my reply to your poll is - BOTH. specially when today's streamers can do more than just stream. I have an Ultra that most of the time just plays directly from my laptop (using DDC) running Feishin app connected to a Navidrome server, which also runs LMS. but every now and then, I'd like to play music that I don't have on the server from Tidal (or Bandcamp) . not to mention the ease of switching currently played music, to play on multiple WiiM devices (with the Ultra's "Sound Profile" beta firmware, i'm able to do that switch with a click of a button on the WiiM's remote)
-
Welcome to ASR. There are many reviews of audio hardware and expert members to help answer your questions. Click here to have your audio equipment measured for free!
Streamer versus Computer?
- Thread starter Gorgonzola
- Start date
khaliss
Active Member
I don't have much experience on HiFi audio streaming boxes, the only one I've tried so far is the Wiim Ultra streamer box. I understood the appeal (simplicity, compact form factor, bang-for-the-buck solution, etc)... however, it just didn't tick all the boxes for me (particularly on versatility)... and I didn't want to spend more. So I ended up returning it, and just sticking with my PC setup (just posted about it in a separate thread):
Computers don't add ambient noise when they are built to be quiet. Mine are far enough away that they are silent from the listening position. I have my own library located on 72tb's worth of HDD space. I access my library from JRiver and my computer is totally silent - it's also water cooled. When I turn everything off - you can't hear it ... I doesn't even pickup on my phone's sound level meter. I also control my library with my android phone (I use Gizmo and JRemote 2 ).
There are plenty of ways to even interpret the question.
Some would argue that a simple Raspberry Pi (frequently utilized as a streamer or Roon bridge) is already a computer. Even further, some would argue that a streamer (streaming DAC or DACless streaming transport) is also a computer.
And then some of us literally use both, such as myself; I use a Mac Mini M1 to house my local library and run both, Plex server and Roon Core (server). I stream my music over my LAN to my speakers via a Roon bridge, even Qobuz which I stream through Roon because it has album-level ReplayGain/volume leveling, so one could argue that the Roon bridge device is a streamer, but a computer is on the opposite end over LAN functioning as music server as well as DSP processor.
As such, I have voted computer simply because universally it would apply, either there's a computer serving the music or the streamer/network bridge itself is also, technically, a computer.
Some would argue that a simple Raspberry Pi (frequently utilized as a streamer or Roon bridge) is already a computer. Even further, some would argue that a streamer (streaming DAC or DACless streaming transport) is also a computer.
And then some of us literally use both, such as myself; I use a Mac Mini M1 to house my local library and run both, Plex server and Roon Core (server). I stream my music over my LAN to my speakers via a Roon bridge, even Qobuz which I stream through Roon because it has album-level ReplayGain/volume leveling, so one could argue that the Roon bridge device is a streamer, but a computer is on the opposite end over LAN functioning as music server as well as DSP processor.
As such, I have voted computer simply because universally it would apply, either there's a computer serving the music or the streamer/network bridge itself is also, technically, a computer.
Id used Mac minis (going back to 2007 when benchmark first released the dac1 usb) and used them (most recently an m1 mini) as my main source until i got an Eversolo a6.
They work fine.
I won’t go back to a pc for Streaming now. The convenience and general lack of management and hassles ( ie software updates etc ) are better for me with a streamer. I also prefer the iPhone and iPad interface of the Eversolo devices over a keyboard or screen sharing.
They work fine.
I won’t go back to a pc for Streaming now. The convenience and general lack of management and hassles ( ie software updates etc ) are better for me with a streamer. I also prefer the iPhone and iPad interface of the Eversolo devices over a keyboard or screen sharing.
Yorkshire Mouth
Major Contributor
There’s an 863 post thread here which suggests a dedicated streamer is better, certainly easier and more reliable:
www.audiosciencereview.com
Frankly, reading through the issues, solutions, loopholes, and difference between different user scenarios, even if the end result is bit perfect, it’s easy to see why someone might want to end up at the same place by just buying a dedicated streamer. Especially when one can be had for <£100.
Ending the Windows Audio Quality Debate
A debate has been long raging in the audio community that Windows isn't fit for use in a high resolution audio system. Sceptics counterargued that bits are bits and Windows audio degradations are a figment of an over-active audiophile imagination. I decided to investigate this as I kept...
www.audiosciencereview.com
Frankly, reading through the issues, solutions, loopholes, and difference between different user scenarios, even if the end result is bit perfect, it’s easy to see why someone might want to end up at the same place by just buying a dedicated streamer. Especially when one can be had for <£100.
I have wasted loads of money on dedicating a computer (iMac) to music playing, I chose it because the big screen could be easily seen and was good for artwork and had a CD drive. Eventually it was no longer new enough to do security updates, maybe I was too cautious but it has to be on my network.
I had tried Sooloos, Roon's predecessor and still have the control 15 but it became obsolete too.
I have gone back to playing CDs and the odd LP and cancelled all my streaming.
I do still play music files from my home ripped library started over 20 years ago, or a file I bought without a hard copy, but not often.
Can't be bothered with anything that may ever require updates or may hang or crash before I can play music.
I had tried Sooloos, Roon's predecessor and still have the control 15 but it became obsolete too.
I have gone back to playing CDs and the odd LP and cancelled all my streaming.
I do still play music files from my home ripped library started over 20 years ago, or a file I bought without a hard copy, but not often.
Can't be bothered with anything that may ever require updates or may hang or crash before I can play music.
In my lounge I have a streaming amp that streams from my RAID/NAS (or phone or USB or SSD, but usually DLNA from the NAS). It also does sound from the TV.
I my home office I have several PCs but the laptop on my desktop does audio over Topping DX1 to either headphones or to one of two speaker & amp setups that happen to be in the office. I generally run FB2K and audio files are either local or come from the NAS over DLNA, usually FLACs. I also run VLC for videos or for music tracks I need to slow down, so as to keep up when playing guitar!
My point is both setups work really well - fine audio, quiet, simple, reliable - and I suspect I get less irritation by not streaming from the internet as such. Easy for me to say, as I have a 55 year collection, all ripped to the NAS since ripping became a thing, so don't need access to an online back-catalogue. I see no end of people complaining about apps used for streaming on various devices, as the platforms keep changing them, and streamer units have to keep updating their built in versions to keep up, and also all the concerns about whether the audio quality is good, or if the platforms have remastered badly etc etc (whether people hear a difference or not). In short, to my ears I get great sound at more or less the convenient touch of a button and it's very reliable and predictable (once you are used to NAS units - I've run one for 15 years without issues, ripping, tagging, backups etc are trivial). Just saying, either way - computer or dedicated units - work well without a fuss....
I my home office I have several PCs but the laptop on my desktop does audio over Topping DX1 to either headphones or to one of two speaker & amp setups that happen to be in the office. I generally run FB2K and audio files are either local or come from the NAS over DLNA, usually FLACs. I also run VLC for videos or for music tracks I need to slow down, so as to keep up when playing guitar!
My point is both setups work really well - fine audio, quiet, simple, reliable - and I suspect I get less irritation by not streaming from the internet as such. Easy for me to say, as I have a 55 year collection, all ripped to the NAS since ripping became a thing, so don't need access to an online back-catalogue. I see no end of people complaining about apps used for streaming on various devices, as the platforms keep changing them, and streamer units have to keep updating their built in versions to keep up, and also all the concerns about whether the audio quality is good, or if the platforms have remastered badly etc etc (whether people hear a difference or not). In short, to my ears I get great sound at more or less the convenient touch of a button and it's very reliable and predictable (once you are used to NAS units - I've run one for 15 years without issues, ripping, tagging, backups etc are trivial). Just saying, either way - computer or dedicated units - work well without a fuss....
Last edited:
melomane13
Active Member
Rpi4 , lyrion with 6 channels camilla dsp ( crossover, room correction, loudness tone controls , crossfeed for headphones)
There is no streamer with the sames features
There is no streamer with the sames features
Tell
Addicted to Fun and Learning
- Joined
- Mar 30, 2024
- Messages
- 893
- Likes
- 1,268
Why wouldn't I use a streamer? I just flip up my phone, open Spotify, choose my music and tell it to play from my Nvidia Shield. Sure sometimes I do want to listen to some MP3s, or maybe game some overwatch while listening to music, then I'll use my computer. But it really is mainly Spotify through my streamer, it's just better that way!
I use only pis (3bs, 4Bs 1-4GB RAM) for roon bridge endpoints, with one exception - one room has an old Chromecast audio puck, one has a display from Ropiee, the others I just use a phone or tablet, or have converted to MIPI DSI lower power non hdmi displays.
LawrenceL
Member
[thought I'd share one of my audio musings from one of my Qobuz Club post. Please enjoy]
lit·mus test
a test for acidity or alkalinity using litmus.
a decisively indicative test.
Litmus—a mixture of colored organic compounds derived from lichen—was discovered in the 14th century to turn red in acidic solutions and blue in alkaline ones, making it a reliable acid-base indicator. Six centuries later, the term "litmus test" took on a figurative meaning. It now refers to any single factor that reveals the true nature of something or determines its classification.
That’s been the value of having my Yamaha receiver with so many so-called bit-perfect sources to stream Qobuz and Tidal through. They’ve allowed me to set up an audio litmus strip—an evaluative framework to test and refine my perceptual acuity.
I have one distinct advantage in discerning differences: DTS Neo:6 Music synthesis in my Yamaha Aventage RX-A880. The DSP processing, which can upsample signals to 24-bit/348kHz, accentuates distinctions between sources. Because DTS applies its “black-box” spatial expansion, these differences become strikingly clear. It’s like having an auditory vectorscope.
Here is my test bench schematic.
My AVR has two optical and two coaxial digital inputs. It also features a rotary selector knob, allowing me to switch between sources in about a second. I disable all inputs except for the two digital ones under test. Then, I rotate the receiver so I can't see which source is active and spin the knob randomly. This lets me switch between sources “blind” to determine which one sounds better.
Studies have shown that fast A/B switching is the most effective method for detecting subtle differences. It serves as a litmus test for my various sources. I continued testing inputs against each other—like a tournament ladder—until I had evaluated all of them.
I created a new System Test Song Playlist. This one really focused on resolving a full range of audio attributes. You may appreciate it. Here's my new playlist description below:
assist from: ChatGPT
I use this list to suss out differences between my different sources. I first tested my DDC Douk Audio U2 Pro. This device, though inexpensive, is pretty impressive. It has three outputs that stream simultaneously: I2S, SPIDIF Optical, and COAX.
Since the Douk U2 Pro has both an optical and a coax output that run simultaneously, I then compared optical to coax, using the blind test methodology. Unmistakably, I could hear the difference between optical and coax, with coax sounding better.
So there is a difference in the performance of optical versus coax SPDIF bitstreams. However, if I ran my system in stereo mode, I could not reliably tell them apart. That may be why some people feel they sound the same.
With these tested, Tested both Qobuz Connect and Tidal Connect streaming using Wi-fi versus Ethernet. My previous test: Wifi? Clean Ethernet? Dirty Ethernet? only tested potential noise injection by Ethernet but now with 2 Wiim Pros and this fast A/B test I can detect subtle differences with two simultaneous streams.
To do this I needed to operate two separate Qobuz accounts and simultaneously run the songs on the Wiim streamers. That way I could hear the difference exactly as the song is playing on the Wiim Pros. One Wiim Pro was placed on Ethernet and the other on Wifi. Then the Wifi and Ethernet were swapped. Result: There was no discernible difference in Ethernet and Wifi in fast A/B switching.
I then used Wifi because there is no difference because I know its fully isolated, and because it has 4 times the theoretical throughput of my 100Mbps Ethernet connections, and well, because I can...
I then compared the Douk Audio U2 Pro to the Wiim Pro using the same fast A/B method. And in careful listening tests, I compared the two and found that in every mode of operation of the Wiim Pro it was bested by the Douk Audio U2 Pro connected over COAX.
Why is that true? Here's some testing I later found that may validate my listening trials. Someone tested the Douk Audio U2 Pro, and here are their findings: powering the Douk Audio U2 Pro from a MacBook yielded measurements of –161 dB to –164.6 dB SNR.
Compare that to Wiim Pro using Wiim Pro Home at -148.0dB SNR [Credit ASR]
Compare that to Wiim Pro using Chromecast at -135.9dB SNR [Credit ASR]
Using Chromecast mode, jitter was tested for both optical and coax were measured at 0.4pS as I read this chart. Well below the suggested audibility threshold of 5-10nS. [Credit ASR]
[Credit ASR]
Findings and Conclusions:
Remember in my last post No Jitter is Not Your Friend... we talked about the effects of jitter. Jitter doesn’t improve imaging. It doesn’t enhance bass response. It doesn’t add crispness to the treble. At its worst, jitter causes crackling and stuttering. None of these sources gave any hint of audible jitter or its effects, at least by my perception. And interestingly, one of the so-called weakest sources—Chromecast—was measured (if I read it correctly) at just 0.4 picoseconds of jitter. That’s well below the commonly accepted audibility threshold of 10 nanoseconds.
Some of these findings defied my expectations. Interesting stuff. Just thought you’d might like to know.
Disclaimer: These are my personal findings. Others may have different experiences. But clearly, with fast A/B switching, even subtle differences can be discerned
lit·mus test
a test for acidity or alkalinity using litmus.
a decisively indicative test.
Litmus—a mixture of colored organic compounds derived from lichen—was discovered in the 14th century to turn red in acidic solutions and blue in alkaline ones, making it a reliable acid-base indicator. Six centuries later, the term "litmus test" took on a figurative meaning. It now refers to any single factor that reveals the true nature of something or determines its classification.
That’s been the value of having my Yamaha receiver with so many so-called bit-perfect sources to stream Qobuz and Tidal through. They’ve allowed me to set up an audio litmus strip—an evaluative framework to test and refine my perceptual acuity.
I have one distinct advantage in discerning differences: DTS Neo:6 Music synthesis in my Yamaha Aventage RX-A880. The DSP processing, which can upsample signals to 24-bit/348kHz, accentuates distinctions between sources. Because DTS applies its “black-box” spatial expansion, these differences become strikingly clear. It’s like having an auditory vectorscope.
Here is my test bench schematic.
My AVR has two optical and two coaxial digital inputs. It also features a rotary selector knob, allowing me to switch between sources in about a second. I disable all inputs except for the two digital ones under test. Then, I rotate the receiver so I can't see which source is active and spin the knob randomly. This lets me switch between sources “blind” to determine which one sounds better.
Studies have shown that fast A/B switching is the most effective method for detecting subtle differences. It serves as a litmus test for my various sources. I continued testing inputs against each other—like a tournament ladder—until I had evaluated all of them.
I created a new System Test Song Playlist. This one really focused on resolving a full range of audio attributes. You may appreciate it. Here's my new playlist description below:
assist from: ChatGPT
I use this list to suss out differences between my different sources. I first tested my DDC Douk Audio U2 Pro. This device, though inexpensive, is pretty impressive. It has three outputs that stream simultaneously: I2S, SPIDIF Optical, and COAX.
Since the Douk U2 Pro has both an optical and a coax output that run simultaneously, I then compared optical to coax, using the blind test methodology. Unmistakably, I could hear the difference between optical and coax, with coax sounding better.
So there is a difference in the performance of optical versus coax SPDIF bitstreams. However, if I ran my system in stereo mode, I could not reliably tell them apart. That may be why some people feel they sound the same.
With these tested, Tested both Qobuz Connect and Tidal Connect streaming using Wi-fi versus Ethernet. My previous test: Wifi? Clean Ethernet? Dirty Ethernet? only tested potential noise injection by Ethernet but now with 2 Wiim Pros and this fast A/B test I can detect subtle differences with two simultaneous streams.
To do this I needed to operate two separate Qobuz accounts and simultaneously run the songs on the Wiim streamers. That way I could hear the difference exactly as the song is playing on the Wiim Pros. One Wiim Pro was placed on Ethernet and the other on Wifi. Then the Wifi and Ethernet were swapped. Result: There was no discernible difference in Ethernet and Wifi in fast A/B switching.
I then used Wifi because there is no difference because I know its fully isolated, and because it has 4 times the theoretical throughput of my 100Mbps Ethernet connections, and well, because I can...
I then compared the Douk Audio U2 Pro to the Wiim Pro using the same fast A/B method. And in careful listening tests, I compared the two and found that in every mode of operation of the Wiim Pro it was bested by the Douk Audio U2 Pro connected over COAX.
Why is that true? Here's some testing I later found that may validate my listening trials. Someone tested the Douk Audio U2 Pro, and here are their findings: powering the Douk Audio U2 Pro from a MacBook yielded measurements of –161 dB to –164.6 dB SNR.
Compare that to Wiim Pro using Wiim Pro Home at -148.0dB SNR [Credit ASR]
Compare that to Wiim Pro using Chromecast at -135.9dB SNR [Credit ASR]
Using Chromecast mode, jitter was tested for both optical and coax were measured at 0.4pS as I read this chart. Well below the suggested audibility threshold of 5-10nS. [Credit ASR]
[Credit ASR]
Findings and Conclusions:
- The DTS Neo:6 Music processing modes even make subtle differences clearly audible.
- There is an audible difference between optical and coax.
- The audibility rankings correlate with the measured S/N data.
- The Douk Audio U2 Pro streamed over coax sounded better than the Wiim Pro in every mode.
- All inputs sounded subtly different, but Ethernet and Wi-Fi sounded identical.
- A computer can be a credible, even excellent source for streaming hi-res Qobuz.
Remember in my last post No Jitter is Not Your Friend... we talked about the effects of jitter. Jitter doesn’t improve imaging. It doesn’t enhance bass response. It doesn’t add crispness to the treble. At its worst, jitter causes crackling and stuttering. None of these sources gave any hint of audible jitter or its effects, at least by my perception. And interestingly, one of the so-called weakest sources—Chromecast—was measured (if I read it correctly) at just 0.4 picoseconds of jitter. That’s well below the commonly accepted audibility threshold of 10 nanoseconds.
Some of these findings defied my expectations. Interesting stuff. Just thought you’d might like to know.
Disclaimer: These are my personal findings. Others may have different experiences. But clearly, with fast A/B switching, even subtle differences can be discerned
Attachments
Last edited:
popej
Addicted to Fun and Learning
- Joined
- Jul 13, 2023
- Messages
- 701
- Likes
- 599
I would expect an seriuos audio listenig test to start with something like: we follow DBT protocol, we level match output with following measurements, we provide statistics to estimate result error. Sorry, can't force mysel to read it all.
OP
- Thread Starter
- #114
I thank @LawrenceL for his posting his lengthy tome which doubtless took him a lot of effort.
Like "Larry" (sorry) I used to believe that fast A-B'ing could reveal the most subtle differences. I'm no longer so sure about that. The reasons are (i) personally at my age my high frequency is only about 8kHz, and (ii) in sighted testing there is always the potential (or certainty) that we will be deceived by our biases.
Nevertheless, unlike some folks here,I'm fine with people posting their impressions as long as they don't insist on their impressions be absolute truth & fact.
Like "Larry" (sorry) I used to believe that fast A-B'ing could reveal the most subtle differences. I'm no longer so sure about that. The reasons are (i) personally at my age my high frequency is only about 8kHz, and (ii) in sighted testing there is always the potential (or certainty) that we will be deceived by our biases.
Nevertheless, unlike some folks here,I'm fine with people posting their impressions as long as they don't insist on their impressions be absolute truth & fact.
@LawrenceL: Hm, I am a bit confused by those digital I/O tests.
Digital I/O is bitperfect, unless some sample conversion (or even rate conversion) is applied. I can imagine those tiny odd harmonics could be e.g. due to conversion int -> float32 -> int + (lack of) rounding which may not be perfectly symmetrical around zero. Clock has no effect on digital I/O.
That relates to my second question - measuring jitter on fully digital path - how does it work? There is no jitter involved, the SPDIF receiver just collects the incoming samples and passed them to the PC in blocks, it does not care about their exact timing.
But maybe there was some D-A-D conversion involved, but the measurements do not show any artefacts of that.
Digital I/O is bitperfect, unless some sample conversion (or even rate conversion) is applied. I can imagine those tiny odd harmonics could be e.g. due to conversion int -> float32 -> int + (lack of) rounding which may not be perfectly symmetrical around zero. Clock has no effect on digital I/O.
That relates to my second question - measuring jitter on fully digital path - how does it work? There is no jitter involved, the SPDIF receiver just collects the incoming samples and passed them to the PC in blocks, it does not care about their exact timing.
But maybe there was some D-A-D conversion involved, but the measurements do not show any artefacts of that.
LawrenceL
Member
@popej
That is definitely a 'TLDR' post. Very much a tome.
Apologies all. Thanks for being polite. And you are right to challenge my process. I tried to keep the length as manageable as possible by omitting some of the finer details.I am a test engineer by training, so the concerns you raised are exactly what I considered before starting. The advantage of my AVR and these units is the ability to level match. I used a sound pressure meter to calibrate the channels before the listening test began. I am aware that even as little as 0.1 dB difference can skew results, so I worked to eliminate that variable from the equation.
In my testing, I also utilized this method to keep the process unsighted and blind:
This is something I spent a lot of time on a few months ago and I only share it in the context of whether a computer is a credible streamer and in my case my computer is my best streamer. These are my personal observations of course, And your mileage may vary.
- Regards All
Last edited:
LawrenceL
Member
@Gorgonzola
It was a tone indeed.
What I found with careful listening was insidious, and I wanted to dig in to understand why I was hearing differences. If you look at my post, I explicitly shared the idea that these are simply my personal observations. I mentioned:Disclaimer: These are my personal findings. Others may have different experiences. But clearly, with fast A/B switching, even subtle differences can be discernedI am an old geezer. While my hearing is good for my age, it is certainly not what it used to be. I have to give the credit to whatever circuit nuances are in my AVR.
However, I believe I have tested carefully enough to eliminate expectation bias, sight bias, and other factors. The results were repeatable enough that I felt they were worth sharing. Again, I wanted to provide this in the context of my specific system, where my computer consistently outperforms the dedicated WiiM streamer.
Last edited:
LawrenceL
Member
@phofman
It is quite all right. I am also a bit confused.
Those measurements, by the way, were not data I took myself. They were sourced from ASR (Audio Science Review) evaluations of these specific devices.As I mentioned in my longer post, these results could definitely be the byproduct of how my AVR and its DTS DSP handle subtle issues like timing, jitter, and noise. These variables can differ between devices and along different signal paths.
I cannot say for certain what causes these differences. What I can say is that in my blind testing, the variations are pronounced enough to be quite repeatable in synthesized 5.1 channel listening. Interestingly, when I switch my system into 2.1 stereo mode, those differences tend to diminish.
- Regards
Last edited:
LawrenceL
Member
Similar threads
