The "Official" Aurender Discussion Thread

[FredericV]

Does the MQA decoder work with both MPD (the opensource software they are using when playing directly from the Aurender, e.g. when using their Conductor MPD app) and the Roon endpoint on the Aurender? Or is it limited to MPD only?

[FredericV]

On 4/11/2018 at 6:13 PM, The Computer Audiophile said:

Be vary careful not to rush to judgement and accept this as fact. I highly recommend listening for yourself and doing some research.

 

When we were still using MPD 0.17 (same open source player as used by Aurender) and later up to 0.19 (these days we don't use MPD):

- the MPD version made a sound difference
- even the flac decoder made a difference, 0.19 had several flac decoders in different libraries

I even used to modify the decoderlist in our MPD builds. Also 0.19 and above is C++ based, the > 7 year old 0.17 is C based. Even the language made a sound difference, with no apparant changes in calling the external flac library, except that 0.19 has object based calls, and 0.17 procedural based.

So WAV vs FLAC vs uncompressed FLAC causes different load patterns on the CPU, and on most streamers this makes a sound difference. CPU's generate microwave in certain repetitive patterns, which differ between these formats.

[FredericV]

On 9/10/2018 at 6:25 PM, Quadiffusor HK said:

OK, so the potential culprit may be an innocent attempt at plugging in a USB cable while the unit or units are powered on.  Does that mean powering down the components are a "must", but only when connecting two components (ie. Aurender and outboard DAC) via USB for the very first time?

 

If one device has a SMPS and the other a linear one, you can get a voltage difference between the two devices. Sbooster even wrote some paper about this issue. Also can happen when each side has a SMPS.

If you connect the USB cable from the server to the DAC, and you see a spark when inserting the USB cable in one, you actually risk damaging the USB output of the server if it has not been designed to handle this issue. Some high-end USB implementations are inert for this issue (e.g. sotm, which is what we use), while other vendors can have the USB card's ports blown each by each. We had one brand with dual outputs and one of our resellers blew USB port after port until we were tired of this issue after replacing the dual USB card several times. In his system, the server was powered linear, but his class-d amps had SMPS supplies. This was introducing the issue in his system.

This issue also occurs when you insert a USB powered HDD into a network player like a dune HDI, which in my home cinema always gives a little spark. If you disconnect all power wires and first attach the USB cable, and then the power cables, you don't have this spark issue, so you have no risk in killing the USB port.

Good designs should be inert to this issue, as most PC's when connected to powered USB devices, also have this issue - and survive it.


 

[FredericV]

Can't no longer edit the previous post, so here's the technical explanation of SMPS leakage, killing USB ports under certain conditions:

https://www.sbooster.com/FAQ/
 

Quote

SMPS (switch-mode power supply/-ies) generate a high leakage current with voltage spikes > 300V AC on the output. This is caused by two capacitors that are internally connected between the Hot/Neutral and output of a SMPS.

Some audiophile USB interfaces are inert to this issue, others are not, and most PC gear seems also inert.

[FredericV]

10 hours ago, jon2020 said:

 

Very tempting prospect.....

 

http://34.227.103.167:8000/acs10/ 

 

I wonder what CPU aurender is using. The never specify it, and recently started to specify the system memory in case of the ACS10.

e.g. the N100 series have 4 GB of ram and a 32 bit linux OS & kernel, but this is nowhere specified - only the SSD and HDD are specified

[FredericV]

8 minutes ago, jon2020 said:

 

Curious as to why you ask about this. Would the choice of CPU make much difference for SQ?

 

It is quite likely, the processing power would have been optimised by Aurender in its selection of CPU. 

Looking at the machine we have in the lab, the design is based on 2 boards: a bottom plate with the IO and psu, and a separate board with the CPU, which they probably use in other models. The top plate gets quite hot, so I'm wondering what the TPD of the cpu is. I've seen and used other designs with 6 ~ 17 Watt TPD CPU's and they never run hot. I also have a lot of old atom boxes which run hot.

For me thermal properties and TPD matter a lot (as a designer).
 

[FredericV]

2 minutes ago, The Computer Audiophile said:

Frederic- You’re a competitor of Aurender. Your behavior here is highly suspect. 

 

They are just technical questions. e.g. if you buy a Roon Nucleus, you know what's inside.

And yes I never hid my affiliation with the industry, so what's the problem?