Good combo of software/hardware for doing audio measurements

[SeanB]

I am looking to upgrade PC audio slightly, so will grab a standard component amplifier from storage, and put into it a Toslink audio output unit I got a while ago ( with the world's worst wall wart PSU, but I knew that already) and then power it off the internal rails with a 5V regulator, leading the optical cable out to an external USB card with toslink output. Should improve things a little by removing the PC noise.

Originally i used a first generation ISA Sound Blaster Pro, and to reduce the noise I put some copper foil shielding on each side of the card connected to the card grounding on the mounting ear. Improved things a lot, though I had to leave a slot open on the one side and put it on the end of the ISA bus.

[_Wim_]

I use the Visual Analyzer software: http://www.sillanumsoft.org/

As for the hardware i made my own ASIO USB sound card with galvanic isolation and the best audio ADC i could get from TI.

Wow, I would be interested in one of these as wel...

[_Wim_]

As to software, I also use ARTA and sometimes Holmimpluse (for loadspeaker measurements only).

[jackenhack]

I'm thinking of getting the PCM4222EVM Evaluation board http://www.ti.com/tool/pcm4222evm and connect it to a USBStreamer http://www.minidsp.com/products/usb-audio-interface/usbstreamer. The specs on the PCM4222 are insane! And feeding it with a really low noise power supply. The PCM4222 Evaluation board costs around $145 and the USB to I2S input costs around $100.

[davorin]

Hmm...XMOS....sounds familiar....I must have some XMOS eval boards lying around here with USB and audio Interfaces (o;

[jackenhack]

Hmm...XMOS....sounds familiar....I must have some XMOS eval boards lying around here with USB and audio Interfaces (o;

Ebay is full of XMOS I2S to USB, but their only for output to DAC, not to ADC unfortunately. I really want to get a really good test rig for audio measurements, but except for the QA400 there isn't anything out there. Of course, I could sell my body to science, and have enough money to buy a cable from Audio Precision. 

[Timur Born]

I am looking to upgrade PC audio slightly, so will grab a standard component amplifier from storage, and put into it a Toslink audio output unit I got a while ago ( with the world's worst wall wart PSU, but I knew that already) and then power it off the internal rails with a 5V regulator, leading the optical cable out to an external USB card with toslink output. Should improve things a little by removing the PC noise.

Yes, putting a non conducting optical link in between gets rid of all the PC noise. I am using this combination myself for daily use, connecting all kinds of stuff (including an old X-Fi) to a Fireface UC and then going optical to a Babyface with its big knob right in front of me for easy volume control.

Originally i used a first generation ISA Sound Blaster Pro, and to reduce the noise I put some copper foil shielding on each side of the card connected to the card grounding on the mounting ear. Improved things a lot, though I had to leave a slot open on the one side and put it on the end of the ISA bus.

Problem nowadays is that the *ground* itself is full of noise. A high performance graphic-card will pollute all grounds on your mainboard and feed its noise to earth over every possible cable connection. Then the combination of how you have your devices connected to the same or different power-strips decides about how much of the noise if audible on 3-prong connected speakers (simple 2-prong desktop ones usually don't suffer). There can be cancellations and there can be summing of the same noise signal traveling via different path. For example:

PC -> DVI -> display -> power-strip 1 -> power-strip 3
PC -> USB -> audio interface -> speaker -> power-strip 2 -> power strip 3

If you are only suffering from noise on the speakers (there can be other issues) then something as simple as breaking pin 1 (shield) on an XLR cable (or S on a TRS) will help. Of course you could risk your life by getting rid of earth, but that really isn't the solution out of the various that one should consider. Notice how the workaround are more or less the same as the ones used to get rid of mains 50/60 Hz hum loops.

Another software I know about: http://www.hpw-works.com/index.php

Btw, it seems that none of these programs offer brown noise generation, except for DFFS3?!

[dom0]

If you are only suffering from noise on the speakers (there can be other issues) then something as simple as breaking pin 1 (shield) on an XLR cable (or S on a TRS) will help. Of course you could risk your life by getting rid of earth

You don't get rid of earth by lifting the shield of a signal cable. You do however remove that earth connection which might improve things, if the differential input amp is good enough.

[Timur Born]

You misinterpreted my sentence, and my writing may not be top of the game sometimes (non-native here).

I meant: the simple solution is to break pin 1 XLR, the dangerous - and different - solution would be to break earth (on the power-plug). So two alternatives in two sentences, no direct connection between the two.

[Berni]

Yes XMOS is the most popular solution for async USB audio right now. Does work pretty well to be fair.

Connecting one of those XMOS interfaces to the PCM4222 eval board would give you a pretty good system if you ask me, but what "Alex Nikitin" has is pretty much as good as you can get. Past about 0.0001% you sort of get in to the range there the internal noise of opamps might start becoming the limiting factor even if you use the best opamps out there as well as things like thermal resistor noise becoming an impenetrable wall.

[Timur Born]

What do you want to measure anyway? Your onboard audio may be sufficient, maybe coupled with Asio4all if you need smaller latencies (and use a software that supports Asio).

And increasing rise times on your graph displays already goes a long way to average out noise (floor) from any (semi-)constant signal that you want to measure.

[jackenhack]

What do you want to measure anyway? Your onboard audio may be sufficient, maybe coupled with Asio4all if you need smaller latencies (and use a software that supports Asio).

And increasing rise times on your graph displays already goes a long way to average out noise (floor) from any (semi-)constant signal that you want to measure.

For me, I just want the test equipment to be better than the DUT. I had a EMU 0404 USB external device that actually was very good, but the drivers for it was crap, and to get a consistant input was almost impossible because of a bad implementation of the rotary encoders for the input. I want to be able to see if my modifications has a positive impact on the tested device. It would be nice to have a calibrated input so real measurements could be done (which excludes RMAA), but that's not my main priority.

[Alex Nikitin]

what "Alex Nikitin" has is pretty much as good as you can get.

As a side note - you don't need quotation marks, it is my real name  .

Cheers

Alex

[Berni]

As a side note - you don't need quotation marks, it is my real name  .

Cheers

Alex

Or so you would want us to believe

Just joking.

[davorin]

Already got feedback from Fast Precision:

Since you're in Europe, I highly recommend you contact our affiliated U.S partner: Quantasylum Inc. for purchasing the audio analyzer (cced here), our FP8400 is simply the Chinese version of QA400 from Quantasylum. With Quantasylum, You'd get a better price, faster shipping time, etc.

thanks and best regards.

S. Chen

[davorin]

Hmm....if I recall correctly....I never saw anything about phase errors in any of the software I tested....
Guess it would be difficult with USB ADC/DAC devices due to latency...

[hayatepilot]

I use a miniDSP UMIK-1 for measuring audio.

It has an almost unbeatable bang per buck since it comes fully individually calibrated. With REW you can even use it as a sound level meter for absolute measurements. 

And since it has the soundcard built in, you don't have to worry about a preamplifier and gainstructure.
If you have a HDMI capable AV-receiver you don't even need an analog audio output since you can adress all channels of even a 7.1 system individually over HDMI. Guide to HDMI measurements using the UMIK-1

For non professional use, this is more than you ever need. 

Greetings

[Timur Born]

If your are into iOS devices then this one combined with their respective software looks neat:

http://www.studiosixdigital.com/audio-hardware/iaudiointerface2/iaudiointerface2-informatio.html
http://www.studiosixdigital.com/audiotools/

[davorin]

I also got the UMIK-1...great device at a great Price :-)

So alread two of those microphones in Switzerland then (o;


Hmm...seems not many Linux Software out there for doing frequency/phase repsonse and THD measurements...

[Tim F]

Using a E-MU 0204 USB for DAC/ADC duties and a Dayton Audio EMM-6 mic for measuring speakers here. I used mostly self written MATLAB scripts for software, but use HOLMImpulse to generate sweeps since it does a good job of automating the sweep generation, playback, recording, recording offset detection and saving the recording to a wav file.


http://feleppa.com.au/speakermeas.html

[davorin]

My QA400 arrived this morning now in Switzerland...ordered it last Wednesday only as I was promised I would be in the first round to get hands on the new QA405 beta device coming this autumn...

So will keep you posted on my experience with this device....

[davorin]

Hooked it up to my home PC now running Win 8.1 x64....

Calibration went through fine...not on VMWare Fusion though....

But a simple FR measurement runs forever and only left Signal and clipping LEDs are blinking...nothing else Happening....

[davorin]

My fault (o;

The first installed 1.055 Version doesn't get roperly replaced by the newer 1.22 Version...also it doesn't complain during Installation...just installs new files and keeps existing ones (o;

So be sure to remove an older Installation before...

[linux-works]

a sound card that I was told is about as good as it gets for pc's is the juli@ by ESI.  the pci is the classic card but there is a pci-e as well.

an audio designer friend of mine swears by this card for its a/d and d/a.

other interfaces known to be high quality: emu 0404usb and m-audio firewire audio interfaces (old by now and no one uses firewire, but its still a good reference audio module).

[davorin]

Well I don't have any computers with PCI or PCIe slots...only small Intel NUCs sitting below the monitors (o;

My dream device would be a rack based device with slots for individual in/out interfaces, either analog...or what I miss...SPDIF/I2S interfaces for measuring also own A/D and D/A designs...

At least I could use my HP 3488A switch with the VHF card to remotely switch analog inputs/outputs.......ah yes...a A/D-D/A card for HP 3488A would be nice...but probably GPIB would be too slow then....