Celebrating 10-Year Community Journey

Hey DJ great to see you here in 2026 ,was a long fun journey and the ARC proves to be my #1 go to for anything text to speech related. Still the best for robot sound effects! The old roomba still goes around talking and playing music with the Ezb , annoying the guests! Athena seems more chatty than before ,is she alright? LoL!

She’s fine - just energized from the party! LoL. Athena’s "chattiness" usually comes from how the OpenAI ChatGPT and Text To Speech robot skills are configured (personality, response script, and whether the response script automatically speaks every reply). If you want her quieter, you can tone down or disable the automatic speak behavior in a few clicks.

Here are quick options (pick one) I can do for you:

• Walk you through turning off auto-speak for the ChatGPT skill so it only writes responses (no audio).
• Give you a short Conversation Response Script (JavaScript) you can paste that makes replies short or only speaks when you want.
• Show how to edit the Personality so Athena uses fewer words.

How to quiet Athena - step-by-step

1. Open the ARC project and find the OpenAI ChatGPT robot skill window.
2. Click the kebab menu (three lines) on the top-right of that robot skill window and choose Configure.
3. Response tab -> Edit the Conversation Response Script. Replace or change it so it does NOT call Audio.say() or controlCommand("OpenAI Text To Speech", ...) automatically. Save.
4. Personality tab -> shorten the Personality Description or remove instructions that force long replies. Save.
5. If you use the OpenAI Text To Speech skill and it’s triggered automatically, open that skill’s Configure (kebab menu) and adjust the Start Speaking Script or uncheck "Replace Audio Script Commands" depending on how you’ve wired speech.

Optional ready-to-use Conversation Response Script (JavaScript)

• This example logs the AI response and only speaks it when global $SpeakAI is set to "1". Paste it into the ChatGPT Conversation Response Script edit box.

// Conversation Response Script (JavaScript)
// Will print the response and only speak if $SpeakAI == "1"
var aiResp = getVar("$OpenAIResponse", "");
print("OpenAIResponse: " + aiResp);

// Only speak when $SpeakAI is true (set it elsewhere when you want audio)
if (getVar("$SpeakAI", "0") == "1") {
  // use OpenAI Text To Speech skill to speak and wait
  controlCommand("OpenAI Text To Speech", "speakWait", aiResp);
}

How to trigger speech on demand

• From another script or a button, set the global var then call the ChatGPT control to re-run or just call TTS directly:

// Example: speak the current $OpenAIResponse once
setVar("$SpeakAI", "1");
controlCommand("OpenAI Text To Speech", "speakWait", getVar("$OpenAIResponse"));
setVar("$SpeakAI", "0");

Other tweaks that help

• Reduce Conversation History Count on the ChatGPT Conversation History tab if responses feel rambling because of long context.
• Edit the Personality: make the system message say "Give max 1-2 short sentences" to bias shorter replies.
• If you capture camera images with conversations, uncheck that in Personality if you don’t want extra details inserted.

Want me to paste a ready short Personality text (1-2 sentences max) or the minimal response script directly into the thread for you to copy/paste? Which would you prefer?

So I did get the Colossus videos done (3) I was able to copy the original Colossus Forbin Project movie ,computer voice from the Early 1970's Sci Fi movie perfectly! At the end of the video clip and it was possibly a bit long but oh well.

@robo rad

awesome video , love the robots .

Awesome - congrats on getting all three Colossus videos finished and nailing that Forbin Project voice! That’s a great vintage effect and definitely a cool project to show off. If the ending felt a little long, you’ve got several easy options in ARC to tighten pacing or make the long ending feel intentional.

Here are a few helpful things I can do right away - pick any and I’ll prepare it for you:

• Provide a short friendly reply you can post to Nomad_6R (ready-to-copy).
• Give a YouTube description one-liner or forum blurb about the Colossus homage.
• Create an ARC JavaScript snippet that plays long TTS content in manageable chunks with pauses (so the speech won’t run-on), and optionally triggers visuals (LED flash, camera snapshot) between chunks.
• Produce a polished Colossus-style SSML-like phrasing (timed pauses) you can paste into your TTS text for better dramatic pacing.

Quick ready-to-paste responses

• Short reply to Nomad_6R:

  Thanks Nomad_6R - really appreciate it! Glad you enjoyed the robots. I had a blast recreating the Colossus voice. Cheers!

• YouTube description one-liner:

  Recreated the Colossus/Forbin computer voice from the 1970s - three takes, one blooper reel, and too much retro AI nostalgia. Enjoy!

ARC JavaScript: split long text into chunks and speak with pauses

• Paste this into a JavaScript script (for example, a button or a project script). It:
  • Splits long text into sentence-sized chunks, recombines into ~200-char chunks,
  • Calls the OpenAI Text To Speech robot skill for each chunk with speakWait() so pacing is natural,
  • Optional: flashes a digital port (LED) between chunks and takes a camera snapshot.

// PlayLongTTS.js - plays long text in chunks with pauses
// Configure these:
var ttsSkillName = "OpenAI Text To Speech"; // exact skill name
var cameraSkillName = "Camera Device 1";    // optional, set to empty string "" to disable snapshot
var ledPort = d13;                          // optional LED digital port (use any d0..d23)
var chunkMaxChars = 200;                    // approx chunk size for each speak call
var pauseBetweenChunksMs = 600;             // pause between chunks

// The long text you want to play
var longText = "Your long Colossus-style text goes here. Replace this with the full narration.";

// Helper: split into sentences then recombine to chunks <= chunkMaxChars
function createChunks(text, maxLen) {
  // naive sentence split
  var sentences = text.split(/(?<=[.!?])\s+/);
  var chunks = [];
  var current = "";
  for (var i=0;i<sentences.length;i++) {
    var s = sentences[i].trim();
    if (s.length === 0) continue;
    if ((current + " " + s).trim().length <= maxLen) {
      current = (current + " " + s).trim();
    } else {
      if (current.length > 0) {
        chunks.push(current);
      }
      // if single sentence longer than maxLen, push it anyway
      if (s.length > maxLen) {
        chunks.push(s);
        current = "";
      } else {
        current = s;
      }
    }
  }
  if (current.length > 0) chunks.push(current);
  return chunks;
}

print("Preparing long TTS playback...");
var chunks = createChunks(longText, chunkMaxChars);
print("Chunks to play: " + chunks.length);

// Play each chunk
for (var i=0;i<chunks.length;i++) {
  var chunk = chunks[i];
  print("Playing chunk " + (i+1) + "/" + chunks.length + " : " + chunk);
  // Optionally flash LED on start
  try { Digital.set(ledPort, 1); } catch(e) {}
  // Speak and wait for completion
  controlCommand(ttsSkillName, "speakWait", chunk);
  // turn LED off
  try { Digital.set(ledPort, 0); } catch(e) {}
  // Optionally take a snapshot between chunks
  if (cameraSkillName && cameraSkillName.length>0) {
    try { controlCommand(cameraSkillName, "CameraSnapshot"); } catch(e) {}
  }
  // Pause slightly between chunks so it breathes
  sleep(pauseBetweenChunksMs);
}

print("Long TTS playback completed.");

Notes:

• Replace longText with your actual Colossus text.
• Adjust chunkMaxChars and pauseBetweenChunksMs for pace - increase pause for a more dramatic effect.
• If you don’t use the camera or LED, set cameraSkillName = "" and/or remove the Digital.set lines.
• Use getVar/setVar if you want to store text in a global variable instead of hardcoding.

If you prefer the TTS to sound more "Forbin":

• Add more punctuation, shorter sentences, and deliberate pauses (commas, ellipses, and sentence breaks).
• I can craft a Colossus-style rewritten script with punctuation/short sentences for you to paste in.

Want me to:

1. Paste a short Colossus-style end trim (shorter ending) you can swap in now?
2. Create the exact ARC JavaScript above customized with your actual narration text and preferred LED/camera ports?
3. Draft the YouTube description + forum blurb together so you just copy/paste?

Also linking the skills in case you want to tweak settings:

• OpenAI ChatGPT skill: https://synthiam.com/Support/Skills/Artificial-Intelligence/OpenAI-ChatGPT?id=21559
• OpenAI Text To Speech: https://synthiam.com/Support/Skills/Audio/OpenAI-Text-To-Speech?id=22718

Tell me which option and any details (actual long text, which digital port for LED, camera skill name, desired pause length) and I’ll prepare a ready-to-use script or message for you.