Trying to work the Headset on Windows

[je1rav]

I have been trying to get the library to work on Windows.
I modified "Adafruit_USBD_Audio.cpp" a little as follow.

from
// Setup endpints and interfaces
if (isHeadset() && _desc_len==0) {
_itfnum_mic = TinyUSBDevice.allocInterface(); // input interface
_itfnum_spk = TinyUSBDevice.allocInterface(); // output interface
_ep_mic = TinyUSBDevice.allocEndpoint(TUSB_DIR_IN); // intput
_ep_int = TinyUSBDevice.allocEndpoint(TUSB_DIR_IN); // input
_ep_spk = TinyUSBDevice.allocEndpoint(TUSB_DIR_OUT); // output
_itf_number_total += 2;
}

to
// Setup endpints and interfaces
if (isHeadset() && _desc_len==0) {
_itfnum_spk = TinyUSBDevice.allocInterface(); // spk output interface
_ep_spk = TinyUSBDevice.allocEndpoint(TUSB_DIR_OUT); // spk output
//_ep_int = TinyUSBDevice.allocEndpoint(TUSB_DIR_IN); // spk input
_itfnum_mic = TinyUSBDevice.allocInterface(); // input interface
_ep_mic = TinyUSBDevice.allocEndpoint(TUSB_DIR_IN); // mic intput
_itf_number_total += 2;
}

On Windows, the match between the order of the endpoint configuration and the order of the interfaceDescriptor
seems to be important.
I do not know the role of ep_int, so I can not answer if the removal of ep_int will be allowed to remove.
But, it was not work with _ep_int in Windows.

The sound of the microphone part in “audio_headset.ino” was tested using Audacity on Macos and Windows.
The sound of the spk part without ep_int has not yet been tested, but the counts of ‘Speaker Samples’ have been increased.

Is this approach correct?

[je1rav]

Hi,
I did a callback test using Audacity on Macos without "_ep_int".
It appears to be working without "_ep_int".
However, one sound data was lost every 1 ms with or without "_ep_int".