Streaming over Ethernet testing

[Anonymous]

So the big question is what was happening to the sound at source (the wav file on the PC) after it had travelled through the 3 methods.

Assuming that technically wired is the most efficient system, no doubt there, is the sound as recorded not as good as it could be, and eg. wifi somehow changes the sound so that certain listeners prefer it?

An analyser on the input to the Olive box might be interesting.

 

[Anonymous]

I mean to say that the test helps to show the difference is not with the means of transmission

In that case, I have a beef with you

"The test shows" (or helps to show) is a plain lie, or a spin. The test shows nothing. An inconclusive test is not a proof to the contrary.

The differences were clear and profound. By the end we could all tell whether set-up 1, 2 or 3 was being used.

You know, that in itself is a notable achievement. I'd think that WHF could make serious waves when they are able to consistently identify differences in digital transmission.

 

[Andrew Everard]

You know, that in itself is a notable achievement. I'd think that WHF could make serious waves when they are able to consistently identify differences in digital transmission.

Actually, our three reader participants did that, so maybe they should be making those waves. But having talked to my colleagues who were involved in the Big Question session in question, it's worth pointing out that the participants didn't know what they were listening to – ie the nature of the test – and were surprised that they'd heard such great differences when it was explained to them (after they had given their views) what was being changed.

 

[AnotherJoe]

The differences are due to how the kit is setup in the test.

Because the laptop is streaming to the streamer (strange setup - but there u are) UDP via RTSP is used. This makes the streaming susceptible to packet loss - so you may hear differences between the 3.

If the streamer was connected directly to the NAS then the protocol used would be TCP/IP and there would be no packet loss, so all 3 would sound the same.

This is one of the potential differences between the 'push' and 'pull' streaming models.

 

[The_Lhc]

The differences are due to how the kit is setup in the test.

Because the laptop is streaming to the streamer (strange setup - but there u are)

Yes I thought that.

UDP via RTSP is used. This makes the streaming susceptible to packet loss - so you may hear differences between the 3.

If the streamer was connected directly to the NAS then the protocol used would be TCP/IP and there would be no packet loss, so all 3 would sound the same.

This is one of the potential differences between the 'push' and 'pull' streaming models.

Ack, someone who knows what they're talking about! I have no idea how to react to this...

 

[Andrew Everard]

If the streamer was connected directly to the NAS

What NAS? The music was stored on the laptop, and streamed from that to the Olive, which was acting as a streaming client/player, connected via analogue to the player.

 

[AnotherJoe]

If a NAS was used instead of the laptop - all 3 would have sounded the same.

(I should have said 'a' NAS - rather than 'the' NAS)

 

[Anonymous]

Progress!

Presumably the wifi link was direct from the Mac into the Olive wireless?

However, how does random packet loss affect the overall sound? And not produce clicks or silence.

http://inst.eecs.berkeley.edu/~ee290t/sp04/lectures/packet_loss_recov_paper11.pdf

 

[Andrew Everard]

Progress!

Hardly. No new information there, assuming you've read the article in question. You have, haven't you?

Presumably the wifi link was direct from the Mac into the Olive wireless?

Yes, in each case, the connection was from the laptop containing the content to the Olive; only the method of transmission changed each time.

 

[Andrew Everard]

If a NAS was used instead of the laptop - all 3 would have sounded the same.

I think that's another triumph of assumption over evidence, but then the 'digits is digits' brigade would have us believe that any digital system will always sound the same, however it's connected. I detect a spot of wriggling on the hook setting in here...

 

[AnotherJoe]

RTSP makes use of error concealment techniques.
Note the word 'concealment' - not correction.

 

[Anonymous]

Progress as in the discussion discussing the technology.

The cable and homeplug used the Olive ethernet port, wifi the Olive aerial? Data into the ethernet port must be different to produce a different sound - agree?

So how are the differences explained technically by engineers?

 

[Andrew Everard]

The cable and homeplug used the Olive ethernet port, wifi the Olive aerial?

Yes, of course. We tried connecting the homeplug via the Wi-Fi aerial, but it didn't seem to work at all. Strange that.

 

[AnotherJoe]

If a NAS was used instead of the laptop - all 3 would have sounded the same.

I think that's another triumph of assumption over evidence, but then the 'digits is digits' brigade would have us believe that any digital system will always sound the same, however it's connected. I detect a spot of wriggling on the hook setting in here...

The test forces use of UDP - Olive direct to a NAS would use TCP/IP.

 

[Andrew Everard]

The test forces use of UDP - Olive direct to a NAS would use TCP/IP.

But we weren't using a NAS. We were streaming from a laptop computer, just as many consumers will.

 

[Anonymous]

Ho ho!

Data into the ethernet port must be different to produce a different sound - agree?

So how are the differences of streaming data explained technically by engineers?

 

[Andrew Everard]

Sent by Desire(2.3.3 custom insertcoin Rom)_______

I think those bits of data might have been sent in error.

 

[Andrew Everard]

Data into the ethernet port must be different to produce a different sound - agree?

So how are the differences of streaming data explained technically by engineers?

You're just posting the same thing over and over again. Is it Groundhog Day all over again?

 

[AnotherJoe]

And I agree with you there.
But the people in this thread were wondering why there could be a difference between the 3.

I'm just explaining why there's a difference - and a solution to it

Sent by Desire(2.3.3 custom insertcoin Rom)_______

 

[professorhat]

And I agree with you there. But the people in this thread were wondering why there could be a difference between the 3. I'm just explaining why there's a difference - and a solution to it

To be fair, you've posted a potential reason for the difference between wireless and ethernet, but not for the homeplugs.

Ignore me, I'm being an idiot.

 

[Andrew Everard]

To be fair, you've posted a potential reason for the difference between wireless and ethernet, but not for the homeplugs.

Ignore me, I'm being an idiot.

On which sage note by the prof, I think we'll stop going round and round the same track.