Trying to test audiocards...the wrong way?
Trying to test audiocards...the wrong way?
Feb 9 2013, 22:35
Joined: 9-February 13
Member No.: 106530
I was trying to make some tests to audiocards that I own (Roland UA-1G, M-Audio 2494, M-Audio Delta 1010LT), in order to do some other testing after that with several CD, DVD players, Media Centers, etc.
My first thought was something like having the outputs of the soundcard connected to the inputs, play one song and record it on another track (I'm using Nuendo).
That's the part of the process where all things worked out as planned, except...
...I was thinking that by doing this, if I phase reversed one of the stereo tracks, I would end up only with the differences between the two (or nothing, on a perfect scenario).
I had to compensate for the delay and any volume difference, but on the best 'phase cancellation' spot, I still hear (mostly) the very high frequencies.
That didn't surprised me on the Roland UA-1G, as it's a USB card with volume control and dedicated to record instruments on the go, so I would find fair enough that it wasn't very flat on high frequencies. That's something that even REW confirms in the 'soundcard calibration' part of the setup.
But having the same behaviour with both the M-Audio made me wonder if there's something wrong with my test basics.
I'm playing and recording at 44.1 and 16 bit (trying to keep it from the original source to the end Wave file), so any input on this theory would be appreciated.
Feb 9 2013, 22:37
Group: Super Moderator
Joined: 23-June 06
Member No.: 32180
Feb 18 2013, 16:34
Joined: 29-October 08
From: USA, 48236
Member No.: 61311
The standard recommendation is a loopback test using RightMark Audio Analyzer.
This page offers some warnings, best practices, and criticisms.
There are some reservations about that page. Going down the list:
(1) MAXIMUM CLEAN OUTPUT LEVEL - RMAA has no concept of absolute levels. It can't measure voltages, power outputs, etc
This is a common problem with using audio interfaces as measurement tools. They are typically very consistent, but they are generally not calibrated and their de facto calibration may not be the same as their specifications. Note that the complainain's software of choice Spectra Lab shares this situation. The best solution is probably to calibrate the test using a regular DVM. Be sure to use a test signal < 400 Hz as common DVMs often don't have anything like flat frequency response above that.
(2) OUTPUT IMPEDANCE - Anything designed for driving headphones including PC's, portable players, headphone amps, USB headphone DACs, audio interfaces, pro gear, etc. has an output impedance.
This can be obtained by making two measurements, one open circuit and one with a relevant load, and then doing some calculations.
(3) DAC LINEARITY – Some DACs exhibit considerable non-linearity at low levels.
This information can be determined a number of different ways. The common form of dynamic range measurement involving a -60 dB signal gives considerable insight into this problem. It is by definition not a problem with the ever-so-commin Sigma Delta DACs that form about 99.44% of all audio DACs in service today.
(4) SQUARE WAVE PERFORMANCE – A 1 Khz square wave reveals a lot of information about analog and digital audio components such as stability, bandwidth, rise time, compensation, and for digital devices, the type of digital filtering used
Virtually every audio interface has a brick wall filter at a maximum of 192 KHz. Square wave testing is highly overrated in the sense that equipment with horrible-looking square waves can sound good and equipment with excellent appearing square waves can sound bad. It is good for looking at things like stability. A 150 MHz bandpass oscillioscope it the tool of choice for this sort of thing. A great deal of audio work can be done without this capability.
(5) • JITTER - Jitter has been proven to be audible in some circumstances
Another highly overrated audio parameter. Other than analog media playback, few if any cases of this being audible have ever been found. Jitter can be determined from a THD test if the noise floor is low enough and the frequency scale is magnified.
(6) • VARIOUS TWIN TONE TESTS – RMAA has a single IMD test that’s similar to the SMPTE test but it’s not clear if it follows the SMPTE standard. And there’s no capability for other important twin tone tests such as the popular CCIF 19 Khz/20 Khz which is very revealing of high frequency performance.
All of the versions of RMAA that I've used had a setup option for changing the test tone frequencies for the SMPTE test to be what I wanted. If memory serves, the analysis of the frequencies I tried was correct and reasonble.
(7) • THD20 – It’s very useful to measure THD at 20 Khz for a variety of reasons
Actually this can be a very misleading test because common brick wall filtering used in digital audio can eliminate most or all of the harmonics. 19 & 20 KHz IM testing measures the same basic nonlinearity and provides useful results in spite of band limiting.
(8) • LOW FREQUENCY DISTORTION – It’s also useful to measure THD+N at very low frequencies such as 5 to 10 hz to expose thermal modulation effects, power supply issues, and more.
It seems like this situation could be addressed by using RMAA's setup feature but I've never tried it.
(9) • POWER vs THD: RMAA cannot perform the classic measurement of output power versus THD.
RMAA can collect the data, but you're going to do it manually. Last time I worked with it, the author's seemingly favorite tool Spectra lab also required a lot of manual work to prepare this report.
(10) •SELECTIVE SPECTRUM TESTS: It’s very useful to perform different spectrum testing with various input signals.
Generating signals on a PC is pretty easy, and RMAA's spectrum analysis tool is there for you to use.
(11) •RESIDUAL ANALYSIS – Analyzing the residual distortion products can be very revealing
RMAA presents a spectral analysis with just about every test and you can manipulate the vertical and horizontal scales as you will. It doesn't get a lot better than this with most programs.
(12) •REAL TIME RESULTS - Many problems may only show up briefly or intermittently. And many adjustments (like finding the clipping point) are best performed in real time.
You can use just about any FFT-based analysis program with quasi real-time display for this purpose. If memory serves, Holme Impulse is good for this and free. Also, you can do some of this in RMAA using its level-setting facility.
(13) •SlEW RATE – RMAA cannot measure slew rate which, especially for audio power amplifiers,
Slew rate is right up there with square waves in the league of potentially misleading tests. It is audible only as it corresponds to high frequency dynamic range which is probably best evaluated using twin tone testing.
(14) • DELAY & LATENCY – RMAA cannot measure delay and latency for digital hardware
There are freeware tools for this, and all of the DAW software I've used can measure this.
(15) • HARDWARE LIMITATIONS – Nearly all PC sound hardware has some severe limitations that restrict RMAA and any other software. For example bridged and certain other amplifiers cannot have any of their output terminals grounded or connected together.
Another problem that is inherent with PC audio interfaces and afflicts any software that uses them to gather data including the author's fave Spectra Lab. I've built some special purpose attenuators for dealing with this.
Bottom line is that if you have access to something like an Audio Precision test set, even an old one, a lot of the above problems can be solved, some by standard features. IME the toughest problems to solve are related to point 15.
|Lo-Fi Version||Time is now: 9th March 2014 - 22:54|