Hi,

I'm using Google Speech API and EP with uniMrcp.
I can use Voice Recognition.

I would like to use Japanese language.
But I can use English.

I took the log using tcpdump command.
Do you know where I should set the Language?

From log
Line 935 - My application sent grammar and language. Language is Japanese.
<grammar mode="voice" xml:lang="ja-jp" maxage="0" weight="1.0" src="builtin:speech/transcribe" type="application/srgs+xml"/>

Line 1008 to 1014 - I think EP sent language info to mrcp.
Channel-Identifier: 75d3daeaceeb452a@speechrecog
content-id: 1
content-type: text/uri-list
speech-language: ja-jp
Content-length: 25
builtin:speech/transcribe

Line 1031 to 1045 - but send some information again to mrcp. Language is en-us.
Channel-Identifier: 75d3daeaceeb452a@speechrecog
confidence-threshold: 0.500000
dtmf-term-timeout: 0
input-modes: dtmf voice
logging-tag: VM-EP72-185-2018103083038-7
n-best-list-length: 1
no-input-timeout: 8000
recognition-timeout: 20000
save-waveform: false
sensitivity-level: 0.500000
speech-complete-timeout: 2000
speech-incomplete-timeout: 3000
speech-language: en-us
speed-vs-accuracy: 0.500000
vendor-specific-parameters: swi.rec.logValue="ANII=30010|DNIS=41918"

Regards,
Kodera

I can make anything out of the log you attached. I am a little bit confused what the problem is. Are you saying you are sending Jp lang code in grammar to EP from the app server but EP seems to change it to en-us when it sends to the speech server?