirement of a PC connection.
The schematics' can be anything....from next to nothing to running a Disney theme park. Just depends on the power of the Core.
Most folks in home audio go with a Core110f. Silent, USB, lots of analog balanced I/O.
My experience is that it's FIR capability is limited 6144 taps per ch, for maybe up to 8 channels. 6 channels for sure. 4k taps works for 8ch.
(All Q-sys is 48kHz only)
Discontinued Cores that use I/O cards like the Core250i, Core500i, and Core510i...can all run 16k taps per channel, which is the current limit allowed in schematic designs. I've used a 500i and a 510i, to run 15 channels of 16k taps per ch. Dunno about a 250i, but I feel pretty confident it can run at least 8 ch of 16K taps.
I have zero reservation with the only recently discontinued card Cores. QSC has made the strategic decision to enable all peripherals, amps included, with network capability, by passing local I/O.
Must admit, it's easier routing audio to my two 4 channel QSC network amps via ethernet, that running XLR lines from Core cards to other regular amps.
I've bought all my qsys gear used on ebay, Cores and amps.
With patience and constant monitoring, incredible buys can be had.
250i and 500i Cores are inexpensive right now. The trick is finding one with the right I/O cards one needs, already in it.
For under well under $1000, a processor can be had that totally smokes anything else I can find regardless of price (other than other newer Qys stuff)
Hi Steffen, I don't have any experience with Acourate other than having read the manual and Mitch's book a few times. I look to Christian here.
Frankly though, Acourate seem overly complicated to me. Maybe it's great for room correction, but I can't see using it for speaker design and measurement.
I think measurements and FIR generators are really two different things....at least from the perspective of DIY speaker building.
Hi again Steffen, I love the 110f, and use it for all kinds of quick experiments...especially checking out electrical filters, etc.
But after moving beyond 3-way syns and a sub, a stereo setup using 8 channels of 4k taps per channel, was the FIR limit.
And I wanted more, especially after adding small mids and needing 10ch for stereo.
Really though, the 110f would be fine for that too, as long as a less steep lin phase sub to main xover is used . (IIR should always be sub hpf imo)
Anyway, I now use a 510i, which I grabbed off ebay for about $1300, with eight I/O cards including 64 ch Dante !
15ch of 16k taps is no problem when running syn10 LCR.
btw Christian , I meant to put some Qsys links in my reply..
Designer software...https://www.qsys.com/resources/software-and-firmware/q-sys-designer-software/
Online Training ......which is absolutely outstanding and comprehensive....a MAJOR feature https://training.qsc.com/mod/page/view.php?id=560
Hardware Core's
110f... https://www.qsys.com/resource-files...p_cores/core_110f/q_dn_core_110f_specs_v2.pdf
(this spec sheet is for V2, a slimmed down V1 version that doesn't include GPIO due to covid chip issue i think...V1 is what's going to be found on ebay.
Make sure any V1 shows front panel screen picts beyond the big QSC startup screen...if not, the SSD drive may be compromised ...$600 repair bill)
250i, 500i ....https://www.qsys.com/resource-files/productresources/dn/dsp_cores/q_dn_core_250i_500i_specs.pdf
510i .....https://www.qsys.com/resource-files/productresources/dn/dsp_cores/core_510i/q_qsys_core_510i_specs.pdf
Thanks a lot!
Will take a Look!
Please let me know, what kind of measurements you like to have… and recording distance 😉
Will take a Look!
Please let me know, what kind of measurements you like to have… and recording distance 😉
I found this on the QSC site:
https://q-syshelp.qsc.com/Content/Schematic_Library/filter_FIR.htm
Say say max 16k Taps…. And they don‘t define any core… so it should work on any?
https://q-syshelp.qsc.com/Content/Schematic_Library/filter_FIR.htm
Say say max 16k Taps…. And they don‘t define any core… so it should work on any?
Good you found the help file on FIR.
The software can allow more than a Core can handle, as it's built for all Cores. You start a schematic by specifying which Core.
The power of the particular Core determines how many taps and channels it can run.
A Core110f will not run over 1 ch of 10,000 taps....
And like said, the larger Cores are good for 15 channels of 16k per channel.
You may want to download free Q-sys Designer. It lets you build schematics on your PC, and play with all the features.
It just won't pass audio without a Core. https://www.qsys.com/resources/software-and-firmware/q-sys-designer-software/
Any and every component you pull into the schematic, like a FIR filter, has an online help file like you found.
This is a great free way to see if Qsys works for ya....i'd suggest starting the free Q-sys Designer training course.
And it's a MUCH easier learn that Crosslite for example....
Your measurements... I'd love to see how the 5"s work on your horn. That is a much longer CD to port build than I think I've seen anywhere.
The raw response of the 5"s, and their relative timing to the CD ,is what i'm most curious about.
Don‘t worry about the delays… will be removed by the DSP later on!
What is not working in the HP6040 ist the distance from the mids to the tweeter… I moved them out of the slot forward to tha mouth, because i had to many problems in the frequency response of the tweeter. I know, I am leaving the lambda/4 rule here… so: no lunch for free 😉
So it seems, that the HP6040 is not the best „playground“ for a meh, but the resulting sound is really better, than expected.
I am sure, this is not my last one 😉
Maybe, I will start to build the next from scratch… but this will be very time consuming 😉
What is not working in the HP6040 ist the distance from the mids to the tweeter… I moved them out of the slot forward to tha mouth, because i had to many problems in the frequency response of the tweeter. I know, I am leaving the lambda/4 rule here… so: no lunch for free 😉
So it seems, that the HP6040 is not the best „playground“ for a meh, but the resulting sound is really better, than expected.
I am sure, this is not my last one 😉
Maybe, I will start to build the next from scratch… but this will be very time consuming 😉
Let me make sure I understand the measurements.....particularly the last one 'Tweeter and Midrange'.
Looks like they are both raw without any xovers in place. ?
(I guess the 150Hz hp and 600Hz xover you mentioned are what you decided to use on the processed posts you showed earlier, but not involved now.)
Looks like there is about a 3.5ms delay between the mids and tweeter.
Is all that correct?
What's the physical distance between the mid ports and tweeter?
Eyeballing the HP6040 drawings, and assuming you tried to get the ports as early in the horn as possible,
I'm guessing they ended up somewhere around 6 cm from tweeter.
That would account for a little over 1ms of delay, but I don't get where the remaining 2.5ms would come from without a low pass in play on the mids,
which doesn't appear to be in place????
Very curious....
Looks like they are both raw without any xovers in place. ?
(I guess the 150Hz hp and 600Hz xover you mentioned are what you decided to use on the processed posts you showed earlier, but not involved now.)
Looks like there is about a 3.5ms delay between the mids and tweeter.
Is all that correct?
What's the physical distance between the mid ports and tweeter?
Eyeballing the HP6040 drawings, and assuming you tried to get the ports as early in the horn as possible,
I'm guessing they ended up somewhere around 6 cm from tweeter.
That would account for a little over 1ms of delay, but I don't get where the remaining 2.5ms would come from without a low pass in play on the mids,
which doesn't appear to be in place????
Very curious....
Hi Mark… please don‘t care about the delays! I have a Crossover at i think 150 Hz in the DSP to protect the tweeter! Pleas only look the responses itself and don‘t compare the delays to each other!
I have moved the mids to the front, because the holes in the slot where destroying the tweeter response! Will post a picture what makes it easy to see 😉
I have moved the mids to the front, because the holes in the slot where destroying the tweeter response! Will post a picture what makes it easy to see 😉