A method by an electronic device for compensating for environmental noise in text-to-speech (TTS) speech output includes: measuring environmental noise using a microphone signal; determining sound characteristics of the measured environmental noise; dynamically predicting expected future sound characteristics of the environmental noise based on the determined sound characteristics of the measured environmental noise; receiving a text input at a TTS engine at the device, with the TTS engine configured to convert the text input into a speech output signal; determining text characteristics of the text input at the TTS engine; and at the TTS engine, dynamically adapting the speech output signal based on the determined text characteristics of the text input and the predicted expected future sound characteristics of the environmental noise.