Sound Tweak Rpi4

[kit1cat]

How do I turn the cache off?

How did you turn it on ?

Using Tim's command (has to be set after each boot) http://moodeaudio.org/forum/showthread.p...9#pid22699


Or Kent's  (permanent)  http://moodeaudio.org/forum/showthread.p...7#pid22767

If Tim's then just restart.  (Caching does not happen unless you rerun the commands each boot)
If Kent's then remove the additions and restart.

Thanks, will give it ago in the morning.

Just tried another quick test,

Base memory 18%
Added 4 tracks 19%
play track 1 26%
play track 2 28%
play track 3 32%
play track 4 47%

play tracks 1,2,3 & 4 again 47%

add track 5 47%
play track 5 54%

memory usage seems to increase in line with the length of the track, track 4 is 24 minutes.

I'm not sure I understand this, is memory usage considered a bad thing?  I have always been under the impression that unused RAM was wasted RAM as loading from RAM is faster than from HD or SD card?
As long as other processes are not starved of RAM & RAM is released allowing future tracks to cache, what is the problem?
I have not tried any of these cache settings being discussed, but if I saw my RAM usage at 90% I wouldn't be concerned.

No problem with memory usage, the test just prove that the tracks are NOT loaded to memory before they are played.

[grasshopper]

How do I turn the cache off?

How did you turn it on ?

Using Tim's command (has to be set after each boot) http://moodeaudio.org/forum/showthread.p...9#pid22699


Or Kent's  (permanent)  http://moodeaudio.org/forum/showthread.p...7#pid22767

If Tim's then just restart.  (Caching does not happen unless you rerun the commands each boot)
If Kent's then remove the additions and restart.

Thanks, will give it ago in the morning.

Just tried another quick test,

Base memory 18%
Added 4 tracks 19%
play track 1 26%
play track 2 28%
play track 3 32%
play track 4 47%

play tracks 1,2,3 & 4 again 47%

add track 5 47%
play track 5 54%

memory usage seems to increase in line with the length of the track, track 4 is 24 minutes.

I'm not sure I understand this, is memory usage considered a bad thing?  I have always been under the impression that unused RAM was wasted RAM as loading from RAM is faster than from HD or SD card?
As long as other processes are not starved of RAM & RAM is released allowing future tracks to cache, what is the problem?
I have not tried any of these cache settings being discussed, but if I saw my RAM usage at 90% I wouldn't be concerned.

No problem with memory usage, the test just prove that the tracks are NOT loaded to memory before they are played.

OK, I clearly misunderstood your tests.  I've started using the testing MPD & the 64bit kernel, just for the heck of it really, I have not viewed any logs as it's played music continually when I have asked it to, via a NAS & Tidal using bubbleupnp, it's not skipped a beat & sounds amazing.  No manual "tweaks" or amendments.

[boucher]

Just tried 0.22. Crazy good. I don’t even want to touch it at this point. I have an rpi4 -> Schiit Modi Multibit over USB - Tidal upnp. I didn’t think I could get straight USB to sound good on the Mimby, given its antiquated USB implementation, but here I am. Soundstage is much wider, veil is lifted, music has more weight and presence, seems more effortless, graininess seems to be gone, better instrument separation. Analog smoothness with digital clarity. Love it.

Moode 6.6, 64-bit, performance.

M P D   S E T T I N G S

Version = 0.22~git
Volume control = disabled
ALSA device = hw:1
SoX resampling = *:32:*
SoX quality = very high
SoX multithreading = on
DSD over PCM (DoP) = no
Replaygain = off
Replaygain preamp = 0
Volume normalization = No
Audio buffer = 8192 (kb)
Output buffer size = 131072 (kb)

[hifinet]

MPD 0.22~git (test) should be the default setting. For first time users of Moode, this is what they need to hear.

[boucher]

MPD 0.22~git (test) should be the default setting. For first time users of Moode, this is what they need to hear.

I seemed to have lost a little bit of goodness with 6.7 and 6.7.1... weird. Anyway, I agree that it is pretty solid.

As a side note, how do you like your Icepower monoblocks. I see parts-express has them deeply discounted right now. Worth building or would you recommend going another direction? Currently running my setup through a Denon AVR, so probably a big jump for my system. Thoughts?

[Tim Curtis]

MPD 0.22~git (test) should be the default setting. For first time users of Moode, this is what they need to hear.

Prolly couldn't do that because some users reported issues with 0.22 and long MPD indexing times, but I might be able to periodically recompile 0.22 to the most recent commit and put up a download link in an "MPD 0.22 test" thread. Something like that.

[hifinet]

MPD 0.22~git (test) should be the default setting. For first time users of Moode, this is what they need to hear.

I seemed to have lost a little bit of goodness with 6.7 and 6.7.1... weird. Anyway, I agree that it is pretty solid.

As a side note, how do you like your Icepower monoblocks. I see parts-express has them deeply discounted right now. Worth building or would you recommend going another direction? Currently running my setup through a Denon AVR, so probably a big jump for my system. Thoughts?

The Icepower are excellent amps. Using diy 70W PP tube monoblocks. Read good things about the Icepower. A lot of interest in Class D these days. Picked up a new pair of 500ASPs off eBay for a very good price. Quite shocked how good the 500ASPs sound. Mids similar to the tube amps, bass of course much better, highs not as clear and extended. Very neutral sounding. Using them as the summer amps. Rock solid. Have another pair of 125ASX2 monoblocks in the family. Sound great and rock solid performance. The 125ASX2 need an unbalanced to balanced converter board if using as BTL monoblocks.

[wkloppen]

Hello gents...

Does the linux audio adjustment script still work with the new moode version running on pi4b with 8 mem?


thx

reg

willem

[wkloppen]

well at least the noop scheduler doesn't exist anymore but could be easily changed to none. Not sure what the exact difference is between the 2 and whether or not it affects the SQ. Doubt it to be honest. btw I'm only running Qobuz. No local files, nasses, disks or any of that stuff.

Code:

#Reduce operating system latency
#echo noop > /sys/block/mmcblk0/queue/scheduler

regards

willem