piper-custom-voices / preview-16speakers.sh
agentvibes's picture
Upload folder using huggingface_hub
aca1385 verified
Raw
History Blame Contribute Delete
1.63 kB
#!/bin/bash
#
# Preview all 16 speakers in the 16Speakers.onnx voice model
#
MODEL_PATH="/home/fire/claude/AgentVibes/mcp-server/voices/16Speakers.onnx"
OUTPUT_DIR="/home/fire/claude/AgentVibes/mcp-server/voices/previews"
SAMPLE_TEXT="Hello, this is speaker number"
# Create output directory
mkdir -p "$OUTPUT_DIR"
# Speaker names from the JSON file
declare -a SPEAKERS=(
"Cori_Samuel"
"Kara_Shallenberg"
"Kristin_Hughes"
"Maria_Kasper"
"Mike_Pelton"
"Mark_Nelson"
"Michael_Scherer"
"James_K_White"
"Rose_Ibex"
"progressingamerica"
"Steve_C"
"Owlivia"
"Paul_Hampton"
"Jennifer_Dorr"
"Emily_Cripps"
"Martin_Clifton"
)
echo "🎀 Generating voice previews for all 16 speakers..."
echo ""
# Generate preview for each speaker
for i in "${!SPEAKERS[@]}"; do
SPEAKER_NAME="${SPEAKERS[$i]}"
OUTPUT_FILE="$OUTPUT_DIR/speaker_${i}_${SPEAKER_NAME}.wav"
echo "[$((i+1))/16] Generating preview for: $SPEAKER_NAME (Speaker ID: $i)"
# Generate TTS with speaker ID
echo "$SAMPLE_TEXT $((i+1)), $SPEAKER_NAME." | piper \
--model "$MODEL_PATH" \
--speaker "$i" \
--output_file "$OUTPUT_FILE" 2>/dev/null
if [ -f "$OUTPUT_FILE" ]; then
echo " βœ… Saved to: $OUTPUT_FILE"
# Play the audio
paplay "$OUTPUT_FILE" 2>/dev/null || aplay "$OUTPUT_FILE" 2>/dev/null || afplay "$OUTPUT_FILE" 2>/dev/null
sleep 1
else
echo " ❌ Failed to generate"
fi
echo ""
done
echo "πŸŽ‰ Preview generation complete!"
echo "πŸ“‚ All previews saved to: $OUTPUT_DIR"
ls -lh "$OUTPUT_DIR"