Daqarta
Data AcQuisition And Real-Time Analysis
Scope - Spectrum - Spectrogram - Signal Generator
Software for Windows
Science with your Sound Card!
The following is from the Daqarta Help system:

Features:

Oscilloscope

Spectrum Analyzer

8-Channel
Signal Generator

(Absolutely FREE!)

Spectrogram

Pitch Tracker

Pitch-to-MIDI

DaqMusiq Generator
(Free Music... Forever!)

Engine Simulator

LCR Meter

Remote Operation

DC Measurements

True RMS Voltmeter

Sound Level Meter

Frequency Counter
    Period
    Event
    Spectral Event

    Temperature
    Pressure
    MHz Frequencies

Data Logger

Waveform Averager

Histogram

Post-Stimulus Time
Histogram (PSTH)

THD Meter

IMD Meter

Precision Phase Meter

Pulse Meter

Macro System

Multi-Trace Arrays

Trigger Controls

Auto-Calibration

Spectral Peak Track

Spectrum Limit Testing

Direct-to-Disk Recording

Accessibility

Applications:

Frequency response

Distortion measurement

Speech and music

Microphone calibration

Loudspeaker test

Auditory phenomena

Musical instrument tuning

Animal sound

Evoked potentials

Rotating machinery

Automotive

Product test

Contact us about
your application!

Sound Card Performance Tests

Introduction:

You probably don't need a new sound card to use Daqarta effectively; the built-in sound of most systems is usually more than adequate for lab-type Daqarta applications that don't require audiophile performance.

Five USB devices are tested here, plus a Realtek ALC883 sound chipset built into an XP laptop for comparison. Prices for the USB units run from US $25 to over $90.

Each unit has its own merits; the 5.1 Channel CM6206 unit ($25) is small, nicely built, and has no hardware controls to get accidentally changed... an important factor for lab use. It also has the lowest output impedance and highest output drive capability, plus the widest Line input sensitivity range. In addition, it supports 6 output channels.

The 7.1 Channel CM6206 unit ($60) includes dual Mic In connectors, plus excellent low-frequency input response (less than 1 Hz). It also supports 8 output channels.

The Behringer UCA202 ($30) has excellent noise and superb distortion specs, plus input and output frequency response (at 48000 Hz sample rate) comparable to units costing 3 times as much.

The Creative X-Fi ($90 or more, if you can find it) shares best-in-test noise and distortion specs with the ASUS Xonar U7. The X-Fi also has excellent frequency response at 48000 and 96000 Hz sample rates when run on Windows Vista or later. But on XP at 48000 Hz, only its input response is excellent; output response is the worst of the test group. Surprisingly, both input and output are excellent at 96000 Hz, even on XP.

The ASUS Xonar U7 ($90 or more, if you can find it) is the best all-around card of those tested here, due to superb noise and distortion specs, best frequency response, the ability to run at sample rates up to 192000 Hz on Windows Vista or later, and 8 output channels.

Note that neither the X-Fi nor Xonar U7 allow Daqarta to change sample rates directly, but there is no obvious indicator of that. You need to set the rate in the custom control panel, or in Windows (Vista and later), and then set the same value in Daqarta. Otherwise, if you only change the rate in Daqarta, you'll just get a simulated rate via sample rate conversion in Windows. (This is probably true of other cards that allow sample rates above 48000 Hz, due to the Windows Driver Model.)

All the USB devices come with USB cables included.


Performance Measurements:

In the Comparison Table measurements below, input and output impedance measurements were made using the methods discussed in Sound Card Impedance Measurement.

Maximum input and output voltages are slightly below the sudden onset of a sharp distortion rise indicative of clipping, as seen in the spectrum (which is much more sensitive than visual detection of clipping on the waveform). These are not necessarily the levels for lowest distortion, but are the highest "useful" levels.

Ideally, when comparing devices for noise and distortion, an input range should be selected such that the signal nearly fills the range, but doesn't clip. The idea is that the signal should be using all the bits of the ADC; A low-level signal that doesn't use all the bits will have increased distortion just due to quantization, even if the ADC itself is perfect.

In many of the tests below, a "loopback" connection was used between the output and input jacks, with the Daqarta Generator creating the test signal. Instead of a simple audio cable, however, separate output and input cables were connected to a junction box where the signal could be monitored via a separate benchtop digital voltmeter for adjusting the absolute level.

Except for frequency response measurements as indicated, all cards were tested using a 48000 Hz sample rate.


Noise Measurements:

In the noise and distortion tests, a 1 kHz, 1.00 Vrms signal was used for all devices except the 7.1 Channel CM6206, where 0.50 Vrms was used because its inputs could not handle 1.0 V. In all cases, an input range was selected such that the signal came within 3 dB of the full-scale ADC range.

Noise was measured in several ways. With the above reference signal present, the 1 kHz noise floor was measured by placing one cursor on the reference peak and one at a representative non-harmonic position in the flat central part of the spectrum. Spectrum averaging was used to find the average value of the noise at that location, and the Delta cursor readout gave the difference from the fundamental peak. The reported noise is thus given in dB below the reference level.

The Loop Floor was then measured by toggling the Generator off but leaving the loopback in place. The floor was again measured, but since there was no reference peak to allow the Delta readout, the cursor reading was manually subtracted from the peak dB used in the above test. The value is thus still in dB below the reference level, even though that reference was not actually present.

The Loop RMS was obtained by toggling the Delta readout to Sigma, and setting one cursor at either end of the spectrum. If there was an obvious DC component present, the low-end cursor was moved up a couple of lines to avoid it. The Sigma value was again manually subtracted from the former peak dB.

Since some of the Loop noise was presumably contributed by the output, the cables were disconnected and Line In Floor and Line In RMS were measured like the above Loop Floor and Loop RMS.


Distortion Measurements:

Distortion was measured at the 1.00 (or 0.50 Vrms) reference level, and also at 10 dB below that reference; the latter is a typical audio industry convention. Note, however, that no weighting was used. (The audio industry often uses "A-weighting", which reduces the harmonics being measured and so gives better-looking results.)

Harmonic distortion is shown for harmonics 2 through 5; there were no significant higher harmonics. The Total Harmonic Distortion (THD) shown is computed from these harmonics and the 1 kHz fundamental only, in both dB relative to the fundamental, and in percent. The computation was done using the method of the RMS "Sum" of dB Values topic, which also discusses use of the the dB_Sum macro to simplify the process.

Below the computed THD are THD+N (THD + Noise) and THD 25 (THD using up to 25 discrete harmonics), measured with Daqarta's THD_Meter mini-app. In addition, these measurements were repeated using the THD_Meter's Wave Average mode to reduce the noise floor. This gives a more accurate pure THD value, while giving a possibly too-good THD+N (due to the noise reduction). If the THD+N value improves greatly with Wave Averaging, it means that the original reading was dominated by noise, which has been reduced. If it doesn't improve much, it indicates that the THD dominates.

(Note that the THD_Meter can also measure the THD value including only harmonics 2-5 as mentioned previously, simply by setting its Max Harmonic from 25 down to 5.)

The reference frequency for all harmonic distortion measurements was actually 984.375 Hz instead of 1 kHz, using the spectral Line lock option of the Daqarta Generator to get coherent sampling: The frequency had an exact integer number of signal cycles in the 1024 samples used for the FFT, so it and all harmonics fell exactly on spectral lines. This eliminated the need for a spectral window function, and allowed measurement of the true noise floor without it being obscured by spectral leakage "skirts" due to the window.

Intermodulation distortion (IMD) was measured with Daqarta's IMD_Meter mini-app using three standard methods: SMPTE, DIN, and CCIF / ITU-R, with both 2- and 4-component readings shown for the latter. (See IMD Modes under the IMD_Meter Help topic for a discussion of each method.) Values were measured using the specified frequencies for each standard, and also using the Spectral Line Lock option for more accurate actual IMD readings. Since Spectral Line Lock uses slightly different frequencies, these measurements aren't technically compliant with the standards, but they are closer to what would be obtained with a "perfect" IMD meter.

IMD measurements were taken at levels based on those used for THD, but not at the same RMS values. The reason is that D/A and A/D converters have absolute amplitude limits, not RMS limits. Since IMD measurements require two primary tones, whose combined amplitudes must never exceed 100% of the converter range, their sum was set to produce the same peak amplitude as the single tone used for THD.

For example, the CCIF / ITU-R test uses equal-amplitude tones, so each tone's Stream Level in the Generator was set to 50% instead of the 100% used for THD. But the identical setting was used in the Volume dialog (F9 key). Thus, where the THD used 1.00 Vrms, the CCIF / ITU-R used sqrt(0.50^2 + 0.50^2) = 0.707 Vrms.

The SMPTE and DIN tests use primaries at a 4:1 ratio, hence Stream Levels of 80% and 20%. The RMS value was thus sqrt(0.80^2 + 0.20^2) = 0.82 Vrms.


Crosstalk Measurements:

Crosstalk (interchannel leakage) was measured at the same 1.00 or 0.50 Vrms reference level as for the noise and distortion tests. A 1 kHz reference-level tone was generated on the Left Output, and a 20 kHz tone on the Right Output.

With a loopback connection as used above, the 1 kHz crosstalk was measured by looking at the Right In channel, which should ideally only contain a 20 kHz peak. The measured value was the difference between that peak and any peak that appeared at the 1 kHz position.

The 20 kHz crosstalk was likewise measured on the Left In channel as the difference between the 1 kHz peak and any 20 kHz peak that leaked through.

Since some of the crosstalk could arise due to leakage between conductors in the loopback cable, the above measurements were repeated with no cables present, and reported as Open values.


Frequency Response:

The Low Frequency -3 dB Response frequencies were determined for the Output by measuring on a separate benchtop oscilloscope in DC mode. The Generator was first set to 1 kHz and the peak-to-peak amplitude was measured on the scope. Then the Generator frequency was lowered until the amplitude was reduced to 70% of the 1 kHz value (-3 dB = 0.707), and that frequency was recorded.

The Line In measurement was done by driving the input with a separate benchtop oscillator, and using the Daqarta Voltmeter in Peak-to-Peak mode. The 1 kHz value was obtained, then the oscillator frequency was lowered until the reading was reduced by 3 dB, as above.

Note, however, that the Daqarta Voltmeter can only give correct peak-to-peak readings if it sees at least one full cycle, such that both a positive and a negative peak appear on each 1024-sample trace. With a 48000 Hz sample rate, the lowest frequency for that to happen would be 48000 / 1024 = 46.875 Hz... much higher than the cutoff for any decent sound card. To deal with that, Decimation was used to reduce the effective sample rate as needed.


The High Frequency - 3 dB Response for the Output was determined using a separate scope, as for the low frequency response. For Line In, a separate oscillator was again used, but not with the Daqarta Voltmeter. Instead, the peak amplitude was read from the Spectrum cursor readout at 1 kHz, then the oscillator frequency was increased until the peak reading dropped by 3 dB.

The high frequency response measurements were repeated using a 96000 Hz sample rate. Only the Creative X-Fi, the ASUS Xonar U7, and the Realtek ALC883 can actually run at this sample rate, though the Realtek output is still limited to less than 24 kHz, the same as when sampling at 48000 Hz.

The ASUS Xonar U7 was also tested at 192000 Hz, the only one of the group that can run at this sample rate.

See also the Loopback Frequency Response Comparison section, below.


Polarity:

The Polarity tests were done using a separate scope. The Generator was set to produce a biphasic Pulse waveform, such that the positive phase preceded the negative phase, and there was a dwell time at zero before the next pulse.

The output had normal polarity if the positive phase appeared first on the scope. A loopback cable was then used to feed the output back to the input, and if the Input waveform seen by Daqarta also showed the positive phase first, the input likewise had normal polarity. This was the case with all devices except the Behringer UCA202, which inverted both the output and input (which means that it could not have been detected with a loopback alone).


Comparison Table:

In the table below. (f) means the value was indistinguishable from the noise floor. * means that further explanation is included under the specific card section. Under Inputs: Mic, M = Mono, S = Stereo, D = Dual (2 jacks).

                          CM6206       Behringer   Creative    ASUS     Realtek
                       5.1       7.1     UCA202      X-Fi    Xonar U7   ALC883
Price, USD             $25       $60      $30        $90*      $90*    (laptop)

Size, inches:
    Width              3.5       5.0       3.5       5.5       5.5        -
    Depth              2.4       2.5       2.0*      3.9*      3.2*       -
    Height             0.8       1.0       0.9       1.0       1.0        -

Line Connectors       3.5mm     3.5mm      RCA       RCA    RCA/3.5mm    3.5mm

Inputs:
    Line                y         y         y         y         y         y
    Mic                 M         D         -         S         S         M
    Stereo Mix          y         y         -         y         -         y

Mic Bias, Vdc:         2.25      1.25       -        5.1       2.5       2.5

Impedance, ohms:
    Line Out            2         7        413        35       700       46
    Headphone Out       -         31        51        35        21        -
    Line In            5.2k      14k       12.7k     13k       49k       15k
    Mic In             1.6k      7.8k       -        3.2k      3.2k      2.3k

Maximum Line Output, Vrms:
    No load            1.50      1.00      1.10      1.54      1.13      1.50
    100 ohms           1.46      0.77      0.09      1.14      0.14      1.00

Maximum Headphone Output, Vrms:
    No load             -        0.92      0.85      1.54      1.34*      -
    100 ohms            -        0.71      0.50      1.14      1.11*      -

Maximum Input, Vrms:
    Line In
        Max sens       0.024     0.015     1.25      1.00      1.00      0.250
        Min sens       1.05      0.66       -        2.00      1.00      1.500
    Mic In
        Max sens       0.024     0.013      -        0.043     0.003     0.025
        Min sens       1.05      0.66       -        0.163     0.330     0.250

Noise, dB:
 1 kHz Ref, Vrms      1.00       0.50      1.00      1.00      1.00      1.00
    1 kHz Floor       -108       -109      -104*     -115      -116      -109
    Loop Floor        -109       -109      -114      -119      -128      -109
    Loop RMS           -82        -83       -87       -96       -97       -81
    Line In Floor     -112       -111      -115      -122      -124      -112
    Line In RMS        -83        -85       -89       -95       -97       -85

Harmonic Distortion, 1 kHz, dB:
    Ref Vrms          1.00       0.50      1.00      1.00      1.00      1.00
        f2             -75       -80       -98(f)    -96       -84       -63
        f3             -84       -95       -86       -99       -97       -79
        f4             -90      -101      -102(f)   -112      -103      -108
        f5             -95      -105       -96      -105       -99      -109
       THD dB          -74       -80       -85       -94       -84       -63
       THD %          0.02%     0.01%     0.006%    0.002%    0.007%    0.07%

       THD+N %        0.051%    0.017%    0.021%    0.005%    0.008%    0.075%
         Wave Avg     0.018%    0.013%    0.007%    0.003%    0.007%    0.074%
       THD 25         0.020%    0.013%    0.007%    0.002%    0.007%    0.074%
         Wave Avg     0.015%    0.013%    0.007%    0.002%    0.006%    0.074%

    -10 dB, same range:
        f2             -76       -89       -96(f)    -98       -95       -72
        f3             -85       -99(f)    -97(f)    -89       -100      -97
        f4             -91       -98       -98(f)    -108(f)   -112(f)   -98
        f5             -91       -99(f)    -98(f)    -95       -110      -99
       THD dB          -75       -88       -91       -88       -94       -72
       THD %          0.02%     0.004%    0.003%    0.004%    0.002%    0.025%

       THD+N          0.054%    0.028%    0.031%    0.014%    0.007%    0.037%
         Wave Avg     0.012%    0.025%    0.005%    0.006%    0.002%    0.026%
       THD 25         0.017%    0.008%    0.006%    0.004%    0.002%    0.027%
         Wave Avg     0.007%    0.008%    0.002%    0.004%    0.002%    0.026%


Intermodulation Distortion, %:
    Ref Level same as THD
       SMPTE          0.144     0.159     0.162     0.143     0.133     0.117
         Spect Line   0.068     0.038     0.035     0.015     0.049     0.168
       DIN            0.156     0.128     0.160     0.147     0.150     0.253
         Spect Line   0.061     0.039     0.033     0.014     0.055     0.171
       CCIF-2         0.011     0.011     0.004     0.009     0.005     0.072
         Spect Line   0.003     0.007     0.001     0.001     0.002     0.052
       CCIF-4         0.062     0.034     0.010     0.047     0.046     0.045
         Spect Line   0.021     0.016     0.009     0.006     0.023     0.008
         Wave Avg     0.018     0.016     0.007     0.006     0.023     0.007

    -10 dB, same range:
       SMPTE          0.157     0.129     0.138     0.147     0.131     0.126
         Spect Line   0.080     0.045     0.047     0.025     0.017     0.069
       DIN            0.181     0.162     0.125     0.129     0.139     0.157
         Spect Line   0.078     0.044     0.040     0.024     0.018     0.070
       CCIF-2         0.025     0.010     0.007     0.007     0.003     0.027
         Spect Line   0.007     0.002     0.002     0.001     0.001     0.019
       CCIF-4         0.061     0.037     0.017     0.051     0.037     0.043
         Spect Line   0.014     0.007     0.006     0.012     0.012     0.008
         Wave Avg     0.004     0.003     0.006     0.012     0.012     0.003


L-R Crosstalk, dB:
    6-ft Loopback
        1 kHz          -91       -95       -90       -78        -93       -90
        20 kHz         -89       -93       -66       -78        -70       -73
    Open
        1 kHz          -94       -86       -72       -108      -120       -94
        20 kHz         -75       -84       -82       -115      -102       -79

Low Freq -3 dB Response, Hz:
        Output          3        0.4       1.7       0.15      0.15*       14
        Line In         41       <1         4         4        0.6*        8

High Freq -3 dB Response, Hz:
    48000 Hz Sample Rate
        Output        21100     21000     22000     16000*    23600     21300
        Line In       21400     21000     23500     23800     23800     21440

    96000 Hz Sample Rate
        Output          -         -         -       47500*   ~47000     21300
        Line In         -         -         -       47500     47800     42000

    192000 Hz Sample Rate
        Output          -         -         -         -       55000       -
        Line In         -         -         -         -       75000       -

Polarity:
    Out                Norm      Norm      Inv       Norm      Norm      Norm
    Line In            Norm      Norm      Inv       Norm      Norm      Norm
    Mic In             Norm      Norm       -        Norm      Norm      Norm

Loopback Frequency Response Comparison:

48000 Hz Sample Rate:

Arbitrary vertical spacing has been added here; otherwise, all traces would have overlapped in low-to-mid frequency regions.

A loopback cable was connected from Line Out to Line In. The Daqarta Generator was set to provide a uniform white noise for a perfectly flat average driving frequency spectrum, then 1024 frames of the response spectrum were linearly averaged. (The fine wiggles in the response plots are due to residual noise; they do not represent actual response variations.)

Note that the X-Fi and U7 each have separate traces for Windows 7 and XP, since their responses differed substantially between the two systems.


96000 Hz Sample Rate:

Of the group tested, only the ASUS Xonar U7 and the Creative X-Fi can run at a 96000 Hz sample rate. The plots here were obtained as for the 48000 Hz responses (above). Both were run on a Windows 7 system, but similar results were also obtained with XP.


192000 Hz Sample Rate:

The ASUS Xonar U7 is the only unit of the test group that can run at a 192000 Hz sample rate. This plot was obtained as for the 48000 Hz and 96000 Hz responses (above) and was run on a Windows 7 system, but similar results were also obtained with XP.

Note that this is a loopback response, using both Line Out and Line In. As noted in the Comparison Table above, the input-only response extends out to 75000 Hz (-3 dB point), while the output-only response only goes to 55000 Hz. The combined response shown here thus rolls off at 54000 Hz.


5.1 Channel CM6206 USB:

Price: US $25

Dimensions: 3.5 x 2.4 x 0.8 inches.

Includes: USB cable, driver mini-CD.

Connectors: USB In, Mic In, SPDIF TOSLINK In, SPDIF TOSLINK Out, Line In, Front Out, Rear Out, Cen/Bass Out.

Controls: None.

This is a small unit, neatly packaged in an extruded aluminum chassis. Although inexpensive, it is a capable performer for many "lab-type" Daqarta applications that don't require "audiophile" specs. It has the lowest output impedance and highest output drive capability, and the widest input sensitivity range of the group.

It is available from multiple suppliers. The actual manufacturer is not stated. You can locate a supplier by searching the Web for "5.1 channel USB sound" (without the quotes) and looking for a matching picture. The test unit was purchased from HDE through Amazon.

Please note: This device has multiple manufacturers who typically use the same case and the same basic CM6206 circuit, but may use different circuit board layouts and different component values. The latter may affect the low-frequency input and/or output response, so your unit may differ from the values reported in the Comparison Table.

This is the same device model (but not the same unit) that was used for the Sound Card DC Input / Output Modification by adding a small circuit inside the case. This model was also used for the Simple Sound Card Unipolar DC Modification.

The device uses the C-Media CM6206 8-channel codec ("coder-decoder" combined ADC and DAC) chip, but only 6 channels are connected. The Front (Left and Right) channels were used for all tests here.

The unit comes with a driver mini-CD, which you should install if you are running Windows Vista or later. But if you are running Windows XP and using multi-channel (5.1) outputs, you may want to avoid that and just use the default driver that is installed when you plug in the USB connector. See the CM6206 5.1 and 7.1 Driver section below for details.

The Mic In is mono-only. There is no advantage to using this instead of Line In unless you are actually using a microphone and need the 2.5 Vdc electret bias... Mic In and Line In have the same sensitivity ranges. If you do use Mic In, note that only the Left input signal (the tip of the stereo plug, which is the white conductor of a standard cable) is connected separately to both Left and Right internal channels. Each channel will respond to its own input level control.

This "feature" might possibly be useful for capturing unrepeatable signals (like wild animal calls) with an unpredictable dynamic range: You could set one channel to a less sensitive range as a backup in case the other is overdriven.

CAUTION: The mini-USB jack on this unit is slightly too large, such that the plug sometimes loosened during use and occasionally caused Daqarta to crash. Deforming the plug slightly to insure a snug fit was helpful, but not a permanent fix. For ultimate reliability, consider removing the jack and hard-wiring the plug end of the cable directly into the unit. See USB Socket Replacement under Sound Card DC Input / Output Modification, as well as USB Cable and Connector Pinouts under Cables and Connectors in the Appendix.

None of the other tested units had this sort of problem, including the Creative X-Fi which uses the same type of mini-USB jack.



CM6206 5.1 and 7.1 Driver:

Both the 5.1 Channel CM6206 USB (above) and the 7.1 Channel CM6206 USB (below) use the same custom driver, installed from a mini-CD that comes with each device. An icon with a small blue-and-yellow musical note will appear in the Windows system tray whenever you plug in the device. Clicking the icon opens a custom control panel that allows selection of various options and enhancements, including separate volume controls for each of the multi-channel outputs. (See CM6206 Multi-Channel Volume Controls, below.)

The custom driver also provides a Stereo Mix input, which allows Windows XP users to view MIDI output waveforms, spectra, or spectrograms (not just the Pitch-to-MIDI note display) by toggling Pitch Track off during a DaqMusiq performance.

(On Windows Vista and later you may not need this, since it typically routes the output to the input by default. If not, see Hidden or Disabled Inputs under Vista / 7 / 8 / 10 / 11 Issues in the Troubleshooting section.)

However, on Windows XP systems the custom driver inappropriately mixes Front channel signals into the other channels, and Back into Side. This defeats the whole point of having multiple outputs, so you should probably skip the custom driver installation for XP, unless you don't intend to use multiple outputs but you do want to use special effects like reverberation to enhance MIDI performances.

Tip: The driver will be used for the specific USB port that the card is plugged into at the time it is installed. You can just plug the device into another port whenever you don't want to use the custom driver.

On XP systems without the custom driver, the outputs remain separate. When Multi-Channel Outputs is active, the F9 key that normally opens the standard Daqarta volume controls instead opens Daqarta's Multi-Channel Volume Controls dialog with full independent control for each of the channels.

If you do use the custom driver on XP, there is a way to get 4 separate output channels with independent volume controls: Don't use the offending Front or Back channels.

First, make sure you have 6 or 8 channels selected under Output Device - Analog Output in the custom driver control panel. Then in Daqarta's Multi-Channel Outputs dialog you can either toggle off the Front Channel Selects, or make sure there are no Left or Right Streams buttons set for them.

Likewise toggle both Back outputs off, leaving only Center, Low Freq, Side Left and Side Right. You can then use the separate custom volume controls to control these.

Note that on the 5.1 Channel CM6206 USB unit the jack marked 'REAR OUT' always receives the Side channels. If you set the driver for 8 channels, the Back channels do nothing. If set for 6 channels then Back and Side are mixed together (along with Front, as noted).

With the 7.1 Channel CM6206 USB there is no output from the 'BS-OUT' jack if set for 6 channels. With 8 channels, this jack receives a mix of Side, Back, and Front, while the 'SS-OUT' jack receives only Side and Front, not Back. In either case, 'C/B-OUT' has Left Front mixed with Center, and Right Front mixed with Low Freq (Bass).

On Windows Vista or later the custom driver works properly, keeping all channels separate and allowing individual volume controls. However, you must first configure the card for multi-channel operation under Windows. Then set the desired channels in the driver control panel.



CM6206 Multi-Channel Volume Controls:

On Windows XP, when Multi-Channel Outputs is active, the F9 key that normally opens the standard Daqarta volume controls instead opens Daqarta's Multi-Channel Volume Controls dialog with full independent control for each of the channels.

With Vista/7/8/10/11 (until the next Daqarta version, with full multi-channel volume support), you need to use the custom driver. To access the individual volume controls for each channel, open the custom driver control panel via the blue-and-yellow note icon in the system tray. It opens on the 'Main Setting' tab.

On the left panel, make sure that 'Output Device - Analog Outputs' is set for 6 or 8 channels.

On the right panel, make sure that 'DSP Mode' and '7.1 Virtual Speaker Shifter' are off (not blue).

Click the 'Digital Volume' button (speaker icon with ascending volume bars). It will turn blue, and tiny volume sliders will appear next to each of the speakers in the room image below it. Each slider has a range from +12 to -96 dB, but they are so small it is hard to make fine adjustments. You can click the '+' or '-' symbols to move up or down in single-dB steps, though these too are very small and require careful pointer positioning. The little dB readouts do not allow direct entry.

Note: You should avoid using any positive dB settings here. They do nothing if the Generator Level for that channel is already at 100%, but can only boost the volume if Level is lower. For example, if Level is 50%, you can get a maximum of +6 dB boost here, despite higher settings.

The 'Reset' button will restore all sliders to 0 dB, which is probably best for most work. Any other setting means that the Generator Level settings won't reflect the true output levels.



Restoring CM6206 Driver Defaults:

Open the custom driver control panel via the blue-and-yellow note icon in the system tray. It opens on the 'Main Setting' tab.

On the right panel, make sure that 'DSP Mode' and '7.1 Virtual Speaker Shifter' are off (not blue). The Virtual Speaker Shifter adds Left-Right channel crosstalk, as well as distributing various amounts of the signals to the additional channels. If you also use this unit for other applications like music listening, you should make sure that this option is off when using Daqarta. (Except perhaps to to enhance MIDI performances.)

Click the 'Mixer' tab. In the upper 'Vol' panel make sure only 'Wave' is selected (button at bottom of slider is all blue). In the lower 'Rec' panel, make sure only 'Line In' is selected.

Click the 'Effect' tab. Make sure the 'More Options' button under 'Environment' is not blue. Ignore 'Environment Size'. Under 'Equalizer' click 'Default' to cancel any other buttons and center all sliders at 0 dB. (It's also OK if no buttons are active.)

Click the 'Karaoke/Magic Voice' tab, and make sure the 'On' button is not blue.



7.1 Channel CM6206 USB:

Price: US $60

Dimensions: 5.0 x 2.5 x 1.0 inches.

Includes: USB cable, SPDIF TOSLINK optical cable, driver mini-CD.

Connectors: USB, SPDIF TOSLINK In, SPDIF TOSLINK Out, Mic(1), Mic(2), Line In, HP Out, F Out, SS Out, BS Out, C/B Out.

Controls: Record Mute button, Play Mute button, Vol Up/Down buttons.

This device is available from multiple suppliers. The actual manufacturer is not stated. You can locate a supplier by searching the Web for "USB 7.1 CH 3D External Sound Pocket" (without the quotes) and looking for a matching picture. Some distributors are Sabrent (model USB-SND8), TigerDirect, and USBgear.

Please note: This device has multiple manufacturers who typically use the same case and the same basic CM6206 circuit, but may use different circuit board layouts and different component values. The latter may affect the low-frequency input and/or output response, so your unit may differ from the values reported in the Comparison Table.

As with the above 5.1 channel device, this unit uses the C-Media CM6206 8-channel codec chip, only here all 8 channels are connected. The Front (Left and Right) output channels were used for all tests, marked F-OUT on the unit.

The unit comes with a driver mini-CD, which you should install if you are running Windows Vista or later. But if you are running Windows XP and using multi-channel (7.1) outputs, you may want to avoid that and just use the default driver that is installed when you plug in the USB connector. See the CM6206 5.1 and 7.1 Driver section above for details.

The overall low frequency response is excellent; the Line In -3 dB point is below 1 Hz.

One quirk of this unit is that the Left and Right channels are reversed from the normal "Red is Right" convention: Here the Right channel is the plug tip (white conductor) and Left is the ring (red conductor). This is true for both outputs and Line inputs, so it's not obvious in loopback testing.

Another quirk is that the two LEDs on top of the unit are mis-labeled: The green LED that is on continuously when USB power is supplied, and which flashes when the output is active, is marked "Record Mute". Conversely, the red LED that comes on when the Record Mute button is active is marked "Operation". (The Play Mute button does not stop the blinking of the green LED, however.)

Nevertheless, these buttons and the red LED can be convenient for some tests. You do need to be careful not to hit Play Mute accidentally, because there is no indication of its state; if you notice a lack of output after toggling an adjacent button, try clicking this button. (If that doesn't restore the output, click the button again... otherwise when you find the real cause of the missing output it will still be off.)

A nice feature of this unit is that there are separate Mic(1) (Right channel) and Mic(2) (Left channel) input connectors. Each is a stereo jack, with 1.25 Vdc electret microphone bias connected to the tip (white conductor) along with the signal.

However, on Windows Vista and later the stereo Mic inputs may not be enabled by default. Before starting Daqarta, right-click on the speaker icon in the system tray and select Recording Devices. You should see separate options for USB Microphone and Line inputs. You must disable Line before you can enable Microphone (and vice-versa). Select and enable Microphone, then click on Properties. Select the Advanced tab and change the default "2 channel, 16 bit, 44100 Hz (CD Quality)" to "2 channel, 16 bit, 48000 Hz (DVD Quality)". Click OK and you are ready to start Daqarta.

Note, however, that during performance testing there were two occasions where the Mic In signals seen by the unit inexplicably dropped to zero and stayed that way. Even after quitting Daqarta and unplugging the USB connector to remove power from the unit, upon reconnection and restart there was still no signal.

If this happens, open the system tray driver icon and click on the little speaker volume symbol near the upper center. Then click the Reset button that appears next to it... the Mic signal suddenly reappears. It's not at all clear what action provoked this behavior in the first place, but possibly plugging cables in and out caused the signal and bias to be briefly shorted. Presumably, in any stationary setup this would not arise.

The unit has a separate headphone jack marked HP-OUT, but it actually has a higher output impedance (31 ohms) than the F-OUT (7 ohms).



Behringer UCA202 USB:

Price: US $30

Dimensions: 3.5 x 2.0 x 0.9 inches. The 2 inch depth measurement does not include 0.35 inches for the protruding RCA connectors.

Includes: USB cable (attached), user manual.

Connectors: L/R Input, L/R Output RCA, SPDIF TOSLINK Out, 3.5mm Phones.

Controls: Monitor On/Off, Volume.

The UCA202 uses the Burr-Brown PCM2902E codec chip.

The unit does not come with a driver disk; simply plugging it in causes Windows to install a generic driver.

The UCA202 has only a Line input (no Mic), and has no input level control. It has a fairly high Line output impedance (over 400 ohms), which limits the clean signal it can deliver to a 100 ohm load to 0.09 Vrms.

There is a separate headphone jack with 50 ohm impedance that can deliver 0.50 Vrms into 100 ohms, and has comparable noise and distortion to Line Out. Note, however, that this output is affected by the small thumbwheel volume control on the case, which operates in tandem with the software volume. If you use the headphone output, you may want to apply tape to the thumbwheel to insure it stays in the maximum volume position, then use only the software controls in Daqarta. That way you can insure repeatable results by always setting the same volume step numbers or dB values.

The UCA202 has excellent noise and superb distortion specs, plus input and output frequency response (at 48000 Hz sample rate) comparable to the Creative X-Fi and ASUS Xonar U7 that cost 3 times as much. (On Windows XP, it's actually much better than the X-Fi, which has frequency response problems at 48000 Hz on XP.)

Note that you should insure that the Monitor switch on the unit is in the Off position for Daqarta operation.

Note that this unit inverts the input and output signals, which may be an issue with pulse, ramp, or similar polarity-specific stimulus or response waveforms. After you run Auto-Calibration, you should perform a Full-Scale Range calibration and enter each Input or Output range as a negative number. This will allow Daqarta to correct the signal inputs and Generator outputs.

In the presence of a tone, the noise floor tilts down on either side of it. With a 1 V, 1 kHz tone, the floor is -99 at 2 kHz, -108 at 22 kHz. With no actual tone present, there may be a low-level "idle tone".

Note: This unit is stereo only. If you use it with Daqarta's Multi-Channel Outputs the channels will be combined into stereo, with the exception of the Low Freq channel which is never used. Instead, the Center channel will appear in the Right output as well as the Left.

If the Channel Selects include either Front channel, all the rest will be at a much lower volume. This is true even if the Front channels have no active Left and Right Stream buttons. With the Front channels off, the volume of the remaining channels will be much louder, but will be reduced proportional to the number of channel pairs selected.



Creative Sound Blaster X-Fi HD USB:

Price: US $90 or more, no longer in production

Dimensions: 5.5 x 3.9 x 1.0 inches. The 3.9 inch depth measurement does not include 0.25 inches on the rear for the grounding lug and RCA connectors, nor 0.3 inches for the front volume knob.

Includes: USB cable, stereo 3.5 mm plug to dual RCA adapter cable, setup sheet, Installation CD, Media Toolbox 6 Installation CD (not needed).

Connectors: USB, SPDIF TOSLINK In, SPDIF TOSLINK Out, L/R Line/Phono In (RCA), L/R Line Out (RCA), Ground, Headphone (1/4"), Microphone (1/4").

Controls: Volume

The X-Fi uses a Cirrus Logic CS5361 stereo ADC chip (as does the ASUS Xonar U7, below), and an AKM AK4396VF DAC chip.

This unit is comparable to the top-rated Xonar U7 in noise and distortion specs, and has much better Line Out drive due to lower output impedance (35 ohms versus 700 for the U7) and higher no-load output voltage (1.5 Vrms versus 1.1 for the U7).

Like the U7, the X-Fi can run at a 96000 Hz sample rate, and it runs well at that rate on Windows XP. (The U7 can run at 96000 and also at 192000, including on XP.)

The X-Fi has excellent frequency response at 48000 Hz when run on Windows Vista or later. But on XP, however, it has the worst output high-frequency response of any unit in the test group. See the Loopback Frequency Response Comparison for details, and look at the traces marked 'X-FI WIN7' and 'X-FI XP'. These are loopback measurements, and it turns out that the droopy XP curve is entirely due to problems with the output; the Line In response is very flat, as shown by the -3 dB value of 23800 Hz in the Comparison Table.

At the 96000 Hz sample rate, however, the X-Fi is superb even on XP. It is highly recommended for those who need to work at high frequencies, and for anyone else who doesn't mind the 2x reduction in frequency resolution due to using the higher rate with Daqarta's fixed 1024-sample spectra. In fact, you can have the best of everything simply by running at 96000 Hz but using 2x Decimation to get an effective 48000 Hz input rate; in that case, the frequency response is absolutely flat over the whole 24 kHz range.

The X-Fi comes with a custom driver and control panel software on the black Installation CD included with the unit. You probably don't need this on Vista and later Windows versions, since the default driver that loads via USB when you plug in the unit handles all Daqarta functions except special effects like reverberation to enhance MIDI performances.

However, note that the X-Fi does not allow Daqarta to control the sample rate directly, so XP users will need to install the custom driver to get access to its control panel to change rates. Left-click the round silver "volume control" icon in the taskbar, then click Creative Audio Control Panel. Under the Performance tab set Sampling Rate to 96 kHz, then set Bit Depth to 16 Bit. Then set the same sample rate in Daqarta.

Vista and later users can right-click the normal Windows speaker icon in the system tray and set Playback devices, then Speaker - USB Sound Blaster HD, then Properties at the bottom of the dialog, then Advanced, and select "16 bit, 48000 Hz" or "16-bit, 96000 Hz". Repeat for Input Devices, setting the same rate, and be sure to set the matching sample rate in Daqarta.

Note that if you simply set the desired rate in Daqarta without setting it in Windows or the custom control panel, you'll just get a simulated rate via Windows sample rate conversion. This is a "feature" of the Windows Driver Model, which also affects the ASUS Xonar U7 and probably other cards that allow sample rates above 48000 Hz.

The custom driver does provide a "What U Hear" input (otherwise known as "Stereo Mix"), which allows Windows XP users to view MIDI output waveforms, spectra, or spectrograms (not just the Pitch-to-MIDI note display) by toggling Pitch Track off during a DaqMusiq performance.

(On Windows Vista and later you may not need this, since it typically routes the output to the input by default. If not, see Hidden or Disabled Inputs under Vista / 7 / 8 / 10 / 11 Issues in the Troubleshooting section.)

One potential problem with the X-Fi is the front-panel volume knob, which is easily bumped and apparently can't be disabled via the driver software. Fortunately (unlike the UCA202), it affects the digital volume controls, so you can at least see in Daqarta (via the F9 key to open Daqarta's calibrated volume controls) if it has been bumped.

To avoid such problems entirely, you can disable this control mechanically via a drop of hot glue between the knob and panel. Or, if you want to easily undo it, you can use masking or painter's tape. The knob position doesn't matter, since the software can control the volume directly.

Note, however, that pushing the knob in will mute the output; a large green 'Mute' will appear on the desktop at the lower right. This is independent of the software mute. Make sure it's not muted before you unplug the unit to glue it.

Note: This unit is stereo only. If you use it with Daqarta's Multi-Channel Outputs the channels will be combined into stereo, with the exception of the Low Freq channel which is never used. Instead, the Center channel will appear in the Right output as well as the Left.

If the Channel Selects include either Front channel, all the rest will be at a much lower volume. This is true even if the Front channels have no active Left and Right Stream buttons. With the Front channels off, the volume of the remaining channels will be much louder, but will be reduced in proportion to the number of channel pairs selected.



ASUS Xonar U7 7.1 USB:

Price: US $90 or more, no longer in production

Dimensions: 5.5 x 3.2 x 1.0 inches. The 3.2 inch depth measurement does not include 0.2 inches on the rear for the RCA output connectors.

Includes: USB cable, SPDIF TOSLINK optical adaptor, Quick Start Guide, Driver / User Manual CD.

Connectors: USB, SPDIF coaxial/optical Out, L/R Front Line Out (RCA), L/R Side, Center/Bass, Rear Line Out (3.5 mm), Mic/Line In (3.5 mm), Headphone Out (3.5mm).

Controls: Output volume knob on top (click to toggle between headphones and speakers), Mic volume +/- rocker switch on front. There is also a tiny USB Audio 1.0-2.0 slide switch on the bottom. All tests were conducted with this switch in the 2.0 position.

The Xonar U7 uses a Cirrus Logic CS5361 stereo ADC chip (as does the Sound Blaster X-Fi, above), and two Cirrus Logic DAC chips: A CS4398 stereo DAC, and a CS4362A 6-channel DAC. All chips are rated for 24-bit, 192 kHz sampling.

Note: The driver must be installed from the supplied CD for proper operation.

This is the best all-around card of the test group: It not only has superbly low noise and distortion, but the best frequency response as well. In addition, it can run at sample rates up to 192000 Hz and has 8 output channels.

However, just like the Creative X-Fi HD discussed above, it does not allow Daqarta to directly control sample rate. To change the sample rate, double-click the orange-and-black Xonar icon in the system tray. On the left side of the control panel, right-click the Speakers icon and select Sample Rate, then click on the desired rate. Then right-click on Line In and set the same rate. Then start Daqarta and set that same sample rate there. You can change the rate while Daqarta is open, but you should toggle the Generator and Input off until the above steps are completed.

Alternatively, instead of the Xonar control panel, users of Vista and later Windows versions can right-click the normal Windows speaker icon in the system tray and set Playback devices, then Speaker - Xonar U7 then Properties at the bottom of the dialog, then Advanced, and select "16 bit" at your chosen sample rate (typically 48000, 96000, or 192000). Repeat for Input Devices, setting the same rate, and be sure to set the matching sample rate in Daqarta.

Note that if you simply set the desired rate in Daqarta without setting it in Windows or the custom control panel, you'll just get a simulated rate via Windows sample rate conversion. This is a "feature" of the Windows Driver Model, which also affects the Creative X-Fi HD and probably other cards that allow sample rates above 48000 Hz.

As shown under the Loopback Frequency Response Comparison section above, the loopback response at the 192000 Hz sample rate is far from flat, rolling off at only 54000 Hz instead of closer to the 96000 Hz Nyquist limit (half the sample rate) as might be expected. This is primarily due to the output response rolling off at 55000 Hz; the input response on its own extends to 75000 Hz.

At very low input and output frequencies (1 Hz and less) there are quirks in both the output and input responses. Line Out has a peak of +4.9 dB at 0.5 Hz without a load, falling to +3.5 dB at 1 Hz with a 1000 ohm load, and +2.3 dB at 100 ohms.

Line In has a smooth response down to 0.6 Hz, but it has strange overload behavior. Though it handles signals up to 1.00 Vrms at most frequencies, it is limited to about 300 mV at 1 Hz. This does not appear to be simple waveform clipping, but may be due to saturation in a 'servo' circuit used to get extreme low-frequency response without large (or any) capacitors in the direct signal path.

Neither of these quirks would be cause for concern under normal use, since the frequencies involved are far below the audible range. But you should keep these in mind if you are trying to make measurements at very low frequencies.

For Multi-Channel Outputs you must set the proper configuration in the driver. Open the driver control panel, right-click on the Speakers icon and select Speaker Settings, then click on 7.1 Surround. (The other options are Stereo, Quadraphonic, and 5.1 Surround.)

Alternatively, for Windows Vista or later you can set the configuration in Windows itself. See Multi-Channel Sound Card Configuration for the needed steps, but the above driver method is easier.

Once you have set the configuration, you can right-click the Speakers icon in the driver control panel and select Volume Control to see sliders for Master, as well as individual controls for each channel. By default the Left and Right sliders are locked together in pairs (Front, Rear, Cen/LFE, and Side). To unlock the separate sliders of a pair, click on the lock icon at the bottom between 'L' and 'R'.

For Windows Vista/7/8/10/11, note that Daqarta's own Master and Wave volume controls (F9-key) only apply to the Front outputs. However, the Master Mute will mute all 8 channels. (Wave Mute only mutes the Front pair.)

For Windows XP, when Multi-Channel Outputs is active, Daqarta replaces the above controls with the Multi-Channel Volume Controls dialog for full independent control for each of the channels.

Caution: The Xonar U7 has a large round volume control that can be pressed to toggle between output sources. However, doing this causes Daqarta to lose control of the volume... even after you click back to the original source. The same thing happens if you manually change output connections between headphone and line out.

Unlike some stereo-only sound cards, when the U7 is configured for Stereo but used with Daqarta's Multi-Channel Outputs, all Left channels (including Center) are mixed together equally, as are all Right channels (including Low Freq).

IMPORTANT: The Xonar U7 identifies its outputs as different devices (Speaker or Headphones) depending on whether the headphone jack is in use. Each such output device must be calibrated separately, just like a separate sound card, which will result in a separate calibration (.ATN) file... see Managing Multiple Devices.

Note, however, that the U7 uses standard 3.5 mm stereo jacks for inputs and headphone outputs, but uses RCA jacks for line outputs. Since auto-calibration uses a "loopback" cable from output to input, you will need a special RCA-to-3.5 mm cable to calibrate the line outputs, whereas only a standard stereo male-male 3.5 mm cable is needed to calibrate the headphone outputs.



Realtek HD ALC883 Chipset (XP laptop):

This is the built-in Realtek ALC883 sound chip of an Acer Aspire 3690 laptop running Windows XP. This is an old mid-priced system, included in this test group to show that you may be able to get decent performance with what you already own.

This chip is specified as capable of operating at a 96000 Hz sample rate, but although this allows an input -3 dB frequency response limit of 42000 Hz, the output limit is only 21300 Hz... the same as when running at the default 48000 Hz rate.

Mic In is mono-only. If both Input channels are on, each receives the identical Left input signal only (the tip of the stereo plug, which is the white conductor of a standard cable). If only one Input channel is active, it will receive half the sum of the Left plus Right input signals. A 2.5 Vdc electret microphone bias is applied separately to both Left and Right channel connectors.

Note: This chipset is stereo only. If you use it with Daqarta's Multi-Channel Outputs the channels will be combined into stereo, with the exception of the Low Freq channel which is never used. Instead, the Center channel will appear in the Right output as well as the Left.

However, unlike the Behringer UCA202 and the Creative X-Fi, the individual channel volumes are not affected by the number of other active Channel Selects.

GO:

Questions? Comments? Contact us!

We respond to ALL inquiries, typically within 24 hrs.
INTERSTELLAR RESEARCH:
Over 35 Years of Innovative Instrumentation
© Copyright 2007 - 2023 by Interstellar Research
All rights reserved