Discussion in 'Pyra News' started by EvilDragon, Mar 11, 2019.
Though I am not a programmer, is there a to-do list somewhere?
Nothing better I've seen than AtC's post just above yours.
I'm trying. Understanding the Kernel for the first time was a much bigger task than I thought it was going to be. I finally got through the 6000 page TRM for the OMAP5432 and it was exciting to see how detailed the TILER information was. How to apply it seemed conceptually easy. Write to memory X, tell the memory manager that memory X is special. Read from Y. How to make that work in Linux, I'm working on.
https://dev.pyra-handheld.com/EvilDragon/pyra-os/issues I thought was a good spot for development issues.
I updated the first post with some more pictures
Ah, well, there's still some low-level stuff to do, for example a proper working heat / limit clock setup and more.
There's also a lot of proper setup needed to be done, which is not really low-level, but need some work as well.
For example, while audio works, the different devices and channels are not named and routed properly.
Right now, the speakers don't play any audio, as they're not properly routed. An earlier test setup from Nikolaus had the speakers running on channel 3 and 4 whereas the headset runs on channel 1 and 2. This could still be the case, I haven't dived in that too much yet.
In the end, the speakers and headphones as well as Bluetooth should appear as different outputs, so you can route any audio stream to any output.
The default configuration for the desktop needs some work, etc.
I'm not that much concerned about any emulators yet - those will appear automatically once we ship it
And - except for the mentioned sound issue as well as the rotation / 3D-conflict, everything is basically ready, just not yet polished.
--- Double Post Merged, Mar 27, 2019, Original Post Date: Mar 27, 2019 ---
I'm not sure how the Pandora did that, whether it was a special touchscreen or driver, but these days, it's not easy to find a resistive touchscreen at all for a reasonable price... so if the hardware can't do it, it won't work :/
--- Double Post Merged, Mar 27, 2019, Original Post Date: Mar 27, 2019 ---
And screen orientation is not a problem in case of 3D, only a transformation of 90 degree before rendering the main content.
Maybe. I am last in line. If any of those do not pass ED's tests at completion...
How do we connect the speakers. We put speakers and sealed them, before we put PCB but I couldn't see them connected anywhere ?
They have spring contacts and automatically connect when the PCB is sitting on them.
So speakers directly connects with PCB but are not connected with each other, that is probably the reason they show up as two different outputs in pulse. I think it will need some software hack for them to work as a single stereo speaker instead of two mono speakers.
I'm sure they're wired in as a single stereo system to the SoC.
https://dev.pyra-handheld.com/EvilDragon/pyra-os/issues/2 please read the last comment by ED.
My own pulseaudio configuration works in terms of stereo outputs of specific devices I have plugged in, so I'm not sure about this channel 1 and channel 2 business. I'm trying to read up on pulseaudio so I have a better understanding of it in general.
Ah, pulseaudio thinks it's dealing with quadrophonic sound! Now I know what he means by channel; the LF, RF, LB and RB signals!
Yes, but that's just a software issue. The things being on different channels I think is just a configuration problem. I don't think it's how it's wired up that's the problem. But I could be wrong.
PulseAudio does not support handling audio channels independently, it expects a typical 2.0/4.0/5.1/7.1 etc channel configuration like in every consumer audio codec. You have to somehow make them appear as individual audio devices to pulse to be able to configure these channel pairs separately.
Don't ask me how this is even supposed to be done, that's one of the reasons why I don't use it. My sound card has 4 individual stereo channels as well and the default pulse configuration is just broken. Just changing the default output channels and mirroring the outputs from the speakers to the headphones in the ALSA configuration beneath pulse should be fairly easy, though - unless the pulse plugin has some usage constrictions I'm not aware of.
Great update photos ED. The prototypes are starting to look a lot more like fully functional units
So on the Pyra, i dont need my tiny Bluetooth Transmitter for Wireless Headphones `??
Sounds great, i know on the Pandora, there was a way to get Bluetooth Headphones to work, but i dint want to bother to get it to work..
If Bluetooth Headphones on Pyra is that easy as connect a Bluetooth Keyboard on Pandora, this wouldbe quite cool..
Please someone correct me if I've got this wrong, but my understanding is that the first production run is to be only 500 units. (600 minus a big margin for rejects, so likely closer to 600 units than to 500.)
For unfortunate souls (such as I) who arrived late to this party, there will be an additional wait of at least 2 months, counting from when the first production run has been shipped, before we can even hope for our units to ship.
Don't forget to add in a month for Easter and a month for the summer holidays!
My reading of that [am I an incurable pessimist?] is that this is because no one is working on it; that everyone with hardware is waiting for... I don't know... newer hardware... someone else (TM) to fix stuff so they can resume testing/developing... I don't know. (Except for hns, or course, who seems to be fighting a constant battle against the entropy of ill-disciplined upstream Linux-kernel developers who seem to be incapable of refraining from making breaking changes, while keeping Letux up-to-date on many more platforms than just Pyra.)
If this logjam is not just a figment of my imagination, I'm not sure how it can be cleared. Perhaps the key log is EvilDragon sending out the newer prototypes; perhaps it is EvilDragon giving the green light for mass production (and hoping for the best even though testing of the existing prototype boards does not really feel very complete). I don't know.
Nonetheless, I'm still optimistic that I might have a Pyra for Christmas.
You're correct. The chip supports all different kinds of configuration, and it's all setup in the driver of the audio chip.
The wiring is correct, but you need to tell the software how you wired stuff up there, basically
Yes, the issue is not related to Pulseaudio, ALSA has the same.
In the world of pro audio, there is no such thing as 'stereo'. Each two-wire output (signal & ground) or mono three-wire (balanced XLR) is a channel. The concept of left/right is speaker positioning and can often be from the perspective of the stage - so, stage-left and stage-right. Tying two or more channels together or creating a 'gang' of channels can result in stereo or surround effect sound.
There is no such thing as a single stereo channel. Stereo being two discrete audio channels of course. You -can- have one jack with multiple connections sharing a common ground/common connection, but what consumer audio labels as left or right is simply an audio channel that has been labeled.
A proper mixer allows a person to take any inbound channel (input - physical or application sourced) and proportionally mix it with other channels to any output destination. Seeing discrete channels is a good thing - they can be ganged into stereo - or flipped left/right - or have 'left' go to one output and 'right' go to another, blended into a common output, etc.
Hope that helps - the above argument/discussions around channels Vs 'stereo' seemed to need a bit of clarification around what a channel is and how it works from a mixer perspective. There is nothing wrong with having the SoC output what the OS sees as 'channels'.
Oh I knew it wasn't a single stereo channel, I was just trying to explain that it was all one system, not completely separate.
Separate names with a comma.