Intermittent loud ticks and no-audio using Spotify

[JasonLG1979]

the CPU cycles are offloaded...are they not ?

Well yes if the DAC is doing the resampling the Pi wouldn't even have knowledge let alone cost any CPU cycles.

downstream preference is subjective

Exactly. Why would you want to color the sound before it even gets to your DAC?

[JasonLG1979]

If sound preferences are subjective the only proper thing to do is output the most objectively transparent signal and let users color the sound if and how they choose with their choice of downstream components.

[DRONE7]

Then giving users the option of a re-sampler or not meets your criteria ?

The upcoming resampling is opt-in so no hit if that is not your preference.....

[CallMeMike]

(fast forward to year 2120) Your average Joe/Joanne gets up in the morning... yawns and reaches for the morning vaccine kit... then takes the all-day nutrition tablet with a little synthetic water and cogitates a bit about participating to the daily "Let's pull down that statue" march... the decision is a swift 'NO'... yawns again and connects to the GWW (Galaxy Wide Web) for the daily "How to eliminate intermittent loud (sound) ticks".

[basmeyer]

PS - I'm curious to know what setting got @basmeyer's

Code:

rate: 44100 (352800/8)

Yeah, me too! I was looking at and fiddling with some settings in Audio of Moode but it did not made it change. SoX resampling is disabled, it that would matter. The formula is correct though.

I'm not too keen on doing a work around because eventually it gets fixed upstream. I should prolly post an issue on shairport-sync repo.

We can certainly try Git master. I'll post back with results but prolly not till next week.

I was already wondering if the whole buffer thing would become proper at some point. If I use HiFiBerryOS on this card I do not experience spike/tick related problems, but I rather stick to Moode for other reasons (that I do not mention in detail but HiFiBerryOS is far from ideal for me and looks and feels like it needs a looooooot of work to get this "v1.0-feel" away).

If not I can always think of a new DAC card or whatever (USB..) DAC and flock this one, including the customized casing btw but hey - I should have done some testing before I did all that. Learn to let go is the song.

If you do need some testing because of the need of Spotify Premium I love to hear if I can help.

[Tim Curtis]

I have no idea what you mean by "the whole buffer thing". MPD, Airplay, Squeezelite, Bluetooth all run well on moOde without any special buffering thing.

Maybe try compiling the Git master version of librespot and see if it works better.

[JasonLG1979]

Why wouldn't you just write a ALSA plugin that uses sox for resampling that you could specify a minimum sampling rate and bit depth, a target sampling rate and bit depth and buffer parmas so that all apps could benefit instead of just MDP?

[the_bertrum]

"just" the four most dangerous letters in programming.

[JasonLG1979]

I have no idea what you mean by "the whole buffer thing". MPD, Airplay, Squeezelite, Bluetooth all run well on moOde without any special buffering thing.

Maybe try compiling the Git master version of librespot and see if it works better.

That really depends on each individual alsa backend. Last time I tried airplay it didn't sound very good at all until I fix the buffer parmas. Similar issues to librespot. What you think runs fine could be the result of not knowing/hearing any better. The only way to find out for yourself is to check to see what each backend is doing and then do an abx comparison to see if changing buffer parmas actually makes an audible and/or performance difference. You would need to do these tests on your lowest spec target device.

[JasonLG1979]

"just" the four most dangerous letters in programming.

It can't be any worse then several different apps with varying quality ALSA backends.