ARTA

Jump to Latest
R

rich08

Member
Joined 2012
Cousin Billy said:
Time to enter this thread.

Ivo. On behalf of numerous diy speaker builders, thank you. Your program if amazing. There are countless measurement options, and the price is very affordable..

You've shown great class with rich08, maybe he needs to take a good laxative.
Click to expand...
Good laxative occasionally help each.
It was only a question as stupid excuse me.
And I agree, Arta is great, nothing you do not.
 
J

JMB

Member
Joined 2004
Question regarding LIMP RLC and Cable Compensation

Hi Ivo,
In the recommendations regarding Cable Compensation, it is recommended to use the values measured at 45 degress phase. For RLC, it is recommended to use the values at 1000 Hz as long as Impedance is less than 100 ohms, which would certainly be the case for cables. Why the difference?
Thanks,
Jay
 
I

iMat

Member
Joined 2014
artalabs.hr
JMB said:
Hi Ivo,
In the recommendations regarding Cable Compensation, it is recommended to use the values measured at 45 degress phase. For RLC, it is recommended to use the values at 1000 Hz as long as Impedance is less than 100 ohms, which would certainly be the case for cables. Why the difference?
Thanks,
Jay
Click to expand...

In the case of measuring very low impedance I assume that least (S/N) error will be when real and imaginary part are of similar value.

Ivo
 
I

iMat

Member
Joined 2014
artalabs.hr
esgigt said:
to what extend does this answer my question? As far as I'm concerned it's an answer to an antirely different question.
Click to expand...

Maybe it was good answer, as largest error ARTA box has in impedance measurements. It comes from low impedance of resistor voltage divider (10k). A better solution will be if we put to active high input impedance buffers before voltage divider.
Or,
If you use professional USB soundcard you can use headphone output to directly drive measured impedance and reference resistor (in this case good value is 47 ohms). Now, there is no need for voltage divider and reference resistor should be connected directly to instrumentation inputs (which have large input impedance: 0.5-1M).

Ivo
 
E

esgigt

Member
Joined 2013
ente said:
... you can find some data regarding the second proposal of Ivo in the "LIMP Tutorial" page 29/30.

http://www.artalabs.hr/AppNotes/LIMP_Tutorial_Version_2_4_English.pdf

Regards
Heinrich
Click to expand...
Thanks, Heinrich... but I already know electronics enough to be over-familiair with measuring methods...

iMat said:
Maybe it was good answer, as largest error ARTA box has in impedance measurements. It comes from low impedance of resistor voltage divider (10k). A better solution will be if we put to active high input impedance buffers before voltage divider.
Or,
If you use professional USB soundcard you can use headphone output to directly drive measured impedance and reference resistor (in this case good value is 47 ohms). Now, there is no need for voltage divider and reference resistor should be connected directly to instrumentation inputs (which have large input impedance: 0.5-1M).

Ivo
Click to expand...
Ivo, I think the clue is in seperating the measuring-department from the registration-department (Arta) using differential amplifiers. Of course that method introduces a new problem: calibration.

I think that one ought to allow the seperate "grounds" (measurement vs. registration) to float relative to each other. That's why I want to use differential amplifiers.

Cheers.
 
maiko

maiko

Member
Joined 2016
Hi everyone, this is a very long thread and it seems many sounds cards have been discussed, I am looking for a new card, either internal or external, what would be the latest suggestions on a best-bang-for-the-buck card for around 100£$€ or a bit more?

Reading a bit regarding Focusrite and Steinberg it seems theres not only some quality issues with low frequency noise, crosstalk etc., but also reports on glitches using USB, is it best to stay away from all kinds of USB cards?
 
Last edited:
R

rich08

Member
Joined 2012
maiko said:
Hi everyone, this is a very long thread and it seems many sounds cards have been discussed, I am looking for a new card, either internal or external, what would be the latest suggestions on a best-bang-for-the-buck card for around 100£$€ or a bit more?

Reading a bit regarding Focusrite and Steinberg it seems theres not only some quality issues with low frequency noise, crosstalk etc., but also reports on glitches using USB, is it best to stay away from all kinds of USB cards?
Click to expand...

Try out Steinberg UR22mkII, for this money probably nothing better. I use Roland Quad Capture...
 
PRTG

PRTG

Member
Joined 2005
maiko said:
Hi everyone, this is a very long thread and it seems many sounds cards have been discussed, I am looking for a new card, either internal or external, what would be the latest suggestions on a best-bang-for-the-buck card for around 100£$€ or a bit more?

Reading a bit regarding Focusrite and Steinberg it seems theres not only some quality issues with low frequency noise, crosstalk etc., but also reports on glitches using USB, is it best to stay away from all kinds of USB cards?
Click to expand...

After disappointment with Presonus Firestudio Mobile firewire and ESI Maya 22 USB cards I sticked to Roland Quad Capture USB card. In loopback mode measurement 192kHz sample-rate its frequency response was linear up to whopping 80kHz. I had also great results measuring with borrowed Edirol FA101 that at 96kHz sample-rate was linear up to about 35kHz.

Regarding USB cards I have noticed that measurement precision may drift towards less after computer is waken up after sleep mode sometimes giving strange results. I'm not sure if problem is specific to particular computer, driver, soundcard, USB type or PC power management standard. After noticing that it can be solved by plugging card out and back after wakeup I made it routine before starting any measurement session.

ASIO drivers are preferrable over WDM in device setup. I have noticed on some cards that if WDM is used with Windows sound control panel and soundcard hardware samplerate settings set at different values weird errors may appear, like 1-st or 2-nd order LPF gets applied at 16kHz (forced antialiasing filter applied?) which may give 3-6dB decline at 20kHz in loopback test.

I prefer Windows 10 over Windows 7. I have noticed that THD measurements differ significantly between these platforms with everything else in setup being the same (computer, card, mic, amplifier). Windows 10 seems to give more reasonable and more stable THD measurement results.

Hope this helps.
 
maiko

maiko

Member
Joined 2016
Thanks a lot PRTG for your reply and valuable information, I will have to think what way to go because I already have a quite good sound card, but it is with an old PCI connector which means I have to use an older mother board, so I am also looking into finding some kind of adapter so I can have the card in its own shielded box outside the noisy computer box, something like a PCI-E to PCI cable would be great.

I have used ASIO4ALL in the past which is still supported all the way back to Win 98SE, but I never did anything advanced with my soundcard so I couldn't tell much, back then I just wanted to improve the playback of music, it's quite long time agoso my memroy is rusty but I think there were some small improvements.
ASIO4ALL - Universal ASIO Driver

What kind of SW did you use to verify the frequency response linearity (80 kHz) of your setup?
 
Last edited:
PRTG

PRTG

Member
Joined 2005
maiko said:
What kind of SW did you use to verify the frequency response linearity (80 kHz) of your setup?
Click to expand...

It is best to use manufacturer's ASIO drivers and using ASIO4ALL only when there are none available.

It is best to use some ultralinear card for reference measurements. If you don't have one, you may stick to assumption that loopback test is good enough to determine soundcard's capabilities. Connect card's input to output using cable with connectors that fits its input and output physical sockets, set up ARTA to use particular cards inputs and outputs (see manual) and then measure frequency response with ARTA.

P.S. If your card has "Mix" knob that regulates mix output balance between "Playback" and "Input" like Quad Capture you have to position the knob towards "Playback", otherwise you may run into oscillation or less precise measurement results.
 
Last edited:
E

ente

Member
Joined 2004
ARTA 1.9.0 published

ARTA release version 1.9.0 is published (ARTA Download).

Changes from version 1.8.5 are:

ARTA
- New command added for automatic spatial IR group recording using rotating turntable *), **)
- New menu command added in Imp window for export of FR for single file IR or spatial IR Group (group export of frd, txt, csv files for simulation)

*) ARTA accepts two drivers for rotating turntables:
--> command line program defined for DIY turntable
--> internal driver for Outline RT 250-3D (this driver is not fully tested, yet).

**) Info DIY-turntable
http://www.artalabs.hr/AppNotes/ARTA-Hardware&Tools - Annex 1 - AutomaticTurntable-GerRev1.3.pdf
http://www.artalabs.hr/AppNotes/AP-9 ARTA-AutomaticTurntable-GerRev1.4a.pdf

- New graph grid labeling algorithm uses grid on .., 1, 2, 5, 10, .. values.
- Now user chooses direction of Magnitude Top spin control function.
- New pink noise filter applied in continuous pink noise generation.
- Separate switches added for frequency compensation in FR and spectrum windows.
- Power spectrum and distortion calculations now accept fr. compensation.
- Number of averaging for noise and swept sine excitation is handled separately.
- Ctrl+A command now works in DFT window.
- Graph Copy in CSD and Burst Decay views accepts different view sizes.
- A bark scale cursor readout corrected.
- A change to lower resolution available for FR difference from overlay.

LIMP
- In LIMP a report for frequency Fs is obtained from optimization.
- In LIMP estimation of Mms uses Small formula if optimization obtained one differs more than 10%

Regards
Heinrich
 
E

ergo

Member
Joined 2001
This is interesting. Especially the turntable project. I have used Kimmo's Arta recorder + manual table but automated would be so nice especially with VituixCAD accepting full polars easily.

I had Google Translate to turn this to English (as best as it was able to)... file is here if anyone else wants to get a bit better idea and does not know German either.
https://1drv.ms/b/s!AvVPh3wl37NzgZMCKMyT9GeyhgRuRw

Seems that the PCB+controller is for sale here for 99EUR
ARTA Turntable
and the other items are also reasonable price - so would make a nice DIY project.
 
You must log in or register to reply here.
Top Bottom