Voice-Activated Audio Recordings
Record every loving conversation and joke of every day. Video is too data intensive for this.
Audio recordings of sweet nothings said with my significant other are some of our most prized possessions. I'd like to have more.
A physical input would also work well, like tapping once for SAVE THE LAST 30 MINUTES, twice for SAVE THE PREVIOUS TWO HOURS, and tap four times for SAVE TODAY. That would give users two smaller chunks of time if they know they want to save something specific. Otherwise they can save the whole day in an emergency. Haptic feedback buzzes or blinking LEDs would assure the user of how much was stored.
The system would then have the challenge of collating the recorded chunks. Minimization of dead air has to sound natural (again, Imesart's iOS app has done a better job of this than most I've tried). The voice-activation program has to already be recording in order to capture the first ticks of human audio just prior to the trigger. And while some conversations depend on dead air for meaning, most just have empty space that can be removed without affecting the flow. It would be a balance for machine learning to estimate what's happening among the voices in order to decide what to cut out. Otherwise you'd store a lot of useless content that would need to be reprocessed once conversation parsing technology advances.
Yep, I'd love to have a CUI-based assistant (conversational user interface) that would record events automatically, and then like the Star Trek computer you could say "Save the last two hours" if something important just happened. Of course that's a few years off and probably has to integrate a large corporate neural network for the conversational aspect, but I'd at least like something less brute-force than carrying my iPhone around in my hand in order to fill the lacuna between now and the future. My dad had a terrible memory, so recording evidence of love and friendship is something I'm becoming more conscious of as I get older.
Josh Armstrong commented
Thanks for getting back to us. Now that's a thorough and amazing response! Thank you! Visions of the far future are always welcome as well!
This has been forwarded to the appropriate persons and we'll consider it for the future. Audio recording is something I'm noticing many of you are asking for, and I really see how it could be useful in my life too. I'll push this and we'll see if capturing your story could evolve into your own personal "audiobook of your life."
Always feel free to post more great ideas, and keep being awesome.
Have a great day!
I have an app called Audio Memos for iOS which has a continuous voice-activated mode that records until the phone runs out of storage or battery. We used it during a vacation recently and captured every day without any failures. It took some practicing ahead of time to get the gain/threshold/decay settings properly calibrated, but in the end it would trigger only when we were talking, even whispering, and from loud non-vocal noises, and we recorded under 2GB of WAV audio per day. I'm not sure what kind of algorithm the developer (Imesart) uses but it's very effective at isolating human speech for the most part. We didn't miss any conversations and we didn't have tons of unwanted noise triggering it.
I really want a continuous and ignorable Narrative device that records everything (I'd wear two Clips, one for audio, one for interval pics), then offload the data to a hard drive for posterity every couple of days. One day when we're eighty we want to remember our youth and share the emotions again without having to put in too much work to capture the events while we're living them. I'm sure there will be data crunching software in the next few decades that can handle big stores of human speech and assemble stories from them on a whim, which would be a nice safeguard against senile depression.
Josh Armstrong commented
Thanks for your suggestion. :)
I really like this idea as well, since I think it could be beneficial for a lot of people.
Would you prefer recordings to be continuous? Or perhaps set to an interval?
When you say voice-activated, would you want to have a command that you say so that the Clip records the last, say, one minute?
Please let us know what you would prefer, and I'll forward this up for consideration. :)
Have a great day, Ryan.