Step 1: NAS commons settings
The step 1 of the OpenNAS wizard allows the user to select the target FPGA-based board in which the NAS will be deployed, as well as the basic options like the number of frequency channels in which the sound will decompose and if its a monaural NAS or a binaural NAS.
This parameter is used to generate the constraints file correctly (according to the board's resources and its characteristics) and to set the Clock Freq. parameter automatically. When "OTHER" option is selected, a constraints file is generated, including the signals but without any pin assignment. The user can modify this file after the whole process to complete the pin assignment with its board specifications.
- AERNODE: to deploy the NAS into the AER-Node board.
- ZTEX: to deploy the NAS into the ZTEX 2.13 board.
- SOCDOCK: to deploy the NAS into the SOCDOCK board.
- OTHER: not specified boards.
This parameter indicates if the NAS has either one or two "ears". For some applications, as pure tone classification, monaural NAS is sufficient. However, for sound source localization applications, a binaural NAS is needed to calculate the difference between the two ears' input audio signals.
- MONO: to generate a monaural NAS. In this case, the left ear is used.
- STEREO: to generate a binaural NAS.
This parameter establishes in how many different frequency channels the NAS will decompose the input audio signal, inspired in how the biological cochlea works. The maximum number of frequency channels will depend on the resources that your target board has.
- Minimum: 2.
- Maximum: 512.
- Recommendable: power of 2 values, i.e., 8, 16, 32, 64, 128, 256, 512.
Depend on your target FPGA-based board. Take care with this parameter, because even if it is set wrong, the NAS will work, but the output will not be correct.
- AERNODE: 50 MHz.
- ZTEX: 48 MHz.
- SOCDOCK: 100 MHz.
If the user selects "OTHER" in the NAS chip field, the user MUST set this parameter properly. Otherwise, the NAS could not work as expected.
