Jump to content
IGNORED

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


Message added by The Computer Audiophile

Important and useful information about this thread

Posting guidelines

History and index of useful posts

Most important: please realize this thread is about bleeding edge experimentation and discovery. No one has The Answer™. If you are not into tweaking, just know that you can have a musically satisfying system without doing any of the nutty things we do here.

Recommended Posts

19 hours ago, austinpop said:

 

So true! If you leave it play through a track, or better yet, an album, it mostly works. Although even there, on occasion it seems to forget how to do gapless playback, which gets annoying with classical pieces. 

 

If you do want to switch mid-song, it helps to hit pause first, wait a spell, then play the next song.

 

If you go full ADD on it, and start switching tracks and sample rate midstream - it will hurt you!

 

 

Yes, it's still a construction zone! Hard hats mandatory.

 

It's funny how experiences can differ; probably as I am using a straightforward LMS/Squeezelite-R2 solution.

 

I find the setup works very stably, with excellent sound quality. I leave the NUC & LMS server on 24/7 and when I return I simply press play - 95% of the time. I have had to reboot my LMS server once.

 

WRT the squeezelite settings:

I have added the -a 50MB switch.

 

I find that the sound is neutral and the bass VERY controlled. For instance, if I listen to a Hans Zimmer soundtrack he uses a lot of deep bass, both as drums and in pulses. These are now resolved so that other details are now clearly heard. But, the sound is neither warm nor toppy, hence my use of neutral.

 

This does not mean that everything sounds great. For instance Queens 'Night at the Opera'. I have two versions on my system, a vinyl rop of the original album, recorded from my LP12 at 9624, and a DVD-A rip of the stereo 9424 track. The former sounds superb. The latter sounds perhaps a tad better in terms of the bass and drums, but some of the none-Mercury sung tracks are rather too sharp.

 

M

Link to comment
2 minutes ago, MrUnderhill said:

It's funny how experiences can differ; probably as I am using a straightforward LMS/Squeezelite-R2 solution.

 

I find the setup works very stably, with excellent sound quality. I leave the NUC & LMS server on 24/7 and when I return I simply press play - 95% of the time. I have had to reboot my LMS server once.

 

I should have made this clear... I have no issues if I run LMS.

 

The problems occur only when using Roon. So the bugginess is in the Roon implementation of LMS emulation, and Room have already declared they will not support software squeezebox players.

 

That said, many of us do not want to give up the Roon user experience, to which we have gotten accustomed, and which is so much nicer than the LMS experience.

 

I want to have my cake and eat it too!

Link to comment
5 hours ago, LTG2010 said:

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

 

What you are saying is streaming audio from Tidal (per example) never can be "perfect" because I'm sure Tidal servers, Internet servers, doesnt have LPSU, clocks, etc. We never will have control from what is happening before our modem. And I can told you for sure, I can streaming Tidal, some tracks, mainly MQA, which are crystal pristine sound (with all limitations of my system).

Link to comment
2 minutes ago, Advieira said:

 

What you are saying is streaming audio from Tidal (per example) never can be "perfect" because I'm sure Tidal servers, Internet servers, doesnt have LPSU, clocks, etc.

 

Please stay on topic. This thread is about direct listening experiences. We're not debating the "how does it work" here. For guidance, please look at the pinned post at the top of this page, and the first section of the first post of this thread.

 

To your point, there are plenty of Rob Watts fans on this thread, and plenty of Blu2, HMS, Dave, and TT2 owners. The fact of the matter is that the upstream chain definitely matters. The mechanisms are not yet fully understood. While most of us would celebrate a DAC that was 100% immune to the quality (or lack thereof) of the upstream chain, we have yet to find one. Yes that includes the Chords.

 

Please focus your participation here on reporting your listening experiences, or reading about others'. There are other threads to debate (endlessly) the mechanisms at work. 

Link to comment
45 minutes ago, LTG2010 said:

With how much RAM are you using this setting?

 

8 minutes ago, MrUnderhill said:

I have 8GB

 

Yes, 8GB. You can't allocate 4GB of buffer space unless you have at least 8GB of RAM on the endpoint. 

 

As part of this, I trimmed down the AL footprint as much as I could. I posted these in my Tips and Tricks post on the AL troubleshooting thread.

Link to comment

Great post Leo!

 

Having been listening to my DNBE your description resonates. I asked my girlfriend to sit and listen to a couple tracks yesterday evening as I'd finally got my DNKE server powered by SPS-500 and DNBE powered by LPS1.2.

 

It's an audiophile cliché but it happened. We both looked at each other at one point and said we"d never heard that before. Could it be expectation bias. I guess so. But the detail is outright spooky and I'm coming from a Zenith SE which ain't exactly a slouch in the details dept.

 

Anyway. Your finding that the DNBE sounded better on stock power to the 7C is very interesting. (Borderline heresy for this forum! ?)

 

One thing you didn't mention. Are both units in the original cases or better fanless cases?

 

I ask because I've got my DNBE running on an LPS1.2 but there are audible pops as we discussed. The pops were every 5 to 20 seconds running extreme mode. Running normal mode they are intermittent. Enough to prevent enjoying the music. Well no. Enough that I'll probably promote the SPS-500 to endpoint duties even if that means using a lower spec psu on the server. Well yes.

 

I've disabled everything I can in the BIOS and am running only one core. Not sure if the pops are because it's drawing too much power still or if it's a Roon thing. Beyond that let's take this part of the convo over to the support thread.

 

Cheers,

Alan

 

 

Synergistic Research Powercell UEF SE > Sonore OpticalModule (LPS-1.2 & DXP-1A5DSC) > EtherRegen (SR4T & DXP-1A5DSC) > (Sablon 2020 LAN) Innuos PhoenixNet > Muon Streaming System > Grimm Audio MU1 server > (Sablon AES) Mola Mola Tambaqui DAC > PS Audio M1200 monoblocks > Salk Sound Supercharged Songtowers

Link to comment
1 hour ago, greenleo said:

7CJYH and i7DNBE as an NAA Endpoint Comparison

 

I want to a share a comparison between 7CJYH(7C) and the i7DNBE(7D) as endpoints running NAA fed by HQP.  In short the 7D sounds better.  Without comparison, 7C is already very good, so good that I can live with it for a very long time.  The following are the details.

 

Physical Setup:

Both of the audio PCs (NAA endpoints) 7C and 7D were running AL headless V6 using the same USB Disk

Both the PCs using exactly 4GB RAM with Kingston RAM with the same model number.

Same USB cable and LAN cable were used.

Control PC (server) running HQP, LXQt v3.02 sent files to the NAA.

 

Power:

7CJYH powered by LPS-1.2

 i7DNBE powered by the SMPS come with 7CJYH, yes an inferior PS


 

How I test:

The LAN cable and the USB cable connected to the 7CJYH were unplugged and then plugged into the i7DNBE and vice/versa for A/B.  The time it took including the sw reset in the Control PC took less than 2 minutes.  This was important because my audio memory is short and nobody would help me for a blind test.  Two sets of A/B were performed, 1st to get an initial impression/verdict and the 2nd to confirm/falsify the impression/verdict.  Two tracks, one live vocal and the other and Tv track, that I'm very familiar with were used for the comparison.

 

When I just finish building the 7D, I listened a few tracks through it and found it sounds better.  Hence I start the audition with the following sequence.

 

1. Start with 7C

     Live vocal track played for 1 min (which I found lesser SQ immediately)

     Tv track played for 1 min (sounded great).

2. Switch to 7D ASAP. 

     Tv track played and jotted down my feeling (I felt it sounded even better)   

     Live vocal track played.  Concrete, tangible differences were found immediately ("conclusion" verified).

3. Repeat 1 and 2 and draw/falsify conclusion (if any)

 

SQ wise:

Some "concrete" evidence:

For the Tv track, the sound of i7DNBE was even more crispy.  This was subjective and my honest feeling is reported here.

 

For the live vocal track, the difference was/is undeniable: I could hear the throat movement (sipping of saliva?) of the signer at 0:39 and 0:41 using the 7D effortlessly.  Those movements may be observed through 7C if one pays attention.  Another tongue movement at 0:51-0:52 could be heard through 7D easily whereas barely noticeable through the 7C.  Please note that these observations were/are something that I disliked/dislike very much because that's due to the work of bad sound engineer and I rather these imperfections weren't in the track.  There should be no placebo effect due to expectation bias.

 

To quote an old statement: the better the system, the more the defects (are revealed), very true.  

 

Atmosphere wise:

No words on the Tv track as I don't know the original sound. 

 

The atmosphere of the live vocal track just sounded truer when compare with my feelings of attending concerts.  I've never attend the concert of the live track but I've attended concert held in the same venue.  Hopefully, this give a better reference of the following sentences.  The being there feeling was stronger/realer.  I would term it as it sounds denser (or higher density as I read from other forums), not sure if that's what others mean the 7C sounds thinner comparatively.  I find using audio vocabulary very difficult here and hope that the readers may perceive what I mean.  The language barrier upon intangible descriptions is certainly highlighted here.

 

Conclusion:

7C is great.  7D is definitely better.  I was surprised that I could tell the differences given that 7D was with an SMPS.  Whether the difference is large enough for the price difference, only an audition and personal value may tell.  If I can afford, I would go for the i7DNBE for the peace of mind and its low cost in this hobby.

 

Help Please: 7D, not mine, will go in 2 days.  If I may, I would like to hear its SQ using LPS-1.2 before it goes but so far couldn't.  For anyone who's done it before, I'd be very grateful if you may tell me how to do it.  Thank you in advance.

Appreciate your taking the trouble to provide this feedback. One observation:

You mention you just finished building the 7D. My experience is that new components will change several times during their initial burn-in. A very excellent LPS I was running-in exhibited several different characteristics before it finally settled down. At one stage it sounded quite overhyped, which on a short listen sounded more detailed but after 30 minutes or more became fatiguing. 2 days later this characteristic was completely gone.  If your 7D wasn’t run in I would expect it to sound different after a few days (in the case of my LPS, weeks)

Link to comment
26 minutes ago, Blackmorec said:

Appreciate your taking the trouble to provide this feedback. One observation:

You mention you just finished building the 7D. My experience is that new components will change several times during their initial burn-in. A very excellent LPS I was running-in exhibited several different characteristics before it finally settled down. At one stage it sounded quite overhyped, which on a short listen sounded more detailed but after 30 minutes or more became fatiguing. 2 days later this characteristic was completely gone.  If your 7D wasn’t run in I would expect it to sound different after a few days (in the case of my LPS, weeks)

Thank you Black.

 

While the 7C has been here for quite a while powered by the LPS-1.2, the 7D was done the night before.  As usual, I played a few songs to make sure it worked.  Then it's left burning in headless extreme mode continuously.  In that mean time, I tired to power the 7D using the LPS-1.2 but failed.  Can't wait any longer as it must go soon.

 

LPS-1.2 is the best and the only LPS I have.  Hence no complete fair footing may be provided, what a shame!  The 7D bested the 7C was highly unexpected and I urge other CAers to have a go and report.

 

I'm still trying to power the 7D with all the cores on.  Given that the thread's current opinion is the number of cores matters and provides better SQ.  Only with the same number of cores on will the comparison be apple and apple.

 

What I'm thinking now is to underclock the RAM and/or underclock the Core speed to control the TDP.  Did google but find no clues on NUC7i7DNBE.  Anybody knows please chime in.

Link to comment
26 minutes ago, greenleo said:

7D in Akasa Plato X7D

 

 

My second Plato X7D arrives today for the DNKE and the weather is lousy this weekend here in the UK. So perfect timing as I'm going to try and compare the Zenith SE to the NUCs.

 

Also want to try both with and without the TX-USBUltra.

 

Cheers,

Alan

Synergistic Research Powercell UEF SE > Sonore OpticalModule (LPS-1.2 & DXP-1A5DSC) > EtherRegen (SR4T & DXP-1A5DSC) > (Sablon 2020 LAN) Innuos PhoenixNet > Muon Streaming System > Grimm Audio MU1 server > (Sablon AES) Mola Mola Tambaqui DAC > PS Audio M1200 monoblocks > Salk Sound Supercharged Songtowers

Link to comment
On 12/13/2018 at 6:19 PM, austinpop said:

Hey all,

 

I've reorganized the 1st post and cleaned it up. Please take a look and encourage others to as well. Suggestions welcome.

Thanks Rajiv, amazing!

Topaz 2.5Kva Isolation Transformer > EtherRegen switch powered by Paul Hynes SR4 LPS >MacBook Pro 2013 > EC Designs PowerDac SX > TNT UBYTE-2 Speaker cables > Omega Super Alnico Monitors > 2x Rel T Zero Subwoofers. 

Link to comment
2 hours ago, BigAlMc said:

 

 

My second Plato X7D arrives today for the DNKE and the weather is lousy this weekend here in the UK. So perfect timing as I'm going to try and compare the Zenith SE to the NUCs.

 

Also want to try both with and without the TX-USBUltra.

 

Cheers,

Alan

Can't wait your report Alan.  But I guess you need a SR7 for the DNBE endpoint to outplay the SE, HaHa.  I must say that I very much hope that I'm wrong as I can't afford the SE.  I'm planning on a 7CJYH as the server and the DNBE as the endpoint, too ambitious I suppose. 

Link to comment
3 hours ago, greenleo said:

Thank you Black.

 

While the 7C has been here for quite a while powered by the LPS-1.2, the 7D was done the night before.  As usual, I played a few songs to make sure it worked.  Then it's left burning in headless extreme mode continuously.  In that mean time, I tired to power the 7D using the LPS-1.2 but failed.  Can't wait any longer as it must go soon.

 

LPS-1.2 is the best and the only LPS I have.  Hence no complete fair footing may be provided, what a shame!  The 7D bested the 7C was highly unexpected and I urge other CAers to have a go and report.

 

I'm still trying to power the 7D with all the cores on.  Given that the thread's current opinion is the number of cores matters and provides better SQ.  Only with the same number of cores on will the comparison be apple and apple.

 

What I'm thinking now is to underclock the RAM and/or underclock the Core speed to control the TDP.  Did google but find no clues on NUC7i7DNBE.  Anybody knows please chime in.

What are your bios settings? How much memory are you running? What is connected to the i7? What is in the m.2 slot? Do you have an Iso Regen?

 

The vbus current requirements of a DAC vary so an ISO Regen isolates the NUC from a high current consumption DAC.

 

Warning - there are variations of behavior between boards, and therefore likely not every board will run on the lps1.2 in quad core mode.

 

 

Pareto Audio aka nuckleheadaudio

Link to comment
2 minutes ago, lmitche said:

What are your bios settings? How much memory are you running? What is connected to the i7? What is in the m.2 slot?

BIOS:

- only 3 USB enabled, 2 at the front USB 2.0 for booting and keyboard, 1 at the back USB 3.0 for connecting to the DAC.

 

- Most onboard device disabled except the LAN.

 

- Things like HDMI CEC and something like s3/s4/s5 on wake left at default value.  It seems that they are boot related hence I dare not to touch it.

 

- RAM frequency set to 16x

 

 

Memory:

4GB RAM Kingston CL17 RAM sued.

 

What is connected:

Only LAN cable to the control PC, USB cable to the DAC and the USB Headless disk are connected.

 

Nothing in the m.2 slot

 

Regards.

Link to comment

How does your DAC use vbus power?

 

With three lps1.2s here, one powers the i7dnbe, one the ISO Regen, and one injects power into the DAC via the vbus.

 

I always use an ISO Regen with my vbus powered DAC, so perhaps that is why the lps1.2 powers the i7 board successfully.

 

Have you tried a different USB drive? Current consumption varies greatly here as well. Smaller is better of course.

Pareto Audio aka nuckleheadaudio

Link to comment
2 minutes ago, lmitche said:

Does your DAC take vbus?

 

Have you tried a different USB drive? Current consumption varies greatly here as well. Smaller is better of course.

vbus is used as a handshake.  My USB is an 8-Yr 16GB Corsair Voyager.  Any recommendation on the USB Disk?  

Link to comment

Create an account or sign in to comment

You need to be a member in order to leave a comment

Create an account

Sign up for a new account in our community. It's easy!

Register a new account

Sign in

Already have an account? Sign in here.

Sign In Now



×
×
  • Create New...