Do all STREAMERS sound the same when used with the same external DAC?

[clearnfc]

Hello. Let’s suppose you have the same app (e.g. TIDAL) being played on a cheap PC, a BluSound Node 2i, an Aurender N10 , a NAD M10 or any other streamer in between. If all those streamers are connect the same way (e.g. optical or usb) to the same external DAC, would the sound be exactly the same? I ask it because as of now I'm using a Node 2i as a streamer with an external DAC but was wondering if there are "better" streamers that could improve the sound quality I'm getting now? Thank you!

Just saw this old thread. The answer is no....

Like amps, speakers etc.... the sound between DACs do vary a little.... So, its not exactly the same (exactly = 100% the same). Also to highlight that no 2 speakers/amp/dac etc... are exactly the same. They are all slightly difference (within a certain tolerance). The sames goes for everything.

 

[Bamyasi]

Just saw this old thread. The answer is no....

Like amps, speakers etc.... the sound between DACs do vary a little....

DACs, maybe in some cases. Digital-only streamers, no.

 

[srkbear]

DACs, maybe in some cases. Digital-only streamers, no.

With DACs I expect differences in distortion and noise specific to errors introduced during the DA conversion process. These are quantified during Amir’s measurements and are reflected in his performance ratings.

With streamers, especially those outputting the signal via asynchronous USB, the DAC clock should be the master, and the digital stream should accumulate in the buffer in the input receiver, obviating any clock issues in the streamer itself. So any claims from manufacturers about precision clocks and “anti-jitter” technology justifying an exorbitant price point make zero sense to me.

 

[srkbear]

Just saw this old thread. The answer is no....

Like amps, speakers etc.... the sound between DACs do vary a little.... So, its not exactly the same (exactly = 100% the same). Also to highlight that no 2 speakers/amp/dac etc... are exactly the same. They are all slightly difference (within a certain tolerance). The sames goes for everything.

I think the same goes for everything except for digital transports, especially streamers. If they don’t have a built-in DAC and they’re just carrying a digital signal via a network or internal hard drive, that’s strictly in the digital domain where “sound quality” is not a valid term.

Not until the signal undergoes the DA conversion process should there be any audible differences, and in the theoretical event that the streamer did introduce any errors in the signal, this should manifest as glitches or dropouts, not a degradation in sound quality. Thus my beefs with the N10.

 

[clearnfc]

I think the same goes for everything except for digital transports, especially streamers. If they don’t have a built-in DAC and they’re just carrying a digital signal via a network or internal hard drive, that’s strictly in the digital domain where “sound quality” is not a valid term.

Not until the signal undergoes the DA conversion process should there be any audible differences, and in the theoretical event that the streamer did introduce any errors in the signal, this should manifest as glitches or dropouts, not a degradation in sound quality. Thus my beefs with the N10.

Not quite, digital signals are subjected to factors like jitter, external interference, noise as well. ITs just that they are less susceptible.

What makes computer related signals "reliable" is because of the error checking and correcting methods/algorithms. Eg, TCP/IP is able to ensure the data received is correct via checksums.... If its wrong, the computer will instruct sender to send again.

http://audio-probe.com/en/documentation/clock-jitter-and-audio-quality/?ckattempt=1

The problem with SPDIF is that there is no way to detect and correct any error. Think of it as UDP, sometimes you may see artifacts or popping noise etc... in a video streaming. ITs because of packets not received or data incorrect.

 

[srkbear]

Not quite, digital signals are subjected to factors like jitter, external interference, noise as well. ITs just that they are less susceptible.

What makes computer related signals "reliable" is because of the error checking and correcting methods/algorithms. Eg, TCP/IP is able to ensure the data received is correct via checksums.... If its wrong, the computer will instruct sender to send again.

http://audio-probe.com/en/documentation/clock-jitter-and-audio-quality/?ckattempt=1

The problem with SPDIF is that there is no way to detect and correct any error. Think of it as UDP, sometimes you may see artifacts or popping noise etc... in a video streaming. ITs because of packets not received or data incorrect.

Well I think we’re basically saying the same thing Although I’ll add that if a streamer is connected to a DAC via asynchronous USB, the DAC clock serves as the master, and the buffers in its input receiver should obviate timing errors. And as you say, TCP/IP has error and packet loss-correcting protocols, and thus the Ethernet stream shouldn’t be a source of errors. So I would argue that the issue of jitter should come down on the quality of the DAC’s input receiver (at least via USB), not the quality of the streamer clock.

But noise? In a pure digital domain? Unless you’re referring to a ground loop, I’m going to have to ask for an explanation on that one—prior to DA conversion the signal is solely compromised of ones and zeros, and I fail to see how noise could be a factor until the DA process. Glitches, pops or dropouts? Sure—because the digital code being processed is adulterated during (theoretical) errors. But digital data is not sound, and it’s not music—I need some clarification on what you mean by noise, which exists in the analog domain.

 

[Beershaun]

Here is an older article from Amir on how an HDMI source affected the performance of his DAC.

A Deep Dive into HDMI Audio Performance

Note: this is a slightly revised version of an article I wrote for the Widescreen Review Magazine. It was published in January 2014 I think. ----- A Deep Dive into HDMI Audio Performance Sometimes I feel that digital audio is the most misunderstood technology around. If I asked you how a TV...

audiosciencereview.com

 

[antcollinet]

Here is an older article from Amir on how an HDMI source affected the performance of his DAC.

A Deep Dive into HDMI Audio Performance

Note: this is a slightly revised version of an article I wrote for the Widescreen Review Magazine. It was published in January 2014 I think. ----- A Deep Dive into HDMI Audio Performance Sometimes I feel that digital audio is the most misunderstood technology around. If I asked you how a TV...

audiosciencereview.com

So - let's add the condition "using the same digital interface" to the thread title.

We could also note that the article is between 6 and 8 years old.

 

[Beershaun]

So - let's add the condition "using the same digital interface" to the thread title.

We could also note that the article is between 6 and 8 years old.

Sure. It clearly demonstrates the answer is yes. Then the new question is to dig deeper into the implementation of both the streamer and the interface to prove the answer is no.

 

[12Many]

Here is my thoughts - no conclusion though. I am of the opinion there should be no difference in sound, subject to the following. Note: When comparing items, it is important to specify the items. Is one item a 35$ streamer and the other a $1000 steamer, are are two $600 streamers being compared?
The streamers has to do 3 things and has basically three general sections to do this.
First, it has to receive and properly decode the Ethernet data. After a long trip over cable, the digital data looks alot like an analog signal and there is some type of signal coding to increase bandwidth. This is usually error free and the error rate is very very small and I believe there is error correction coding built it. However, if the streamer is really terrible or the incoming data is terrible, it is possible have a high bit error rate - usually this will crash the stream though, making it non-functional.
Second, processing occurs on the decoded data so that it converted from Ethernet protocol output to whatever format is going to be send out over the digital output of the streamer. This should be error free, but again, if this is ******, there could be errors.
Third, the stream needs to be transmitted over the digital output, so there is more processing such as a USB output driver to driver the signal over the usb output port and cable.
IMO, all of these things typically occur with such a low error rate, in a properly built and functioning system, to have the data output essentially be the same.
Finally, I want to mention that it is my opinion that errors in the digital data can cause a different sound. For example, the digital value 10011, if experiencing a error, could be after the error 10001. Absent error correction, I don't think the DAC is going to know there was an error but will blindly convert the signal to an analog signal, which would be different then intended. This is just one sample of many though. I think Ethernet has error correction, however, I can not comment if bit errors that could occur inside the streamer or as part of the transmission from the streamer to the DAC will get caught. Thus, I think it could change the sound, but the chance of an error is tiny. I read about a USB data transfer test and if I recall correctly, there were gigabytes of data transferred without any errors. If I am wrong on any of this, please provide correction. Thank you.

 

[antcollinet]

Here is my thoughts - no conclusion though. I am of the opinion there should be no difference in sound, subject to the following. Note: When comparing items, it is important to specify the items. Is one item a 35$ streamer and the other a $1000 steamer, are are two $600 streamers being compared?
The streamers has to do 3 things and has basically three general sections to do this.
First, it has to receive and properly decode the Ethernet data. After a long trip over cable, the digital data looks alot like an analog signal and there is some type of signal coding to increase bandwidth. This is usually error free and the error rate is very very small and I believe there is error correction coding built it. However, if the streamer is really terrible or the incoming data is terrible, it is possible have a high bit error rate - usually this will crash the stream though, making it non-functional.
Second, processing occurs on the decoded data so that it converted from Ethernet protocol output to whatever format is going to be send out over the digital output of the streamer. This should be error free, but again, if this is ******, there could be errors.
Third, the stream needs to be transmitted over the digital output, so there is more processing such as a USB output driver to driver the signal over the usb output port and cable.
IMO, all of these things typically occur with such a low error rate, in a properly built and functioning system, to have the data output essentially be the same.
Finally, I want to mention that it is my opinion that errors in the digital data can cause a different sound. For example, the digital value 10011, if experiencing a error, could be after the error 10001. Absent error correction, I don't think the DAC is going to know there was an error but will blindly convert the signal to an analog signal, which would be different then intended. This is just one sample of many though. I think Ethernet has error correction, however, I can not comment if bit errors that could occur inside the streamer or as part of the transmission from the streamer to the DAC will get caught. Thus, I think it could change the sound, but the chance of an error is tiny. I read about a USB data transfer test and if I recall correctly, there were gigabytes of data transferred without any errors. If I am wrong on any of this, please provide correction. Thank you.

Ethernet is error free. Error detection is essentially perfect, and if the receiving device gets an errored packet, it simply requests it to be resent. Essentially all you are doing is transferring an audio file from the service to the streamer over the internet. If we were not able to to that consistently for billions of terabytes daily, the modern world would stop working.

Errors in internal processing? Certainly not from an "error rate". Just about every electronic device on the planet uses microcontrollers. If these were not able to process data completely error free, they could not work. They'd be continually crashing. Yes, I know - some are, but it is rare, and in order for it to change sound, it would have to be continuous. For example - when there is an error in my firetv stick, it doesn't change the picture or the sound - it simply bombs out of the app it is playing, and playback stops altogether. Further, these sorts of problems are almost always due to programming errors, rather than the processor device getting internal data transfer errors.

Transport via USB, Toslink etc. Again a solved problem for decades. If it doesn't work (essentially - error free) then the device is broken. More importantly even if there is a high error rate that can have audio impact - the effect is not subtle. It will never result in a reduction in sound stage, or subtle "timbre" or "tone" or "texture" effects (or whatever other mumbo jumbo the deluded like to use to descibe differences between digital kit). Errors in digital audio result in gross "breaking" of the sound in the form of dropouts, pops, clicks etc. If you just get these at a very low error rate within the specified range of 1 every few days, you're probably not going to notice it. If the process breaks such at the error rate is high - it is immediately obvious that the process is broken, and the music becomes un-listenable.

So feel safe to make a conclusion. If the streamers are not broken, and regardless of price, they sound the same - all other things being equal**.


** equal here includes the nature of the interface into or out of the device. For example, balanced - or galvanically isolated. Ie the same care taken to eliminate ground noise (which is a system characteristic, not a device characteristic)

 

[SMc]

Here is my thoughts - no conclusion though. I am of the opinion there should be no difference in sound, subject to the following. Note: When comparing items, it is important to specify the items. Is one item a 35$ streamer and the other a $1000 steamer...?

I've done this one! In a casual comparison there's no difference between my Chromecast Audio optical output and my Bryston BDP-1. Each have their use and have different user experiences but I wouldn't repeat the Bryston purchase.

 

[srkbear]

Here is my thoughts - no conclusion though. I am of the opinion there should be no difference in sound, subject to the following. Note: When comparing items, it is important to specify the items. Is one item a 35$ streamer and the other a $1000 steamer, are are two $600 streamers being compared?
The streamers has to do 3 things and has basically three general sections to do this.
First, it has to receive and properly decode the Ethernet data. After a long trip over cable, the digital data looks alot like an analog signal and there is some type of signal coding to increase bandwidth. This is usually error free and the error rate is very very small and I believe there is error correction coding built it. However, if the streamer is really terrible or the incoming data is terrible, it is possible have a high bit error rate - usually this will crash the stream though, making it non-functional.
Second, processing occurs on the decoded data so that it converted from Ethernet protocol output to whatever format is going to be send out over the digital output of the streamer. This should be error free, but again, if this is ******, there could be errors.
Third, the stream needs to be transmitted over the digital output, so there is more processing such as a USB output driver to driver the signal over the usb output port and cable.
IMO, all of these things typically occur with such a low error rate, in a properly built and functioning system, to have the data output essentially be the same.
Finally, I want to mention that it is my opinion that errors in the digital data can cause a different sound. For example, the digital value 10011, if experiencing a error, could be after the error 10001. Absent error correction, I don't think the DAC is going to know there was an error but will blindly convert the signal to an analog signal, which would be different then intended. This is just one sample of many though. I think Ethernet has error correction, however, I can not comment if bit errors that could occur inside the streamer or as part of the transmission from the streamer to the DAC will get caught. Thus, I think it could change the sound, but the chance of an error is tiny. I read about a USB data transfer test and if I recall correctly, there were gigabytes of data transferred without any errors. If I am wrong on any of this, please provide correction. Thank you.

A streamer is nothing other than a basic computer, that serves to carry an Ethernet digital signal from my router to my DAC via a USB connection—with some metadata retrieved and presented on a GUI along the way. It’s been an accepted fact that desktop and laptop computers do a perfectly fine job of transmitting digital audio streams to a DAC, and nobody fusses over whether it’s a cheap Dell Laptop, the latest iMac or a tricked out custom PC tower doing the job. After all, the very reason why we attach external DACs to PCs is to overcome the limitations of their internal sound cards.

It has also been shown that a simple RPI can handle this task with the highest possible fidelity as well—the only variables involved are the maximum sampling rates the hardware can handle, whether it can transfer native DSD signals, whether it has MQA passthrough capabilities, is equipped with an internal Roon Bridge, etc.

The majority of costly streamers are charging you for a fancy GUI, for an internal DAC that is probably inferior to the one you have already, or for a bunch of Byzantine menu features that you’ll likely set once and never change again.

For those of us who use Roon, a hardware GUI becomes superfluous—there isn’t much metadata out there that Roon’s software GUI doesn’t already provide in an ingenious and elegant format. This is why, after auditioning a slew of streamers at a range of price points (including a monstrous Cambridge Audio option, an inscrutable Cocktail Audio piece, and a feloniously priced NAIM offering), I settled on iFi’s Zen Stream—which for $399 saves me the hassle of building an RPI in favor of a pre-fabricated Volumio-based micro computer, with a Roon Bridge baked right into the motherboard.

It works seamlessly with Roon, turning my Topping d90se into an effective Roon Ready device—the only limitation being a maximum PCM sampling rate of 384khz and a native DSD limit of 256. Of course iFi is now offering the Neo Stream, which can handle PCM 786 and DSD 512–but that’s not the explanation for the extra $800 asking price. It’s about the internal DAC that comes included and an orgy of connector types that for me will go to waste.

In my view a streamer is the element of any system on one should spend the least money. The DAC, which is also essentially a computer in terms of its function, should be expected to cost a little more, although not much. It’s my amplifier and headphones on which I invested the most, in increasing order—because material costs and the complexity of design actually translate to the greatest gains in the end result going to my ears. A few wise audio gurus taught me this investment strategy early on, and their guidance has served me well.

 

[sq225917]

The answer is that they 'should' sound the same, and 99% of them do. A rpi3 and my m1 imac sound identical, as does the digital out of my fiio m15 and 90s vintage audiolab cd transport.

 

[SylvainQc]

The signal through a streamer stays entirely in the digital domain. Barring signal degradation due to a Bluetooth connection leg or uncorrectable jitter errors or gross downsampling or ground loop hum or buzz, streamers and most modern DACs leave no room for audible improvement of the sound quality. Unless the DAC you have was intentionally designed to color the sound. The only things that would matter are features, music format support, price and build quality/longevity. Inaudible tiny further improvements in the specs of DACs is possible, but why care about it? The "best" in terms of specs is a moving target that is pointless to chase.

Instead, get better loudspeakers. Get an amp with a SINAD of 90 dB or better at the power levels your speakers need to sound enjoyably loud in your room. In addition to the room treatment you are considering, software EQ the music for your room and speakers. Find better music to listen to. Protect your hearing from damage. Train yourself to listen more attentively. Listen to some glorious music recorded in the 1940s and 1950s: enjoying it despite the incurable lower fidelity of most of the recordings will train your brain to focus on the essentials of music enjoyment instead of fretting about almost imperceivable differences. Health, mood, shared listening with family and friends with similar music tastes, dancing, study of music, can all help get more enjoyment out of music listening.

Hi guys. I don't want to make no ennemys here, just to exchange ideas, experiences.. Among other sources, I use a first gen. Bluesound streamer. I equally use it to read FLAC files on hard disk (from my ripped CDs). It sounds great. I happened to have bought an old Enlightened Audio Designs Dac in a pawn shop last fall.. and guess what? I'm stuck to my speakers for hours, enjoying how more lively, palpable the music is when I use this DAC.. and there's that guy on Youtube (Steve..) who says that his $10k streamer gets so much more detail and limpidity to the music than his Bluesound Node streamer.. it leaves me doubting.. I was telling myself that 'a bit is a bit'.. until I heard this EAD Dac in my system..
And I'm not mentioning how (in winter..) my system gets a notch or two better 3D sound when I reinstall my beloved class A Quad II tube amps, over my ss amps.. but that's just me.
p.s. j

 

[ahofer]

Hi guys. I don't want to make no ennemys here, just to exchange ideas, experiences.

Welcome to ASR. Generally we aren't that interested in exchanging uncontrolled, time-distanced, subjective experiences here, as they are basically useless for evaluating gear. See, for instance, my signature quote from Alan Shaw about "loudness as a bias" in uncontrolled comparison.

In your case, you seem to be conflating DACs with streamers as well. DACs can indeed sound different, especially some high end designs that deliberately EQ the output for a signature sound. Also, there can be impedance mismatching between components that changes frequency response. At ASR we tend to value linear, "straight wire (with gain)", response from our electronics, as opposed to distorted or colored. The good news is that extremely accurate streaming, conversion, and amplification is available at a very reasonable cost, leaving the objective audiophile more money to invest in speakers, EQ, and room treatment, which can actually have a large and beneficial effect on sound quality.

I'd suggest you read through more of this thread and some of the newcomer material before/instead of digging in on this, otherwise you probably won't enjoy this place too much. I hope you stick around and give it a try.

 

[antcollinet]

Hi guys. I don't want to make no ennemys here, just to exchange ideas, experiences.. Among other sources, I use a first gen. Bluesound streamer. I equally use it to read FLAC files on hard disk (from my ripped CDs). It sounds great. I happened to have bought an old Enlightened Audio Designs Dac in a pawn shop last fall.. and guess what? I'm stuck to my speakers for hours, enjoying how more lively, palpable the music is when I use this DAC.. and there's that guy on Youtube (Steve..) who says that his $10k streamer gets so much more detail and limpidity to the music than his Bluesound Node streamer.. it leaves me doubting.. I was telling myself that 'a bit is a bit'.. until I heard this EAD Dac in my system..
And I'm not mentioning how (in winter..) my system gets a notch or two better 3D sound when I reinstall my beloved class A Quad II tube amps, over my ss amps.. but that's just me.
p.s. j

If you are listening sighted - then certainly cognitive biases can be at play* changing what you percieve in the sound even though the actual signals to the speakers to not change.

Giving your youtuber the benefit of the doubt (and not assuming he is just shilling for the manufacturer) he is also subject to this. Simply knowing that he is listening to expensive gear is enough to bias his perception to hear a difference even when non exists. This is why we don't pay much attention to sighted listening impressions here.


* Here is an example. Even when you know how your brain is being fooled - it is still fooled.

Even though this example is speech, the same sorts of biases can impact you when listening to music.

.

 

[ahofer]

btw, is this your unit? Or is it substantially similar?

Enlightened Audio Design DSP-1000 Series III D/A processor Measurements

Sidebar 2: Measurements

www.stereophile.com

That could certainly sound different. Notice the channel noise imbalance and the even-order distortion (which many people find pleasant, allegedly).

 

[krabapple]

Welcome to ASR. Generally we aren't that interested in exchanging uncontrolled, time-distanced, subjective experiences here, as they are basically useless for evaluating gear. See, for instance, my signature quote from Alan Shaw about "loudness as a bias" in uncontrolled comparison.

In your case, you seem to be conflating DACs with streamers as well. DACs can indeed sound different, especially some high end designs that deliberately EQ the output for a signature sound.

But more importantly to impress on a newbie: such DACs are extremely rare in 'DAC space'. As they should be. From one POV, they represent either willfully or mistakenly 'bad design'.

Also, there can be impedance mismatching between components that changes frequency response.

Or there can simply be difference between two DACs in the output (voltage) level at the analog stage. Which is equivalent to just turning the 'volume knob' up or down. It means nothing to the intrinsic 'sound quality' of the DAC. It is why close level matching is one of the crucial aspects of a fair comparison.