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

[Poldi]

I hope you guys are gonna be patient with a slow-minded one like me

 

I spent the last few evenings reading the most important posts of this massive thread (great work with the index!) which was still quite a long read.

 

So this is what I got:

So regarding the NUCs it is advised to follow a two-box-model with a low-power endpoint and a high-power server.

 

And regarding endpoints several people proved that using a NUC with more power results in better sound. The same was proven for NUCs as a server.

 

Regarding endpoints the following models were mentioned very often and ranked according to sound quality: NUC7CJYH < NUC7PJYH < NUC7i7DNBE

Is there a reason why these specific models were advised? Or can you just buy any NUC7-Board and the more powerful it is the better for the sound? (I am asking because some of the models mentioned are already sold out in my country)

 

And regarding server there is just the advice to buy a NUC with as much power as you can, right? (especially if you want to DSP-stuff)

 

Did anyone do comparisons between NUC7 and NUC8 boards regarding sound quality? I suppose there won't be a big difference but you never know...

[Poldi]

Thanks all of you for this helpful input!

 

Then I'll try to stick to one of the models mentioned for the endpoint.

 

And regarding server: As my server will also be standing in the living room I don't want to have a fan spinning if I can avoid it. And it would be easier to get a fanless case for a 10W or 15W TDP NUC and I would avoid thermal throttling.

 

In that case the NUC7i7DNBE would seem suitable. But is it strong enough for room correction? I haven't tried this so far but would like to do so in the future. Also it only has one ethernet port so bridging won't be possible. Hm... after reading the reports about bridging, I wouldn't want to miss it really.

[Poldi]

Many thanks @Bricki! I wasn't aware of this possibility.

 

Regarding the server I have one more remark:

Although I found several posts trying to explain the reason why NUCs are so special regarding sound quality, I got the impression that this applies especially to NUCs as endpoints, right?

 

Because according to this post (and other posts) of @romaz: https://audiophilestyle.com/forums/topic/30376-a-novel-way-to-massively-improve-the-sq-of-computer-audio-streaming/?page=409&tab=comments#comment-892124

 

... he got very good results using non-NUC-servers together with NUC-endpoints. And his guess seems to be that having more cores and (perhaps a bigger smart cache) in the server has a positive impact on the sound. So for the server side it doesn't have to be a NUC as long as you can run Audiolinux on it in RAM?

Because if that is true, people who (like me) want to have a fanless server could also look for other solutions than NUC.

 

 

[Poldi]

@rickca: Is thermal throttling the same as "Thermal Monitoring Technologies"? If so then the 15W TDP  i7-8650 also has thermal throttling:

https://www.intel.com/content/www/us/en/products/processors/core/i7-processors/i7-8650u.html

 

Or is it more unlikely for a 10W /15W TDP design to experience thermal throttling because it doesn't get as hot as a 28W TDP design? Hope it's not a stupid question to ask.

 

and thanks @the_doc735 for your recommendation!

[Poldi]

Hey @rickca thanks man for explaining!

 

So that means either stick with the 10/15W models.

 

Or if you want more power (and the findings of @austinpop and @romaz and others suggest the more power/cores the server has the better) you need server with a fan and/or a good enclosure and/or bigger dimensions than the NUC (to better dissipate the heat) to avoid thermal throttling.

[Poldi]

thanks again @rickca!

 

The server @bobfa is building also looks interesting. Can't wait to hear his results. Although his built is probably too hard for me as a beginner to duplicate.