Ok, this is a big release. I'll fix any issues with this and then hopefully release an official beta.
Download link as usualThere are two main changes:
1) autocomplete/tooltips/"goto definition" for the script editor
2) audio support for messages
Autocomplete---------
This is on by default, but it might be off if you already had SCI Companion on your machine (as I'm sure you do). In that case, go to Tools->Preferences, and check "Enable browse info"
Now you should have autocomplete (e.g. "intellisense") as you type. You can also hover the mouse cursor over variables/functions to get more info. Or right-click on them and "goto definition". You can also quickly navigate through the script using the combo box at the top of the script editor.
Hopefully this is all reliable and fast now :-)
It does take a few seconds when you open a game for the "browse info" to be built (status should be on bottom left)
Message audio (speech)--------------
- If you have a microphone attached to your computer, you can now just click on a message entry and press the record button to record audio for that message. Or you can load a wav file.
- If you record your own audio in SCI Companion, I keep around the original recording as a wav file. That way you can re-process it with different settings (using the little button with a waveform icon). The original recording is 16-bit audio, but the default for message audio is 8-bit, which is fairly noisy - no way to get around it. You can play around with different compression settings, or noise gates, etc..
- Since the resource.aud file can get really huge (typically around 400MB for Sierra games), I don't rebuild it every time you save a message (with audio) resource. Instead, the audio resources are kept as individual files in the audiocache folder. Whenever you run the game, it checks to see if anything has been changed, and it will automatically repackage everything into resource.aud. It should be mostly transparent, i.e. you shouldn't need to think about it. You can also force things to be repackaged using the icon that looks like a compact disk (in case you're not using SCI Companion to run/test your game).
- You can automatically generate lip-sync data for a message entry with audio. "Quick Lip Sync", or more options in the Advanced dialog. The lip-syncing can be a bit hit or miss. Obviously the more clear you speak the better. If you've used SCI Companion to record the audio, I feed it the post-processed 16-bit audio (even if you've specified the actual audio resource to be 8-bit). I'm not too familiar with the wrapper library I'm using (which uses Microsoft SAPI underneath), but it seems to be hard-coded to US English. It probably wouldn't be too much effort to have this configurable for different accents and languages, but I'm not sure (I suspect the lipsync results will still be somewhat reasonable for other languages anyway).
- You have the option of using the message text to aid in the lip-syncing. Sometimes this makes it worse (it gets the right mouth shapes, but they don't line up with the audio - whereas just using the audio, you get things lined up nicely, but maybe they are the wrong mouth shapes). If you get bad results with one option, try the other.
- You can use the "sample" mouth to check how good/bad the lip-syncing is. But to actually get it working in the game for an actual mouth view, you need two things:
- you need to say which view/loop is the mouth for a given talker number (this should be whatever your Talker instance uses)
- you need to come up with a mapping of the phonemes to particular cels in that mouth view/loop
Let me know if there are any questions! Hopefully it's not too hard to figure things out.
I'll release my "explore SCI1.1" game soon, with full speech and lip-syncing in it.