Two Questions

[Lttlwing16]

Hey @Tim Curtis @TheOldPresbyope

I know you guys are busy working on the 8.x release bugs, but just wanted to circle back and let you know I changed out my Allo Digione for a Audiophonics DIGIPI+I2S HAT this weekend and I am still getting I2S sync errors in the /var/log/kern.log albeit less frequently and with the Audiophonics HAT, I'm having no stuttering during 24/96 playback. Not sure if the I2S sync error is a global kernel error, or is specific to my setup, or something that needs to be looked at.

As an aside -- Allo support did confirm my Digione board "needs repair or modification" .. just after 2 years and wanted me to pay round trip shipping from the US to India which would've cost almost as much as a new Digione. I kindly let them know I'd be taking my business elsewhere if this is the longevity of their products.

The Audiphonics sounds great with the MoOde!

Cheers,

David

[Tim Curtis]

Very odd to have I2S sync errors. Maybe try contacting Audiophonics to see if they have tested the device with moOde 8 releases. It could be some sort of hardware or compatibility issue?

[Lttlwing16]

Very odd to have I2S sync errors. Maybe try contacting Audiophonics to see if they have tested the device with moOde 8 releases. It could be some sort of hardware or compatibility issue?

Will do. I actually already have a service ticket open with them about the PSU jumper pins on the DIGIPI+I2S not behaving as configured on their website. 

That said, since this is occurring across two different HATs and at least 7.x and 8.x versions of MoOde what is the likelihood the 3B+ is the culprit? I believe I have a spare around the house, so I can it out test later. 

Additionally, there is at least some evidence the error is harmless and occurs when the bcm2835-i2s is set as clock slave. I believe you were even involved in the conversation at github.

[Tim Curtis]

I suppose if there are no issues with Audio then the errors might be harmless and can be ignored :-)

[Lttlwing16]

I suppose if there are no issues with Audio then the errors might be harmless and can be ignored :-)

Currently there are not.. 

Did a little more digging, and came across below from this page 3rd post down. 

The I2S sync error messages appear to be a normal result of starting or restarting the ALSA read interface. Hence I get a sync error in dmesg every time the application gets a read error and calls snd_pcm_prepare().

The read errors turned out to be an overrun. The application is running real time priority but every so often writes to a log file on the SD card. This can sometimes be a little slow and blocks long enough for a read overrun. Moved the log file to RAM disk and it has run for over 12 hours now with no errors.

How would I go about moving the log file to the RAM disk? Or do you the above would be unrelated?

[Tim Curtis]

It's a very old post and sounds like the guy's application was poorly written.

I2S sync error is very odd. I don't see them with any of my audio devices.

[Lttlwing16]

It's a very old post and sounds like the guy's application was poorly written.

I2S sync error is very odd. I don't see them with any of my audio devices.

Is the Sync error between the Rpi>HAT or HAT>DAC? or both?

[Tim Curtis]

I2S is a communications interface on the Raspberry Pi. The error is apparently occurring when the OS is trying to communicate with the HAT board over the I2S interface.

[Lttlwing16]

I2S is a communications interface on the Raspberry Pi. The error is apparently occurring when the OS is trying to communicate with the HAT board over the I2S interface.

Thanks Tim, I'll try to use a different Pi board tomorrow, but reading around on the internet, this seems to be a very common kernel error with the bcm2835 board. And as we've discussed, may be harmless if there isn't also audio issues or DMA error as well.

[Lttlwing16]

Just to follow up am still getting the same I2S sync error with the Allo Digione with a different Rpi 3B+. It does lineup now to whenever music stops for a while and then I begin playing something.. 24/96 playback however is no longer stuttering, SO FAR... will update as I move along..