Bill Morrow Posted December 11, 2016 Posted December 11, 2016 I used to use wireless audio with my X10 and HomeSeer to make voice announcements of the motion sensors outside. This was really useful and worked well. Bur since it was wireless, it got a little annoying when the neighbors started watching porn wirelessly at 2:00 AM. I have a few cheap audio ICs that let you record a few seconds of voice and let you play different messages based on the input. I was going to use these with an EZIO, but if someone has already built one.
MWareman Posted December 11, 2016 Posted December 11, 2016 I use voice announcements via my Elk system. Expensive if that's all you need, but if you have an Elk already it may be an option.
Bill Morrow Posted December 11, 2016 Author Posted December 11, 2016 I use voice announcements via my Elk system. Expensive if that's all you need, but if you have an Elk already it may be an option. Thanks, but I already have a few evoxcon boards I bought from SmartHome when they were selling out. I was just hoping to save some time. They still have them on eBay for a scary price.
KeviNH Posted December 11, 2016 Posted December 11, 2016 If you have the Network Module and a spare rPi, I built a simple voice announcement program for Unix which uses free Text to Speech software to take a string sent via the network module and speak it. It can also playback any MP3 file. Since so many free Internet-based text to speech services have become unreliable, my solution uses a local speech synthesis program (The free 'espeak' open source application). espeak takes a second or two to convert text to speech, so I cache each phrase locally, this way the next time it needs to say the same thing, playback is nearly instantaneous. I have mine working with an Apple airplay speaker. I think it is possible to make the announcements play over a google device (Chromecast/CCast Audio) but not via Amazon Echo/Dot/Tap. It should also work with Sonos (needs a Sonos device with line in).
Bill Morrow Posted December 11, 2016 Author Posted December 11, 2016 If you have the Network Module and a spare rPi, I built a simple voice announcement program for Unix which uses free Text to Speech software to take a string sent via the network module and speak it. It can also playback any MP3 file. Since so many free Internet-based text to speech services have become unreliable, my solution uses a local speech synthesis program (The free 'espeak' open source application). espeak takes a second or two to convert text to speech, so I cache each phrase locally, this way the next time it needs to say the same thing, playback is nearly instantaneous. I have mine working with an Apple airplay speaker. I think it is possible to make the announcements play over a google device (Chromecast/CCast Audio) but not via Amazon Echo/Dot/Tap. It should also work with Sonos (needs a Sonos device with line in). Very Kewl. My file-server runs 24/7 so it could host eSpeak. I also have a few Pi laying around.
Jimbo.Automates Posted December 11, 2016 Posted December 11, 2016 Yes that is pretty cool. Currently I use old Android tablets to speak pushover notifications sent by the ISY. But if this can send to chromecast I need to switch since I could send it to my Google Home's! Sent from my Pixel C using Tapatalk
KeviNH Posted December 12, 2016 Posted December 12, 2016 Yes that is pretty cool. Currently I use old Android tablets to speak pushover notifications sent by the ISY. But if this can send to chromecast I need to switch since I could send it to my Google Home's! Should be straightforward to playback the audio through a single chromecast from Linux by using something like CastNow. Not sure about sending to a group of Google speakers simultaneously, but I could see enhancing the script to allow selecting which cast device to target when playing. 1
MWareman Posted December 12, 2016 Posted December 12, 2016 It should also work with Sonos (needs a Sonos device with line in). You can actually get the Sonos to play arbitrary .mp3 files without needing to hook a Rpi up to the line in of a Sonos, purely with the network module and a SOAP request pointing Sonos to the url of the .mp3 file. I tested this a few months ago, but the delay was about 2-3 seconds on top of any conversion delay. Too much for my application.
KeviNH Posted December 12, 2016 Posted December 12, 2016 If the delay is acceptable, then my script already caches the generated audio as MP3, so tying it to Sonos would be trivial to code. I am looking at the line-in option because I also want to connect the line-out from my weather radio.
kurelgyer Posted December 14, 2016 Posted December 14, 2016 I'm a former HAL2000 user. For a time, I used HAL and ISY together. ISY handled the Insteon automation. HAL could respond to scenes and execute text to speech or play sound files. HAL was hooked up to my home theater system for audio output. Example: When an outside motion sensor triggered, and I was not home, an mp3 file of a vicious dog barking played. And then a light would turn on. When an inside motion was triggered and I was not home, all the lights would turn on, and a TTS player notifying the intruder that a photo was taken and the police were called. I'm looking for a way to replicate this without having to have HAL running on a dedicated PC 24/7. 1
larryllix Posted December 15, 2016 Posted December 15, 2016 I'm a former HAL2000 user. For a time, I used HAL and ISY together. ISY handled the Insteon automation. HAL could respond to scenes and execute text to speech or play sound files. HAL was hooked up to my home theater system for audio output. Example: When an outside motion sensor triggered, and I was not home, an mp3 file of a vicious dog barking played. And then a light would turn on. When an inside motion was triggered and I was not home, all the lights would turn on, and a TTS player notifying the intruder that a photo was taken and the police were called. I'm looking for a way to replicate this without having to have HAL running on a dedicated PC 24/7. Will HAL run on a small RPi that you could dedicate to your ISY expansion?
Thanatar Posted December 15, 2016 Posted December 15, 2016 I've been using a Pi loaded with openelec for the past year to play my announcements. I have a lot of different audio files that I call with a network resource. I can even adjust the volume before an announcement. You can use a set of powered speakers or a hifiberry amp on top of the Pi. An added benefit is I can use it to AirPlay to from my phone.
Teken Posted December 16, 2016 Posted December 16, 2016 I've been using a Pi loaded with openelec for the past year to play my announcements. I have a lot of different audio files that I call with a network resource. I can even adjust the volume before an announcement. You can use a set of powered speakers or a hifiberry amp on top of the Pi. An added benefit is I can use it to AirPlay to from my phone. Would you be able to offer a step by step guide how this was accomplished with screen shots of the network resources to do so? I have two RPi's sitting idle that already have openelec preloaded. Any guidance in this area would be greatly appreciated!
Thanatar Posted December 16, 2016 Posted December 16, 2016 (edited) Sure. Links below for others that would like to try. Get the disk image from here: http://openelec.tv/get-openelec/category/7-raspberry-pi2-builds Follow the Wiki to get it installed onto a memory card: http://wiki.openelec.tv/index.php/Raspberry_Pi You’ll need a monitor to set it up initially. On the Pi go to Settings>System>Audio and ensure it is set to your audio device. Most likely Analog, Unless you use a HifiBerry Amp. I used a set of amplified computer speakers for a while, but nearby cell phones would cause interference. So I went with a HiFiBerry Amp+. https://www.amazon.com/HiFiBerry-LYSB014BYS1AQ-ELECTRNCS-AMP/dp/B014BYS1AQ/ref=sr_1_2?ie=UTF8&qid=1481890112&sr=8-2&keywords=hifiberry+amp It is an onboard 25wx2 amplifier. Crystal clear audio without any noise. It stacks on top of the pi. If you go this route be sure to add it to the Pi by following this set up guide. https://support.hifiberry.com/hc/en-us/articles/205311302-Configuring-OpenElec-with-device-tree-overlays You will also have to attach the Pi to your network. I used a wireless USB adapter on a Pi2, the Pi3 comes with wireless already. Under the Openelec menu set the IP address to a static IP so that it doesn't change. Make these changes as well System>OpenELEC>Services>Enable Samba & SSH System>Settings>Remote Control> Turn on both options. System>Settings>Web Server>Allow remote control>On System>Settings>Music>Playback>turn off Play next song Place your audio files under the Music folder on the Pi. I used this website to create my voice announcements. http://www.oddcast.c...ple.php?sitepal There is no way to download the audio files, so I used audacity http://www.audacityteam.org/download/ to capture the audio. Create a network resource with the following info. /jsonrpc?request={%22jsonrpc%22:%222.0%22,%22id%22:%221%22,%22method%22:%22Player.Open%22,%22params%22:{%22item%22:{%22file%22:%22storage/music/Mail.mp3%22}}} All you have to change is the filename in bold to match what your filename is. Keep in mind that filenames are case sensitive. Here is a NR to change the volume to 68. Any value 0-100 is valid. /jsonrpc?request={"jsonrpc":"2.0","method":"Application.SetVolume","params":{"volume":68},"id": 1}} You can also use variables in the NR call to have different responses. You can just append a number onto the end of your filename and use a program to generate a random number. Be sure the number you generate will have an available audio file. My variable is an integer, number 11. My filename is ex Hello1, Hello2, Hello3. /jsonrpc?request={%22jsonrpc%22:%222.0%22,%22id%22:%221%22,%22method%22:%22Player.Open%22,%22params%22:{%22item%22:{%22file%22:%22storage/music/Hello${var.1.11}.mp3%22}}} Playback is almost immediate, I use it as my doorbell (The Jetson’s) and other various alerts. I can have as many audio files as my card will hold. Currently I have around 40 different things that say or do something. Hopefully this helps, If I missed anything let me know and I will update this. Edited December 19, 2016 by Thanatar 6
Teken Posted December 16, 2016 Posted December 16, 2016 Simply fantastic can you offer the link to the TTS? I wanted to use a voice which was as natural as possible which isn't anyone in my home. I think hearing someone in the family speaking over a PA / speaker would be crazy weird! Thank You! ========================= The highest calling in life is to serve ones country faithfully - Teach others what can be. Do what is right and not what is popular.
Thanatar Posted December 16, 2016 Posted December 16, 2016 Simply fantastic can you offer the link to the TTS? I wanted to use a voice which was as natural as possible which isn't anyone in my home. I think hearing someone in the family speaking over a PA / speaker would be crazy weird! Thank You! ========================= The highest calling in life is to serve ones country faithfully - Teach others what can be. Do what is right and not what is popular. I had thought about having the wife make the recordings, But she says I don't listen to her now, So I figured I would never hear the announcements. Here is the link: http://www.oddcast.com/home/demos/tts/tts_example.php?sitepal The cool thing is it doesn't have to be speech. I have the Mario fail sound play for my daughter when it is time to leave for school. I got extra WAF points when I added the AOL "You've got mail" sound for our mailbox. Also one more cool Openelec usage. I have several IP cameras around the house. I placed a few monitors in key areas with a Pi mounted to the back. A simple stream file loaded to the Pi allows me to view the camera feeds using Alexa to say which one is displayed. 1
Teken Posted December 16, 2016 Posted December 16, 2016 I had thought about having the wife make the recordings, But she says I don't listen to her now, So I figured I would never hear the announcements. Here is the link: http://www.oddcast.com/home/demos/tts/tts_example.php?sitepal The cool thing is it doesn't have to be speech. I have the Mario fail sound play for my daughter when it is time to leave for school. I got extra WAF points when I added the AOL "You've got mail" sound for our mailbox. Also one more cool Openelec usage. I have several IP cameras around the house. I placed a few monitors in key areas with a Pi mounted to the back. A simple stream file loaded to the Pi allows me to view the camera feeds using Alexa to say which one is displayed. You Sir, must really work for Santa Clause because this has been a fantastic gift for this X-MAS! Thank you so much for sharing because this new solution will round out my existing HA for two other ongoing projects that can use such a feature. Merry Christmas and the very best to you and yours in 2017!
Teken Posted December 16, 2016 Posted December 16, 2016 Thanatar, Do you have a recommendation for a MP3 sound capture suite to use? I thought there would be some kind of option on that site to define the test to speech and than down load the text string.
Thanatar Posted December 16, 2016 Posted December 16, 2016 You Sir, must really work for Santa Clause because this has been a fantastic gift for this X-MAS! Thank you so much for sharing because this new solution will round out my existing HA for two other ongoing projects that can use such a feature. Merry Christmas and the very best to you and yours in 2017! Glad to give back. I ended up buying the Dashbox and GEM because you and many others spoke so highly of it. I only wish I would have bought it sooner. Once you have tested it, I can't recommend enough the HiFiBerry Amp. It made all the difference in the system and with two speakers placed in different areas of my house I can hear it easily. Let me know how it all works out for you and Merry Xmas to you and yours.
Thanatar Posted December 16, 2016 Posted December 16, 2016 Thanatar, Do you have a recommendation for a MP3 sound capture suite to use? I thought there would be some kind of option on that site to define the test to speech and than down load the text string. I used Audacity: http://www.audacityteam.org/download/windows/ I never did find an online TTS that didn't sound like a robot and would let you download the file. So I used Audacity with it set to record the speech as it played. Then I just trimmed the file and adjusted it so they all sounded the same.
Teken Posted December 16, 2016 Posted December 16, 2016 (edited) Glad to give back. I ended up buying the Dashbox and GEM because you and many others spoke so highly of it. I only wish I would have bought it sooner. Once you have tested it, I can't recommend enough the HiFiBerry Amp. It made all the difference in the system and with two speakers placed in different areas of my house I can hear it easily. Let me know how it all works out for you and Merry Xmas to you and yours. Ha - That is so awesome to hear and know! The Alpha / Beta team has been working hard to offer the masses the very best user experience from the DB / GEM package. There are many more features yet to come so keep your eyes open in my install thread and in the general announcements at the Brultech site. I don't believe it was a very hard sell to make me consider that HiFi Berry Amp - LOL, SOLD! I used Audacity: http://www.audacityteam.org/download/windows/ I never did find an online TTS that didn't sound like a robot and would let you download the file. So I used Audacity with it set to record the speech as it played. Then I just trimmed the file and adjusted it so they all sounded the same. Too funny that was one of the hits I saw before I posted my ask to you on Google! To offer you some insight the over all goal is to create a proof of concept which I hope can be shared to the DB / GEM forum members in hopes of offering proactive / reactive voice announced power level, energy level, temp level, pulse level conditions and values. This is my round about way to get it going before Amazon offers two way feedback from Echo, Dot, Tap. Edited December 16, 2016 by Teken
Thanatar Posted December 16, 2016 Posted December 16, 2016 (edited) Ha - That is so awesome to hear and know! The Alpha / Beta team has been working hard to offer the masses the very best user experience from the DB / GEM package. There are many more features yet to come so keep your eyes open in my install thread and in the general announcements at the Brultech site. I don't believe it was a very hard sell to make me consider that HiFi Berry Amp - LOL, SOLD! Too funny that was one of the hits I saw before I posted my ask to you on Google! To offer you some insight the over all goal is to create a proof of concept which I hope can be shared to the DB / GEM forum members in hopes of offering proactive / reactive voice announced power level, energy level, temp level, pulse level conditions and values. This is my round about way to get it going before Amazon offers two way feedback from Echo, Dot, Tap. That was my goal as I started this project. I wanted a way to know if the garage or shed doors were opened. It has grown into alerting the wife when the washer and dryer are done (downstairs and impossible to hear the buzzer) My doorbell, and many other noises and alerts. I don't know that I would use Alexa at this point with the flexibility my setup offers. Plus you can play audio from an iOS device to it with airplay. Edited December 16, 2016 by Thanatar
Teken Posted December 16, 2016 Posted December 16, 2016 That was my goal as I started this project. I wanted a way to know if the garage or shed doors were opened. It has grown into alerting the wife when the washer and dryer are done (downstairs and impossible to hear the buzzer) My doorbell, and many other noises and alerts. I don't know that I would use Alexa at this point with the flexibility my setup offers. Plus you can play audio from an iOS device to it with airplay. I agree just a cursory review of what you outlined here indeed offers much more than what Alexa can at the moment. Have you been able to link a state variable to a voice command? In the interim I would have to accept a defined value would have to be met and the very same speech phrase would than be triggered.
Thanatar Posted December 16, 2016 Posted December 16, 2016 (edited) I agree just a cursory review of what you outlined here indeed offers much more than what Alexa can at the moment. Have you been able to link a state variable to a voice command? In the interim I would have to accept a defined value would have to be met and the very same speech phrase would than be triggered. Unfortunately, no. Because I do prerecorded audio files, there is no way to have it do a random number. It would have to be already created first. i believe KeviNH has a system that will do a random number. Edited December 16, 2016 by Thanatar
Teken Posted December 16, 2016 Posted December 16, 2016 Unfortunately, no. Because I do prerecorded audio files, there is no way to have it do a random number. It would have to be already created first. i believe KeviNH has a system that will do a random number. I will review what KevinNH has offered up above in a few minutes. But I have to crawl before I can walk then haul aszz using KevinNH's method later. Ha . . .
Recommended Posts