That is awesome! Any chance you could post some showcased css to let us know how you did it. Best example I’ve seen yet!

I would be happy to do so, but I have been struggling to find (good & cheap) port expanders that allow multiple pHATs - ideally 3. If you could recommend some that would be greatly appreciated.

“This was using the already existing repo Mirror Mirror On The Wall, with the associated skill, which was why I hadn’t posted the skill itself. This had very kindly been developed by joanaz in the past, but the instructions have not been updated in a while so it’s somewhat more challenging than I had anticipated to get through.”

Sorry, wrote this without seeing who it was that posted. Basically, I don’t think this requires new code per se, but would just need new instructions (ReadMe), and for me at least, a lot of css-ing.

Hi everyone, I am just wondering whether anyone can give me some advice please on having multiple pi systems altering the same mirro. I don’t really know how to go about setting this up, but as is a consequence of having run out of GPIO space really. Therefore, I’d like to have two rip inputs to the same mirror space. Anyone got any suggestions? Am I being daft, can I in fact just do this like the remote, through localhost:8080?

I wrote this in a different thread not too long ago actually, but I noticed a vast increase in detection/response rate when moving from a cheap USB mic to a decent webcam microphone. I would say the response now is on the order of 1-2 secs, with effectively no errors in voice interpretation. I’m sure this is helped immensely because the clarity of the audio into the rpi is much clearer - the cheap USB mic had terribly noisy audio in.

I didn’t know where to post this, but I figured I should put this here.
As of 5 minutes ago, I now have a working version of Alexa on the mirror, with Google and Youtube API’s working, with full module control.
This is using the new Alexa skill set up, which does not match the instructions. It has only taken 2 weeks of puzzlement.
If there is enough interest I will write out instructions, but it’ll take a bit, so please voice (pun intended) interest if you wish this working. I should add that the one thing to note is that the css is terrible for this, and there are definitely some wrinkles to iron out, but it works!!!
I should stress this in advance, if you want module control only this is NOT as good (voice recognition not as good) as MMM-Voice, nor as convenient (takes much longer to effect modules with default setup).

Hi Mykle, it’s not a module but simply the vocal recognition.
I was saying ‘hide modules’ which was being interpreted as ‘flights modules’ over and over again. I turned on the debug options on MMM-voice to try and work out why it was working so poorly, and what I discovered was simply that it couldn’t interpret what i was saying! The words displayed on the mirror were nothing like the words I was aiming for. Having altered the dictionary I’m very happy thanks! But yes, I am using your great module.

Just a quick note here, I tried a Logitech 9000 series webcam today and it is superb with MMM-voice. Not a cheap solution, but really worthwhile. The audio recognition is excellent, and actually the response from MMM-voice is much, much faster. I think the clarity of the audio going in greatly reduces the processing time down to near-instantaneous. I don’t know if it helps that the webcam is described as being a ‘fast’ webcam. Really good!

Thanks for your advice! I’ve actually just been trawling through some software installs for automated Sphinx dictionary generation, all of which I have decided are needlessly complicated for MMM purposes. Instead I have now modified the dictionary within the MMM-voice folder, and added around 5 more verbal definitions for all of the key words I want, based on my accent. I would say now the recognition for key words like ‘hide’ is not far shy of 100%. This dictionary based approach, adding word where required, seems incredibly useful, and for me greatly increases the usability of the module. I had been so frustrated with the response rate from the mirror, but now am actually incredibly impressed.

This is slightly tongue-in-cheek, but I have recently discovered that the MMM-voice module does not deal well with my accent. At all. No matter how much I want to be able to say ‘Hide’ all that is ever displayed is ‘Flights’. Therefore I would like to increase the dictionary to account for my silly accent. Does anyone have any experience with making this work? Thanks in advance!

Thanks for that, that might also be an option. I thought a webcam might be the answer and I’ve tried out a few of my logitech versions, but there’s something about running it through ALSA that makes the sound really crummy. This isn’t a multiple pickup type of affair though, so maybe this is the answer. Regardless, many thanks for your advice.

Hi Mykle, Thanks for the info. I have been pretty extensively reading your posts, and it was some of the posts from you and cowboysdude that persuaded me to try out the MMM-voice. I’ve spent so long trying to get the Alexa integration working with the MirrorMirror skills and just can’t get it running. Hello Lucy installed really nicely so thanks ever so much for your work on that. The latency isn’t really an issue, but some combination of my accent, the mic and the code mean that it takes me several attempts to get a voice response. I had read some info about being able to use a software boost through ALSA for a USB mic, but I’ve been struggling to find the way of really doing it. I would agree I’d like the mirror to work well so I’d be happy to upgrade the mic if I could be sure of a result. The mirror will ultimately also be ‘frameless’ and so I will mount a mic somehow on the glass edge. Again, my thanks for your advice. Perhaps I’ll contemplate a laptop board.

Hi folks, just writing to ask if you have any ideas about boosting microphone input volumes, or whether there is a good mic choice for this sort of project. I have a USB mic installed at the moment, and it works fine. I have MMM-Voice up working and so on, but I’m a bit bothered by the range of the mic. I have the Pi mounted behind the screen, and it cannot pick anything up from in front. I have to be more or less chewing the microphone for it to recognize what I’m saying and it’s making the voice modules struggle profusely. Any ideas/suggestions would be greatly appreciated.

I am currently using this module, and so I can confirm it works. Things I have altered compared to your config.js (in the UK bus sections) are as follows: I have deleted the two ‘show…’ terms. They default to your responses and so aren’t needed, and the limit I think should be in quotation marks or whatever you currently use in your config file to define values. Then I think this should work. I tested your atcocode on the website and it returns sensible information, so I think if you use those tweaks all should go well.
Hope that helps!

I’ve recently been battling with this module, but this bit should be OK. Just make sure you’ve cloned the module within the ‘modules’ folder and then inside the MirrorMirrorOnTheWallSkill folder you should find the additional folders you need. I don’t know if this helps, but these folders definitely existed when I tried this a few days ago. I should add, I don’t think it ever actually tells you to clone the module, but it’s taken as obvious I think that this should be done. Hope this helps.

NB Sorry, I should add that this procedure is very complicated these days, because the instructions no longer correspond to the way the Alexa skill system works online. I would think hard before continuing on this path, and contemplate switching over to MMM-Voice. Trust me, this has been my bane for the last week.

I’m sorry to be a nuisance, but I was looking for a small amount of help with this module, simply because I feel I’ve got so far I’d rather not give up. I have been through both of the Github pages and am now at the stage where Alexa gives me a response when I ‘start’ the MM module. However, when I boot the mirror, I get an error: ‘Could not find main module js file for MirrorMirrorOnTheWallSkill’. In addition, and I assuming linked to this, is the fact that the skills associated with the module don’t work, I simply get the standard ‘there is a problem with the requested skill response’. I tried pretty hard to battle through the new Alexa skill building as best I could to get this going, and am hoping a small tweak that you may suggest will finish this off for me. I’ll then see if I can write an updated procedure for creating the skill - if anything I feel it’s a bit more of a pain than the instructions suggest! Many thanks in advance for any advice.