Streaming quality
Discussion
Qualitatively, how much weight should you give (or is in-bedded) in the different stages of streaming music.
Taking the following as a basic set-up:
Storage->network->DAC->AMP->Speakers.
AMP/Speakers we can view as passive for this discussion. However, what I'd like to focus on is how important is the quality of you hard-drive or CAT5/Wireless. Which DACs cope better with this set-up through reclocking/buffering trickery? What is the tolerance of the digital signal to its origin in this context?
I know its a broad question, but I stopped listening to CDs years ago and frankly wont return.
Edited as the post got truncated.
Taking the following as a basic set-up:
Storage->network->DAC->AMP->Speakers.
AMP/Speakers we can view as passive for this discussion. However, what I'd like to focus on is how important is the quality of you hard-drive or CAT5/Wireless. Which DACs cope better with this set-up through reclocking/buffering trickery? What is the tolerance of the digital signal to its origin in this context?
I know its a broad question, but I stopped listening to CDs years ago and frankly wont return.
Edited as the post got truncated.
Edited by Ultra Violent on Monday 18th January 16:49
UV
Your hard drive and network connection will be fine, otherwise your Word documents would be all jumbled up every time you opened one. The main criteria with hard drives is to buy decent ones and make sure you have a backup.
I have several Squeezebox devices on my home network, running happily from an Atom based Tranquil server.
I have over the years experimented with many differenct DACs on the back of a Squeezebx, the best one (and the one I currently use) is my Meridian processor, which has a reclocking FIFO buffer and the option to upsample.
Meridian was one of your options from previous threads, and if fed by SB or Sonus or Sooloos should (will) sound the same.
davidy
Your hard drive and network connection will be fine, otherwise your Word documents would be all jumbled up every time you opened one. The main criteria with hard drives is to buy decent ones and make sure you have a backup.
I have several Squeezebox devices on my home network, running happily from an Atom based Tranquil server.
I have over the years experimented with many differenct DACs on the back of a Squeezebx, the best one (and the one I currently use) is my Meridian processor, which has a reclocking FIFO buffer and the option to upsample.
Meridian was one of your options from previous threads, and if fed by SB or Sonus or Sooloos should (will) sound the same.
davidy
Yeah, I've been speaking to the guys at Meridian, and planning to go up when I next head over that way.
I guess my point is that network protocols are typically designed to guaranty delivery of data packets, but with music you obviously have a timing requirement. In simple terms there must be a point where you give-up trying to send a packet and just move onto the next one in an effort to keep in 'sync'. This is obviously more important with wireless where interference can easily inject delays. In bad cases this will cause drop-outs, but there must be some middle ground where you are effectively inducing Jitter.
There is the Cullens Mod for the Sonos, and this is available as a digital only upgrade (i.e. just changes the digital output circuits). I'm not sure I understand the point of that. We must also, by association, be saying that CD transports are basically all the same (from a music point of view)....
I guess my point is that network protocols are typically designed to guaranty delivery of data packets, but with music you obviously have a timing requirement. In simple terms there must be a point where you give-up trying to send a packet and just move onto the next one in an effort to keep in 'sync'. This is obviously more important with wireless where interference can easily inject delays. In bad cases this will cause drop-outs, but there must be some middle ground where you are effectively inducing Jitter.
There is the Cullens Mod for the Sonos, and this is available as a digital only upgrade (i.e. just changes the digital output circuits). I'm not sure I understand the point of that. We must also, by association, be saying that CD transports are basically all the same (from a music point of view)....
Don't most streamers buffer the data so delivery is not infact instant?
I always assumed this was to allow packets that turn up in the wrong order can be stored and rearranged via the inbuild error correcting in data networking. aka Streaming isn't realtime. VoIP is realtime and relies on QoS configurations to give commercial quality. I guess you could apply QoS in this environment but streaming probably doesn't need it due to the built in buffering.
I always assumed this was to allow packets that turn up in the wrong order can be stored and rearranged via the inbuild error correcting in data networking. aka Streaming isn't realtime. VoIP is realtime and relies on QoS configurations to give commercial quality. I guess you could apply QoS in this environment but streaming probably doesn't need it due to the built in buffering.
OK, can someone explain how this can improve Jitter if a network is buffered in the digital domain?
http://www.cullencircuits.com/webapps/site/67005/7...
http://www.cullencircuits.com/webapps/site/67005/7...
Edited by Ultra Violent on Monday 18th January 20:33
But the DAC is a separate device. The Sonos is just the transport. I would spend a lot more on a standalone DAC if I was going that way and i would expect the reclocking/upsampling to be done there to limit Jitter and reduce unnecessary signal paths.
So my point remains the same (i think). Is everything upstream of the DAC equal (within sensible financial limits). Is the only difference between a Mac Book/Sonos/SB/Sooloos/PC/Media Server... the DAC section. Or put another way, what is the threshold of error for the pure digital signal? For how long can I use cheap electronics before I 'have' to spend mega bucks on cables et al, to propagate the musical signal?
So my point remains the same (i think). Is everything upstream of the DAC equal (within sensible financial limits). Is the only difference between a Mac Book/Sonos/SB/Sooloos/PC/Media Server... the DAC section. Or put another way, what is the threshold of error for the pure digital signal? For how long can I use cheap electronics before I 'have' to spend mega bucks on cables et al, to propagate the musical signal?
I am both an audiophile and audio engineer.
I am audiophile enough to spend £3000 on a disc transport and audio engineer enough that when I could stream CDs from a hard drive out of a £400 Mac Mini I did a proper test.
I really, really wanted to convince myself that the previous £3000 transport with Apogee clock and top power supply etc would make a huge difference. Therefore any bias was toward the high end system.
However, in a lot of detailed testing I could not tell the difference in a blind test. Further to that, I couldn't even tell the difference in an open test where I could remote control both sources from the listening position back and forth. One caveat I should point out is that my DAC has a very good clock and buffer in it.
At the moment, my music is streamed in real time through iTunes/Frontrow from CDs ripped as WAVs on an external drive connected to my router which wirelessly streams to my living room to my MacBook Pro and optically to my DAC.
Doing this is no way inferior to the £3000 transport. Once in a blue moon there will be a total drop out - this system will either work, and work through a few buffers or totally fail. No loss of quality in between.
With regard to the upgrade questioned, if your DAC is particularly poor with jittery signals there will be a minor improvement. However, if your DAC is RAM buffered or retimed before conversion then I cannot see how it will make a positive difference at all.
I am audiophile enough to spend £3000 on a disc transport and audio engineer enough that when I could stream CDs from a hard drive out of a £400 Mac Mini I did a proper test.
I really, really wanted to convince myself that the previous £3000 transport with Apogee clock and top power supply etc would make a huge difference. Therefore any bias was toward the high end system.
However, in a lot of detailed testing I could not tell the difference in a blind test. Further to that, I couldn't even tell the difference in an open test where I could remote control both sources from the listening position back and forth. One caveat I should point out is that my DAC has a very good clock and buffer in it.
At the moment, my music is streamed in real time through iTunes/Frontrow from CDs ripped as WAVs on an external drive connected to my router which wirelessly streams to my living room to my MacBook Pro and optically to my DAC.
Doing this is no way inferior to the £3000 transport. Once in a blue moon there will be a total drop out - this system will either work, and work through a few buffers or totally fail. No loss of quality in between.
With regard to the upgrade questioned, if your DAC is particularly poor with jittery signals there will be a minor improvement. However, if your DAC is RAM buffered or retimed before conversion then I cannot see how it will make a positive difference at all.
System would be Sonos 90, Naim DAC, Naim pre/power/hi-cap etc, Meridian Sooloos, or Meridian + Sonos (no Sooloos obviously).
I already a load of Sonos/Naim and like it, but thinking about changing a few things. This may include switching to Sooloos with or without Meridian speakers.
So are you saying, if I have a good DAC, the Cullens mod is pointless on the Sonos?
I already a load of Sonos/Naim and like it, but thinking about changing a few things. This may include switching to Sooloos with or without Meridian speakers.
So are you saying, if I have a good DAC, the Cullens mod is pointless on the Sonos?
Edited by Ultra Violent on Monday 18th January 23:27
Ultra Violent said:
For how long can I use cheap electronics before I 'have' to spend mega bucks on cables et al, to propagate the musical signal?
I think this is the point. There is no musical signal in the network, just data. The file on the server is data, it's passed to the sonos as data and then translated into music. TCP has built in error correction. From wiki - Reliable protocols, such as the Transmission Control Protocol (TCP), guarantee correct delivery of each bit in the media stream. However, they accomplish this with a system of timeouts and retries, which makes them more complex to implement. It also means that when there is data loss on the network, the media stream stalls while the protocol handlers detect the loss and retransmit the missing data. Clients can minimize this effect by buffering data for playback.
I can't see how audio quality is effected by streaming, the only situation is if the buffer fails to keep up. This is likley to totally stall the stream not degrade quality. The only way to avoid this is have high bandwidth dedicated to the music data with QoS enabled. Total overkill, I can stream DVD quality video with 5.1 sound from a server in my office to the PS3 over wi-fi with no issue.
Ultra Violent said:
System would be Sonos 90, Naim DAC, Naim pre/power/hi-cap etc, Meridian Sooloos, or Meridian + Sonos (no Sooloos obviously).
I already a load of Sonos/Naim and like it, but thinking about changing a few things. This may include switching to Sooloos with or without Meridian speakers.
So are you saying, if I have a good DAC, the Cullens mod is pointless on the Sonos?
Yes.I already a load of Sonos/Naim and like it, but thinking about changing a few things. This may include switching to Sooloos with or without Meridian speakers.
So are you saying, if I have a good DAC, the Cullens mod is pointless on the Sonos?
Edited by Ultra Violent on Monday 18th January 23:27
I would challenge anyone with a DAC which buffers the data to pick out the difference between clocks before the data was buffered.
Put it this way,in the very detailed test I detailed above it was impossible to tell the difference whether the source was a £3000 spinning disc player or a £400 Mac running data from a hard disc. If a factor of almost 10 in cost (when you consider the digital cable on the disc player retails at almost the cost of the Mac itself) makes no difference whatsoever, I am willing to bet that a few hundred quid of mods is going to make even less difference.
Spend the cash on a good DAC that retimes the signal and you can feed it anything within reason. That is what it is designed to do.
UV
I have done the same tests as Justin, and happily use a Squeezebox as a source. I too have tried reclocking/dejittering devices between the SB and my Meridian Processor and I could hear no difference.
I still that Sooloos is a big investment, and whilst it has a great user interface it is very expensive, better sound quality IMO would come from keeping the Sonus and spending the Sooloos money on the rest of the kit.
davidy
I have done the same tests as Justin, and happily use a Squeezebox as a source. I too have tried reclocking/dejittering devices between the SB and my Meridian Processor and I could hear no difference.
I still that Sooloos is a big investment, and whilst it has a great user interface it is very expensive, better sound quality IMO would come from keeping the Sonus and spending the Sooloos money on the rest of the kit.
davidy
JustinP1 said:
I would challenge anyone with a DAC which buffers the data to pick out the difference between clocks before the data was buffered.
Is this not a statement of the obvious? Properly implemented buffering and re-clocking will leave no trace of the original timing, so provided the bits are correct (which they will be for any transport) all transports must sound the same.tybalt said:
JustinP1 said:
I would challenge anyone with a DAC which buffers the data to pick out the difference between clocks before the data was buffered.
Is this not a statement of the obvious? Properly implemented buffering and re-clocking will leave no trace of the original timing, so provided the bits are correct (which they will be for any transport) all transports must sound the same.
Although it is obvious to those in the know, I thought it answered the OP's question exactly.
DavidY said:
UV
I have done the same tests as Justin, and happily use a Squeezebox as a source. I too have tried reclocking/dejittering devices between the SB and my Meridian Processor and I could hear no difference.
I still that Sooloos is a big investment, and whilst it has a great user interface it is very expensive, better sound quality IMO would come from keeping the Sonus and spending the Sooloos money on the rest of the kit.
davidy
Out of interest, have you tried the Apogee Big Ben?I have done the same tests as Justin, and happily use a Squeezebox as a source. I too have tried reclocking/dejittering devices between the SB and my Meridian Processor and I could hear no difference.
I still that Sooloos is a big investment, and whilst it has a great user interface it is very expensive, better sound quality IMO would come from keeping the Sonus and spending the Sooloos money on the rest of the kit.
davidy
I think I may have mentioned it previously, and gets rave reviews in the pro-audio world.
That said, it suffers from the same problems as a lot of subjective testing and internet reports:
1) Those who should loudest and post a review have the most extreme views
2) The device by design admits to working better in some systems than others therefore how effective it is totally changes, therefore;
3) 30% say it works due to science. 30% say it can't work due to science. 40% say it is 'snake oil' and marketing. 95% have never tested one with their own ears
I replied to this earlier, but the post got lost in the ether...
I don't think it is necessarily obvious. It very much depends on what the DAC actually does (in terms of clock sync/upsample/buffering/voodoo). My question was really to determine that whilst in the digital domain you are pretty much isolated from signal degradation. Clearly different DACs will sound different, but the important spec is really how it manages the buffer, reclocking, and up-sampling.
I haven't seen (or heard) the other DAC...
I don't think it is necessarily obvious. It very much depends on what the DAC actually does (in terms of clock sync/upsample/buffering/voodoo). My question was really to determine that whilst in the digital domain you are pretty much isolated from signal degradation. Clearly different DACs will sound different, but the important spec is really how it manages the buffer, reclocking, and up-sampling.
I haven't seen (or heard) the other DAC...
Looking at the situation logically, £ saved on a transport, mods to a transport or 3rd party reclocking, could simply be spent on a DAC which happens to retime the signal.
One I would definitely recommend auditioning is the Chord DAC64. Although there is a newer model out the original is still superb. £2500 new, there is a buy it now at £1400 on ebay and one at £750 with no bids with a couple of days to go.
This unit has a RAM buffer as we have discussed. Even more interesting is that there is a switch on the back which turns the buffer off, the buffer with a one second delay and a 4 second delay. In my experience of using the unit over a few years the switch demonstrates the opening of the soundstage in both width and depth.
One I would definitely recommend auditioning is the Chord DAC64. Although there is a newer model out the original is still superb. £2500 new, there is a buy it now at £1400 on ebay and one at £750 with no bids with a couple of days to go.
This unit has a RAM buffer as we have discussed. Even more interesting is that there is a switch on the back which turns the buffer off, the buffer with a one second delay and a 4 second delay. In my experience of using the unit over a few years the switch demonstrates the opening of the soundstage in both width and depth.
Gassing Station | Home Cinema & Hi-Fi | Top of Page | What's New | My Stuff