Bat detector

Hi Grizzly,

That is simply too far fetched for this processor. We have experimented with a "crude" approach to recognizing species by looking at the peak frequency, rythm etc but this did not give a reliable output. We could - but I wonder how usefull that would be - create a database with information on the frequencies used by different species but what use would that have during either a live walk around or during a recording ?

cheers
Cor
 
A species suggestion is something that is quite tricky and complex. Most of the devices that can do that fail quite a number of times. Software that has reasonably accurate results is quite complex and uses a lot of computing power so this is not something easy to do. Things do get better when you have only a few species to determine. Difference between a common and Nathusi Pipistrelle are reasonably obvious most of the time. But Kuhls look a lot like Nathusii.
Well anyway, if it was easy it was already in there....
I am sure it is a wish for a long time from Cor and many others.

Maybe a suggestion for autorecording would be an option that calculates the sun set sunrise time for the next recordings depending on the found GPS coordinates/times. I am recording the year round and do handle the detector quite often so it is not a big deal to change the deepsleep times for autorecording but it would be nice if I do not have to check the settings every time.
 
Hi Edwin,

Your wish is ... possible. I allready have a sunrise/set code onboard but it needs some testing.

cheers
Cor
 
Hi,

Most of the features that have been requested are added to the upcoming update. I hope to share this within 1-2 weeks, will announce in this forum.

cheers
Cor
 
The new update will
  • allow adding a prefix to each recording to recognize individual TB-machines
  • allow using the last stored GPS location in recordings without having the GPS active (saves batteries in deepsleep)
  • allow the user to directly go to the SETTINGS menu by longpressing the L-Encoder
  • allow direct replay of the last recording by longpressing the R-Encoder
  • allow switching to autoRecord by longpressing the L-Pushbutton
And the new update has a setup that allows the user to create and use 10 locations with stored location-data that can be used during recordings. The data is stored as a textfile on the SDcard and can be edited manually also. For those that do not have a GPS module installed this allows you to setup locations also manually. For each location DaylightSavingTime can be specified (none, Europe or America) and the time-offset from UTC is also stored.
In conjunction with the new location system we also calculate sunset/sunrise for any location you are using. These can be used automatically in deepsleep allowing a teensybat to be set to monitor a location based on local sunset/sunrise.

Final tests are currently running.

cheers
Cor
 
Hi Cor,

Thanks for the new software version. I just updated my Teensy Bat.
I noticed that the files published on github with version 1.7 do not contain source codes.
The archive names are a bit misleading, but they do not contain source code.
Is there a chance to publish the codes for version 1.7?

Krzysztof
 
Hi,

I am currently holding the source code until its clear that the code is without issues. And if you are interested in the code, maybe you can share what you would like to change in the code. The code of v1.6 has all the "crucial" algorithms in place and they have not been changed.

Those archive names you mention are autogenerated by Github, I cannot avoid that. They do NOT contain any sourcecode.

cheers
Cor
 
Hi,

I tested software versions v1.7 on my TB. And a few problems I found.:
#1 - Battery status in the settings menu sometimes shows strange values:
20240209T164012.jpg
20240209T164144.jpg
20240209T164448.jpg

#2 v1.7 I found strange option in the Left Encoder menu "Freq.AUTO"
20240210T092759.jpg

#3 - My TB sometimes hangs when I try to enter the settings menu. Then I have to turn it off and on again. I don't know if it's a software problem or if I have some hardware problem with my TB.

Krzysztof
 
Hi Krzysztof,

Thanks for that report. On the battery, the % is read directly by the hardware and based on the changes (drop in voltage) TB estimates the time "left" in the battery. But your screen indeed shows that the hardware reports the % rather oddly. I havent seen that myself thusfar. Is the battery a "normal" Li-Ion ?

On the freeze, we have seen that during testing but thusfar no clear "pattern" has been reported that triggers this behaviour. If you get a clear idea on when this happens (sequence of actions that always creates a freeze) that will be helpfull.

EDIT: forgot to comment on the Freq.AUTO ... can you tell me how you got this ? Start the TB and then ...

Thanks for reporting !
Cor
 
Hi Krzysztof,

Thanks for that report. On the battery, the % is read directly by the hardware and based on the changes (drop in voltage) TB estimates the time "left" in the battery. But your screen indeed shows that the hardware reports the % rather oddly. I havent seen that myself thusfar. Is the battery a "normal" Li-Ion ?
My battery is (LITER 606090). I purchased the kit with mainboard v0.9 and battery at https://www.teensybat.com/parts
On the freeze, we have seen that during testing but thusfar no clear "pattern" has been reported that triggers this behaviour. If you get a clear idea on when this happens (sequence of actions that always creates a freeze) that will be helpfull.
This happened several times when I tried to enter the settings. By long or short press of the left encoder. This is what the screen looked like:
IMG_20240211_094033441.jpg

EDIT: forgot to comment on the Freq.AUTO ... can you tell me how you got this ? Start the TB and then ...
Now it's hard to say how it happened.
Thanks for reporting !
Cor
For now, I have changed the SD card to another one. I know it should be Class 10.
So far I have been using the Kingston microSDHC 32GB C4 (SDC4/32GB) I had at hand.
But I changed to SanDisk 16GB microSDHC Ultra A1 C10 and I'll see how it goes.

Thanks,
Krzysztof
 
Hi,

#1 - Thanks to the Class 10 SD card, my Teensy Bat works OK. So far, freezing when entering the status menu has never happened.
#2 - But on software version 1.7, the battery status menu still displays strange battery life values. Previously, in software version 1.6, I did not notice this behavior.
#3 - And I found another repeatable method for hanging the TB (for software versions 1.6 and 1.7). It always happens in the GPS status menu. Follow these steps: 1) turn on TB (GPS off), 2) go to settings, 3) go to the menu with GPS status, wait 1 s, 4) quickly switch the menu using the right menu encoder once to the left and once to the right, 5) go to point 3).

Krzysztof
 
Hi Krzysztof,

Thanks for those follow-ups. I just now see them because a new entry in this thread was not automatically send to my mail-adress.
On #2, the odd thing is that we have not changed that section during the update. The battery info on life also should be seen as "experimental" as it calculates life based on changes in the battery-voltage. It estimates how fast/slow the battery depletes during current usage. As an example: If you for instance switch GPS on/off this will have an effect. The gps "pulls" the battery-voltage down probably, when switching it of the voltage might increase.
#3 You mention "quickly" ... so this only happens if you rotate the encoder fast ?

Cor
 
Hello, I am trying to change the sample rate on the teensy 4.1 so I can fully take advantage of the PDM Microphone I am using (https://www.elecrow.com/digital-mems-microphone.html). The microphone can operate in ultrasonic mode, so it is capable of "hearing" between 80 and 80 kHz, but to operate in this mode he needs the frequency clock to be between 3.072 and 4.8 Mhz. I tried to use setI2S function (https://github.com/alfa95139/pt8211-with-Teensy-4.1/blob/main/setI2SFreq.cpp) to change the sample rate, but when change it to 192000 kHz frequency generated by BCLK with "setI2SFreq(192000)" is around 12.30 Mhz, and nothing is detected by the teensy board. I have created my own thread about this https://forum.pjrc.com/index.php?th...-mems-pdm-mic-to-apply-pdm-in-bearings.74629/ , but this thread seemed to be on topic. Is there a simple way to change the sample rate and adjust the clock frequency without the shield? Or without any other tool?
 
Hi,

I have no experience with PDM microphones and their usage. I see you use an example library from frankB but I have no idea how he used that in a real application with ultrasonic sounds. The microphones we use in our batdetector also use knowles MEMS (SPU0410LR5H) microphones but they do not use PDM.

cheers
Cor
 
Hi,

I have no experience with PDM microphones and their usage. I see you use an example library from frankB but I have no idea how he used that in a real application with ultrasonic sounds. The microphones we use in our batdetector also use knowles MEMS (SPU0410LR5H) microphones but they do not use PDM.

cheers
Humm, MEMS (SPU0410LR5H) microphone is an analog microphone, right? I am afraid to use an analog due to the higher noise associated.
 
We have been using this microphone for several years to listing to bats and they work fine. We have clean recordings using these.
 
We have been using this microphone for several years to listing to bats and they work fine. We have clean recordings using these.
Ok, maybe it was a wrong assumption. If I buy one MEMS (SPU0410LR5H), can I use it with Teensy 4.1 with a sample rate of 192000 kHz without using any shield or extra board?
 
We have been using this microphone for several years to listing to bats and they work fine. We have clean recordings using these.

Ok, maybe it was a wrong assumption. If I buy one MEMS (SPU0410LR5H), can I use it with Teensy 4.1 with a sample rate of 192000 kHz without using any shield or extra board?
I have no idea how that would work, the audio-shield is needed (as far as I know) to convert incoming audio signals into the digital domain for processing. But you can search the forums to see if anyone has allready done that. The audio-shield works fine and isnt expensive.
 
I have no idea how that would work, the audio-shield is needed (as far as I know) to convert incoming audio signals into the digital domain for processing. But you can search the forums to see if anyone has allready done that. The audio-shield works fine and isnt expensive.
Ok, I will definitely buy the shield, I already read a lot of foruns here and almost all of them use it. With the shield, is it possible to do it easily?
 
There are lots of examples around on the forum. Just study those. You can study the code of our bat-detector (V1.6 code is available) but thats tailored to a very specific design . Look at the beginning of this thread to see a more "down to earth" code.
 
There are lots of examples around on the forum. Just study those. You can study the code of our bat-detector (V1.6 code is available) but thats tailored to a very specific design . Look at the beginning of this thread to see a more "down to earth" code.
Ok, thank you
 
There are lots of examples around on the forum. Just study those. You can study the code of our bat-detector (V1.6 code is available) but thats tailored to a very specific design . Look at the beginning of this thread to see a more "down to earth" code.
Just one more question, sorry, you use a MEMS (SPU0410LR5H) with I2S, right?
 
Hi,

A few days ago we got a report from a user that had issues with version 1.7. The issues were related to the resulting wav-file behaving different in Batexplorer. It turns out that this was due to the newly added option to use a serialnumber in the filename. Edwin and I have therefore created/test v1.71 which is since today available on github.

cheers
Cor
 
Back
Top