24 Jun
2016
24 Jun
'16
10:51 a.m.
Hi there!
I want to display the frequency response (amplitude, impedance and
maybe phase) of my self-build speakers on my homepage. I like the clean
look of gnuplot and I wonder if anyone knows about a manual for linux-
users to make "clean" speaker measurements (with room influences
eliminated as far as possible) to get an output in a style like this:
https://3.bp.blogspot.com/-Zsqjk1N1Sws/V2xYWnunOrI/AAAAAAAAAWE/aYPZLa2x
ItYkpFLA3f1zO25IN_AjPY_PACLcB/s1600/eq_smooth.png
(just 1 amplitude, of course, and with a dB range of 80 to 120, f.e.)
I have some experience in speaker measurement, but unfortunately I'm
not that smart to write my own octave-scripts. I'm just a linux-audio-
USER, you know.
Greets!
Mitsch
24 Jun
24 Jun
12:17 p.m.
Michael Jarosch, Jun 24 2016:
Hi there!
I want to display the frequency response (amplitude, impedance and
maybe phase) of my self-build speakers on my homepage.
Just a starting point,
Michael, and I'm not sure how much help it will be. There is Aliki, which can record
frequency responses and, of course, there are other graphical analyser tools, whose screen
you might capture. Then there is DRC, digital room correction. Maybe DRC has everything in
it, that you need. Just thinking aloud...
...
Ta-ta,
----
Ffanci
* Homepage: https://freeshell.de/~silvain
* Twitter: http://twitter.com/ffanci_silvain
* GitHub: https://github.com/fsilvain
1:33 p.m.
Am Freitag, den 24.06.2016, 12:17 +0200 schrieb Ffanci Silvain:
Michael Jarosch, Jun 24 2016:
Thanks for your response!
As long as I know, the tools you mentioned are not made for my purposes
as the room is explicitly included in the measurement, because in the
end speaker AND room is what to be linearized, frequency wise. I need
something different: The speaker itself, in the best case without the
influence of a room surrounded.
Once I used DLSA, a measurement system that could be plugged to the
printing port of (old) windows-PC. There you could configure a time
window to eliminate early reflections of the room. But now, I don't use
windows - and if I did, my laptop doesn't have any printing port. :)
There is JAPA, JAAA and of course "Room EQ Wizard" REW. But I guess,
they can not export any data to create a diagram that fits my needs.
Would be more or less a screenshot solution and I don't know if I could
be happy with that.
Greets!
Mitsch
Hi there!
I want to display the frequency response (amplitude, impedance and
maybe phase) of my self-build speakers on my homepage.
Just a starting point,
Michael, and I'm not sure how much help it
will be. There is Aliki, which can record frequency responses and, of
course, there are other graphical analyser tools, whose screen you
might capture. Then there is DRC, digital room correction. Maybe DRC
has everything in it, that you need. Just thinking aloud... 
3:17 p.m.
Hi Mitch,
On Friday 24 June 2016, Michael Jarosch wrote:
Am Freitag, den 24.06.2016, 12:17 +0200 schrieb Ffanci
Silvain:
5 years ago I modified jaaa (version 0.6.0) to be able to store
measured data into a file. You could launch it with an additional
flag "-f Filename" and than anytime you freeze the display the data
gets written into the same file so you have to rename that file
between different freezes.
Those patches never found the way into Fon's version (he was in the
last stage of an rewrite and thus the patches did not fit with his
actual version). He obviously does not have any need for storing the
measured data.
I've done something similar (more quick & dirty) for japa (same
version) that is still available.
If you are interested please send me a private note. I can send you
executables as well as the patched sources for jaaa and japa.
Gerhard
Michael Jarosch, Jun 24 2016:
> Hi there!
>
> I want to display the frequency response (amplitude, impedance
> and maybe phase) of my self-build speakers on my homepage.
There is JAPA, JAAA and of course "Room EQ Wizard" REW. But I
guess, they can not export any data to create a diagram that fits
my needs. Would be more or less a screenshot solution and I don't
know if I could be happy with that. 
3:39 p.m.
On Fri, June 24, 2016 6:33 am, Michael Jarosch wrote:
As long as I know, the tools you mentioned are not
made for my purposes
as the room is explicitly included in the measurement, because in the
end speaker AND room is what to be linearized, frequency wise. I need
something different: The speaker itself, in the best case without the
influence of a room surrounded.
The traditional way of doing that is by measuring outdoors in a field.
You still get a reflection from the ground even if there are no trees or
fences nearby, so you either try to minimize that by placing the
microphone very close to the ground, or place both speaker and microphone
on a stand a meter or two tall so that the ground reflection is attenuated
a little.
--
Chris Caudle
5:07 p.m.
On 24/06/2016 14:38, Chris Caudle wrote:
On Fri, June 24, 2016 6:33 am, Michael Jarosch wrote:
Or you can put the speaker on (or in) the ground pointing up, and
suspend the microphone.
>As long as I know, the tools you mentioned
are not made for my purposes
>as the room is explicitly included in the measurement, because in the
>end speaker AND room is what to be linearized, frequency wise. I need
>something different: The speaker itself, in the best case without the
>influence of a room surrounded.
The traditional way of doing that is by
measuring outdoors in a field.
You still get a reflection from the ground even if there are no trees or
fences nearby, so you either try to minimize that by placing the
microphone very close to the ground, or place both speaker and microphone
on a stand a meter or two tall so that the ground reflection is attenuated
a little.
-- Chris Caudle 
5:10 p.m.
On Fri, 24 Jun 2016 16:06:54 +0100
james <james(a)mansionfamily.plus.com> wrote:
On 24/06/2016 14:38, Chris Caudle wrote:
I don't think that's right, as surely the ground would then act as an extended
baffle.
--
Will J Godfrey
http://www.musically.me.uk
Say you have a poem and I have a tune.
Exchange them and we can both have a poem, a tune, and a song.
On Fri, June 24, 2016 6:33 am, Michael Jarosch
wrote:
Or you can put the speaker on (or in) the ground pointing up,
and
suspend the microphone. >As long as I know, the tools you mentioned
are not made for my purposes
>as the room is explicitly included in the measurement, because in the
>end speaker AND room is what to be linearized, frequency wise. I need
>something different: The speaker itself, in the best case without the
>influence of a room surrounded.
The traditional way of doing that is by
measuring outdoors in a field.
You still get a reflection from the ground even if there are no trees or
fences nearby, so you either try to minimize that by placing the
microphone very close to the ground, or place both speaker and microphone
on a stand a meter or two tall so that the ground reflection is attenuated
a little.
-- Chris Caudle
5:36 p.m.
On 24/06/2016 16:10, Will Godfrey wrote:
I suppose you're right, I normally think primarily about driver measurement.
Or you can put
the speaker on (or in) the ground pointing up, and
>suspend the microphone.
I don't think that's right, as surely the
ground would then act as an extended
baffle.
7:39 p.m.
On Fri, June 24, 2016 10:06 am, james wrote:
Or you can put the speaker on (or in) the ground
pointing up, and
suspend the microphone.
As Will pointed out, that will not be the same as true free field response
because any sound which would have radiated to the rear (typically just
low frequencies with a box style speaker) will be reflected back to the
front.
Possibly you could compensate for it, but measurements of output vs polar
angle become difficult (e.g. to measure output at -180 degrees you would
have to put the speaker face down on the ground).
To the original poster: don't forget to spin the speaker around and
measure at various angles off axis. The frequency response of the sound
which is emitted off the main listening axis and gets reflected from
nearby walls has a very big influence on the perceived frequency balance
of a speaker.
--
Chris Caudle
8:31 p.m.
On Fri, Jun 24, 2016 at 01:33:03PM +0200, Michael Jarosch wrote:
Am Freitag, den 24.06.2016, 12:17 +0200 schrieb Ffanci
Silvain:
Whether the room is included or not doesn't depend on the tools,
but only on how you use them. If you don't want to include the
room, you need a measurement setup that avoids room reflections
during the valid part of the impulse response. For speakers, one
method (already mentioned) is to do the measurent outdoors, with
the speaker on the ground and mic suspended above it. The 'ground
effect' can be accounted for later, and anyway most speakers are
use close to a wall anyway so it's not entirely wrong.
I have some reason to suspect that the plot you referred to
was actually based on a measurement using Aliki.
To measure impedance you need to capture speaker voltage and
current at the same time. This can be done with any good stereo
soundcard and some simple electronics (a few resistors is all
that's required).
An alternative to using Aliki is pyjacktools, in particular
if want to process the data using python/numpy/matplotlib.
Ciao,
--
FA
A world of exhaustive, reliable metadata would be an utopia.
It's also a pipe-dream, founded on self-delusion, nerd hubris
and hysterically inflated market opportunities. (Cory Doctorow)
Just a starting point, Michael, and I'm not
sure how much help it
will be. There is Aliki, which can record frequency responses and, of
course, there are other graphical analyser tools, whose screen you
might capture. Then there is DRC, digital room correction. Maybe DRC
has everything in it, that you need. Just thinking aloud...
As long as I know, the tools you mentioned are not made for my purposes
as the room is explicitly included in the measurement, because in the
end speaker AND room is what to be linearized, frequency wise. I need
something different: The speaker itself, in the best case without the
influence of a room surrounded. 
25 Jun
25 Jun
3:26 a.m.
Le 24/06/2016 10:51, Michael Jarosch a écrit :
Hi there!
I want to display the frequency response (amplitude, impedance and
maybe phase) of my self-build speakers on my homepage. I like the clean
look of gnuplot and I wonder if anyone knows about a manual for linux-
users to make "clean" speaker measurements (with room influences
eliminated as far as possible) to get an output in a style like this:
https://3.bp.blogspot.com/-Zsqjk1N1Sws/V2xYWnunOrI/AAAAAAAAAWE/aYPZLa2x
ItYkpFLA3f1zO25IN_AjPY_PACLcB/s1600/eq_smooth.png
(just 1 amplitude, of course, and with a dB range of 80 to 120, f.e.)
I have some experience in speaker measurement, but unfortunately I'm
not that smart to write my own octave-scripts. I'm just a linux-audio-
USER, you know.
Greets!
Mitsch
Hello,
I am using qloud to do that. it compiles over Qt4, Qwt and jackd.
sources are available at http://gaydenko.com/qloud/
Hth,
- Ben
2:35 p.m.
On Sat 25 June 2016, Ben Rouits wrote:
Hi Ben,
thanks for the hint. I just tried to compile qloud myself and failed
with message:
----------- snip ------------------
ExcitForm.cpp:98:20: error: ‘class QwtCounter’ has no member
named ‘setSingleStep’
this->fMinCnt->setSingleStep(1);
^
ExcitForm.cpp:119:20: error: ‘class QwtCounter’ has no member
named ‘setSingleStep’
this->fMaxCnt->setSingleStep(1);
----------- snip ------------------
Have you been able to compile latest version (1.0) without problems?
I've installed all the mandatory development packages so far. I might
have to contact the author.
Thanks
Gerhard
<small rant>
I love Linux a lot and are exclusively using it for 15 years now - but
it's a real pity that you can not launch old executables at all later
distributions easily. Otherwise authors could offer a binary
additionally to the sources that would save possible users a lot of
headaches. In Windows mostly all old *.exe programs from XP area are
running fine under Win7 or above and if you try to get the latest
e.g. gnuplot versions running (fill in program of your choice) that
is not yet in your distribution it's much more easy for a Windows
user to do so as executables are available than for a Linux user..
<\rant>
I want to
display the frequency response (amplitude, impedance and
maybe phase) of my self-build speakers on my homepage. I like the
I am using qloud to do that. it compiles over Qt4, Qwt and jackd.
sources are available at http://gaydenko.com/qloud/
3:37 p.m.
On Saturday 25 June 2016, Gerhard Zintel wrote:
On Sat 25 June 2016, Ben Rouits wrote:
Hi Ben,
thanks for the hint. I just tried to compile qloud myself and
failed with message:
----------- snip ------------------
ExcitForm.cpp:98:20: error: ‘class QwtCounter’ has no member
named ‘setSingleStep’
this->fMinCnt->setSingleStep(1);
^
ExcitForm.cpp:119:20: error: ‘class QwtCounter’ has no member
named ‘setSingleStep’
this->fMaxCnt->setSingleStep(1);
----------- snip ------------------
I solved the above mentioned problem: My latest Mint installation does
not have libqwt6-qt4-dev package available thus I had to compile and
install qwt by myself. Unfortunately the install routine does not
install qwt properly but might be due to my lack of understanding.
I was able to compiled qloud properly but launching it, it bombs out
with:
"symbol lookup error: bin/qloud: undefined symbol:
_ZN10QwtCounter8SetRangeEdd
Now I'm lost :-(
Gerhard
I want to
display the frequency response (amplitude, impedance
and maybe phase) of my self-build speakers on my homepage. I
like the
I am using qloud to do that. it compiles over Qt4, Qwt and jackd.
sources are available at http://gaydenko.com/qloud/ 
26 Jun
26 Jun
1:40 p.m.
Hello Gerhard,
I could compile and run it flawlessly.
I think you installed a wrong version of lib Qwt, you need Qwt6 compiled
against Qt4.
Qwt comes with at least two versions (sometimes 4!) depending on your
distro. I'm using also another program that needs Qwt6, linked against
Qt5 though, so have more pkgs installed:
headers files:
ii libqwt-dev 6.1.2-5 NEEDED
ii libqwt-headers 6.1.2-5 NEEDED
ii libqwt-qt5-dev 6.1.2-5 (not needed for qloud)
library files:
ii libqwt-qt5-6 6.1.2-5 (not needed for qloud)
ii libqwt6abi1 6.1.2-5 (i don't know wth is it)
Hope this helps.
- Ben
Le 25/06/2016 14:35, Gerhard Zintel a écrit :
On Sat 25 June 2016, Ben Rouits wrote:
Hi Ben,
thanks for the hint. I just tried to compile qloud myself and failed
with message:
----------- snip ------------------
ExcitForm.cpp:98:20: error: ‘class QwtCounter’ has no member
named ‘setSingleStep’
this->fMinCnt->setSingleStep(1);
^
ExcitForm.cpp:119:20: error: ‘class QwtCounter’ has no member
named ‘setSingleStep’
this->fMaxCnt->setSingleStep(1);
----------- snip ------------------
Have you been able to compile latest version (1.0) without problems?
I've installed all the mandatory development packages so far. I might
have to contact the author.
Thanks
Gerhard
<small rant>
I love Linux a lot and are exclusively using it for 15 years now - but
it's a real pity that you can not launch old executables at all later
distributions easily. Otherwise authors could offer a binary
additionally to the sources that would save possible users a lot of
headaches. In Windows mostly all old *.exe programs from XP area are
running fine under Win7 or above and if you try to get the latest
e.g. gnuplot versions running (fill in program of your choice) that
is not yet in your distribution it's much more easy for a Windows
user to do so as executables are available than for a Linux user..
<\rant>
_______________________________________________
Linux-audio-user mailing list
Linux-audio-user(a)lists.linuxaudio.org
http://lists.linuxaudio.org/listinfo/linux-audio-user
I want to
display the frequency response (amplitude, impedance and
maybe phase) of my self-build speakers on my homepage. I like the
I am using qloud to do that. it compiles over Qt4, Qwt and jackd.
sources are available at http://gaydenko.com/qloud/
5:48 p.m.
On Sunday 26 June 2016, Benoc3aet Rouits wrote:
I could compile and run it flawlessly.
I think you installed a wrong version of lib Qwt, you need Qwt6
compiled against Qt4.
Qwt comes with at least two versions (sometimes 4!) depending on
your distro. I'm using also another program that needs Qwt6, linked
against Qt5 though, so have more pkgs installed:
headers files:
ii libqwt-dev 6.1.2-5 NEEDED
ii libqwt-headers 6.1.2-5 NEEDED
ii libqwt-qt5-dev 6.1.2-5 (not needed for qloud)
library files:
ii libqwt-qt5-6 6.1.2-5 (not needed for qloud)
ii libqwt6abi1 6.1.2-5 (i don't know wth is it)
Thanks Ben,
those versions are not in my repositories so I had to compile libqwt
by myself from source. Qloud works now after I used your above
mentioned libqwt version - thank you.
Nonetheless I have not found a way to export the measured and
post-processed data though.
On Friday 24 June 2016, Michael Jarosch wrote:
There is JAPA, JAAA and of course "Room EQ
Wizard" REW. But I
guess, they can not export any data to create a diagram that fits
my needs.
Therefore qloud will not help Michael either.
Gerhard
27 Jun
27 Jun
2:20 p.m.
Le 26/06/2016 17:48, Gerhard Zintel a écrit :
[...]
Nonetheless I have not found a way to export the measured and
post-processed data though.
Yep, qloud has just an integrated viewer (using qwt),
but not even printing support. I just need it to print the response
curve. So what i do is a screenshot then print the image.
But if you need post-processing, maybe you can look at the generated
files ? they might be parseable i guess. Or you can contact the author
for feature request ?
Best regards,
- Benoît
28 Jun
28 Jun
1:05 a.m.
Am Montag, den 27.06.2016, 14:20 +0200 schrieb Benoît Rouits:
Le 26/06/2016 17:48, Gerhard Zintel a écrit :
Or you can contact the author for feature request ?
Mmmh, I guess, the author would have ported his project to qt6,
already, if it was something he's still interested in.
Unfortunately, most projects that delivered an opensource solution for
speaker development ("kfilter", "gspeakers") haven't seen a long
time
of activity - at least not long enough that I noticed it before it was
an actually dead project.
Hopefully, qloud isn't dead. Although, it doesn't seem to be an
application to get a nice diagram for displaying, it has really
interesting features like showing 3rd and 5th harmonics.
Building from source failed on my machine, too, with
/usr/bin/ld: cannot find -lqwt
collect2: error: ld returned 1 exit status
Makefile:199: die Regel für Ziel „../bin/qloud“ scheiterte
make[1]: *** [../bin/qloud] Fehler 1
make[1]: Verzeichnis „/home/mitsch/src/qloud-1.0/src“ wird verlassen
Makefile:39: die Regel für Ziel „sub-src-make_default“ scheiterte
make: *** [sub-src-make_default] Fehler 2
Sorry, it's german. :)
Guess, I should really contact the author.
Greets!
Mitsch
4 Jul
4 Jul
1:36 p.m.
Am Dienstag, den 28.06.2016, 01:05 +0200 schrieb Michael Jarosch:
Guess, I should really contact the author.
And that's what I did.
Don't know if it would have happened anyway, but he released a new
version of qloud exactly the same day I asked him for help:
http://gaydenko.com/qloud/download/qloud-1.1.tar.bz2
Unfortunately, I still have problems building it. But I will tell you
if and how I got around those issues, once it is done.
Greets!
Mitsch
7:21 p.m.
On 07/04/16 07:36, Michael Jarosch wrote:
Am Dienstag, den 28.06.2016, 01:05 +0200 schrieb
Michael Jarosch:
This is the first one to build for me since learning about it in this
thread :)
A comment on Arch Aur Qloud page says its compatible with "current gcc".
I'm running gcc-multilib 6.1.1-2. Qwt is 6.1.2-3
I ran:
qmake-qt4
make
executable seems to run fine.
--
www.the-temp-agency.com/lollipop-factory
Guess, I should really contact the author.
And that's what I did.
Don't know if it would have happened anyway, but he released a new
version of qloud exactly the same day I asked him for help:
http://gaydenko.com/qloud/download/qloud-1.1.tar.bz2
Unfortunately, I still have problems building it. But I will tell you
if and how I got around those issues, once it is done.
Greets!
Mitsch
_______________________________________________
Linux-audio-user mailing list
Linux-audio-user(a)lists.linuxaudio.org
http://lists.linuxaudio.org/listinfo/linux-audio-user
7:24 p.m.
On 07/04/16 13:20, Tweed wrote:
On 07/04/16 07:36, Michael Jarosch wrote:
btw, I built from above mentioned tarball not Aur.
--
www.the-temp-agency.com/lollipop-factory
Am Dienstag, den 28.06.2016, 01:05 +0200 schrieb
Michael Jarosch:
This is the first one to build for me since learning about it in this
thread :)
A comment on Arch Aur Qloud page says its compatible with "current gcc".
I'm running gcc-multilib 6.1.1-2. Qwt is 6.1.2-3
I ran:
qmake-qt4
make
executable seems to run fine.
--
www.the-temp-agency.com/lollipop-factory
_______________________________________________
Linux-audio-user mailing list
Linux-audio-user(a)lists.linuxaudio.org
http://lists.linuxaudio.org/listinfo/linux-audio-user Guess, I should really contact the author.
And that's what I did.
Don't know if it would have happened anyway, but he released a new
version of qloud exactly the same day I asked him for help:
http://gaydenko.com/qloud/download/qloud-1.1.tar.bz2
Unfortunately, I still have problems building it. But I will tell you
if and how I got around those issues, once it is done.
Greets!
Mitsch
_______________________________________________
Linux-audio-user mailing list
Linux-audio-user(a)lists.linuxaudio.org
http://lists.linuxaudio.org/listinfo/linux-audio-user
4:34 a.m.
On 07/04/16 19:04, Michael Jarosch wrote:
Am Montag, den 04.07.2016, 13:23 -0400 schrieb Tweed:
hmm, I have one debian stretch machine (gcc 5.4.0-4). I'll give that a
try.
--
www.the-temp-agency.com/lollipop-factory
btw, I built from above mentioned tarball not Aur.
Yeah - it seems to work fine with Arch. But unfortunately, I'm using
debian testing.
Greets!
Mitsch
_______________________________________________
Linux-audio-user mailing list
Linux-audio-user(a)lists.linuxaudio.org
http://lists.linuxaudio.org/listinfo/linux-audio-user
7:29 a.m.
On 07/04/16 22:33, Tweed wrote:
On 07/04/16 19:04, Michael Jarosch wrote:
Was able to build on debian
stretch.
installed libqwt-dev 6.1.2-5 (i think this version takes care of your ld
-lqwt error)
added "QMAKE_CXXFLAGS += -std=c++11" to src/src.pro (found after
googling -std=c++11 warning)
qmake-qt4
make
executed binary bin/qloud
captured sine sweep and plotted.
--
www.the-temp-agency.com/lollipop-factory
Am Montag, den 04.07.2016, 13:23 -0400 schrieb
Tweed:
hmm, I have one debian stretch machine (gcc 5.4.0-4). I'll give that
a try.
--
www.the-temp-agency.com/lollipop-factory
_______________________________________________
Linux-audio-user mailing list
Linux-audio-user(a)lists.linuxaudio.org
http://lists.linuxaudio.org/listinfo/linux-audio-user
btw, I built from above mentioned tarball not Aur.
Yeah - it seems to work fine with Arch. But unfortunately, I'm using
debian testing.
Greets!
Mitsch
_______________________________________________
Linux-audio-user mailing list
Linux-audio-user(a)lists.linuxaudio.org
http://lists.linuxaudio.org/listinfo/linux-audio-user
3091
days inactive
3102
days old
linux-audio-user@lists.linuxaudio.org
24 comments
10 participants
participants (10)
-
Benoît Rouits -
Chris Caudle -
Ffanci Silvain -
Fons Adriaensen -
Gerhard Zintel -
james -
jonetsu＠teksavvy.com -
Michael Jarosch -
Tweed -
Will Godfrey