CONNECTHelp Center
All guidesDownload Connect

Best practices for Connect users

Build a repeatable routine for natural delivery, dependable audio, and accurate translation before, during, and after every call.

5 min readUpdated September 6, 2026Every call

Before the call

Prepare the physical headset, Connect configuration, languages, and calling application before the conversation begins. A repeatable 30-second pre-call check catches incorrect routing and poor microphone placement before they affect another participant. Keep the last known-good settings recorded so an unsuccessful adjustment can be reversed quickly.

Connect the headset firstOpen Connect after the physical device is attached so it appears in the list.
Confirm both languagesDo not rely on yesterday's source, target, or language hints.
Activate the right contextUse a saved profile for specialist calls, names, or product vocabulary.
Preview your voicePress How I Sound? and listen for clipping or a wrong voice selection.
Route the calling appSelect Connect Mic and Connect Speaker inside the app's own settings.
Set the boom positionPlace it two finger widths from the corner of the mouth and keep it fixed.
Start each work session with the same baseline. Connect the wired USB headset directly, confirm both physical and virtual devices, disable competing speech enhancement, and make a short test call before the first live conversation.

During the call

Protect the known-good setup while the conversation is active. Keep Connect open, leave the boom microphone in place, and avoid changing several controls while someone is speaking. When a correction is necessary, pause the conversation and change only one variable so the result remains easy to judge.

  • Keep Connect open. Minimizing the window is fine, but closing the app stops its audio processing and routing.
  • Keep the boom fixed. Moving it mid-sentence changes the input level and can make the beginning and end of a phrase sound different.
  • Leave a turn-taking pause. Avoid speaking over another participant, especially in bidirectional mode.
  • Say names deliberately. Add recurring names, brands, and acronyms to a pronunciation dictionary.
  • Watch the service state. If Connect shows waiting or unrouted, pause and check the call and audio-device settings.
  • Preserve the baseline. Do not enable several noise, gain, or enhancement controls during the same live call.

Use clear voice habits

Connect does not require an artificial speaking style, but it benefits from consistent and complete speech. Use a natural conversational rhythm, finish each word, and maintain a steady level from the beginning to the end of the sentence. Clear delivery improves the source audio available to every downstream feature without making the conversation sound rehearsed.

  • Natural paceSpeak at your normal conversational speed. Rushing reduces clarity, while speaking unnaturally slowly makes the exchange feel disconnected.
  • Clear articulationPronounce complete words and avoid mumbling or trailing off. Take a natural pause between sentences instead of stretching individual words.
  • Consistent volumeUse a comfortable call level and hold it steady. Shouting, whispering, or fading at sentence endings makes the signal harder to process consistently.
  • Controlled breathingBreathe between sentences rather than through the middle of a phrase. Keep the microphone beside the mouth so it captures speech without direct breath noise.
Connect can process the audio it receives, but it cannot repair every source-language error. Correct pronunciation, complete sentence structure, and an accurate pronunciation dictionary remain important for names, products, acronyms, and specialized terms.

Choose the right mode

Select the mode that matches who needs translation and how the conversation is expected to flow. Keep the original setting noted so an unsuccessful adjustment can be reversed quickly. This makes the outcome easier to evaluate, explain, and reproduce across future calls. Use the guidance here to make one deliberate change at a time and preserve a clear baseline.

SituationRecommended setup
You speak one language to an audienceUnidirectional translation
Both sides need translationBidirectional translation with clear turn-taking
Speed is the priorityInstant Mode with short, complete sentences
Several people share the conversationSpeaker Labeling and expected-language hints
Specialist terminology is expectedActivated Conversation Context plus a pronunciation dictionary

After the call

Complete these final checks before leaving the session or returning your audio settings to their normal configuration. Use the guidance here to make one deliberate change at a time and preserve a clear baseline. Confirm the result in a controlled test before moving to the next decision or configuration step.

  • Press CONNECT-OFF before closing the calling app.
  • Return the calling app to your normal audio devices if the Connect route was temporary.
  • Review transcription history only when it is enabled and appropriate for your account and conversation.
  • If quality changed during the call, record the time, platform, selected devices, and visible service state before contacting support.

Avoid these common mistakes

These issues are easy to miss and can reduce clarity, interrupt routing, or prevent translated audio from reaching the call. Most are caused by a changed device, an extra processing layer, or a skipped pre-call check rather than the translation itself. Compare the current setup with the recorded baseline and reverse one difference at a time.

Do not start with every enhancement enabled. Begin with the simplest working path, then add Noise Suppression, Speaker Labeling, Emotion Transfer, or other options only when the call needs them.
  • Starting Connect after the call and expecting the calling application to switch devices automatically.
  • Using an external webcam microphone, a laptop microphone, or loud open speakers instead of the approved USB headset.
  • Setting only the operating-system default instead of selecting Connect Mic and Connect Speaker in the calling application.
  • Placing the microphone directly in front of the lips, beneath the nostrils, or against the face.
  • Changing source or target language while someone is speaking.
  • Resetting virtual audio as a routine step instead of a recovery action.
Best habit: run the same 30-second pre-call check every time. Consistency removes more audio problems than last-minute troubleshooting.
+

Frequently asked questions

These answers clarify the habits that make Connect calls more consistent from one session to the next. Use them with the pre-call checklist rather than changing several controls during a live conversation. If quality changes, compare the current setup with the last known-good baseline before troubleshooting further.

Do I need to speak slowly for Connect?

No. Speak at your natural conversational pace, pronounce complete words, and leave a short pause between turns. Speaking unnaturally slowly can make the conversation feel disconnected, while rushing or trailing off reduces the clarity of the source audio.

What should I check before the first call of the day?

Connect the approved USB headset, confirm its microphone and speaker inside Connect, verify the language direction, and select Connect Mic and Connect Speaker in the calling application. Finish with How I Sound? and a short internal test call.

Can I close Connect after the call has started?

No. You may minimize the window, but closing Connect stops its processing and virtual-audio route. Keep it running until the conversation is complete, then press CONNECT-OFF before leaving the call.

Should noise suppression be enabled in both Connect and the calling app?

Start with one primary processing layer. Disable or minimize overlapping noise suppression, automatic gain, and voice enhancement in the calling application or headset software, then enable only the Connect controls required for the call.

Support

Need help? Get in touch with our Support Team for assistance. Include the calling application, operating system, selected microphone and speaker, translation direction, and the exact step that failed. A short description of the expected result and what happened instead helps the team respond more quickly.

The following articles continue this workflow with setup instructions, troubleshooting checks, and feature-specific guidance. Start with the resource that matches your next task, keep the current configuration available for comparison, and validate each change in a short test call before applying it to a live conversation.