Streamers with no DAC: worth it?

[Steffen]

Toslink and even coax transport-to-DAC links can be riddled with issues, because they often feed the bit stream – in real-time! – straight into the DAC, so things like timing jitter, pulse width and edge rise times come to bear in full force. This is because D/A conversion is an entirely analog process, with the analog quantity being the time between pulse edges. The data stream, that was impervious to considerations like this, suddenly becomes an analog signal upon hitting the DAC.

 

The only way to control this is for the DAC to store a sufficient number of words (16 bits, 24 bits or whatever it operates on) ahead of time in its own buffer, and fetch them from there for conversion to the beat of its own clock signal. This will diminish any timing and edge rise issues that the bit stream may have suffered during transmission to the DAC. Oh, and as the cited articles mention, it needs its own pristine power supply, away from the nasties of any digital circuitry, because in order to synthesise analog output voltages it must rely on a stable reference voltage.

[Snoopy8]

18 hours ago, awayward said:

Definitely not what OP really wanted, but slightly relevant. I’m a firm believer in one piece of equipment for one task. A jack of all trades cannot perform to the level of an expert in the specific trade IMHO

I had 6 in the chain: streamer + PSU, USB regen + PSU, DAC + PSU. Could have gotten even more e.g. dedicated network switch, external clocks etc.  The complexity & expense of the setup drove me to simplify. 

 

Hence my earlier statement that, in hindsight, I should have gotten a high end all in one....

[rand129678]

On 15/08/2019 at 2:00 PM, o2so said:

I would like to know if, how and why expensive streamers that do NOT incorporate a DAC can improve (or even affect in any way) sound quality.

 

A post by one of the better DAC designers on the planet (see 3rd party measurements of his Dave and Qutest DACs for example).

 

This is very much in line with the John Swenson links I shared earlier. It all comes back to ground plane noise potentially affecting the analogue electronics, as Rob mentions.

 

Edited August 19, 2019 by rand129678

[o2so]

2 hours ago, rand129678 said:

 

A post by one of the better DAC designers on the planet (see 3rd party measurements of his Dave and Qutest DACs for example).

 

This is very much in line with the John Swenson links I shared earlier. It all comes back to ground plane noise potentially affecting the analogue electronics, as Rob mentions.

 

Interesting however this looks like it refers to DAC, not digital streamers

[rand129678]

1 hour ago, o2so said:

Interesting however this looks like it refers to DAC, not digital streamers

 

The very opening words are "the reasons why sources and digital interconnects sound different". And then he gives his nutshell explanation.

 

Streamers are sources.

 

You do have to read through all this stuff very carefully, to see how it all links together (if you're interested in this stuff).

 

Especially John Swenson's 3 part interview I linked. I think I had to read over it 3 times and I'm sure if I read it again now I'd pick up something I missed previous times.

 

So it's all linked and comes back to the ground plane noise John Swenson discusses.

 

As I mentioned in a previous post, when I'm listening to Tidal I don't care about ground plane noise in their overseas servers. But streamers (sources) are typically electrically coupled to your DAC... Unless you have a streamer with optical output.

 

Edited August 19, 2019 by rand129678
Added more info

[o2so]

1 hour ago, rand129678 said:

 

The very opening words are "the reasons why sources and digital interconnects sound different". And then he gives his nutshell explanation.

 

Streamers are sources.

 

You do have to read through all this stuff very carefully, to see how it all links together (if you're interested in this stuff).

 

Especially John Swenson's 3 part interview I linked. I think I had to read over it 3 times and I'm sure if I read it again now I'd pick up something I missed previous times.

 

So it's all linked and comes back to the ground plane noise John Swenson discusses.

 

As I mentioned in a previous post, when I'm listening to Tidal I don't care about ground plane noise in their overseas servers. But streamers (sources) are typically electrically coupled to your DAC... Unless you have a streamer with optical output.

 

apologies, I missed it

[davewantsmoore]

On 15/08/2019 at 2:00 PM, o2so said:

So how can something expensive be more perfect than bit perfect?

"Bit perfect" is sufficient for digital to digital communication.

 

For digital to analogue conversion .... the entire digital signal waveform  (not just whether it is 01010111 in the right order)  becomes critical.

[firedog]

On 15/08/2019 at 11:33 AM, rand129678 said:

 

John covers many of the potential technical mechanisms involved. Showing all of this at the analogue output of a DAC (measurements) is much more difficult.

 

But if you send those 3 links to the designer/s of your own DAC/s, there is a high chance they will agree with everything John is saying there, in terms of potential technical mechanisms at play. Unless they say their DACs are immune to all the issues discussed.

 

It may be worth asking them (the designer/s of your DAC/s) for their thoughts.

Yes, but no one has ever demonstrated that those potential issues actually effect the output of any properly made DAC. See Archimago's response to Swenson: 

https://archimago.blogspot.com/2019/08/musings-demo-why-bits-are-bits-lets-not.html

 

That said, I can think of 4 reasons to get a proper streamer: a) you want to insure against any potential negative effects - just in case and for peace of mind; b) your server/DAC don't have the best, cleanest USB connections, and the streamer does (related to (a)); c) the streamer involved adds flexibility to your setup. For instance the Sonore streamers and their OS give you multiple ways to connect to your system over the network that you might not be able to do otherwise; d) multi room streaming and especially identically timed multi room streaming is easier with some streamers and the software they run (Roon, LMS, etc)

[rand129678]

3 minutes ago, firedog said:

Yes, but no one has ever demonstrated that those potential issues actually effect the output of any properly made DAC.

I hinted this  in that 2nd sentence you quoted... "Showing all of this at the analogue output of a DAC (measurements) is much more difficult."

 

John has never claimed he has measurements to show these potential technical mechanisms. He has said he is  building his own high resolution A-to-D converter, to hopefully show some of these differences...

 

 

Edited August 19, 2019 by rand129678

[Jonny]

Check out John Darko's YouTube channel. He's compared heaps of streamers on there and he's good at explaining why they sound different. 

[firedog]

20 hours ago, Jonny said:

Check out John Darko's YouTube channel. He's compared heaps of streamers on there and he's good at explaining why they sound different. 

He also does sighted listening, which pretty much means we can't really depend of his results.
How many times has he said the less expensive streamers are equal to, or better than a more expensive one? 

[blocky]

I haven't read all the replies on this thread, but I think this is a very good explanation: see below

 

The point to remember is that even so a streamer is digital in and digital out, the actual signal is always analog. We only infer it as being digital because we assume it will be interpreted that way and that the interpretation is perfect. 

 

Being an engineer with a Telecom background, what I find confusing is why don't DAC's report their errors. There is (apparently) enough redundancy in a PCM signal for a DAC to KNOW it has made an error. If a DAC would flash an LED every-time it made a decoding error in realtime, we could use that to know when to stop spending money on cables! 

 

I guess I just answered my own question.

 

 

 

[o2so]

On 19/08/2019 at 2:00 PM, rand129678 said:

 

The very opening words are "the reasons why sources and digital interconnects sound different". And then he gives his nutshell explanation.

 

Streamers are sources.

 

You do have to read through all this stuff very carefully, to see how it all links together (if you're interested in this stuff).

 

Especially John Swenson's 3 part interview I linked. I think I had to read over it 3 times and I'm sure if I read it again now I'd pick up something I missed previous times.

 

So it's all linked and comes back to the ground plane noise John Swenson discusses.

 

As I mentioned in a previous post, when I'm listening to Tidal I don't care about ground plane noise in their overseas servers. But streamers (sources) are typically electrically coupled to your DAC... Unless you have a streamer with optical output.

 

So have these differences in the ground plane noise been measured? If they can be heard, they can be measured, right? Surely if they are the key difference between cheap and expensive streamers, the manufacturers of expensive streamers would have all interests in proving that there is a point in spending 10 times more and would have made accurate measurements?

[rand129678]

14 minutes ago, o2so said:

So have these differences in the ground plane noise been measured? If they can be heard, they can be measured, right?

 

Very early in the thread, I replied to you:

 

"John covers many of the potential technical mechanisms involved. Showing all of this at the analogue output of a DAC (measurements) is much more difficult.

 

But if you send those 3 links to the designer/s of your own DAC/s, there is a high chance they will agree with everything John is saying there, in terms of potential technical mechanisms at play. Unless they say their DACs are immune to all the issues discussed.

 

It may be worth you asking them (the designer/s of your DAC/s) for their thoughts."

 

Have you asked the designer/s of your own DAC/s? Have you sent those 3 links to them? If it interests you, it would be worth knowing their thoughts maybe...

 

And just a few posts above I already mentioned:

 

"John has never claimed he has measurements to show these potential technical mechanisms. He has said he is  building his own high resolution A-to-D converter, to hopefully show some of these differences..."

 

 

 

 

 

Edited August 21, 2019 by rand129678
Typo

[kukynas]

On 21/08/2019 at 4:04 AM, blocky said:

I haven't read all the replies on this thread, but I think this is a very good explanation: see below

 

The point to remember is that even so a streamer is digital in and digital out, the actual signal is always analog. We only infer it as being digital because we assume it will be interpreted that way and that the interpretation is perfect. 

 

Being an engineer with a Telecom background, what I find confusing is why don't DAC's report their errors. There is (apparently) enough redundancy in a PCM signal for a DAC to KNOW it has made an error. If a DAC would flash an LED every-time it made a decoding error in realtime, we could use that to know when to stop spending money on cables! 

 

I guess I just answered my own question.

 

 

 

Not sure if you answered your own question but it’s simple, DAC chip as well as receiver chip prior are one way devices, there’s no feedback loop between them or between them and transmitter to say I received error please send the packets again so even if you know about it you can’t do anything with it, and even if you could once the chip clock is locked to the signal it starts processing signal no matter how many error it contains coz signal stream is flowing real time, more expensive DACs contain buffering to fix potential errors in clocking (jitter) but even than they can fix only certain amount of problems coz once the signal contains errors from transmitter prior to receiving it they can’t send message back hey I received error send it again like in case of telecom protocol, DAC chips these days don’t produce much errors by them self but are more prone to input signal quality so designers needs to take into account how clean the signal will be prior to DAC chip to get maximum out of it - just to simplify the overall problem with digital audio processing

 

plenty of info on this or other forums as well as on the web about digital audio signal processing and conversion and all troubles related to it if you are interested

[blocky]

On 24/08/2019 at 2:50 AM, kukynas said:

Not sure if you answered your own question but it’s simple, DAC chip as well as receiver chip prior are one way devices, there’s no feedback loop between them or between them and transmitter to say I received error please send the packets again so even if you know about it you can’t do anything with it, and even if you could once the chip clock is locked to the signal it starts processing signal no matter how many error it contains coz signal stream is flowing real time, more expensive DACs contain buffering to fix potential errors in clocking (jitter) but even than they can fix only certain amount of problems coz once the signal contains errors from transmitter prior to receiving it they can’t send message back hey I received error send it again like in case of telecom protocol, DAC chips these days don’t produce much errors by them self but are more prone to input signal quality so designers needs to take into account how clean the signal will be prior to DAC chip to get maximum out of it - just to simplify the overall problem with digital audio processing

 

plenty of info on this or other forums as well as on the web about digital audio signal processing and conversion and all troubles related to it if you are interested

Hi Daniel, thanks for responding.

I understand that the realtime nature of audio precludes any error correction without the introduction of some form of latency due to the buffering required.

I assume with the majority of DAC's, if the DAC is aware of an  error in the decoding it, will not attempt to correct the error but will probably extrapolate the output from previous samples (or something of that nature). 

My point is that if we are aware of when the DAC does achieve zero-errors, then that is valuable information. At that point the digital part of the whole audio-reproduction equation can do no better. And we can stop stressing about that better Ethernet cables or that audio-friendly-switch that's only $500 more. 

 

I do understand that digital audio can (and I think often is) flawed and not 'bit-perfect'. But it is reasonable to assume that it can be, if implemented correctly. I'm just wondering how rare a bit error is in a typical middle of the road audiophile system? Judging bit-errors (due to jitter, signal degradation, interference from the power supply, whatever) using your ears seems ludicrous when we could know this information. 

[kukynas]

5 hours ago, blocky said:

Hi Daniel, thanks for responding.

I understand that the realtime nature of audio precludes any error correction without the introduction of some form of latency due to the buffering required.

I assume with the majority of DAC's, if the DAC is aware of an  error in the decoding it, will not attempt to correct the error but will probably extrapolate the output from previous samples (or something of that nature). 

My point is that if we are aware of when the DAC does achieve zero-errors, then that is valuable information. At that point the digital part of the whole audio-reproduction equation can do no better. And we can stop stressing about that better Ethernet cables or that audio-friendly-switch that's only $500 more. 

 

I do understand that digital audio can (and I think often is) flawed and not 'bit-perfect'. But it is reasonable to assume that it can be, if implemented correctly. I'm just wondering how rare a bit error is in a typical middle of the road audiophile system? Judging bit-errors (due to jitter, signal degradation, interference from the power supply, whatever) using your ears seems ludicrous when we could know this information. 

Yup

[ikhuong]

 

On 15/08/2019 at 2:00 PM, o2so said:

 

 

I read arguments about reduced noise floor. How can this be possible? A streamer transfers series of 1 and zeroes. How is the noise floor transferred?

 

Thanks to anyone who will contribute.

 

I could hear significant difference from upgrading a desktop -->  Macbook pro with battery -->  of Sonore OpticalRendu with Sbooster Linear power supply. The noise floor is a real deal in digital audio reproduction. If your system is good enough, you definitely hear the change.

 

The 1's and 0's parts are actually carried under analog wave form. Noise floor causes wave form inference and DAC chip can misinterpret them to incorrect values, also no real bit-perfect transferred from streamer to DAC (Bit perfect only happens on data network part where check-sum and resend happens), there is no check-sum design currently in any DAC input protocol (tosink/coaxial/usb/you name it), so if there is a missing bit from streamer --> DAC, it would be a missing information.

 

Bad noise also travels all the way from streamers , passing to DAC chip and go to the analog output of the DAC.

 

So an excellent streamer would have excellent power supply, noise filters implementing in every stage, low-noise processors/CPU itself and excellent clock to control the timing of bitstream transfer

 

 

 

 

 

 

[ikhuong]

On 21/08/2019 at 12:04 PM, blocky said:

 

Being an engineer with a Telecom background, what I find confusing is why don't DAC's report their errors. There is (apparently) enough redundancy in a PCM signal for a DAC to KNOW it has made an error. 

 

 

I think the whole hifi industry does not have enough talents or resource to re-define a new  DAC input protocol that have checksum/resend similar to data/Ethernet protocol, both current PCM/DSD protocols are actually designed by giant IT/electronic firms (Sony/Phillips)  who do not really target for high end audio. All DAC manufactures relies on those prototols on their design

We have some genius DAC designers like Rob Watt or Ted Smith in hifi but there are thousands of thousands genius in IT/Telcom industry. So that is why I would say very limited talented designers worldwide in hifi vs IT industry

Edited September 6, 2019 by ikhuong

[o2so]

I just bumped into this review of the Chromecast audio digital output from audiosceincereview:

https://www.audiosciencereview.com/forum/index.php?threads/review-and-measurements-of-chromecast-audio-digital-output.4544/

 

The conclusions are that:

 

Conclusions
There are a few here:

1. The CAST audio functionality of Google Chrome is horrid. There is no excuse for it to be butchering even simple 16-bit signals as it did. While audibly it is not as dire as it looks, I still would avoid it if you can.

2. Roon's implementation of Chromecast streaming is superb. It is bit accurate up to 24 bits and 48 kHz that I tested. Congratulations to Roon for job well done. I assume they received support from Google to implement it as the protocol otherwise is not open to the public.

3. The Chromecast output has more jitter than an audiophile/instrument grade Toslink output. This is evident when used with low quality DACs like Schiit Modi 2 Uber.

4. Using a well-designed Dac like the Topping D50, there is no difference at all between Toslink from Chromecast or higher fidelity sources. All the jitter is filtered out resulting in the performance of the DAC itself being the limit.

#4 is a great news here. It means that if you have a good DAC and use Roon, you can turn your DAC into a streamer/renderer using the Chromecast Audio. For just $35, that is a superb addition. As such, the combination of Roon and Chromecase audio is highly recommended! 
 

 

So, it appears that, whilst being mildly jittery, the jitter is filtered out by any decent DAC resulting in the chromecast audio performing just as well as a multi million dollar streamer, when used with Roon and a decent DAC. 

 

Can anyone explain how jitter is filtered out by the DAC in layman terms?

[Bush thug]

4 hours ago, o2so said:

I just bumped into this review of the Chromecast audio digital output from audiosceincereview:

https://www.audiosciencereview.com/forum/index.php?threads/review-and-measurements-of-chromecast-audio-digital-output.4544/

 

The conclusions are that:

 

Conclusions
There are a few here:

1. The CAST audio functionality of Google Chrome is horrid. There is no excuse for it to be butchering even simple 16-bit signals as it did. While audibly it is not as dire as it looks, I still would avoid it if you can.

2. Roon's implementation of Chromecast streaming is superb. It is bit accurate up to 24 bits and 48 kHz that I tested. Congratulations to Roon for job well done. I assume they received support from Google to implement it as the protocol otherwise is not open to the public.

3. The Chromecast output has more jitter than an audiophile/instrument grade Toslink output. This is evident when used with low quality DACs like Schiit Modi 2 Uber.

4. Using a well-designed Dac like the Topping D50, there is no difference at all between Toslink from Chromecast or higher fidelity sources. All the jitter is filtered out resulting in the performance of the DAC itself being the limit.

#4 is a great news here. It means that if you have a good DAC and use Roon, you can turn your DAC into a streamer/renderer using the Chromecast Audio. For just $35, that is a superb addition. As such, the combination of Roon and Chromecase audio is highly recommended! 
 

 

So, it appears that, whilst being mildly jittery, the jitter is filtered out by any decent DAC resulting in the chromecast audio performing just as well as a multi million dollar streamer, when used with Roon and a decent DAC. 

 

Can anyone explain how jitter is filtered out by the DAC in layman terms?

Awesome.

 

I also found this.

 

https://darko.audio/2018/07/letters-to-the-editor-3-ways-to-juice-your-chromecast-audio/