A novel way to massively improve the SQ of computer audio streaming

[LTG2010]

6 hours ago, the_doc735 said:

At the moment I have a noctura fan on my 218+ and it doesn't blow hot air on other components

The NUC has a tiny case, especially on the more powerful processors, the fan does not work so well.

[LTG2010]

16 minutes ago, the_doc735 said:

Uh! ~ I was talking about my synology NAS 218+, not a NUC?

Well the discussion was based around NUC and the benefits of removing it from its case and using an Akasa fanless one. Temps dropped more than 10 degrees. As well as improvements to sound quality, mentioned by quite a few posters.

[LTG2010]

7 minutes ago, mourip said:

It seems like most of the folks who gravitate to this thread are of the school " If it sounds good and measures poorly then you are measuring the wrong thing." and it's corollary " If it measures great and sounds bad ...then why would you listen to it." ? 

Good design measures good and sounds good. It gets to a stage where the lowest /best measurements don't automatically give you the best sound quality because there are many variables as to component choice. A serious amount of listening needs to accompany measurements. Also you need to be sure that what you are measuring is relevant eg. a poor measurement does not automatically mean a bad product as there are a lot of compromises  and trade offs involved.

[LTG2010]

39 minutes ago, MrUnderhill said:

My DAC is the Audionote 2.1. This famously measures poorly; and yet I would say it is well designed and sounds GREAT.

Its an exceptional example of compromise where the designer has knowingly chosen a component or technique based on sound quality, its unlikely to measure badly in every respect.

A lot is spoken here of power supplies, even taking a simple component as a diode, each has its own sound signature. Shottky's measure well and have a low voltage drop and are used in the  best PS designs. You can however get better measurements using a mosfet diode bridge a lower voltage drop and more efficient design less heat dissipation and more efficient use of transformers. Therefore which is better?

Then we can move on to capacitors, resistors, regulators, transformer design, etc, etc. Compromises need to be made as there is also a budget to consider.

Therefore when a reviewer with a bit of technical knowledge sticks his  reviewerscope on the end of a carefully designed component, it doesn't always give a complete picture.

[LTG2010]

1 minute ago, mourip said:

So I am now very far behind and even skipped a step. Are folks still using NIC bridging as their default setup or have later discoveries superseded/eclipsed it?

Set up a bridge in audiolinux on my server here, my motherboard has the advantage of 2 ethernet ports clocked by an sclk-ex. NUC as endpoint. Works pretty well and is an improvement over endpoint connecting via my netgear gs105. Would like to try wireless connection next, to compare. would be nice if these latest NUC's had dual ethernet ports.

 

[LTG2010]

9 minutes ago, lmitche said:

At this point in time a Linksys RE9000 network extender connects one of it's 5gz radios to my main wifi router with a wired connection to my server

Hi Larry,

Is your server now the big i7 7700 machine or a second NUC?

Alec

[LTG2010]

20 hours ago, mikicasellas said:

CORSAIR Vengeance Performance SO-DIMM DDR4 2400 (PC4 19200) CAS Latency 16

Very good with NUC7C/P

[LTG2010]

6 hours ago, sunny_time_99 said:

The same with  lt3045 Regulators

In general yes with a lot of the DIY offerings, but they work very well in LPS 1.2.

[LTG2010]

24 minutes ago, austinpop said:

hen it works though ... aah!

Exactly, I found turning MQA support off calmed things down a bit but not totally.

With MQA support total nightmare.

[LTG2010]

14 minutes ago, Advieira said:

But if HMS buffers the bits and with powerful algorithms massively upsample the signal, all which is behind it have no impact anymore, clocks for sure, maybe ground loops not. Its make sense and thats what Rob Watts told us at Head-fi.

Buffering is not fool proof, there is no mechanism to check that all the bits have arrived correctly, without any distortion or variation to the signal. It just ensures the equipment is more immune to such distortion.

Taken to the extreme:

If you upsample cr&&p, you get upsampled cr&&p.

If you upsample from a cheap laptop/ poor power supply against good server you will hear the difference

[LTG2010]

9 hours ago, austinpop said:

• OPTIONS="-n sq-NUC -o hw:CARD=Interface -b 2097152:2097152 -a 52428800:4:: -D"

 

With how much RAM are you using this setting?

[LTG2010]

21 minutes ago, rickca said:

Can LMS/Squeezelite work in a server/streamer two box configuration like Roon server/endpoint?

Yes, LMS on Server either windows or linux, squeezelite on renderer.

[LTG2010]

17 minutes ago, rickca said:

Thanks.  So if I install LMS on my Windows 10 machine and I want my NUC to be the renderer, can I still leave an instance of LMS installed on my NUC? 

You can leave LMS on your NUC just don't activate it, just start up squeezelite on the NUC, the only running audio service on the NUC should be squeezelite and on the server LMS.

LMS is pretty straight forward to setup and will find squeezelite. Ipeng should communicate with any compatible server including LMS.

[LTG2010]

18 minutes ago, EnjoyTheMusicNow said:

With a Chord DAC (Qutest) and the Hugo M-scaler, does an SOtM USBultra still add value?

Your sms200Ultra already has the sCLK-EX clocking board in the the tx-USBultra, so I don't believe there is any added benefit.

[LTG2010]

3 hours ago, greenleo said:

Also, an NUC7i7DNBE powered by the SR7

The only problem for most of us is that the SR7 is unobtainable, only a lucky few can obtain one.

[LTG2010]

15 hours ago, Neal.Audio said:

What I'm really asking is for Roon users: Is volume control retained when using a homemade NUC renderer?

I have 3 options here: DSP Volume, Fixed Volume and Device volume.

[LTG2010]

24 minutes ago, bobfa said:

I think that this is DAC dependent.  Research the DAC to see if it has USB Volume control.  

That was my point, probably not made clear the NUC won't affect device / DAC volume.

[LTG2010]

4 hours ago, RickyV said:

Yep, I would like too know.

A comparison between Euphony and AudioLinux would be even better.

That would be useful, they both run on archlinux, so shouldn't in theory be very different sonically.

[LTG2010]

52 minutes ago, RickyV said:

Mmmm I don't know @bobfa sounded very enthusiastic about Euphony at first listen. We'll give him a few days.

I see I missed that, looking forward to @bobfa results.

[LTG2010]

1 hour ago, dctom said:

Considering building a new machine based on a supermicro X10SBA-L  running linux, using an external PS. Does anyone know if the built in cpu would have enough power to run HQplayer or jriver?

It can run both fine but no heavy DSD upsampling, so best to go for a bigger processor/ board, if you need that feature.

[LTG2010]

1 minute ago, valveboy said:

so, it's a case of "try it and see"?

 

 Then tell others your findings on here?

If you parallel LT3045's you lower the noisefloor tenfold, but if you them in series, which is the suggestion, the noisefloor could theoretically be reduced one hundred fold. 

But theories are theories and need to be tested, as regulators have their own sound signatures.

[LTG2010]

20 minutes ago, sandyk said:

Re " the noise floor could theoretically be reduced one hundred fold.  "

Michael Paschenda from MPaudio details, the effects of ripple reduction in his circuits and the theoretical benefit of placing them in series a lot of DIY users have had succes this way, reporting sound quality improvements. Its the noise present on the supply line and its ability to reject it by regulation. LT3045's have variable characteristics, regarding voltage drop and can also  heat up very quickly, so their is a benefit of a series circuit eg. placing 2 @ 1.5v drop as opposed to 1 @ 3v drop. Regarding their use in audio some manufacturers such as HDplex and Uptone audio are using them in their designs as well as the DIY suppliers so there seems to be a benefit which can ultimately be judged by listening but they seem to be very popular eg LPS 1.2.

I think the discussion was focussed on dual stage regulation and its benefits, adding a LT3045 circuit to an exisring power supply. Generally speaking I think dual regulation does increase the power supply rejection ratio and rectification noise rejection in a circuit. See for example Paul Hynes design for SR7 double regulation, which is at the top of his designs list.

 

[LTG2010]

On 3/20/2019 at 8:56 AM, RickyV said:

I have ideas of putting a neutron star on my nuc7i7 in the same case.

That would be the best, I agree with @sunny_time_99 the entire circuit is important especially the power supply.

I think you could use a single Neutron star clock board @25Mhz for the lan and system/usb 24/25mhz, I believe this is the way its done with the DS1 streamer, a single OCXO board.

Heres an alternative to the Neutron star and there is a fitting service if of interest, there are also some nice regulator and power supply designs.

https://www.fidelityaudio.co.uk/c4-low-jitter-uber-clock-fitted-3989-p.asp

[LTG2010]

12 minutes ago, RickyV said:

If the CPU clock is 24Mhz I can't supply it with 25Mhz, can I???

Yes, it depends on the chipset if its compatible, I think thats how its done on the DS1 and that uses a NUC7PJYH but is available with your board as an option. I think @austinpop would know more as he has/had one.

The neutrino should be a big improvement over the standard clocks but if the Neutron star has 2 outputs your good to go

Also Pink Faun have an OCXO clock board that might work for you.

[LTG2010]

20 minutes ago, mourip said:

21 minutes ago, mourip said:

Very nice. Question. Does the card automatically disable power input from the bus when the SATA input is powered? How about bus power ground? 

 

I dont think so the 'external' power is auxiliary, I would cut or just tape over the 12v line on the PCIE that would give a clean supply from the external power.