From Khoa Nguyen (1/27/04):
Cool, put all my 3 kids for a test drive and it was fun. We have karaoke at home and my kids love turning up echo on the karaoke machine and have fun.
Anyway, here's how it break down with the ages:
- 5 yr old - so so, she thought it's another karaoke (3 minutes of playing, mostly just singing).
- 3 yr old - absolute blast, play & giggle for 10-15 minutes before giving up.
- 1 yr old - Confused but love watching her big brother play with it. Understandable since this one is still confused about "why does Daddy's voice come out of the phone?".
I think it'd be great if you can add just a pure echo mode (just small delays + decay) for the fun factor. Another fun factor is pitch change, but this may scare some kids.
From Steve Nordquist (25/04/04,) linking in off Lockergnome: This is just what I ordered as a replacement for the reel-to-reel I loved as a kid of 7, and wanted later for voice and breath training. (You can talk fast better, tune up a tin voice, reinvent throat-singing, all that.)
Improvements, godforbid (hence why not then) this isn't just sucked into ALSA userland stuff:
- If the input isn't already u-law, get the log of it and display THAT or use whatever's handy to render a block of log(sample(n))
- Pick the windows function to deliver filtered input, if not already done. Last I read, there was a raw input for people who want to use their soundcard as an oscilliscope, and another with some basic mic filtering that goes a long way.
- Handy thumbwheel to adjust buffer size or click back in it to replay from a given point
- Doublebuffer to permit more moderate fade of old buffer sound
- erm...noise reduction, and mic remediation junk.
Sure, it's not the plugin API extravaganza that some (mic'd) webcam software is, but I can't redo lighting to indulge like I can redo sound. Even then (laptop fan whirs throatily...it's a TUVA fan. :) )
27Apr04 - Brian Theado - Does anyone have a link to this Lockergnome post and/or contact information for Steve Nordquist? I don't fully understand the above ideas and I'd like to find out more. The only reference on Lockergnome to the babbleback machine I could find was http://channels.lockergnome.com/windows/backissues/20040423.phtml#20040423_3
2004-06-14 - Jason Clinton, KC, MO - Steve Nordquist is patently insane. He's been posting on the Internet since at least 1994. Some of us on the KULUA mailing list have theorized that he's really a government AI experiment gone wrong. There's a guy out there somewhere collecting quotes of Steve's and adding them to a Fortune database. If you're interested in looking at some of his postings to our mailing list: 1
2004-11-12 - Chris Dorna, The Netherlands - We want to use babbleback to simulate a speech defect. The problem is that the delay is to long and the steps (interval) is to large. Is het possible to decrease the steps (0.1 sec) and to start with a delay of 0.1 sec.
Hope someone can sent a mail to chris/at\dorna.nl
2004-11-17 - Brian Theado
Depending on which aspects of the babbleback machine are important to you, I think I can help you with this one way or another. The underlying implementation of the current babbleback doesn't lend itself very well to having very small delays. The small step value isn't a problem its just the small delay value that is a problem.
I don't have a microphone readily available right now to test this out, but here is something you can do to see if it works:
Hit <F2> to bring up the console window and type in the following command
.delay configure -from .1 -to 5 -resolution 0.1
Now the slider bar should range from .1 to 5 seconds with a step value of .1. I recommend disabling the waveform display before changing the delay to anything less than 1 second.
I have a feeling this might not work very well even if you do disable the waveform display. Please let me know.
I have another idea that might work better for you. There is a new, cool programming language called chuck (http://chuck.cs.princeton.edu) that makes it simple to implement the core functionality of the babbleback machine (no graphics--just the continuous recording and playback with delay). The code is just 4 lines long and the underlying implementation is stream based and won't have the same problems as my original babbleback machine implementation. See ChucK babbleback machine implementation