Has anyone checked out the WiPhone. It’s basically a esp32 sip phone. One of the first daughter boards they have made/are making is a LoRa board. Any idea how possible it would be to use the meshtastic protocol and communicate on a meshtastic mesh with the wiphone?
Back in the 90s, I worked on an audio telephony product and had an in house variant of the h.723.1 codec at just under 5kbit / sec. It worked by decomposing speech into phonemes with additional data for the pitch and tone of the audio.
I’m sure lighter and better quality codecs exist today where you’re not limited to a Pentium 66MHz processor for compression.
Excluding overhead, that may fit into the “Medium range” setting in the case of PTT audio.
Oh wow interesting. I hadn’t even considered audio/voice over LoRa (I think that is what you are talking about). I was just thinking of the current implementation of text messages over LoRa.
Nothing much ever came from the wiphone for meshtastic. I believe there’s on developer that has done a little work in their personal GitHub but that’s about it.
There is some support for audio over LoRa using the sx1280 on 2.4 ghz but that’s still experimental. You can see more here: Audio Module Configuration | Meshtastic
So the way I see it working as a user-facing feature given the technical limitations is that you can record audio “notes” which you then send to each other. You’ve no doubt seen the messaging apps that have this feature already - signal does, I think Whatsapp at least as well.
It would be necessary to have a download progress indicator for each note, however, given the use cases for meshtastic as a long distance low power comms thing.
V1 in meshtastic is more like a walkie talkie with a ptt button, once that works ok then some mesh features will be looked at, need to do a ton of tinkering.
The bandwidth required for the communication of two points in the network structure is too much. At the midpoint it becomes necessary to have two lora modules for data transfer.
Range increases at 2.4kbs. With 2.4kbs speed, all waypoints need to be reserved for a call. If there are 20 midpoints between point 1 and point 2, these 20 points will be reserved for a single call and will ignore any other communications.
LoRa modules in this project are not capable of responding to multiple calls simultaneously on multiple frequencies. This situation can only be resolved by adding more lora modules to the network. Frequency shift technique in GSM structure can be used. In GSM antenna placement technique, each base station uses a different frequency from the nearby base stations. Many antennas are used to increase speed without frequency overlap.
The frequency of each nearby module is determined by location, signal strength, and signal coverage tests. It is necessary to think very well for communication protocols, collision avoidance structures.
As a different solution, SX1302 modules can be used in the project. But SX1302 modules are more expensive. It has multi-channel listening feature.
When i got my really early Wiphone Device the firmware source was available. I approached them to integrate meshtastic instead of the simple Lora based SMS service. They told me they were going for a dual licensing approach that would be incompatible with our GPL license, so i wrote the bindings instead. I haven’t messed with it for a long time, BUT IIRC the screen and the firmware/lora worked. I lost interest before including the keyboard and other peripherals. That’s why its not included in the proper meshtastic firmware but must be included from my personal github. Feel free to pick up on that work
Also it would need a T9 implementation to support it for a standalone device. T9 uses disctionaries and we’re really tight on flash space already.
Yeah there is no t9 in the wiphone defaut firmware, just multi tap, and that is good enough. I just got mine, but they forgot to send me the source code which is annoying. I also dont have enough knowledge and experience to integrate an app for meshtastic, but id really love to.
