advertisement


Streaming transports SQ differences?

Why are thinking that data itself is place where is the trouble. To make and maintain conection there are several levels under data and protocols where is also quite possible to run in trouble. That is still electrical connection. There is no zeros and ones in physical level, they are only in your head.
 
gints said:
Why are thinking that data itself is place where is the trouble. To make and maintain conection there are several levels under data and protocols where is also quite possible to run in trouble. That is still electrical connection. There is no zeros and ones in physical level, they are only in your head.
Click to expand...

There are zeros and ones at the output of the receiver at the end of the connection. If you use an optical connection (S/PDIF, toslink) there is no electrical connection.
 
Julf said:
Yes, in extreme cases you do indeed get transmission/data errors, but I don't think that is what we are discussing.
Click to expand...

I am not really thinking of bits being flipped or lost here. I am thinking of more progressive deleterious effects like jitter and wander.

When digital audio is packetised and sent over a network, both the sample rate and the samples are encapsulated in the bits. Typically those bits are sent using TCP/IP. The TCP part ensures that all of the packets are received intact or an error is reported. When digital audio is sent as PCM or DSD from a streamer to a DAC, only the samples are encapsulated in the bits themselves. The sample rate is determined by the rate at which bits flow between the streamer and the DAC. Left unchecked, variances in the timing of the bits will alter the analogue waveform which the DAC outputs. It has been argued that a competently designed DAC will deal with these issues, but the reality is that some DACs are better than others at recovering from these errors.

I would argue that it is best not to introduce such errors to your playback chain at all, if you can help it. You can do that by adding a suitable HAT to your Pi or by investing in a purpose built streamer like the Auralic Aries G1.
 
Strictly Stereo said:
When digital audio is packetised and sent over a network, both the sample rate and the samples are encapsulated in the bits. Typically those bits are sent using TCP/IP. The TCP part ensures that all of the packets are received intact or an error is reported. When digital audio is sent as PCM or DSD from a streamer to a DAC, only the samples are encapsulated in the bits themselves. The sample rate is determined by the rate at which bits flow between the streamer and the DAC. Left unchecked, variances in the timing of the bits will alter the analogue waveform which the DAC outputs. It has been argued that a competently designed DAC will deal with these issues, but the reality is that some DACs are better than others at recovering from these errors.
Click to expand...

Yes and no. The data clock is not necessarily tied to the audio clock even with USB (it is with S/PDIF). Older, more primitive DACs often derived their audio clock from the data clock (using a phase-locked loop) to avoid running out of synchronisation, but with modern ASRC technology, the clock domains can be kept totally separate.
 
Julf said:
There are zeros and ones at the output of the receiver at the end of the connection. If you use an optical connection (S/PDIF, toslink) there is no electrical connection.
Click to expand...
Zeros and ones are only for human perception, so he can visualize and work with it. For toslink it is light pulses, for cd pits, for hdd magnetic charge and so on. Yes, toslink does not have electrical connection, but still light transmission can be not perfect and it has to converted to electrical signal to work with. Ethernet also is galvanic isolated, that does not make it perfect. There is more to data transmission than just data. Even firmware and software used can give different results.
If engineer do not understand why there should be differences between units, it does not mean that there is not. Even if somebody claim a difference, proper scientisc would like to know why. All human knowledge did not drop on our heads one day, things we did not know yesterday, we know today. Maybe tomorrow will bring something new.
 
The post-truth era comment earlier sparked a vision for myself. A couple of posters here are doing the equivalent of wearing a sandwich board with a message telling others of the errors of their ways - "the end is nigh" if we don't reform. It's a modern internet keyboard warrior version but much the same.
 
When choosing a streaming transport you obviously have to look at your priorities; mine is that it must sound as good as possible streaming Tidal (so not locally stored FLAC or WAV files) over Wi-Fi (not connected Ethernet). Just about all my music is sourced via Tidal nowadays - thus making any UI simple enough to work with even the oldest streamers (aka the Marantz NA7004 - because I rate it as a transport - streaming Tidal to it via BubbleUPnP).

I'm however a simple example. I don't bother ripping and storing files (although I am quite capable of doing so).

I've not tried Roon, but other than that I rate Bluesound's UI as the best and most stable).
 
indus said:
Thanks.
Perhaps I didn't explain my circumstances earlier or perhaps it got lost in the thread.

Its a bit boring but please have a read.

I already have a Simaudio 380d streaming dac. I recently acquired a Trinnov ST 2 HiFi for room correction. The 380d has a digital monitor loop for incorporation of RC devices and thats what I was using so

380d digital out into Trinnov, digital out of Trinnov back into 380d for D to A. My CDT also plugs into the 380d.

However I suspect the monitor loop on the Moon isn't that well implemented. The CDT sounds better if plugged straight into the Trinnov and then allow it to do D to A.

So bottom line is I no longer want to use the dac in the 380d /use its monitor loop. I can plug my cdt and a streamer straight into the Trinnov.

The 380d has the older MiND 1 which is ok but not great as an app. It is also non dsd (not sure if I even need this) and non Roon. I gather MiND2 is much better.

So hence my idea to trade the 380d for a new streamer with all the bells and whistles.

However, the 380d can be upgraded to MiND2 with Roon (but not dsd) for circa £800.

Initial enquiries suggest my 380d might only be worth circa £1k for trade in. So if I were to buy a £2k streamer I'd have £1k to pay. The alternative is that I upgrade the 380d for £800 and use its digital output and not its dac.

I would need to test its digital output to make sure its ok. I will have a dac which I don't need right now but who knows what the future may bring. It also has a multitude of inputs inc usb etc.

So either way I'm looking at £800 to £1k.

Now that I've written it all out perhaps the obvious answer is to upgrade the 380d and bypass its dac....once I've convinced myself that its digital output is well implemented.

Thanks for reading all that!
Click to expand...



Guys, only Amber Audio has actually given his opinion on these two options. The greater debate is interesting but I would like some opinions on this if possible.

Thank you
 
gints said:
Zeros and ones are only for human perception
Click to expand...

No, zeros and ones are for any digital circuit. No matter what the analog value on the transmission line is, it is either below or above the threshold value between "0" and "1", and the receiving circuit will determine that.
 
Strictly Stereo said:
Noise absolutely can be and is digitised, packetised and sent over Ethernet networks. I think what you mean to say is that the process of moving it across an Ethernet network does not add noise. Remember also that sending digital audio from a streamer to a DAC is very different from sending it from a NAS or streaming service to a streamer.
Click to expand...

You are wrong on several counts there and your post shows a lack of understanding.
 
gints said:
Why are thinking that data itself is place where is the trouble. To make and maintain conection there are several levels under data and protocols where is also quite possible to run in trouble. That is still electrical connection. There is no zeros and ones in physical level, they are only in your head.
Click to expand...

That's like saying the wind doesn't exist or gravity , cos yo can't see it - you can only observe its affects.

This is incorrect. All the computers on the planet, be them in washing machines or poncey hi fi, operate on a 1 or Zero be nary code stream. (Well some do hexadecimal but let's keep it SI ole) Data moves in streams, chunks or packets from storage devices to CPU or chipsets to peripheral devices and onto access/ listening devices, using electricity either optically or via wire.
 
guest said:
That's like saying the wind doesn't exist or gravity , cos yo can't see it - you can only observe its affects.

This is incorrect. All the computers on the planet, be them in washing machines or poncey hi fi, operate on a 1 or Zero be nary code stream. (Well some do hexadecimal but let's keep it SI ole) Data moves in streams, chunks or packets from storage devices to CPU or chipsets to peripheral devices and onto access/ listening devices, using electricity either optically or via wire.
Click to expand...
Is SI ole a Spanish computer language?
 
guest said:
You are wrong on several counts there and your post shows a lack of understanding.
Click to expand...

I don't think so but you're welcome to offer any corrections you think are necessary. When you digitise analogue audio, the converter cannot tell the difference between the audio you want and the noise you don't want. All of it is digitised. You can then split it up into packets and send it somewhere else via Ethernet. The process of sending it over an Ethernet network won't add more noise to the digitised audio, thanks to error detection and correction mechanisms which guarantee that either bit perfect audio or an error message is received. When you send digital audio from a PC, streamer or other digital transport to a DAC over USB or S/PDIF, it has to be formatted differently. There is no error correction mechanism here only error detection.
 
Julf said:
No, zeros and ones are for any digital circuit. No matter what the analog value on the transmission line is, it is either below or above the threshold value between "0" and "1", and the receiving circuit will determine that.
Click to expand...
As per gints post zeros and ones are for human perception. Digital/computers are usually dual-state machines. We ascribe zero and one for the two states for our convenience. There is the possibility to have multiple-state machines and these have been considered. Clearly in that case we'd need to develop out of zeros and ones.
 
clivem2 said:
As per gints post zeros and ones are for human perception.
Click to expand...

Not really. They are for the logic gates of the processor and DAC.

Digital/computers are usually dual-state machines. We ascribe zero and one for the two states for our convenience.
Click to expand...

Not just our convenience, but also to be able to use boolean algebra.

There is the possibility to have multiple-state machines and these have been considered. Clearly in that case we'd need to develop out of zeros and ones.
Click to expand...

Not only has it been considered, but it has actually been done. Many times. There have been decimal computers and octal computers, and probably other number bases. Many newer flash memory chips use 4-state cells. A friend of mine is working on ternary logic for quantum computers.
 
Julf said:
Not really. They are for the logic gates of the processor and DAC.



Not just our convenience, but also to be able to use boolean algebra.



Not only has it been considered, but it has actually been done. Many times. There have been decimal computers and octal computers, and probably other number bases. Many newer flash memory chips use 4-state cells. A friend of mine is working on ternary logic for quantum computers.
Click to expand...
1 or 0 can also be described as on or off, it frankly matters little what names we use, we could use A and B. Boolean algebra again is a human construct, it's the way we choose to manipulate our our human defined 1 and 0.

Anyway enough from myself...the thread has already been pruned once.
 
indus said:
One side insists its all ones and zeros so no difference if using a half decent dac. On the other hand there are reviews from people like Donnie Darko where he clearly states that the Auralic Aries G1 sounds better than a Blue sound and that the Innuous Statement sounds better than the G1.
Click to expand...
in this Darko video at 10mins he states that when he compared the dCS Bartok's internal streaming section to the Bartok DAC fed by an Auralic streamer he heard no difference (via headphones, he claims that he did not try it via speakers.)
 
You must log in or register to reply here.


advertisement


Back
Top