Starting the audio analyzer by default?

VoltVisionFrenchy · Thu Dec 11, 2014 4:58 pm

Hello!
I am doing an exhibit using rPi and QLC+. I have a microphone hooked to up to an rPi USB audio input device and I have it working great controlling the lights with great low latency. My problem is that each day when the entire power gets cycled and the unit boots up, the audio analyzer is not enabled by default. I have to walk out the trail.....log into it with my phone....and manually enable the Audio Analyzer widget. Is there a way to enable this automatically? Thx!

PS-I have an rPi running 3 different exhibits at my show. Everything is working with the typical beta-type glitches. I am using a "Solo Frame" on one exhibit to select one collection at a time for the different seasons, but there is a glitch where some of the RGB matrices misfire the first time and then work the 2nd time (I saw this problem discussed in another thread). Also, the webserver crashes repeatedly so I need to keep rebooting things.

Thanks and talk soon!
-frenchy (Steve French)
www.voltvision.com

VoltVisionFrenchy · Thu Dec 11, 2014 5:02 pm

Things also crash repeatedly when loading a new project...

VoltVisionFrenchy · Thu Dec 11, 2014 9:22 pm

Wow, what a chinese torture puzzle this is! I have to log into the QLC+ webserver to manually enable the Audio Analyzer widget, then when I put my phone away QLC+ crashes. Then I have to reboot and do it all again. It happened to me 5 times in a row. I finally got it to work by enabling the audio widget, then shutting down my WiFi completely before closing my phone. It is still running (I hope).

I am NOT complaining. I know I am using bleeding-edge beta-type stuff here. I am just reporting from the trenches. Speaking of trenches, here is the Raspberry Pi / QLC+ drum controller in action. There is an XLR microphone cable coming out going to a drum and also XLR DMX cable going to the IP65 DMX lights...
http://imgur.com/viiED9l

Thanks and talk soon!
frenchy (Steve French)
www.voltvision.com

VoltVisionFrenchy · Fri Dec 12, 2014 4:12 pm

Also, Here is one of my IP65 DMX lights in action...
http://imgur.com/Igl2nnW
(I will make a coherent writeup sometime soon for the show and how I am using rPi/QLC+.)

respect,
frenchy (Steve French)
www.voltvision.com

david garyga · Fri Dec 12, 2014 4:23 pm

Hi there,

- The crash when exiting web access should be fixed in the next version

- The solo frame glitches too

- As of now, it is not possible to have audiotriggers started by default. This seems like a nice feature to have, I'll try to work on it.

- Nice green snow you have there

Cheers

VoltVisionFrenchy · Fri Dec 12, 2014 8:52 pm

David,
Thanks for your response! Is it somewhat straightforward to build from source so I could test the changes that you mention? I could do this on Sunday or Monday. I have built OLA from source many, many times for the BeagleBone Black, so there is a chance I could build QLC+ from source for the Raspberry Pi. Thanks!
-frenchy (Steve French)
www.voltvision.com

Massimo Callegari · Fri Dec 12, 2014 11:17 pm

Steve, you asked this question before and I already replied.
Building QLC+ on the Raspberry Pi image that I provide is no piece of cake.
Otherwise, if everybody could do it, I wouldn't have asked for a contribution.

Comparing OLA to QLC+ makes no sense at all. OLA doesn't depend on Qt5, doesn't do multimedia and moreover doesn't display any graphics.

The image I tailored is optimized to the bone to avoid wasting any Raspberry Pi resource.
If you are a GIT maniac and can't wait the releases that I provide, then:

- download a Raspbian Wheezy
- follow these instructions: https://github.com/mcallegari/qlcplus/w ... uild-HOWTO
- wait 2 hours and a half
- enjoy QLC+ on a X Server slow like hell

After that, please don't come back here with performances issues.
Good luck.

VoltVisionFrenchy · Sat Dec 13, 2014 4:34 pm

Massimo,
Thanks for your response! Indeed I know nothing about QT, so thanks for clarifying. To me OLA and QLC+ are both lighting frameworks that can be run on embedded linux devices at boot-time. They both have a simple web interface that I dont really need because I tend to run my exhibits and installations headless and leverage hardware inputs and hardware outputs that are enabled by default. I am curious if your comment about "enjoy QLC+ on a X Server slow like hell" means that all of QLC+ will run slow OR only the webserver/visual portion will run slow...

Regardless, I am not trying to annoy you...sorry about that. I will switch over to using Arduino/Teensy solutions ([with my DMX/RDM shield](http://www.voltvision.com/products_page ... no_shield/)) for now and patiently wait for the next QLC+ rPi version to be released. (As more of a hardware guy, I was excited to program shows at a higher level in QLC+ instead of programming at the C level in Arduino.)

Thank you for all of the work that you do! Last night was the first night of our show and the feedback was overwhelmingly positive. At some point I will document the portions that are running with QLC+. As of last night it was 1/2 Arduino/Teensy and 1/2 QLC+ with rPi and one QLC+/laptop.

respect,
frenchy (Steve French)
[www.voltvision.com](http://www.voltvision.com/)

EthanDavis · Thu May 28, 2015 5:53 pm

Greetings!

Just wanted to check in on this function (starting the audio analyzer by default) to see if there were any updates regarding this functionality. We are trying to run a light box with a headless rPi and are trying to figure out how to get the audio analyzer started on bootup without any interaction. One idea was assigning a Input Key and then using something like xdotool to mimic a keypress 45 seconds after booting.

Anyone else had any luck or with initiating the audio analyzer this way or any other ideas on how something like this might be possible?

Thanks in advance for any help.
Ethan

EthanDavis · Fri Jun 12, 2015 4:47 am

UPDATE:

I have been able to mimic the 'enable' command from the virtual console web interface for an audio analyzer on bootup. If anyone is interested in this process I can post the information at a later date.

My issue now is that the 'enable' action (by clicking the link from the virtual console web interface) does not fully enable the audio analyzer if it has not been enabled from a keyboard and mouse hooked directly to the Pi first. The checkbox in the upper left-hand corner of the audio analyzer turns green, but the frequency/volume meters stay at zero. After the 'enable' attempt, any further attempts even by directly accessing the Pi through the keyboard/mouse fail. The checkmark will show it is turning on and off, but the frequency/volume meters stay at zero. If I click the green checkmark on the audio analyzer first with the mouse hooked directly to the Pi (before trying to use the enable/disable from the virtual console web interface), the audio analyzer starts as expected reading the frequency/volumes.

I've tried multiple projects with multiple brand new images of QLC+ Raspberry Pi version 20150412 on multiple microSD cards on multiple Pis.

Can anyone help on this?

The only workaround I can see now is enabling a Input key on the audio analyzer and setting up an Arduino as a V-USB HID to mimic a keypress X seconds after receiving power. Hackey hack hack.

Any help would be greatly appreciated!

Thanks again,
Ethan

plugz · Fri Jun 12, 2015 8:29 am

Hi,

I have been able to mimic the 'enable' command from the virtual console web interface for an audio analyzer on bootup. If anyone is interested in this process I can post the information at a later date.

Maybe the source of the issue lies in there. What process did you put up ?

EthanDavis · Sat Jun 27, 2015 3:38 pm

Hi plugz,

Thank you for responding. I'll post what I'm trying to do, but my first issue (when I load the program from scratch onto the Pi with no modification) is the same issue chka posted in the Raspberry: Sound to Light with Line In? thread viewtopic.php?f=7&t=8458&start=10#p36848

But i have to start the audiotriger on the gui, over the webpanal it will not start

. Massimo had posted a response back in April of

You might have found some issue with the web interface. I'll check it as soon as I can

but there was no more mention.

My hope is that someone else can test the following and let me know their experience:

That someone else can verify they are having issues "Enabling" the audio trigger when ONLY the web console is used (If the audio analyzer is enabled from the gui first then it works with the web console, but if you try to Enable the audio trigger after boot with the web console only -- without touching the gui -- the checkmark shows green but there is no movement in the volume bars. Also, if you try to enable it from the web console first after boot, and go back to the gui, clicking on the checkmark in the gui will only highlight/unhighlight the checkbox, but will not actually engage the audio trigger function [no volume bar movement]).

Any help in this verification would be greatly appreciated!

Thanks in advance for any help anyone can provide!

Ethan

EthanDavis · Sat Jun 27, 2015 4:29 pm

I just tested the qlcplus_raspbian_wheezy_20150627 image with QLC+ version 4.9.0 and I am still experiencing the issue above.

Ethan

mcallegari · Sun Jun 28, 2015 9:01 pm

I finally had the chance to test this.
It's confirmed. If an audio trigger is activated from the web interface only, it doesn't work.

I already have a suspect, but I need to check the code first.

mcallegari · Mon Jun 29, 2015 10:56 am

Just fixed on GIT.
I will provide a network upgrade package soon.