Streaming music using Google Chromecast Audio

To me it is sillyness that we judge networks based on their Gbps and not latency.

However I've not invested any time in all the kind of problems around on streaming, I feel that this is the problem with many streaming implementations: the incapability to cope with variation of incoming traffic - therefore need for buffering / reclocking.

Good that Naim is going to rely on Google Cast etc ...

Huge posted:

Usually the lack of a clean digital edge puts uncertainty into the timing (source jitter and buffer timing uncertainty), and add uncertainty into the power supply noise (broadband noise).  Both these reduce the ability of a DAC to produce a clean analogue signal.

This digital signal is still "bit perfect", as "bit perfect" doesn't take into account errors in timing or sub digital level errors (which affect timing and noise) - it only accounts for deviations that reach the full bit level, even though smaller errors can also degrade the resultant analogue signal.

I guess this is very much DAC dependant. Not sure this is an issue with Chord DACs as everything gets re-timed against a low jitter clock, so source jitter is completely removed. I remember reading the AP test equipment Rob Watts uses actually measures lower jitter levels on Hugo optical than the coaxial inputs.

Is not noise from coax more of an issue on balance? Very much source and cable dependant, unlike optical.

Halloween Man posted:

Huge posted:

Usually the lack of a clean digital edge puts uncertainty into the timing (source jitter and buffer timing uncertainty), and add uncertainty into the power supply noise (broadband noise).  Both these reduce the ability of a DAC to produce a clean analogue signal.

This digital signal is still "bit perfect", as "bit perfect" doesn't take into account errors in timing or sub digital level errors (which affect timing and noise) - it only accounts for deviations that reach the full bit level, even though smaller errors can also degrade the resultant analogue signal.

I guess this very much DAC dependant. Not sure this is an issue with Chord DACs as everything gets re-timed against a low jitter clock, so source jitter is completely removed. I remember reading the AP test equipment Rob Watts uses actually measures lower jitter levels on Hugo optical than the coaxial inputs.

It's a fundamental problem with ALL DACs.

The underlying problem is mathematical in nature; you can reduce the effects but not eliminate them.  Then there's also the power supply coupling issue, again this can be reduced, but with current technology it can't be eliminated. 

I would love to see that Naim releases a classic series streamer only - without Dac. To be paired with the nDac or Dac V1.

There is currently a reference level nDac, but no brother for it to take care of the streaming part (whispering microRendu here).

Huge posted:

Halloween Man posted:

Huge posted:

Usually the lack of a clean digital edge puts uncertainty into the timing (source jitter and buffer timing uncertainty), and add uncertainty into the power supply noise (broadband noise).  Both these reduce the ability of a DAC to produce a clean analogue signal.

This digital signal is still "bit perfect", as "bit perfect" doesn't take into account errors in timing or sub digital level errors (which affect timing and noise) - it only accounts for deviations that reach the full bit level, even though smaller errors can also degrade the resultant analogue signal.

I guess this very much DAC dependant. Not sure this is an issue with Chord DACs as everything gets re-timed against a low jitter clock, so source jitter is completely removed. I remember reading the AP test equipment Rob Watts uses actually measures lower jitter levels on Hugo optical than the coaxial inputs.

It's a fundamental problem with ALL DACs.

The underlying problem is mathematical in nature; you can reduce the effects but not eliminate them.  Then there's also the power supply coupling issue, again this can be reduced, but with current technology it can't be eliminated. 

+ the inprecision of floating point numbers.

Huge posted:

Halloween Man posted:

Huge posted:

Usually the lack of a clean digital edge puts uncertainty into the timing (source jitter and buffer timing uncertainty), and add uncertainty into the power supply noise (broadband noise).  Both these reduce the ability of a DAC to produce a clean analogue signal.

This digital signal is still "bit perfect", as "bit perfect" doesn't take into account errors in timing or sub digital level errors (which affect timing and noise) - it only accounts for deviations that reach the full bit level, even though smaller errors can also degrade the resultant analogue signal.

I guess this very much DAC dependant. Not sure this is an issue with Chord DACs as everything gets re-timed against a low jitter clock, so source jitter is completely removed. I remember reading the AP test equipment Rob Watts uses actually measures lower jitter levels on Hugo optical than the coaxial inputs.

It's a fundamental problem with ALL DACs.

The underlying problem is mathematical in nature; you can reduce the effects but not eliminate them.  Then there's also the power supply coupling issue, again this can be reduced, but with current technology it can't be eliminated. 

It seems to me that timing effects could be completely eliminated if DACs would start converting a stream only after it has been completely stored in memory. This would not work with internet streams and it would imply perceivable delays of "start playing" events on LAN data, of course. But it would make timing errors in the incoming stream irrelevant. Of course, converting the digital data to analog signals would still result in unavoidable timing errors but this is another story, I think.

Correct me if I'm wrong but 

Joshua Huang posted:

Recently I bought a Google Chromecast audio out of curiosity while it was on sale, but as soon as I tried it with my headphone (AKG Q701, Sennheiser HD598, Senneiheiser Amperior...) using Google play music streaming, I can't believe how good the sound quality and how much power it has! I know part of the reason is the music is directly streaming from Google play music sever without going to my smart phone (Sony Xperia Z3) first, but the quality of the streaming music is just too good to believe, so smooth and so much details. I am surprised by the power of the Chromecast audio as well, to be honest, it sounds it has much more power than my trusted FIIO E12 and even sounds better, especially when push my AKG Q701, it really makes this headphones sing out loud! So I decided to buy a second Chromecast audio just for my Naim SuperNait for listening to the streaming music, wow! almost feels like listening to a CD!  I first tried connect to the SuperNait's AUX input, sound really good, then tried connect to the DIN input 1 using a 3.5mm/DIN cable from Chord, even better! the last I tried is using a toslink cable manufactured by DH Labs, also good as well. I haven't tried more expensive streaming device like SONOX connect or the streamers from Naim, and wondering how Google Chromecast Audio can compare to them? Anyone has tried?

Correct me if I'm wrong, but if you're using Google Play streaming, then you've subscribed/downloaded from there right? In which case Google Play is only 320kbit MP3. And I have to say some albums I bought sound really good - better than anything on iTunes/Apple Music, as far as lossy goes.

Huge posted:

Halloween Man posted:

Huge posted:

Usually the lack of a clean digital edge puts uncertainty into the timing (source jitter and buffer timing uncertainty), and add uncertainty into the power supply noise (broadband noise).  Both these reduce the ability of a DAC to produce a clean analogue signal.

This digital signal is still "bit perfect", as "bit perfect" doesn't take into account errors in timing or sub digital level errors (which affect timing and noise) - it only accounts for deviations that reach the full bit level, even though smaller errors can also degrade the resultant analogue signal.

I guess this very much DAC dependant. Not sure this is an issue with Chord DACs as everything gets re-timed against a low jitter clock, so source jitter is completely removed. I remember reading the AP test equipment Rob Watts uses actually measures lower jitter levels on Hugo optical than the coaxial inputs.

It's a fundamental problem with ALL DACs.

The underlying problem is mathematical in nature; you can reduce the effects but not eliminate them.  Then there's also the power supply coupling issue, again this can be reduced, but with current technology it can't be eliminated. 

Chord and Rob Watts claim their pulse array dacs are immune to source jitter and recommend optical over coax in their portable dacs as it's fully isolated, unlike coax which is very much cable and source dependant. They obviously do not consider source jitter or any power supply coupling issues in their designs to be an important factor when choosing between coax or optical.

With my Hugo TT dac and battery powered laptop I'm unable to distinguish between USB with galvanic isolation and optical.

Halloween Man,

Please look into and get to understand 'sampling theory', digital gate operation at the analogue level (individually and en-mass), and FIR filters - the problem is an information one and is inherent in the fundamental nature of the information processing.

Optical or digital coax has no specific bearing on this aspect of information processing, they are simply implementation technologies for the data transport.  Although both transport technologies have different limiting characteristics that result in non-ideal implementation of the transport, and hence cause different down-stream concerns in the engineering implementation of the DAC.

Hello Huge, it's not my job to be concerned with such matters. However, Mr Watts, whose job it is to be concerned by such matters, considers optical is not the inferior option for his portable designs.

Anyway, we digress, this topic is about Chromecast Audio. Apologies to the op.

Incidentally I never claimed that a well implemented S/Pdif optical interface can't be superior to an inferior electrical S/Pdif interface, just that it can't be immune from timing issues (e.g. jitter, poor rise-times, frame rate variation [an internal effect of longer term jitter] or noise in the signal).  N.B.  Claiming 'immunity' requires to result to be completely unaffected no matter how severe the jitter in the incoming signal.

Actually HM, it is relevant to the discussion about CCA, because if true, then there's no point is using a streamer any more sophisticated than a CCA, as the Chord DAC's input will sort out ALL the non-idealities in the optical signal.  Thus there's no point in having an NDX, Auralic, microRendu etc, just use a CCA and a Chord DAC, connect it via optical and all the problems disappear.  Furthermore, the only CD transport you'll ever need is a cheap DVD player - same logic applies.

Although I think the CCA is outstanding value for money, I don't think it's a universally perfect solution (even with a Chord DAC), I still think that some other streamers can (and do) outperform it.

I do wonder if Chord consider this to be a non issue though - whilst i agree Toslink does remove a whole host of issues it doesn't make the DAC source agnostic (as you rightly say). Given RW's DPA background, i'm surprised that Chord haven't gone back to something like the Deltec Deltran sync system where the DAC provided the clock to the transport (all via Optical) - IIRC Linn did a similar thing with the Numerik drive and converter albeit with coax.  

Huge posted:

Incidentally I never claimed that a well implemented S/Pdif optical interface can't be superior to an inferior electrical S/Pdif interface, just that it can't be immune from timing issues (e.g. jitter, poor rise-times, frame rate variation [an internal effect of longer term jitter] or noise in the signal).  N.B.  Claiming 'immunity' requires to result to be completely unaffected no matter how severe the jitter in the incoming signal.

Actually HM, it is relevant to the discussion about CCA, because if true, then there's no point is using a streamer any more sophisticated than a CCA, as the Chord DAC's input will sort out ALL the non-idealities in the optical signal.  Thus there's no point in having an NDX, Auralic, microRendu etc, just use a CCA and a Chord DAC, connect it via optical and all the problems disappear.  Furthermore, the only CD transport you'll ever need is a cheap DVD player - same logic applies.

Although I think the CCA is outstanding value for money, I don't think it's a universally perfect solution (even with a Chord DAC), I still think that some other streamers can (and do) outperform it.

Hi Huge, if I remember correctly, immune was the word RW used, I don't think he meant it literally. My interpretation was that jitter levels were so comparably extremely low whether it was optical, coax, or usb, What was of greater importance was noise getting injected into the DAC. Optical is completely isolated from noise hence his recommendation for his portable DACs where no USB with galvanic isolation is available. RW does go on to say that for his DACs with USB input that have galvanic isolation the sq by a tiny margin is better than optical when using a battery powered laptop due to isochronous asynchronous USB mode allowing the FPGA chip to supply timing data to the laptop.

True, I suppose this discussion is relevant as Google Chromecast Audio has an optical output. I'm in the camp where there's no point in using a streamer any more sophisticated than a CCA if connected via optical to a well designed DAC such as my own Hugo TT and the data is bit perfect. Not bad for £30.

I've owned the Chromecast Audio for a year now, and while I would not dispute that it is remarkable value for money, and a really convenient thing to use on occasions, I find the SQ is notably less refined and clumsy somehow feeding a DAC-V1 via optical, than the dedicated Windows PC I have feeding the same DAC via USB.  That's using Qobuz lossless streaming for comparison.  I wouldn't describe the Chromecast as awful, or even bad, as some have on similar threads, but definitely sounds much less capable and less enjoyable than the solution I've become used to for Qobuz streaming.  As a result mine has had very little use, but has been handy to have on a couple of occasions.  The reason why it sounds so different I don't know, but said difference is obvious enough I don't feel the need to worry about whether I am imagining it, unlike some other comparisons I've tried to make in the weird world of digital transports.

Songstream, can you be sure that the audio is bit perfect to the V1 from the optical out of the CCA? I know the V1 has a bit perfect test. Would be interesting to know if both USB and optical pass the bit perfect test. CCA only handles certain formats bit perfect, others it will not play at all or convert. Maybe worth checking.

If USB and optical sound different when both are sending the same bit perfect data to V1 then one could argue one of the reasons may be down to the V1 being very sensitive to jitter via optical. From my listening tests with Hugo TT it's impossible for me to reliably tell optical and USB apart.

Halloween Man posted:

Songstream, can you be sure that the audio is bit perfect to the V1 from the optical out of the CCA? I know the V1 has a bit perfect test. Would be interesting to know if both USB and optical pass the bit perfect test. CCA only handles certain formats bit perfect, others it will not play at all or convert. Maybe worth checking.

If USB and optical sound different when both are sending the same bit perfect data to V1 then one could argue one of the reasons may be down to the V1 being very sensitive to jitter via optical. From my listening tests with Hugo TT it's impossible for me to reliably tell optical and USB apart.

It's a shame, because I would be curious to test that very thing, but the bit-perfect test on the DAC-V1 is only for the USB input.  I know the PC passes all the tests, but all I know about the Chromecast is that it does pass the data over in the native bit-depth and sample rate of the stream, within its hardware boundaries anyway.  Correct sample rate etc alone does not make it for bit-perfect output necessarily, of course.  While I know the PC is bit-perfect capable, and passes with a couple of different players that I have used, there are some applications that fail the test even though the source bit-depth and sample rate are maintained, so the test function is still an interesting and useful thing to have, but limited to USB sources.  Some applications must have software DSP processing that alters the final output it seems.  I guess there is no reason why some similar processing couldn't take place within a chromecast, which may even be why there is difference in SQ, though that difference might be engineered and preferable to some.  Who knows?

I've picked one up but whilst the sounds rather good it seems a bit limited.(Its plugged into a headphone amp/headphones as an alternative to an impractical and long lead from the mainsystem) Theres a samsung tablet here aswell as an ipad. It's day one and little time to tinker but I can play off an iplayer radio app or spotify app but I cant seem to get anything to play off stored music on either or off the nearby mac mini..... I guess I have to try to add files to spotify or something similar but none of this seems that simple. Its not as plug and play as i thought!

Musicstreamer lets my my iPhone find my NAS and point it at the Chromecast. It has always worked well for me. 

SJW,

Using the Android tablet...

For music on the device, cast the screen and use whatever music player you normally use.

For music on a DLNA server you can use BubbleUPnP to send it to the CCA, or you can cast the screen and use any Android UPnP player.

Thanks - in fact what I've done is ask my teenage son who uses google play on his phone...I've uploaded 1004(!) albums onto the my library section of google play and seems to play very well into a schitt headphone amp /audio technica headphones. Uploading 260 gb of music took a couple of days. Its to listen to music when he's supposed to be doing his homework....

At 30 quid its a real bargain and I'm very pleased with the end result.

I don't know if others are finding this but in our house its him (17) that's telling me (56) to turn the music down through speakers. His generation seems to just use headphones!

sjw posted:

I don't know if others are finding this but in our house its him (17) that's telling me (56) to turn the music down through speakers. His generation seems to just use headphones!

Yes the same with my 19 and 22 year old... I thinks it's a generational thing... and also perhaps it's easier / cheaper to get an acceptable louder fuller frequency replay on headphones than with speakers.

Simon

its a bloody nuisance when you're yelling upstairs to tell them suppers ready!

does anybody try to feed the Cca with a Linear Power Supply ?

I use a battery/power bank from Amazon, costing about £15.

Also don't cast Tidal from the screen - use the cast icon in the Tidal app.

Dozey posted:

I use a battery/power bank from Amazon, costing about £15.

with a significant improvement ?