Bluesound Node Icon (Quick User Measurements)

[Rja4000]

As others have pointed out before bluOS is in principle open source and Lenbrook are legally obliged to supply the source to anyone who asks. I'm pretty sure there used to be a Github repo for bluOS but it seems to have disappeared ...

I'll have a look.
Thanks

 

[GXAlan]

Best buy ever at this hobby is something who let you see the performance,doesn't need to be elaborate like an AP,just a decent interface who can detect some basic problems.

+1. The E1DA Cosmos ADC

On my NAD C658, which also uses BluOS, you can point a browser at <IP address of your device>/diag?print=1 and see the software diagnostics. It gives an idea of what's going on.

Super cool. You definitely see ALSA and the XML that really does suggest that it is using ALSA’s built in bass/treble. That means we just need someone to fix ALSA

Here’s the most interesting part to me:

I assume the subwoofer is what’s running the PCM512X.

CPU is the i.MX8MM. So probably:

i.MX8M Mini

The i.MX 8M Mini is NXP’s first embedded multicore applications processor with advanced 14LPC FinFET process technology for more speed and improved power efficiency.

www.nxp.com

Oh, and as an aside, the unit does playback DSD128.

 

[Sokel]

I assume the subwoofer is what’s running the PCM512X.

It's specs report a -93dB THD+N 20Hz-20kHz at -1dBFS with 20kHz BW.
(I d0n't understand the "max" -83dB THD+N,so be it)
Not far at all from what you measured.

 

[GXAlan]

It's specs report a -93dB THD+N 20Hz-20kHz at -1dBFS with 20kHz BW.
(I d0n't understand the "max" -83dB THD+N,so be it)
Not far at all from what you measured.

View attachment 429322

I guess it’s possible that the RCA is running the PCM5102 compared to the XLR running the ESS, but that would suggest that the N132 is better for RCA. I guess we would need @Rja4000 to see what his diagnostics logs shows.

All in all, the Node Icon is working very reliably. The SINAD loss from Dirac isn’t accurate in the sense that Dirac isn’t just a PEQ and the “harmonics” may be what is needed as a result of overall presentation. Even though I say don’t adjust gain/delay, there is a very big difference in imaging with Dirac enabled.

With Dirac disabled the phantom center is ultra precision. This is maintained with Dirac based off the single point electrical measurement. With Dirac enabled, there is a more diffuse center image which is also matched with a wider lateral left and right sound stage. Different not better. Dirac might be more spacious while native response (again on my DSP phase linear Meyer X40) is that extreme phantom center. (More measurements here)

All in all, having bought the Fosi ZD3 then the WiiM Ultra and now the Node Icon, the Icon is clearly the superior overall product as long as you don’t have a ground loop and are not bothered by the cost premium.

The QRONO d2a works well to my ears, but I suspect it will get a shrugging panther when Amir reviews it due to the leaky filter even if 1 kHz SINAD will be pretty good.

 

[pogo]

I guess it’s possible that the RCA is running the PCM5102 compared to the XLR running the ESS

The implementation should be similar to the M66, PCM5102 only for sub:

 

[GXAlan]

The implementation should be similar to the M66, PCM5102 only for sub:

View attachment 429431

I wonder how similar or different the Node Icon is to the M66. On my Dirac enabled test, I got 128.3 dBA which is actually the threshold of the E1DA Cosmos, Grade A (rated at 128-129 dBA)

That was the third test I ran so it could have been the E1DA coming up to temperature

The Dirac Mode doesn’t hit the full 4V but it too is pretty much at the threshold of my instrumentation.

How crazy would it be if the Node Icon was the M66, minus DLBC when running as a DAC/digital streamer.

Edit: The high FFT makes the graphs look pretty but the calculations for Noise do in fact sum the noise across all of the bins, so barring user error, it really does seem that the Node Icon reaches the measuring limits of the E1DA Cosmos ADC. Impressive!

 

[somebodyelse]

Super cool. You definitely see ALSA and the XML that really does suggest that it is using ALSA’s built in bass/treble. That means we just need someone to fix ALSA

What built in bass/treble would that be? Someone would probably need to dig into the firmware to find out exactly what they're using in order to replicate it for testing.

 

[GXAlan]

What built in bass/treble would that be? Someone would probably need to dig into the firmware to find out exactly what they're using in order to replicate it for testing.

Getting Root On Linux Amplifier Adds New Inputs

We remember when getting Linux on your average desktop computer was a tricky enough endeavor that only those with the most luxurious of graybeards would even attempt it. A “Linux box” i…

hackaday.com

<alsa usbDac="1">
            <conf file="asound.conf.tone_analog_bm_dirac_spdif_vol" card="soundanalog" spdif="imxspdif" outputPcm="analog_6ch"
                sub2="0" rateConverter="speexrate_medium" mqaRenderBlob="RenderingBlob_bDAC_Max8x"
                dirac="yes" diracBMExpectLFE="yes" />
            <eq id="dirac" displayName="Dirac Live" icon="/images/settings/ic_dirac.png" muteDelayWrap="0.3" control="iface=MIXER,name='Dirac'" class="list" default="0">
                <value name="0" displayName="Off"/>
            </eq>
            <eq id="switch" displayName="{A_TONE_CONTROL_LABEL}" icon="/images/settings/ic_eqonoff.png" control="iface=MIXER,name='Tone Fixed Gain'" default="off" type="onoff" volumeOffset="6" muteDelayWrap="0.15" />
            <eq id="treble" displayName="{A_TREBLE}" icon="icon_treble.png" control="iface=MIXER,name='Treble Playback Volume'" default="12" units="dB" type="treble">
                <dependsOn name="switch" value="ON" />
            </eq>
            <eq id="bass"   displayName="{A_BASS}"   icon="icon_bass.png"   control="iface=MIXER,name='Bass Playback Volume'"   default="12" units="dB" type="bass">
                <dependsOn name="switch" value="ON" />
            </eq>
            <eq id="subwoofer" displayName="{SB_SUBONOFF}" icon="icon_subwoofer.png" control="iface=MIXER,name='BM Subwoofers'" class="boolean" default="0"/>
            <eq id="crossover" displayName="{A_CROSSOVER}" icon="/images/settings/ic_crossover.png" control="iface=MIXER,name='BM Frequency'"
                class="range" min="40" max="200" step="10" offset="40" default="80" units="Hz"><dependsOn name="subwoofer" value="ON" /></eq>
        </alsa>

 

[somebodyelse]

We'd need to see the alsa config (probably asound.conf.tone_analog_bm_dirac_spdif_vol) to know where those mixer tone controls come from. They could be from the device driver giving control to something in the audio hardware, or provided by a plugin. ALSA doesn't have any built in tone controls (unless I've missed some development), so the user has to use a plugin like alsaequal which can provide a mixer interface to LADSPA plugins. The alsa config defines what plugins are used and how they are connected.

Note that the config would probably be covered by the BluOS EULA which doesn't allow this. What I've seen so far reinforces my distrust of audio companies' software capabilities though, especially the (lack of) reaction to a report of a remote root exploit.

 

[Sbrtoy]

Really awesome stuff GXAlan, my Icon is hopefully next in the queue from Amir so we will see his results to compare.

When listening to the Icon vs Wiim the biggest difference to me was the layering of soundstage information. I have KEF R3’s which do this well and there is much more specific depth and positioning with the Icon. Is there a specific measurement that would point to this difference or could it be the filter?

 

[pogo]

See QRONO Technical White Paper - It’s About Time

 

[mr_bill]

In actual listening tests are any of you finding the rca outputs and xlr outputs performing about the same?

 

[Rja4000]

I would occasionally use it in a studio setting so latency is important.

So I measured RTL on the Node N132:
RME ADI-2/4 Pro SE SPDIF Opt. Out > Node Opt In > Node Opt Out > RME Opt. In

Measured with RTL Utility 1.0.8.

Of course, I measured the RME SPDIF Opt. Loopback and deducted from the above.

The best I can measure is 249ms (Node only).
That's with the max buffer size on the RME ASIO.
Anything with a lower buffer size gives me an error.

(In loopback, the RME measures 1.7ms min @192kHz)

This is without Dirac and with Tone control disabled.

 

[GXAlan]

Really awesome stuff GXAlan, my Icon is hopefully next in the queue from Amir so we will see his results to compare.

Nice. Hopefully it’s similar.

When listening to the Icon vs Wiim the biggest difference to me was the layering of soundstage information. I have KEF R3’s which do this well and there is much more specific depth and positioning with the Icon. Is there a specific measurement that would point to this difference or could it be the filter?

No idea. That’s the thing. The WiiM Ultra measures really well. Positioning should be a combination of phase and channel separation but we don’t measure phase that often.

The QRONO d2a filter is definitely detectable with test patterns but I cannot prove it’s audible without more formalized blind testing. (can do it with a 16/44 through QRONO versus a 16/44 upsampled to 24/192 with a traditional filter I suppose).

The best I can measure is 249ms (Node only).

Is there an AV Mode or or off for the optical input? Pogo said 250ms for AV mode off and 50-150ms AV mode on.

 

[pogo]

50-150ms AV mode on

This is adjustable for each input
Default is AV mode off = 250ms.

 

[pogo]

Here is my solution from LINDY (Lip Sync Connector) to also delay the rear speakers to compensate for the minimum BluOS delay of 50ms for my 4.0 system:

 

[Sokel]

Not yet out,but it seems that Wiim also suffers the child diseases of half-baked EQ down low.
Let's wait a little to see.

It's like all these companies are outsourcing their filters to the same builder

 

[Rja4000]

Is there an AV Mode or or off for the optical input? Pogo said 250ms for AV mode off and 50-150ms AV mode on.

Yes. And in that mode, I measured a minimum of 50ms, indeed.

 

[bluefuzz]

It's like all these companies are outsourcing their filters to the same builder

Well, most of these streamer OS's are built on variations of Linux and presumably utilise the available audio toolchains as far as possible. Otherwise why use Linux in the first place? So it's not surprising they have similar issues ...

 

[Mnyb]

Well, most of these streamer OS's are built on variations of Linux and presumably utilise the available audio toolchains as far as possible. Otherwise why use Linux in the first place? So it's not surprising they have similar issues ...

it does not necessarily need to be bad . I suspect some complex interaction with their drivers for their hardware and /or what they actually compiled . They may even have some libs without floating point support .

But Linux is beyond casual understanding for me it can literarily be anything ?