Is it just me, or do you end up installing raspbian over and over again on a complex project?

It sounds very frustrating to say the least.

When I used Google Assistant with a Pi3B, I used a Google AIY Voice Hat to generate the sound, and there were lots of detailed installation instructions available. (but I still had some trouble getting it all working - fortunately I didn't end up reinstalling the operating system)

I haven't tried your specific hardware, so I don't know what issues will occur with that setup, but your project sounds interesting. I hope you get it working.

There are a few of us Element14 minions that have struggled with Pi audio.

The following post was made to resolve a bluetooth audio issue. It contains some links to reference material to better understand Pi's audio setup.

https://www.element14.com/community/message/280804/l/route-all-pi-audio-to-bluetooth-speaker#280804

I found the .asoundrc file key to getting an keeping audio working. As I made changes I kept saving the file contents. The minutes things went for a shite, I restore the file to the last know good configuration.

There are other posts for audio issues. I seem to recall someone made a post on the process they developed to establish audio across a number of applications. Unfortunately I am unable to find it in the time limit my wife gave to play on the computer before guest arrived.

On a complex RPi project I usually start with 2-8G and 2-16G SD cards.

Load all 4 with the latest raspbian image.

load all "extra software and drivers" on the first 8G card and make sure to write down each step you used during the installs.
Try and duplicate the system on the second 8G card, if it works OK, then save that image on a HD. This is your backup BASE image.

Try and add some code, and every time you get a major feature working duplicate it on the second 8G.

At the end of the day, make a copy of the working image.

When you get to a point where you need more storage, switch to the 16G cards.

Scott

The original AIY voice worked fine with the Google card, but I didn't really like that hardware setup and the googl assistant library as used with that was deprecated for the sdk earlier this year

I definitely need to spend more effort on installing a piece, testing, backing up working configuration and then installing the next piece

In the past two months, I've found that GitHub is my biggest buddy to fight off development woes.  For those who haven't embarked on GitHub, yet - the power of it comes with its use at the very start of a project.

In the past, I'd just throw my finished project in there which could benefit others - but, I didn't realize how I was missing the boat for it helping me out in the development phase.  It provides easy backup and retrieval to readily recover your project to a known good state.  In addition, it has you log a quick summary of any changes between syncs between your local project and the cloud.  I've found these features invaluable as a Maker.

Here's my M.O.:

1. At the start of the project, create a GitHub Repository.  Edit the README.md file to lay out the project summary.
2. On your Pi, do a 'git clone repository_address'.  You will now be in sync between a Pi directory and your cloud repository.
3. As you make 3D Printable files or collect sounds and images, use the web interface to organize them in your repository.  Then do a 'git pull' on the Pi to sync it.
4. As you discover required packages and learn tricks for setting up the Pi, update that README.md file accordingly.
5. As you update code on your Pi, do a "git add ., git comit, git push" to sync back to the cloud repository.
6. You can make a bash script that will allow for "install" of your project on a clean pi as well and house it in the repository.  Or simply update the README.md with instructions.

Now, within seconds, you can deploy your project to an infinite number of PIs.  You never lose track of where you left off or the magic .asoundrc file that was working for you.

Here's an example project I have brewing:  Ghost-Catcher-Cam.  The README.md at this state isn't written for others - it's to keep me straight on all the complexities of the build from audio, to display, to OpenCV, mouse pointer hiding, taskbar hiding, etc.

In the end, you'll have a perfect repository allowing you to share or remake your project from scratch without having to remember a thing.  You'll also be able to quickly leverage an old project into a new.

Happy New Year!

Sean

I got some time to find the link another member posted on Pi audio tribulations.

Alsa no audio (glitch) issue Pi 4 fix

I suggest you review the post and examine the links. I'm confident it will increase you knowledge level on the subject. Hopefully it provides you the insight to resolve your issue.

When troubleshooting audio issues I start with a clean install for the Pi and backup the .asoundrc file after confirming audio is working. As I make software changes I keep track of what is happening in .asoundrc. You can see the changes and can compare that to what was working.

I confess consistent audio setup has never been easy on the Pi. Just when you think you got it nailed down it evolves and you are left picking up the gauntlet of audio not working once again.

I very much doubt that it is just you.

I have the latest version of Raspbian installed on my pi 4. Have hardly had any time to use it for a long time now, but try at least to check for updates and install them every week. Yesterday i found time to do some Python programming on it and noticed that it is really screwed up! No one else has access to it and it is shut down when i am not using it, but now the settings are waco - probably due to changes made by an update. I have been using Raspberry Pi's for quite a few years now and have at least 7 of them of various models and have never had problems configuring them properly. The first thing i do when i start a session with any of them is to check for updates and install them. Nonetheless, i have experienced that they occasionally experience so many problems that i believe it is best to just reinstall and reconfigure them.

But i have never before that my configuration settings have been so drastically changed as on my pi 4. Although i live in Norway, i do not believe in setting up anything i have (pc's or pi's) that are not designed in Norway for the Norwegian version. I learned a long time ago that it is wise to always use the native version since that is probably tested most rigorously, and only change the things that are necessary (like for example keyboard and wifi configuration). But suddenly lots of my settings are corrupt and the graphical interface is in a mixture of Norwegian and English! When using the Geany editor for Python programming the "Save" menu is in Norwegian ("Lagre"), while the "Save All" menu is in English!

Very frustrating for me every time things like this happen - and it happens far too often on all systems! Even a huge developer like Microsoft releases updates and products without (in my opinion) adequately testing first.

S**t happens - and keeps happening, unfortunately.

I agree with Sean about Github for version control, I also use GitLab when there are multiple people working on a software project.

I Had kind of a similar nightmare with that during the Pi Chef Challenge. I was integrating the 4D Systems touchscreen and kept bricking the Pi and I had to reinstall. I lost a week or two on it. I finally set it down and came back to it later. Eventually I realized that because the screen had a custom kernel, that every time I tried to run updates it would plow over the custom stuff and basically brick it. So I ended up with an image which is locked in time and can’t be updated, but everything worked once I figured that out. I found later that it is mentioned kinda sorta in passing to not run updates, but I read right over it. Wish it was stated clearer.... So double check all the documentation.

https://www.element14.com/community/community/design-challenges/pichef/blog/2018/02/13/pi-chef-challenge-smart-range-hood-initial-pi-power-up-config