Do high-end streamers make an audible difference?

[reignofchaos]

There is a *huge* difference that I've experienced. On my main rig, I've tried the following streamers in increasing order of sound quality. In all cases, the software used was Roon.

 

1) Wiim Pro Plus - Just terrible sounding overall as a digital source - even with a linear power supply. Dynamics suck, sound is grainy and lacks resolution. Edges are blurred, timing is completely off.

2) Bluesound Node - Sounds about alright but it is slightly warm and some detail is definitely missing. However it doesn't have the speed and timing of the higher end streamers and feels a bit slow. Still a significant upgrade over Wiim

3) SMSL SD9 - More detailed than the above two and also seems to have much better timing and slam. This was tested using airplay mode as it doesn't support RAAT and can't be used as a native endpoint for Roon. So it could actually be better if used with other software.

4) Sonore Microrendu - This used to be my streamer for a long time and it sounded really good - near high end with very clean good resolution and nice dynamics. However it had hang issues and the software was unstable.

5) Lumin U2 Mini - This was in for a short duration so take this with a grain of salt. I found this highly rated streamer to sound not that great. It is exceptionally detailed but seems to have more low level noise than the others. I'm not sure if this was a faulty unit or just how it is. Sounded like power supply noise/buzzing.

6) SOTM SMS 200 - Exceptional for the money - very clean, very organic but not the most detailed.

7) Eversolo DMP A6 ME - So here's the issue with this one - the USB sounds fantastic but but the coax sounds meh. If you wish to use Coax or any other such interface, it is best to use a DDC (I used the SD9 as a DDC)

8. Opticalrendu + Ultracap 1.2 PSU - At the similar level as the 200 Ultra but a bit dry. Some may like this over the SOTM. However could not get this to work stably in my setup - Would always hang at startup and would boot maybe once in ten tries

9) SOTM SMS 200 Ultra Neo + SOTM PSU + TX USB Ultra Reclocker - My current streamer. Sounds really really good - can't really ask for more. Exceptional detail, dynamics, superb tonality. Unlike the rendu, this is exceptionally stable. Never required a reboot.

10) Holo Red - Easily the best one. Everything sounds similar to the SOTM other than one thing. It has a much wider and deeper sound stage. Also if you use gentooplayer, the custom kernels allow you to tune the sound the way you like. Also allows use of Diretta which improves sound over standard RAAT or NAA. I'll probably get this in future.

 

From this list, it should be pretty clear that cost of the streamer has almost nothing to do with the quality of output. It is a minefield and if one doesn't hear a few of them, they might land up on the wrong streamer.

Edited January 29 by reignofchaos

[rand129678]

49 minutes ago, Addicted to music said:

Digital receivers are immune to most noises such as RF, EMI etc  Unless the noise falls into the specification where the receiver identifies and responds  is rare.   And if it does to interrupt and cause errors, the receiver will know as the checksum indicate it's in correct.   As in previous posts, your DAC doesn't respond to EMI RF and even if the noise floor is increased it doesn't care as long as it sees legible data is what you will hear.

 

This isn't correct. A DAC can be susceptible to ground/RF/leakage currents via digital inputs...

This can measurably and sometimes audibly affect a DAC's analogue output.

 

3 ways to counter this:

 

1) Buy DAC which has ground isolated digital input. TOSlink always is. Ethernet always is as long as you don't ruin it with shielded ethernet cables. USB sometimes can be. Digital coax can be. WiFi input always is.

 

2) Buy a streamer which has ground isolated digital output (these do exist but my preference is #1 above).

 

3) Buy a digital isolator which sits in between streamer and DAC. For USB there is Topping HS02 or Intona Isolator.

 

So digital interface is not just about just 0's and 1's - they are electrical connections which can be subject to ground/RF/leakage currents and affect the analogue outputs. 

 

This can't be disputed and can be easily measured.

Fortunately there are solutions (as I listed above).

 

[Addicted to music]

6 hours ago, rand129678 said:

 

This isn't correct. A DAC can be susceptible to ground/RF/leakage currents via digital inputs...

This can measurably and sometimes audibly affect a DAC's analogue output.

 

3 ways to counter this:

 

1) Buy DAC which has ground isolated digital input. TOSlink always is. Ethernet always is as long as you don't ruin it with shielded ethernet cables. USB sometimes can be. Digital coax can be. WiFi input always is.

 

2) Buy a streamer which has ground isolated digital output (these do exist but my preference is #1 above).

 

3) Buy a digital isolator which sits in between streamer and DAC. For USB there is Topping HS02 or Intona Isolator.

 

So digital interface is not just about just 0's and 1's - they are electrical connections which can be subject to ground/RF/leakage currents and affect the analogue outputs. 

 

This can't be disputed and can be easily measured.

Fortunately there are solutions (as I listed above).

 

 

 

Do you ever read what you quoted?

 

Where is the digital noise that's effected the DAC in this pic?

The noise floor is down to below -130db!    Way beyond human hearing unless you have bat ears.  

If noise you  referring that can effect the electrical hygiene it would surely show up like you said.  That said you can have the best clock, PSU grounding arrangement possible but still manage to F__k it up because of bad implementation or a component has gone faulty.   Digital Fundamentals is precise, it's either on or off, zero or one, nothing in between, regardless of imaginary marketing FUD, it doesn't care or respond to anything inbtween.  

Please refer to @The Mad Scientist post before this 

Edited January 29 by Addicted to music

[rand129678]

49 minutes ago, Addicted to music said:

Where is the digital noise that's effected the DAC in this pic?

The noise floor is down to below -130db

 

It is always system dependent.

These are electrical systems not just living in the world of 1's and 0's 

Everything ends up in the analogue domain...

 

I actually do measurements myself and I can take a great DAC (with no digi input isolation) and use RCA outputs and setup the measurement electrical layout in a way to ruin the analogue output.

 

This is how your favourite online reviews will properly test the effectiveness of digital isolators...

 

Using balanced XLR connections helps a lot but a lot of people still use unbalanced RCA connections.

 

My previous post shared solutions.

 

To just look at 0's and 1's and not acknowledge the electrical nature of our systems is not accurate. Everything is electrically connected and ends up in the analogue domain.

 

Anyway this is my last post in this thread.

 

All the best.

 

Edited January 29 by rand129678

[NonPlayableCharacter]

10 hours ago, Addicted to music said:

Because 2 states: zero and ones are used its known as binary operation.  The 2 state zero and one disregards any levels in between, so it doesn't care what is between unless its a zero or a one value.  By implementing this it disregards any values in between to a certain specification.   

 

This holds true for 100Mbps Ethernet transmission but not for the more common Gigabit Ethernet standard which has 5 states possible using 5 different voltages to signal accordingly.

[Addicted to music]

4 hours ago, rand129678 said:

 

It is always system dependent.

These are electrical systems not just living in the world of 1's and 0's 

Everything ends up in the analogue domain...

.

 

4 hours ago, rand129678 said:

 

 

 

Using balanced XLR connections helps a lot but a lot of people still use unbalanced RCA connections.

 

 

 

 

 

Here is a pic of RCA output of the same DAC.  The noise level is literally the same below -130db.   At this level its clean of anything FUD that you can dream up.  And its a under $400 DAC headphone amp!   At this level its absolutely clean.   And where is the noise at the analog output of the DAC you say?  Like I said, either you have a bad implementation or a faulty component if you are measuring junk at the output in the audio human hearing range.

The distortion is well under 0.01%  Jitter is under -130db too!

 

Edited January 30 by Addicted to music

[Addicted to music]

1 hour ago, El Tel said:

 

This holds true for 100Mbps Ethernet transmission but not for the more common Gigabit Ethernet standard which has 5 states possible using 5 different voltages to signal accordingly.

 

It gets to a point where you need to implement other markers in the carrier to get that speed.  There's nothing new here,  Companies that produced faximile machines were pushing 30,000bps down a standard 600ohm standard copper phone cable when Telecom can only guarrantee 9600bps.

[rand129678]

16 minutes ago, Addicted to music said:

And where is the noise at the analog output of the DAC you say? 

 

You are too focused on only 0's and 1's and you are not understanding the point about electrical layout.

 

Is your home system layout like Amir's measurement layout?

Is mine? Is mine like yours? Is yours like the OP's? Is mine? 

As I already wrote:

4 hours ago, rand129678 said:

I actually do measurements myself and I can take a great DAC (with no digi input isolation) and use RCA outputs and setup the measurement electrical layout in a way to ruin the analogue output.

 

This is how your favourite online reviews will properly test the effectiveness of digital isolators...

 

OK that is actually my last post here lol

Edited January 30 by rand129678

[NonPlayableCharacter]

43 minutes ago, Addicted to music said:

 

It gets to a point where you need to implement other markers in the carrier to get that speed.  There's nothing new here,  Companies that produced faximile machines were pushing 30,000bps down a standard 600ohm standard copper phone cable when Telecom can only guarrantee 9600bps.

 

Some may be unfamiliar with how networking functions in the majority of domestic installations and the fact that binary voltages have been superseded. As you showed, it is easy to explain the principles based on old school tech.

 

Even simple and easily understandable mechanisms used at low transmission speeds morph into much more complex mechanisms as we push the speeds up (for example, Manchester Encoding on 10Mbps as a means of embedding the timing signal into the transmission has become a hugely complex nested encoding process using 8B1Q44 and 4D-PAM5 at 1Gbps). Notwithstanding, the more complex mechanisms and checksum processes still result in bit-perfect data transfer.

 

But now we're way off topic even if this side-bar was intended to help a member identify flaws in their understanding caused by uninformed or misinformed sources.

Edited January 30 by El Tel
Edited for clarity

[PKay]

On 27/01/2025 at 6:58 PM, The Mad Scientist said:

A computer which most of us own some form of and a simple USB mic can get you started with a free download of REW. If you're a Roon user, you can add convolution filters via Muse. There are options like CamillaDSP. Also free and available on Linux, Mac and Windows. You can even run it on a Raspberry Pi.

 

It needn't cost much at all and the results are profound.

Are there any REW guidelines/tutorials you can recommend for a rank novice? Also any microphones that are reasonably priced?

Edited January 30 by PKay

[BugPowderDust]

2 hours ago, PKay said:

Are there any REW guidelines/tutorials you can recommend for a rank novice? Also any microphones that are reasonably priced?

 

get a minidsp umik-1 calibrated micu

Edited January 30 by BugPowderDust

[The Mad Scientist]

3 hours ago, PKay said:

Are there any REW guidelines/tutorials you can recommend for a rank novice? Also any microphones that are reasonably priced?

Here's another once you've grasped the basic setup:

It's a little more taxing than the very basic guide, but I managed this as a newbie to REW. There's a comprehensive pdf download in the description which I worked through along with the video. My results were very good with this approach.

 

 

Edited January 30 by The Mad Scientist

[Addicted to music]

6 hours ago, El Tel said:

 

Some may be unfamiliar with how networking functions in the majority of domestic installations and the fact that binary voltages have been superseded. As you showed, it is easy to explain the principles based on old school tech.

 

If you do digital fundamentals, I'm sure they will tech the old school basics so you have an understanding before they dive into modern complex systems that defy the maths and electrical limitations.  i find it fascinating how they get around limitations and yet the implementation gets the job done.

 

6 hours ago, El Tel said:

 

 

Even simple and easily understandable mechanisms used at low transmission speeds morph into much more complex mechanisms as we push the speeds up (for example, Manchester Encoding on 10Mbps as a means of embedding the timing signal into the transmission has become a hugely complex nested encoding process using 8B1Q44 and 4D-PAM5 at 1Gbps). Notwithstanding, the more complex mechanisms and checksum processes still result in bit-perfect data transfer.

 

 

And this is the thing,  with faster speeds comes higher sensitivity to whatever noise  that has the potential to interrupt transmission,  Yet the engineers  would have taken this into consideration which is what many don't understand and keep on drumming up  about "interferences "  where a marketed product "fixes" the issue; a solution looking for a problem.   As you said " bit-perfect data transfer"   even at gigabit and above transfer!  More than enough for any streamer or DAC.

[POV]

10 hours ago, Addicted to music said:

Here is a pic of RCA output of the same DAC.  The noise level is literally the same below -130db.   At this level its clean of anything FUD that you can dream up.  And its a under $400 DAC headphone amp!   At this level its absolutely clean.   And where is the noise at the analog output of the DAC you say?  Like I said, either you have a bad implementation or a faulty component if you are measuring junk at the output in the audio human hearing range.

The distortion is well under 0.01%  Jitter is under -130db too!

 

10 hours ago, rand129678 said:

You are too focused on only 0's and 1's and you are not understanding the point about electrical layout.

 

Is your home system layout like Amir's measurement layout?

Is mine? Is mine like yours? Is yours like the OP's? Is mine?

 

 

This is I believe (another) cross-purposes discussion. The point that @Addicted to music is making is 100% right, and I don't agree that this is a "0's and 1's" focus, but rather an accurate observation that well-designed DACs are broadly impervious to distortion of the digital signal, as this is inherent in the design. In fact, for distortion of the digital signal to manifest in audible issues, it will be gross error issues (i.e., skipping, stuttering, dropouts, etc.) rather than some kind of unquantifiable deleterious effect on audio fidelity.

 

The broader point that @rand129678 is making, though, that it is technically feasible that ground loops, RF interference, and/or leakage currents could introduce noise that materially decreases signal-to-noise ratio via the cable (or cable shielding) is also correct. An important distinction here, though, is that the audibility threshold of signal-to-noise ratio is extremely well understood and easily measured. If in fact the signal-to-noise ratio is decreased to the point that noise is in the audible range, it will absolutely manifest as audible noise or hum, and not as some type of unquantifiable deleterious effect on audio fidelity.

 

Where I personally feel these threads/discussions stray into very clear misinformation is when they start to suggest that:

• there may be something going on that we do not understand or know how to measure; and/or
• that there may be some kind of unexplainable and/or unquantifiable deleterious effect to audio fidelity happening, or that we should just believe that reduction of the noise floor (even) further below the audible range somehow improves audio fidelity when it verifiably cannot. Noise is either audible, or it isn't. These mysterious "veil of blackness" that gets described without ever being explained or quantified is, I think, fairly and squarely a red herring.
• The introduction of another component and associated connections is somehow going to defy very well understood electrical/electronic/digital engineering principles.

 

I also think it's a somewhat reductionist approach to discussion to accuse people that question some of the tenuous explanations provided for totally subjective observations that defy well-understood and studied concepts as trying to mount a 'bits are bits' or 'it's just 0's and 1's' argument when in general folks are simply engaging in the subject matter discussed and seeking clarity on claims that are impossible to substantiate and frankly very difficult to accept.

 

I continue to hope that we can have discussions without them turning personal and people taking offence to informed contrary views.

Edited January 30 by POV

[Addicted to music]

8 hours ago, POV said:

 

 

 

This is I believe (another) cross-purposes discussion. The point that @Addicted to music is making is 100% right, and I don't agree that this is a "0's and 1's" focus, but rather an accurate observation that well-designed DACs are broadly impervious to distortion of the digital signal, as this is inherent in the design. In fact, for distortion of the digital signal to manifest in audible issues, it will be gross error issues (i.e., skipping, stuttering, dropouts, etc.) rather than some kind of unquantifiable deleterious effect on audio fidelity.

 

The broader point that @rand129678 is making, though, that it is technically feasible that ground loops, RF interference, and/or leakage currents could introduce noise that materially decreases signal-to-noise ratio via the cable (or cable shielding) is also correct. An important distinction here, though, is that the audibility threshold of signal-to-noise ratio is extremely well understood and easily measured. If in fact the signal-to-noise ratio is decreased to the point that noise is in the audible range, it will absolutely manifest as audible noise or hum, and not as some type of unquantifiable deleterious effect on audio fidelity.

 

Where I personally feel these threads/discussions stray into very clear misinformation is when they start to suggest that:

• there may be something going on that we do not understand or know how to measure; and/or
• that there may be some kind of unexplainable and/or unquantifiable deleterious effect to audio fidelity happening, or that we should just believe that reduction of the noise floor (even) further below the audible range somehow improves audio fidelity when it verifiably cannot. Noise is either audible, or it isn't. These mysterious "veil of blackness" that gets described without ever being explained or quantified is, I think, fairly and squarely a red herring.
• The introduction of another component and associated connections is somehow going to defy very well understood electrical/electronic/digital engineering principles.

 

I also think it's a somewhat reductionist approach to discussion to accuse people that question some of the tenuous explanations provided for totally subjective observations that defy well-understood and studied concepts as trying to mount a 'bits are bits' or 'it's just 0's and 1's' argument when in general folks are simply engaging in the subject matter discussed and seeking clarity on claims that are impossible to substantiate and frankly very difficult to accept.

 

I continue to hope that we can have discussions without them turning personal and people taking offence to informed contrary views.

 Injecting noise into any system to find its limitation when it will cease operations will usually be given by the manufacturer so you know your operating limitations.  Most products would have been subjected to extreme testing especially when you use it with public telecommunication infrastructure as it has to pass certain accreditation.  To actually inject noise and then say the product is not immune to post to be augmented without data is what I’ll question every time.   At what point did the level of noise interrupted the operation?   Under normal operating conditions,  cloud service to streamer,  just remember the data is travelling into numerous media including fibre optics, coax and Wifi to get to that streamer,  are you hearing the noise obtained through these  digital transmission media?  The answer will be no if the product is under normal operation, 
 as it’s bit perfect.   Specifications such as Cat5e should not be more than 100m run,  so before it gets to 100m you set up a repeater (switch) etc  or if the operating conditions saids for best results the environment should be 10-25 Celsius,  and you have problems operating at under or above this temperature range is very predictable.  Anyone can test the limitation of any device with obvious predictive results.   
When I go out for walks and live stream,  the data is transmitted via air to my phone and then Bluetooth to AirPods,  and I can tell you it’s seamless and all the  musical details are there without any noise accumulated from any of these transmission media. 

Edited January 30 by Addicted to music

[sir sanders zingmore]

Thread was briefly on post approval.

Re opened now. Please take the time to read or reread the conditions under which we reopened it last time

 

 

[ENIGMA]

 

[playdough]

Late to the party,,,Most can hear differing character between 2 types of CD players (noted this in the past myself) running purely digital to the system DAC.

Would expect the same with streamers.....?

[The Mad Scientist]

I think I've said all I have to say on the matter of digital sources, the robustness of digital data transfer and the necessary controls to eliminate subjective bias. I have a fully active x-over with room correction to work on where the measurable (and audible) differences truly are night and day, so I'm bowing out of this discussion.

Edited January 30 by The Mad Scientist

[The Mad Scientist]

7 minutes ago, playdough said:

Late to the party,,,Most can hear differing character between 2 types of CD players (noted this in the past myself) running purely digital to the system DAC.

Would expect the same with streamers.....?

Again, precise level-matching needs to be incorporated within any comparison. Attenuation in the digital domain is a factor. Not all digital sources max out at 0 dBFS.

 

That's the last word from me in this topic.

Edited January 30 by The Mad Scientist

[aussievintage]

10 minutes ago, playdough said:

Most can hear differing character between 2 types of CD players (noted this in the past myself) running purely digital to the system DAC.

 

Quite a big claim.  Any research backing it up?

[playdough]

7 minutes ago, The Mad Scientist said:

Again, precise level-matching needs to be incorporated within any comparison. Attenuation in the digital domain is a factor. Not all digital sources max out at 0dBFS.

 

That's the last word from me in this topic.

Fair call. THe Mrs and I were discussing 2 different CD players we have been testing this week, both feeding the same DAC and the character difference between the players. Same speakers/DSP, same spdif input.

 

3 minutes ago, aussievintage said:

Quite a big claim.  Any research backing it up?

Id be happy to let you discuss that with her,,,,,no problem. She can do the cable and CD swapping...

 

CD Spinners, an OPPO and A Fiio portable, both with spdif output

Edited January 30 by playdough

[aussievintage]

Just now, playdough said:

Fair call. THe Mrs and I were discussing 2 different CD players we have been testing this week, both feeding the same DAC and the character difference between the players. Same speakers/DSP, same spdif input.

 

Id be happy to let you discuss that with her,,,,,no problem. She can do the cable and CD swapping...

 

That truer statement is probably that most people haven't even heard " 2 different CD players we have been testing this week, both feeding the same DAC and the character difference between the players".

 

and under test conditions there would be little statistically significant variation from 50% correct.

[playdough]

1 hour ago, aussievintage said:

and under test conditions there would be little statistically significant variation from 50% correct.

Yes it's not my fault she likes the sound of the Fiio, over the Oppo,,,,,,,,,,I leave it at that.  

This is what happens if you install, calibrate a system and acoustically treat the zone in the same room as the kitchen,,,,

Edited January 30 by playdough

[Almaz]

I laughed when you said she can do the CD and cable swapping.

Lost count of how many times over the years I asked my wife to come and listen and tell me what she thinks, or can she hear the difference. I think she has wised up and just says “ sounds good Love “ Or I might start talking tech,

and she give me that strange look. What would I do without her.