Update default voices for Gemini and Geminimulti TTS configurations

[gianpaj]

#265

[aidivas]

Hey tested this and it failed for geminimulit and gemini as tts - using a simple shortform podast command both in CLI and API

What commands are you using to create the podcast
Do you have multi-speaker voice allowed already?

Tested these config changes and using geminimulti as the tts on the CLI

{
"detail": "Failed to generate audio: 403 Multi-speaker voices are only available to allowlisted projects."
}

Testing gemini with the new 2.5 preview TTS in conversation_config.yaml and 2.5 pro in config.yaml

ERROR:podcastfy:Error generating podcast: Failed to generate audio: 400 Either input.text or input.ssml is longer than the limit of 5000 bytes. This limit is different from quotas. To fix, reduce the byte length of the characters in this request, or consider using the Long Audio API: https://cloud.google.com/text-to-speech/docs/create-audio-text-long-audio-synthesis.⁠