Friday 30 September 2016

Seminar Anouk Sangen 26 October 2016

  Bekijk de webversie  
 
KU Leuven
Invitation Holy Hour
ExpORL, Dept. Neurosciences
 

Children with single sided deafness 

Auditory, language and cognitive development
with and without a cochlear implant

Anouk Sangen

Research Seminar in framework of doctoral education

When?

Wednesday, 26 October 2016, 16:00h

Where?

Seminar Room HP2 (04.227)
O&N II - KU Leuven
Herestraat 49
B-3000 Leuven

Please reply to this email if you are planning to attend the presentation.

 
 
   

Wednesday 21 September 2016

Technical comparison of sound cards: Presonus AudioBox USB/FocusRite Scarlett 2i2


At ExpORL we've standardised on RME sound cards (in particular the Hammerfall DSP Multiface II and Fireface UC) for all our behavioural and objective experiments. While these perform very well, they're also very expensive. We therefore explored whether two much cheaper sound cards, the Presonus AudioBox USB and FocusRite Scarlett 2i2, would be suitable for certain experiments with less strict requirements, such as speech intelligibility in noise.

We concluded that the Presonus AudioBox USB is not suitable for faithful playback nor recording because there is too much stereo crosstalk. The FocusRite Scarlett 2i2 is suitable for faithful playback or recording of audio. However, one has to be aware of the following constraints:
  • Its output dynamic range (around 96 dB) might might be to small for e.g. tone audiometry for hearing impaired people.
  • The sound card is not suitable for playback and recording simultaneously, since the inputs are directly routed to the main output, even if direct monitoring is switched off (for which there is a hardware switch)
  • Input and output gains are adapted with analog hardware knobs (as opposed to the RME sound cards that are set within the driver software). We advise to set the monitor level at the maximum for optimal performance and to avoid ambiguous calibration; input gains should be adjusted such that no clipping occurs.

We analyzed the sound cards using RightMark Audio Analyzer (RMAA) and used the RME Fireface UC as a reference. We used stereo sound cables to obtain balanced input/output. For each test, we used 44 kHz, 32-bit (float) audio. Our test results can be found below.


Complete sound card tests (input coupled to output)

Summary

Test
AudioBox USB Audio
Focusrite Scarlett 2i2 USB
RME Fireface UC
Frequency response (from 40 Hz to 15 kHz), dB:
-0.20, -0.52
-0.01, -0.07
+0.03, -0.03
Noise level, dB (A):
-94.4
-95.8
-109.9
Dynamic range, dB (A):
94.8
95.6
109.9
THD, %:
0.0048
0.0009
0.0016
IMD + Noise, %:
3.776
0.0071
0.0019
Stereo crosstalk, dB:
-42.9
-95.5
-109.5

Frequency response

Spectrum graph

Noise level

Spectrum graph

Dynamic range

Spectrum graph

THD + Noise (at -3 dB FS)

Spectrum graph

Intermodulation distortion

Spectrum graph

Stereo crosstalk

Spectrum graph

Inputs only (output of RME)

Summary

Test
AudioBox USB Audio
Focusrite Scarlett 2i2 USB
Frequency response (from 40 Hz to 15 kHz), dB:
+0.04, -0.17
+0.05, -0.33
Noise level, dB (A):
-98.6
-98.2
Dynamic range, dB (A):
100.4
98.1
THD, %:
0.0031
0.0018
IMD + Noise, %:
0.027
0.0042
Stereo crosstalk, dB:
-99.3
-98.6

Frequency response

Spectrum graph

Noise level

Spectrum graph

Dynamic range

Spectrum graph

THD + Noise (at -3 dB FS)

Spectrum graph

Intermodulation distortion

Spectrum graph

Stereo crosstalk

Spectrum graph

Outputs only (input of RME)

Summary

Test
AudioBox USB Audio
Scarlett 2i2 USB
Frequency response (from 40 Hz to 15 kHz), dB:
+0.06, -0.22
+0.01, -0.08
Noise level, dB (A):
-96.9
-96.9
Dynamic range, dB (A):
96.7
96.6
THD, %:
0.0047
0.0011
IMD + Noise, %:
0.012
0.0044
Stereo crosstalk, dB:
-95.6
-89.7

Frequency response

Spectrum graph

Noise level

Spectrum graph

Dynamic range

Spectrum graph

THD + Noise (at -3 dB FS)

Spectrum graph

Intermodulation distortion

Spectrum graph

Stereo crosstalk

Spectrum graph
We could not eliminate low frequency crosstalk for Presonus AudioBox USB. High frequency crosstalk disappeared for the FocusRite Scarlett 2i2 with the MONITOR-knob turned at its maximum: