Adding weather forecast to max/msp
I’m a student from holland and i’m a max/msp beginner. I love it, and i’m making a interactive controller for my own little project.
It reacts to my voice and knows what time it is and reacts accordingly. I’m going to connect it with an arduino so i can control light, tv en other stuff in my room. I was thinking how great it would be if she would also now what the weather so she could warn me if it’s raining etc.
My question is:
Is it posible to send some realtime information from a weatherforecast site to max/msp ? If so, where do i start? I would greatly apreciate anyone helping me out a little.
Thanks in advance, greetings from holland!
I’ve just experimented by grabbing the latest tweet from my local area (from Manches_weather) and put that straight into [aka.speech]…it was weird.
Out of interest, what are you using for speech recognition and what are you using to get your Max patch to speak?
Did you just copy the tweet ? haha, yes that would sound weird. My idea was to translate the "tweet" or weather forecast to a variable. Like rain is 1 and if it’s 1 she will say watch out is rainy !!
but the problem is getting the weather forecast in max.
I use dragon naturally speaking, text edit, tosymbol, select. And i don’t use a max patch yet. I use http://www.acapela-group.com/text-to-speech-interactive-demo.html so far. But i’m thinking if i find a good patch ( or create one myself ) using max for it.
tell me if i’m not making sense. I can be a little confusing at times.
I used this:
I also recommend using [jit.uldl] if you’re wanting to get information online, probably in conjunction with [jit.str.regexp].
The aka objects are all I’ve tested for speech in and speech out, it’s something I’d like to explore further.
Thanks a lot Dickmedd!
I;m not sure if the aka. objects work for me because i use Win. The aka.listen didn’t work because of that.
Unfortunately I think they are OS X exclusive, I find aka.listen to be quite difficult to use anyway though. Perhaps your solution would work well for now and I’d consider filling it with your own samples for audio feedback. All you need is a bunch of numbers, time-related terms and a bank of different weather variations and you could probably get it working as you want.
I’m tempted to experiment with something similar myself now actually, if I ever get around to it I’ll let you know!
It’s working !
I used [jit.uldl] to download the rss feed. And is used [sadam.rapidXML @symbolOutput1] ( you can find it on the forums )to read it. and some other stuff to get the right information out of the entire feed.
Now i just need to convert the info to something my "controller" understands..