Tidal stutter using LMS

[mherger]

> Isn’t that rather the issue with high compression rate of flac that does
> not work well with SB3, diet and Boom? We added a “forced transcode�
> option that plugins could set, but AFAIR, we activated it in Deezer
> only. I might be wrong, but isn’t tidal doing flac as well now ?

Oh, that could be true, too... See
https://github.com/Logitech/slimserver/pull/466/files - but that's for
TIDAL, not WiMP.

TIDAL actually was the first of the "official" services to support FLAC.

[Redrum]

Isn’t that rather the issue with high compression rate of flac that does not work well with SB3, diet and Boom? We added a “forced transcode” option that plugins could set, but AFAIR, we activated it in Deezer only. I might be wrong, but isn’t tidal doing flac as well now ?

Yes they are (Tidal -> FLAC).

I was surprised when "persistent" was suggested (thank you again btw), because the previous posts in this thread inferred the issue was the inability of the receiver, et all, to handle the transcoding, so having LMS do it was the answer.

Having said that, I have streamed tidal for a few hours in persistent and the receiver has not stuttered. But, it didn't happen all the time anyhow...it seemed it could be media dependent.

Jim

[philippe_44]

Yes they are (Tidal -> FLAC).

I was surprised when "persistent" was suggested (thank you again btw), because the previous posts in this thread inferred the issue was the inability of the receiver, et all, to handle the transcoding, so having LMS do it was the answer.

Having said that, I have streamed tidal for a few hours in persistent and the receiver has not stuttered. But, it didn't happen all the time anyhow...it seemed it could be media dependent.

Jim

If this is that compression problem, it is totally media dependent and some have been, for no reason, compressed using high ratio which uses blocking methods that are probably not implemented on older players.

[bpa]

I've just been testing Tidal Flac with new Muse speaker on an CORE32 8.2.1 - 1632112486 system (default File Types settings) and I got jumps and other issues with a track (Rhapsody in Blue 17min). Jumps and funny silent loop 45 sec before end. Problems were reproducible (i.e. no clean playback of track) and so I tested with Receiver and Squeezeamp - same sort of problems.

Then tested with recent 8.3.0 on Ubuntu - same track - no issue with Receiver or Muse. If the issues was only compression, this shouldn't be the case.

I'll try to test with 8.2.* on Ubuntu.

[bpa]

Then tested with recent 8.3.0 on Ubuntu - same track - no issue with Receiver or Muse. If the issues was only compression, this shouldn't be the case.

I'll try to test with 8.2.* on Ubuntu.

Test of Muse & receiver with 8.2.1 on Ubuntu - no problems.

So my issue is probably not compression.

My problem track is 17min long which perhaps "forces" the problem on my smaller CORE32 system which short track avoid. I'll retest tomorrow to see if issue is reproducible as it was this evening.

[bpa]

Repeated test this morning. The Rhapsody in Blue track consistently had problems on DAC32 using LMS on CORE32, no problem with DAC32 with LMS on Ubuntu - LMS 8.2.1 version. Problems often started after about 5 min.

No problems on CORE32 after Settings/Network/Streaming mode for HTTP(S) set to "Persistent Mode"

[Redrum]

Repeated test this morning. The Rhapsody in Blue track consistently had problems on DAC32 using LMS on CORE32, no problem with DAC32 with LMS on Ubuntu - LMS 8.2.1 version. Problems often started after about 5 min.

No problems on CORE32 after Settings/Network/Streaming mode for HTTP(S) set to "Persistent Mode"

was this the track you had problems with?


If so, I try later this morning, both normal and persistent (receiver and touch on pcP). To be clear, I am on a "HiFi" trial in case that's relevant.

I wondered if this was related to tracks that are "Masters" (MQA - up to 24/96) but I believe I have played quite of few of these selections with no problems

Jim

[bpa]

was this the track you had problems with?

No. 2nd item in playlist "Classical at the movies" under Genres/Classical. Also on Hifi Trial so no access to MQA.



I think the main issue is the duration - it is 17 mins long. "Persistent Mode" solves similar problems for users with podcast (i.e. long continuous track).
A secondary issue is probably encoding - which also affects the memory required for buffering (these are TCP/network buffers). I had problem often shortly after 5mins - perhaps with m4a track it might happen at 10-15mins. If that is the case probably a track longer than 25 mins would be needed.

[Redrum]

No. 2nd item in playlist "Classical at the movies" under Genres/Classical. Also on Hifi Trial so no access to MQA.

Sorry I wasn't clear, I am on Hifi "Plus" Trial.

I'll try this track later this morning. I am wondering if we are hearing the same thing. I can't recall exact tracks, but I have heard stutter on shorter tracks (say under 5min). What I have heard can start immediately in the beginning of a song.

Having said all that, running persistent for about 3 hours yesterday - no problem. Since if was a Tidal assembled "playlist" it was all different media.

Jim

<edit - for clarity, what I am hearing occurs only on a receiver while synced with a touch. So, while the touch is happily playing, the receiver is struggling, and the resulting sound is bizarre when in earshot of both >

[bpa]

I am wondering if we are hearing the same thing. I can't recall exact tracks, but I have heard stutter on shorter tracks (say under 5min). What I have heard can start immediately in the beginning of a song.

I'm pretty sure NOT hearing the same symptoms. The symptoms I heard

1. Somewhere mid track (usually after 5 mins) - player will skip to next track.
2. Skip a chunk in the middle of a track but not replaced by silence - as if a chunk of music was lost.
3. At end of track usually between last 45-15 secs - silence but time clock counts down at x secs to go, then at zero resets back to x secs to go and counts down again & again. With this issues, occasionally player may eventually progress to next track.