simes_pep wrote:
It was a John Westlake design, he was on the DIY Audio forum for a while
talking about the PT products. However I'm not sure if a 24-bit stream
was available when they were designed? It was predominantly a 16/44.1
feed from a CD transport/player, or a 16/48 from DAT, DCC or
SBGK wrote:
after hearing the sound quality of my nad m51 via hdmi from my laptop i
am now resigned to the fact that the touch is a dead end in terms of
sound quality improvements in my system. EDO is just a distraction it
has higher resolution, but poor sound quality.
i have not heard
Zombie wrote:
Can't find the app. Is it removed?
No - its still on the 3rd party list. You should look on the player
Apps Gallery, 3rd Party Apps
Triode's Profile: http://forums.slimdevices.com/member.php?userid=17
SBGK wrote:
with my large buffer settings I can get several minutes playback of
16/44.1 stored on the touch after the music has loaded via the ethernet.
So I can switch off the laptop and disconnect the ethernet while the
music is still playing - the sound doesn't change when I do this, it
adamdea wrote:
Leaving that to one side, have you been able to make any measurements of
the S/PDIF out which could demonstrate and quantify the effect of the
EDO changes (and other mods). I was surprised to read on a PFM thread a
post from John Westlake saying that the S/PDIF out of the
SBGK wrote:
I am still interested in understanding why the usb irq is being called
2000 times a second, is this something that can be reduced or is it
dependent on the usb device used ? I don't think it can be helping sound
quality.
Not having interrupts would have a dramatic impact on
magiccarpetride wrote:
I must say I'm baffled by your posts here. Didn't you originally claim,
when you announced your EDO, that your intention was ONLY to enable
24/192 processing on the Touch, and that you were certain that this mod
could not possibly affect the sound quality otherwise?
SBGK wrote:
why not have the default code use the 0.6 method and then if it senses
internet radio at the start of the track make it branch to a separate
codepath that allow the internet radio format to play.
.
This is what the code is intended to do - so I want someone to prove
that 0.7 is
cfraser wrote:
^ lol, I forget why I even said that now. I guess it's obvious you'd
need a proper/good logic analyzer to monitor an actual design. I was
probably thinking along the lines of the differences between running
slightly different code is sufficiently significant that it can be
SBGK wrote:
So basically the binaries are the same as jive_alsa for 0.6 and 0.7,
just recompiled.
I tried the recompiled 0.7 and it had the same issue as the old one ie
harsh treble compared to 0.6, this is on a new dac so can't blame the
benchmark dac1 for that.
back to 0.6 and very
SBGK wrote:
The outputs are all the same between 0.6 and 0.7, but I guess you knew
that. So I tried 0.5 and it gave the same results, I think most people
would agree that 0.5 sounds significantly different to 0.6 and 0.7, so
must be down to something else.
So now the change is between 0.5
SBGK wrote:
Triode, I have just confirmed to myself that 0.6 sounds different to
0.7, listening to hells bells on ac/dc back in black album the first
notes from the lead guitar are particularly piercing with 0.7 and sound
totally different with 0.6 - more mellow with greater depth.
Look
Sysagent wrote:
As the title says :)
How can I go back to the previous version of the EDO plugin please?
I am on the current new version and to my ears it does sound quite harsh
to listen to for a prolonged period of time, if somebody could post a
step by step way of how to go to the
SBGK wrote:
that's the problem Triode, you ask for feedback and so far you have not
believed what people tell you and , in fact, have belittled the comments
about 0.6 vs 0.7, why should we try your different binaries if you don't
believe what people say and also if you can't hear
SBGK wrote:
no point unless you want a more analogue type sound backed up by
empirical testing by about 10 people.
oh well, there must be some other explanation for the difference in
sound caused by setting large buffers.
Maybe one day we'll get to understand.
Have you tried larger
magiccarpetride wrote:
Actually, 'revealing system' indeed has a solid definition, so there is
absolutely no need to try and relativize it and 'subjectivize' it. It is
quite easy to compare two audio systems and provide quantifiable metrics
as to why is one system more revealing than the
SBGK wrote:
agree that the index change is probably expectation bias, but the other
changes such as buffer sizes, fab4.lua setting etc have a definite
effect in my system.
Changing buffer size where the actual buffer size used changes could
easily have an effect. Changing it beyond the
SBGK wrote:
With spdif out I was using a period count of 4. Setting the buffer time
to 9 gave period and buffer sizes of 1024/4096 for all sample
rates.
With lower buffer time the buffer size and period size would be smaller.
The exact buffer cut off that allows this behaviour is
guidof wrote:
But here is the return from the squeezeplay restart command (this with
EDO 0.6, buffer at 9, period count at 16):
Code:
buffer_size : 4096
period_size : 256 --- lower than default
period_time : 5804 --- lower than
guidof wrote:
When you state that my current settings force the period size to be 256
frames at 44.1 sample rate, are you suggesting any consequences other
than limiting the effective buffer time to 92864?
No consequences other than that - I am interested though if other
settings with the
guidof wrote:
Restarted squeezeplay with above command, which returned the following:
I'm not sure this is what you wanted. If not, please let me know.
See post #54 in this thread - you should see the set of alsa params
which are actually used - as per the details below. The
SBGK wrote:
Triode, this is the command that I use to check the buffer sizes, it
certainly shows different values between 10, 40 and 9
For SPDIF output, TXRX device? (Yes I agree for usb, but as per post 54
I don't believe it will for spdif
SBGK wrote:
good point, was wondering why the EDO mods didn't get the vinyl like
effect of the spdif out, but the extra detail makes up for it.
So do you think 40 is the maximum value ?
Read post 54 - 10 is already too big, no point making any bigger for
spdif with period count of
SBGK wrote:
hmm, misread your post, you are saying it does affect the buffer size
for usb and not for spdif. I noticed the effect on spdif long before I
tried EDO, so think there is definately an audible effect via spdif. In
fact it is much more pronounced than using EDO. Think it is also
SBGK wrote:
have set the alsaPlaybackDevice=default in file SqueezeboxFab4.lua,
this was an SBGK mod, default is 'default'
As I posted above 0 this value is totally irrelevant when EDO is used -
it will have no impact on sound quality.
--change the permissions to allow the index
SBGK wrote:
To conclude, the Touch is just too noisy to live in revealing system.
The EDO mod shows up the inadequacies of the spdif out, but the EDO mod
introduces noise itself, not least the 2500 per second usb interrupts.
I would suggest you need to be careful making blanket statements
guidof wrote:
I have done extensive comparisons of EDO 0.6 and 0.7 and for the most
part agree with your impressions.
Can you confirm the same is true between the two jive_alsa's files I
attached (which are the same as in the two downloads)
For me and with my system, EDO 0.6 + TT3.0 +
rgro wrote:
And, Johnthat particular line is placed where? In the tcp/ip
section, priorities, or kernel??
I guess my point is I think it will be hard to set nrpacks with EDO at
present as you will need to change that kernel module parameter before
the module detects the usb dac. As EDO
cfraser wrote:
It might be convenient if SBGK put some instructions on how to install
older versions of EDO in his blog, like where to copy the files. You are
welcome for me volunteering you. :) I don't have a blog etc. or else I'd
do it.
I get the impression that the instructions might
Ok so here's details of the difference between EDO 0.6 and 0.7 and the
two binaries for people believing there is a difference.
1st - a diff between the two install zips:
Code:
diff -dur 0.6/EnhancedDigitalOutputApplet.lua
0.7/EnhancedDigitalOutputApplet.lua
---
cfraser wrote:
Things appear to be working. However, I wouldn't mind hearing how I
could manually install an EDO version so that it looks (from the SBT
menus) the same as if I had installed the latest EDO version by the
regular method. If there is no (easy) way to do this, and if the
it)
2) New code in jive_alsa which overrides the buffer setting in the case
that resampling is involked. The change to the code is here:
https://github.com/triode/squeezeplay/commit/198f7818c95261bf3fd14110df1004a4e7c35c9a
In short I don't see a reason for people trying to install 0.6 rather
than
It has been reported elsewhere that editting
/etc/squeezeplay/userpath/settings/SqueezeboxFab4.lua to alter buffer
tuning and output device parameters has an impact on sound quality. I
do not believe this to be true. The EDO applet is designed to run
before SqueezeboxFab4 applet and hence
It has been reported elsewhere that tuning the kernel module parameter
nrpacks may have an impact on sound quality with EDO when using a usb
dac.
This can be achieved by:
echo NUMBER /sys/module/snd_usb_audio/parameters/nrpacks
Please note that if you try this tuning parameter there are two
Thought I'd post something to show what the maxium buffer sizes are to
avoid people trying to tune too far.
To see what the actual audio output setup is, login to touch and restart
the application with audio.output logging set to debug:
Code:
#
I promised to publish the source to the modified jive_alsa - I've put it
on github as a forked version of squeezeplay:
https://github.com/triode/squeezeplay
This is the commit which adds the randomisation - any comments on this
doing anything would be interesting, as would views on alternative
JJZolx wrote:
How does jitter get through an asynchronous interface? Isn't that a bit
like talking about jitter on an Ethernet connection?
John is probably talking about non async usb dacs (of which there are
still many) - adding an issolator to these could easliy make it worse.
In the cause
EDO 0.6 contains an experiemental menu (Settings Advanced Digital
Output Buffer Tuning) to allow some simple output buffer tuning to be
performed without need to install TT3 and log into Touch. I make no
claim that buffer tuning will change sound quality, but did want to
speculate on what
rgro wrote:
Hi Triode,
Just to clarify, am I correct in that for your app to function when
music is playing, the Display Off option should be selected under
ScreensaversWhen Playing?
Thanks,
Yes, and probably set the timeout to 10 seconds. After the 10 seconds
timeout
lake_eleven wrote:
Once turned off, how to turn on if needed?
By touching the screen or using the remote control - it acts as any
other screensaver in that it only triggers after a timeout on
inactivity.
Triode's
Covenant wrote:
Triode wrote:
I will post this here as its clearly unproven and something to discuss
on the audiophile board
[As an aside - I note that with the EDO app installed and the internal
digital out selected so that the analog output is not enabled, if you
lake_eleven wrote:
Can this be installed over EDO and TT3.0
Yes, but not the screen off part of TT3 as it is intended to to that
function (and slightly more) tied to the screensaver activation.
Triode's Profile:
I will post this here as its clearly unproven and something to discuss
on the audiophile board
I've just posted a new app to the repro for Squeezebox Touch - this
provides an alternative to the built in Screen Off screensaver which
disables some of the processing associated with the screen
tonyptony wrote:
Does anyone know if Inguz uses the alsa plugin layer or the jive_alsa
process? It would be unfortunate if this interesting new app is not able
to run with Inguz in line. I'll have to try it out and see.
Inguz is all server side and this is all done on the Touch player so
There has been some discussion on the Touch forum about the EDO app and
the potential benfits to audio quality from the spdif output at 96k and
below. This is not the primary purpose of the enhanced digital ouput
app but people have debated whether any possible change in audio quality
exists. I
JohnSwenson wrote:
I have a HRT Music StreamerII (not the +), it works very well with
Triodes plugin and a hub. Adding TT3.0 significantly improves the sound,
with this combination it's getting astonishingly good. (it's not the
best, my homebuilt DAC still blows it away). The II with TT3.0
://www.diyaudio.com/forums/digital-source/28814-asynchronous-sample-rate-conversion.html
--
Triode
Triode's Profile: http://forums.slimdevices.com/member.php?userid=17
View this thread: http://forums.slimdevices.com/showthread.php?t
of running the background process which is part of my blind tester app.
As I said in that thread I can see the amount of time and frequency of
running the idle process causing differences in power consumption of
the cpu, but thus far I think there are any results showing this...
--
Triode
players should be more accurate in recent builds.
--
Triode
Triode's Profile: http://forums.slimdevices.com/member.php?userid=17
View this thread: http://forums.slimdevices.com/showthread.php?t=93879
their results
for test 2 are.
--
Triode
Triode's Profile: http://forums.slimdevices.com/member.php?userid=17
View this thread: http://forums.slimdevices.com/showthread.php?t=91322
Soundman;690607 Wrote:
It's not suitable to test TT-mods. As Triode writes: The applet only
supports changes which can be made on the fly without restarting
Touch. The TT-mods always require a restart (except for TT -k). Comes
to it that the TT-mods should be tested all together, which
file rather than 320 - you can often see this if you have logging for
the helper app set to debug. Other than setting a preference for 320k,
there are no api calls to influence sound quality.
--
Triode
Triode's Profile
I think the latest version of portaudio supports WASAPI - I've not tried
though. Another reason to look at this though...
--
Triode
Triode's Profile: http://forums.slimdevices.com/member.php?userid=17
View this thread
The BBCiPlayer plugin uses the new 96 kbit/s live streams if you are in
the uk.
--
Triode
Triode's Profile: http://forums.slimdevices.com/member.php?userid=17
View this thread: http://forums.slimdevices.com/showthread.php
that the difference between pausing and unpausing a track
and pausing, powering off, powering on and unpausing is minimal - just
a few different grfe (graphic i.e. display frames). This is the reason
for the arguement that nothing special is happening by powering off...
--
Triode
I was trying to say its much more than 50ms - its of the order of 10secs
or more depending on the data rate you are streaming at.
[put I liked to do demos which always worked]
--
Triode
Triode's Profile: http
Buffer size is implemented in the player firmware and is not user
adjustable I'm afraid.
--
Triode
Triode's Profile: http://forums.slimdevices.com/member.php?userid=17
View this thread: http://forums.slimdevices.com
.
--
Triode
Triode's Profile: http://forums.slimdevices.com/member.php?userid=17
View this thread: http://forums.slimdevices.com/showthread.php?t=31833
___
audiophiles mailing list
-variation, loop-control,
mechanical resonance of the disk spinning etc...
So in the case of adding a Superclock to improve performance - do we
replace the PLL clock with a fixed phase clock? I'm struggling to
understand how that avoids compromising the data.
--
Triode
insulation around the earth connection - see here
(thanks to google!):
http://sosnick.uchicago.edu/BNC_50_75.html
--
Triode
Triode's Profile: http://forums.slimdevices.com/member.php?userid=17
View this thread: http
.
--
Triode
Triode's Profile: http://forums.slimdevices.com/member.php?userid=17
View this thread: http://forums.slimdevices.com/showthread.php?t=27946
___
audiophiles mailing list
]
--
Triode
Triode's Profile: http://forums.slimdevices.com/member.php?userid=17
View this thread: http://forums.slimdevices.com/showthread.php?t=27445
___
audiophiles mailing list
Good to hear this John. You will be glad to know that following some
debate, 6.5 will default to replaygain turned off for new players - so
audiophiles should not get tripped over by this in future...
--
Triode
Triode's
Should be OK with alien as the mplayer config resamples it to 44.1.
Also, all you need to do is lift one end, so it is can be restored if
necessary..
--
Triode
Triode's Profile: http://forums.slimdevices.com/member.php
into the power rails
--
Triode
Triode's Profile: http://forums.slimdevices.com/member.php?userid=17
View this thread: http://forums.slimdevices.com/showthread.php?t=25138
___
audiophiles
and 0 db all other times. However having
looked at the server code this is harder to achieve - so is the
suggestion above useful? If so can someone raise an enhancement
request and note it here??
--
Triode
Triode's
the equivalent plots?]
--
Triode
Triode's Profile: http://forums.slimdevices.com/member.php?userid=17
View this thread: http://forums.slimdevices.com/showthread.php?t=22301
...
--
Triode
Triode's Profile: http://forums.slimdevices.com/member.php?userid=17
View this thread: http://forums.slimdevices.com/showthread.php?t=22301
___
audiophiles mailing list
chip a few years ago] includes:
Wide-frame jitter margin built-in 32K RAM.
This is needed to buffer the datastream from variations in rate of
reading the disc and allow a cheap motor + servo to manage its speed.
--
Triode
paper) suggests audibility of jitter goes much lower than 20ns...
--
Triode
Triode's Profile: http://forums.slimdevices.com/member.php?userid=17
View this thread: http://forums.slimdevices.com/showthread.php?t=18116
potentially cause
it to slow down reading the file. This may be your problem.
Have you tried running slimserver on the server machine itself (as then
the files are local and slimserver can access them imediately).
--
Triode
was noticable with this.
More complex dacs attenuate jitter much more (in the digital domain
using async resampling or by secondary plls with much tighter loop
filters). With these any jitter on the spdif line should be much less
noticable...
--
Triode
. flac ~1000 kbps]
--
Triode
Triode's Profile: http://forums.slimdevices.com/member.php?userid=17
View this thread: http://forums.slimdevices.com/showthread.php?t=18199
From that description it sounds like a relatively simple input circuit
(nothing wrong with that). I would expect it will work fine with SB1
or SB2/3, but as it does not include any internal jitter rejection, the
SB2/3 will potentially sound better.
--
Triode
further reduce the jitter on the digital output.
In short SB3 (and 2) is definately technically better for the digital
output than SB1. Whether you can hear this depends on how well your
external dac hides the influence of jitter from the transport.
--
Triode
24/96.
--
Triode
___
audiophiles mailing list
audiophiles@lists.slimdevices.com
http://lists.slimdevices.com/lists/listinfo/audiophiles
Andrew,
Any views on the noise spectrum of Nicads/NMHi batteries? I am
speculating about a battery psu for the HCU04. As I am only interested
in digital out I am wondering whether this is a way to go. I'm
Currently using a simple linear reg for this.
Adrian
--
Triode
1.2V cells to the HCU04
--
Triode
___
audiophiles mailing list
audiophiles@lists.slimdevices.com
http://lists.slimdevices.com/lists/listinfo/audiophiles
posts on the http://www6.head-fi.org/ computers
as sources forum about modded RMEs and DAC1s - I think the conclusion
was that the DAC1 was preferred. [Though I have never heard one myself
and would want to before buying]
--
Triode
___
audiophiles
on AA's computer audio asylum [but I see you posted there
too!]
Adrian
--
Triode
___
audiophiles mailing list
audiophiles@lists.slimdevices.com
http://lists.slimdevices.com/lists/listinfo/audiophiles
line with
a flipflop just before the output buffer. Is the output of your Xilinx
clocked anyway to avoid any benefit of this?
Adrian
--
Triode
___
audiophiles mailing list
audiophiles@lists.slimdevices.com
http://lists.slimdevices.com/lists/listinfo
of this [even if we believed the first assersion]
Now if this thread had been about the visualizer or scrolling text
impacting the sound quality then it would be more interesting
[withdrawing quietly to see if this sparks some more comparison
threads...]
--
Triode
82 matches
Mail list logo
