What Does a Streamer Actually Do?

I’ve heard a difference between a $10k streamer/server and Aurender N20 in my system. It was huge. Much bigger, than between my DAC and MSB Reference, which is 4 times more expensive, than my dac. Who cares how Aurender does it? Probably better power supply, USB output etc. It’s like with any other component: just listen and choose what sounds best to you providing you can afford it:))
 
I have two main setups.
1. My desktop computer t receives a wireless wi- fi signal from my internet router. It is then transported to my stereo via a USB cord.
2. I can receive a signal form my smartphone either through my home wi-fi or directly from my smartphone carrier. then transport the vi Bluetooth to my dac/amp.
Don't get me wrong . I am not advising anyone not to use a streamer.
So you have 3 devices acting as a streamer: your computer, your phone and your amp/dac, since the bluetooth stream also needs to be received/decoded/passed along to the dac. In your first case you have one streamer, case two you are daisy chaining two streamers (phone, streaming amp/dac).

How do they do that?

Notice that these 3 streamers you are using, none of them have any particular care with noise/isolation/maintaining bit perfect data.

- Have you ever heard noise in your usb headphones when your PC starts doing something? 10 years go it was common to get noise when your GPU started working more intensively. Well, PCs usually share power rails between I/O (your usb port), processing, PCI, etc. The power is coming from switching (many times more noisy than we would want) power supplies. The usb ports will not be galvanically isolated in 99.999% of PCs. You are now exclusively counting on your DAC to filter, isolate and reclock your data. Most modern DACs are really good at this, some rarefied group of available DACs don't do it at all because of reasons.

- If you use Bluetooth, you probably don't retain bit-perfect data. Most Bluetooth codecs have limitations on bandwidth and some operating system drivers will automatically resample your data so it fits. This will be transparent for you, nothing you can do, or monitor, without major involvement on your end.

So this is what I meant when I laid down my minimum viable digital transport features, these effects will be mitigated/avoided/nullified.
How do they do it? Using linear or super low noise power, separate IO power rails and buses from everything else, isolated and possibly reclocked outputs, dedicated operating systems with minimal processes and no DSP (unless explicitly asked for), at a minimum.

Again, most of these upstream hygiene concerns are not in your chain, but I'm not surprised you have perceptually undistinguishable high quality sound, the level of noise consumer electronics are putting out is going down exponentially, for a number of reasons, and your dac is probably cleaning and reclocking a lot of stuff at input as well.
 
  • Like
Reactions: MasonStorm
Thanks everyone who has replied so far. Lots of useful information, though I've still not quite grasped what makes one competent streamer different from a "high end" one.

I see absolutely that the "streamer" facility in my PC or phone will be poor, but why should a 10K streamer be better than a 1K (or perhaps less) one?

With DACs and Class D amps, there are obvious ways to determine the likely performance. Nearly all DACs use one of a relative handful of widely available chips and there is a recognised hierarchy - the vendors of these DAC chips describe the attributes of each. Likewise with Class D amps - we don't now consider an amp based on Tripath, but probably one based on a Purifi module. Is there "streamer module" or a particularly technology that may make one streamer significantly better than others?

This is really the gist of my opening post. What specifically (apart of course from a good power supply, components, sockets, wiring, etc) makes for a first rate streamer? Is it in fact an area where hype and wool-pulling over our eyes is actively used to justify a huge price? I'm not yet convinced it isn't!
 
  • Like
Reactions: Iving and RCanelas
When InnuOS 2.0 released, you could find a plethora of overly positive reviews regarding improvements in sound quality. I experienced it myself and felt it was about a $6-800 performance boost in terms of value, at least on the ZenMini.

What improvements in sound quality did you experience?
 
As with a lot of boxes in our systems, it depends on your priorities and how you expect to consume digital music.

The minimum viable streamer for me is akin to a digital transport:
- grabs the data from somewhere (abstracts the data location) and makes it available at a digital output in a bit perfect manner
- this can be from a service stream, internet radio, nas in your network, usb connected harddrive, doesn't matter
- provides a minimum amount of buffering so your output stream is steady
- provides effective isolation from the outside noise chaos. Galvanic isolation at I/O. Recloking for non async outputs (spdif, i2s, aes).
- provides an effective interface to configure, integrate outside services, navigate and select media.

So think of it as a gateway and a noise breaker. Given these specs I require one that is not integrated with other functions. These are small format, mini-computers, and one-boxing them with amplification and dacs is not my jam. I would always opt for a separate component as you listed, but I wouldn't expect performance differences if these minimums that I listed are covered. Since I use roon, I have another device running the service, and this is my endpoint, my noise breaker. If you want to use another app, that would be my greatest differentiator when choosing one of those: do they expose a decent app, is it maintained, can I use 3rd party apps etc.

Thank you for this answer.

Am I correct in distilling your reply as follows:

— If I don’t need Wi-Fi (if I’m fine to connect an iPad to a DAC with a USB cable)

— If I don’t care about multi music service functionality (if I use only Qobuz, and I am happy with the Qobuz GUI)

— If I do not want to use Roon

Then the only remaining function a streamer is doing for me that I care about is re-clocking the incoming digital stream?

Is this correct?

If this is correct, then why don’t I just got a dedicated re-clocker with an outboard LPS and put it between the iPad and the DAC as way possibly to improve the sound quality?
 
Thank you for this answer.

Am I correct in distilling your reply as follows:

— If I don’t need Wi-Fi (if I’m fine to connect an iPad to a DAC with a USB cable)

— If I don’t care about multi music service functionality (if I use only Qobuz, and I am happy with the Qobuz GUI)

— If I do not want to use Roon

Then the only remaining function a streamer is doing for me that I care about is re-clocking the incoming digital stream?

Is this correct?

If this is correct, then why don’t I just got a dedicated re-clocker with an outboard LPS and put it between the iPad and the DAC as way possibly to improve the sound quality?
What is left is output isolation and reclocking, yes.

That is really cool, because if you place a reclocker between your tablet and your dac, you effectivelly made an exploded streamer. There really isn't a lot of magic to it. You have a lightweigth computational unit that exposes a competent UI, grabs data from 'somewhere' and places that data on a clean output, your reclocker/isolator. About it sounding better, it should, but you'll have to try it out!

I'm taking my own rules to heart, I'll be releasing a streamer early next year through Cinnamon that fulfils these and does a bit more. I can honestly say my digital streams do sound better, but what really made me try this is the ease of getting music from anywhere and knowing I'm not compromising anything along the way.
 
Thank you.
 
What improvements in sound quality did you experience?

Lower noise floor, better separation, more musically involving, much, much cleaner and less digital sounding. It felt like they took noise out of the hardware, which might have happened by optimizing performance better for lower internal noise generation. This was apparent both as a transport and utilizing the onboard DAC.

My quoted improvement value stands only with the Zenmini as the overall improvements InnuOS 2.0 brought might be better realized on a unit higher up the chain. I have the LPSU upgrade as well. The performance boost elevated the Zenmini greatly in my mind as a value proposition.
 
Thank you for this detailed explanation, Skip.
 
I have been slightly hammered on this topic by several friends over the last 48 hours, which I greatly appreciate.

Although I am not yet persuaded of the theoretical benefits of a streamer in my particular, idiosyncratic situation, I have discovered an independent reason to eventually move away from the iPad-as-streamer system I am using presently: there is some noise glitchy-ness audible through the speakers when playing with the iPad. No like-y that!

Also, after a week of listening, I think that I am hearing — and this is totally speculative — some kind of digital sameness or homogenization across recordings. Perhaps this is because the current iPad set-up is imparting — or importing — some kind of noise?

So I will be ordering an Innuos PULSEmini and an outboard LPS from Linear Tube Audio designed for the PULSEmini.
 
Last edited:
  • Like
Reactions: RCanelas
I feel your anx RonPULSEmini_solidwhite_1080x1080_Front_Fascia__00893.jpg
 
  • Like
Reactions: Ron Resnick
I have discovered an independent reason to eventually move away from the iPad-as-streamer system I am using presently: there is some noise glitchy-ness audible through the speakers when playing with the iPad. No like-y that!
Yes, as I mentioned to an earlier poster, I'm entirely not surprised at you disappointment with any PC (PC, iPad, phone, etc) used as a streamer. Bluntly put - not built for audio.

I'd be interested to hear your comments on how much better the Innuous Mini turns out to be. In my original post where I asked what a streamer actually does, I've had various explanations though little has led me to conclude that a relatively inexpensive (but audio specific) streamer won't be as just good as a 10K one as far as sound quality is concerned.

Personally I think that a streamer without a first-class control app to easily search for and select music, create playlists, provide access to artist bios and other works, etc is worth avoiding! I use a NAD streamer that has the huge advantage of the excellent BluOS app and I'll only move to a "better" streamer if it genuinely offers better sound and it has an equally user-friendly app. I'm not holding my breath!
 
  • Like
Reactions: RCanelas
Personally I think that a streamer without a first-class control app to easily search for and select music, create playlists, provide access to artist bios and other works, etc is worth avoiding!

For sure! If the primary point of digital for many is convenience, then make sure you find an app that you find to be intuitive and convenient.

I find my friend’s Aurender GUI to be not terribly intuitive. I find the Innuos Sense to be better. I think the Qobuz GUI is good.
 
While some folks love to hate Roon, this already very good interface continues to get better and more feature rich (latest update was yesterday).

Perhaps because the Grimm MU1 was built around Roon, there are no problems (to my ears) with the sound. It is consistently excellent and musically engaging.
 
Everybody hates the middleman {management|. That is always the problem. If an add-on is causing problems, do you remove it or add a fix? The fix then creates its's own problem. Isn't life fun?
 
While some folks love to hate Roon, this already very good interface continues to get better and more feature rich (latest update was yesterday
This worries me! Adding features inevitably makes things more complicated and therefore less intuitive and longer to learn.

I had Roon for a trial period but decided that it offered little more than BluOS apart from its sometimes useful "radio" feature. I stuch with BluOS.

If only apps offered Beginner, Intermediate and Expert levels, so users can easily get to know the software by using basic Beginner features - choosing source, searching for and selecting music, adding to a Playlist, adjusting volume - not much more is needed for an entertaining music evening! Then as we become more envolved and adventurous we can switch to Intermediate or even Expert with extra features such as access to artist bios, similar music, etc. When we have friends round, we can engage Beginner and pass them the iPad and they will easily understand how they can select their own music choices.

I was playing with Rose software a couple of weeks ago and frankly it has put me off their streamers. I'm a simple man and like to keep things simple!
 
This worries me! Adding features inevitably makes things more complicated and therefore less intuitive and longer to learn.

I had Roon for a trial period but decided that it offered little more than BluOS apart from its sometimes useful "radio" feature. I stuch with BluOS.

If only apps offered Beginner, Intermediate and Expert levels, so users can easily get to know the software by using basic Beginner features - choosing source, searching for and selecting music, adding to a Playlist, adjusting volume - not much more is needed for an entertaining music evening! Then as we become more envolved and adventurous we can switch to Intermediate or even Expert with extra features such as access to artist bios, similar music, etc. When we have friends round, we can engage Beginner and pass them the iPad and they will easily understand how they can select their own music choices.

I was playing with Rose software a couple of weeks ago and frankly it has put me off their streamers. I'm a simple man and like to keep things simple!
I like simplicity too (as my What's Good setup illustrates). Roon is truly easy to use once it is installed. Installation on my computer, when I used it that way, went smoothly. As did installation on the Grimm MU1 -- just followed the MU1 manual step by step. But we have our preferences...
 
This advice may not apply to everyone who already has a DAC they enjoy but if your currently in the market for a new DAC I would not buy one today unless it had an Ethernet Input. With an Ethernet Input at the DAC itself there is no need for extra boxes or their associated expense (ie..a physical streamer) in the signal chain.

I've seen some say to avoid a PC as the source but I disagree. My current DAC is Ethernet based only (Merging HAPI) and I feed it directly from my music server using a dual Network Card setup on the server. One NIC pulls from the Network to pickup files stored on a NAS or from streaming services and the other NIC is directly connected to my DAC via an Ethernet cable. No middle man in the way and I couldn't be happier with the sound or almost as importantly, its reliability.

But, short of having a DAC with Ethernet Input though, and assuming your only interested in 2 Channel audio, I have found some "Dumb" external streamers to do an excellent job (from a SQ standpoint) compared to USB direct from a computer or whatever else to the the DAC.

I've tried a tiny ALIX computer with 256MB of RAM running from a laboratory grade external power supply at like 9volts running MPD Linux, to a much higher dollar Bricasti M5 and several other options in between that price range. I personally dont see the need for the fancy/expensive GUI based models. You could just use Roon's well fleshed out interface to control any of these lower cost choices. The simpler the better IMO.
 
  • Like
Reactions: Lee
I can't answer what's, in the box, what makes them better, but streamers are like any other component in that they improve generally as you get better quality units.

I began with a Denon HEOS, then Bluesound Node, then NAD M50.2 and now my Aurender. The Aurender is in an entirely higher league when it comes to sound quality, The prior 3 are in the order of quality IMO. The Aurender was the most expensive but the improvement is some of the best money spent. I mean the Aurender is like hi fi making the rest mid-fi, in the case of he HEOS and Node not even Mid-fi.

My friend had a dedicated computer with all unnecessary stuff removed and tailored for just streaming, he tried an Aeries G1, the improvement was so much he took out the computer and bought a G2.

People thought a $200.00 CDP was as good as any, it's all 1's and 0's, until they found different. If you had a $200 CDP and never heard/tried anything better..... Speculate all you want but if you try, experience will show you the truth.

I too will be interested in hearing Ron's experience with the Pulse.
 
  • Like
Reactions: joaovieira
I try to answer the question of what makes a streamer, a high quality one.
There are many technical characteristics that differentiate the quality of a streamer:
a very important element is the power supply since the DNA of these machines is based and developed in any case on an operating system which is nothing more than a skein of applications and therefore of programs in continuous necessary execution, these compete for time with the CPU which will never be able to devote itself exclusively to the application that manages the audio player. Paradoxically, everything goes as if the only running program, which is what interests us, was running on a less powerful machine than its theoretical potential.
That's why the linear power supply in high performance streamers needs to be taken care of: I recently heard one that has three custom transformers that provide four main power rails from which all the separate voltages are regulated for each part of the streamer. The reasons that lead to the use of three (smaller) transformers instead of perhaps a larger one are justified precisely in the "unstable nature" of the system, in fact the smaller transformers are more stable in the presence of high current oscillations (typical of PCs) and have less hum during peaks. In machines built with "intelligence" it is ensured that each transformer is used for its dedicated operating area: One for the processor, one for the motherboard and one for the peripherals. Since a streamer power supply has a high peak current load and all the currents alas add up in the core of the transformer interacting with each other, separating the areas of action decreases the interaction and improves the quality of the final sound.
Precisely from this point of view, in these products, the use of high inductance power transformers characterized by a low field saturation in the core is aligned and the application of filters to keep the peaks to a minimum and therefore the noise... . how to say the less trash enters and the less I have to filter at the end!! Obviously the same design philosophy is pursued for the loads that join the power source and by interacting produce harmonics and intermodulation noise.

This is the reason that drives "some manufacturers" to implement separate tracks for dedicated areas and, in the best cases, a separate regulator for each load. Of course, regulators also have noise which is a factor in the final power supply output signal. But actually what is more "important" are the noises of the main power supply and the noise generated by the rectifier and the inevitable current peaks of the capacitors.
All these sources of noise must be mitigated in some way (since they cannot be completely eliminated..) and then also in this case adequate and well-made solutions are needed such as the use of special low inductance transformers, Schottky diodes, localized current capacitors, shielded power planes on PCBs, wiring, etc,etc.
Obviously digital without watches does not exist! Its role is FUNDAMENTAL and the same attention goes in particular to phase noise. To maximize performance, the streamer motherboard can be equipped with two clocks, the system clock (chipset) and the processor clock. The system clock (chipset) is required to synchronize all components/clocks on the motherboard. This clock also synchronizes the clock for the CPU, but the CPU clock is usually only used on the chip (PROCESSOR) itself. Since the CPU has to do more things at a time than the motherboard, the clock frequency will be multiplied in the CPU.
Again the general criterion in well-designed machines is to separate using a clock on well-chosen key areas whose implementation has the purpose of lowering the mentioned noise consequently making the jitter lower in the whole system considerably increasing the quality of the listening experience .

Last but not least is the software: particularly noticeable are the heavily modified versions of Arch Linux (low latency / realtime kernel / headless). The advantage of using Linux is that you start from scratch and configure only what you really need for audio playback.
The reasons that determined the choice of some operating systems, mainly reside in the "guarantee" of greater stability.

Linux is a very stable and reliable operating system that can be modified substantially. Furthermore, and most importantly, with Linux it is possible to achieve incredibly low latency (audio latency and processor latency) and latency is one of the key elements to obtain a sound really fluid and natural; the lower the latency, the better the overall playback quality. That's why the buffer size is always turned off or set to the lowest possible setting in the playback software (always dealing with a bunch of applications running all the time!!)
So what differentiates a high-end streamer from other products is the quality of the project capable of intelligently minimizing the weak points of these machines which derive from their very nature or that of being in fact PCs dedicated to music. It is therefore inevitable that the problems of these implementations involve issues such as the player used or the CPU not "able" to follow the rhythm imposed by the DAC, or not perfect in respecting the deliveries of the samples to the DAC in the timing imposed by the frequency of sampling, not to mention the injections of jitter on the USB data, therefore of signal variations that have roots in problems that specifically concern the hardware of these configurations.
Not all streamers on the market achieve these objectives or rather, obtain results such as to obtain an excellent level of reproduction quality. This is inevitably paid for in the final price.
 

About us

  • What’s Best Forum is THE forum for high end audio, product reviews, advice and sharing experiences on the best of everything else. This is THE place where audiophiles and audio companies discuss vintage, contemporary and new audio products, music servers, music streamers, computer audio, digital-to-analog converters, turntables, phono stages, cartridges, reel-to-reel tape machines, speakers, headphones and tube and solid-state amplification. Founded in 2010 What’s Best Forum invites intelligent and courteous people of all interests and backgrounds to describe and discuss the best of everything. From beginners to life-long hobbyists to industry professionals, we enjoy learning about new things and meeting new people, and participating in spirited debates.

Quick Navigation

User Menu