Fidelizer & SBT.. Why should it work?

[chill]

But that test is not necessarily defintive, it only tells you if the processing on the Touch is the issue. There are other possible paths from server to ears such as EMI radiated from server, noise injected on power mains etc. Something which is changing the whole operating environment of the server COULD be changing something which gets transfered through one of these indirect paths, even when audio data is not actively being transmitted to the Touch. Now if you unplugged the the Touch from the ethernet AND unplugged the power form the server at the same time, THEN you would have a more convincing test.

John S.

John

I can't tell if you're just being mischievous, but if not, are you seriously suggesting that the mere presence of a powered-up computer in the same house as the hifi will have an audible effect? That's a bit of a blow for the whole computer-based audio industry. But moreover, and returning to the OPs point I suppose, are you suggesting that the changes in EMI and mains noise between a computer that's running Fidelizer and one that's not will be audible?

I realise that your post is describing a theoretical possibility only (hence the capitalised 'COULD'), but what is the likelihood of these effects being audible in the real world? How bad would your components have to be for such minuscule things to have an impact? IMO, the plug-pull test is already convincing enough.

[SBGK]

John

I can't tell if you're just being mischievous, but if not, are you seriously suggesting that the mere presence of a powered-up computer in the same house as the hifi will have an audible effect? That's a bit of a blow for the whole computer-based audio industry. But moreover, and returning to the OPs point I suppose, are you suggesting that the changes in EMI and mains noise between a computer that's running Fidelizer and one that's not will be audible?

I realise that your post is describing a theoretical possibility only (hence the capitalised 'COULD'), but what is the likelihood of these effects being audible in the real world? How bad would your components have to be for such minuscule things to have an impact? IMO, the plug-pull test is already convincing enough.

lol, this really is the tar pit of audiophile forums.

OP, I would try somewhere else for an answer. The author of fidelizer posts in a few eg jplay forums, you can ask him directly there.

[darrell]

John

I can't tell if you're just being mischievous, but if not, are you seriously suggesting that the mere presence of a powered-up computer in the same house as the hifi will have an audible effect? That's a bit of a blow for the whole computer-based audio industry. But moreover, and returning to the OPs point I suppose, are you suggesting that the changes in EMI and mains noise between a computer that's running Fidelizer and one that's not will be audible?

I realise that your post is describing a theoretical possibility only (hence the capitalised 'COULD'), but what is the likelihood of these effects being audible in the real world? How bad would your components have to be for such minuscule things to have an impact? IMO, the plug-pull test is already convincing enough.

lol, this really is the tar pit of audiophile forums.

OP, I would try somewhere else for an answer. The author of fidelizer posts in a few eg jplay forums, you can ask him directly there.

In what way is chill's reply worthy of this comment? It seems very polite and balanced to me. Is it because the only posts you can countenance are ones which confirm your own prejudices? In any case, the possibility of expectation bias, already mentioned, surely must be considered at least as likely as possible changes in the electrical environment, to all but the most blinkered subjectivist.

The possibility that there is a difference between bits streamed to the SB buffer by a computer running Fidelizer and one not running Fidelizer, *after* the ethernet cable is pulled is mind blowing - do you understand what you are suggesting? Please consider explanations based on psychology, which are orders of magnitude more likely than explanations based on physics.

As far as asking the developer of Fidelizer for the answer to this, it would be like relying on Bob Diamond to tell you whether Barclays Bank was an upright and moral organisation!

[Apesbrain]

As far as asking the developer of Fidelizer for the answer to this, it would be like relying on Bob Diamond to tell you whether Barclays Bank was an upright and moral organisation!

Translation for our members in the USA:

"...it would be like relying on Jaime Dimon to tell you whether JP Morgan Chase was an upright and moral organization!"

[chill]

lol, this really is the tar pit of audiophile forums.

OP, I would try somewhere else for an answer. The author of fidelizer posts in a few eg jplay forums, you can ask him directly there.

SBGK

The fact is that there aren't many places that an explanation of Fidelizer's effects can be hiding. You're obviously reluctant to accept expectation bias, so let's exclude that for a moment. So as I see it the possibilites are:

i) It directly affects the bits that get buffered by the Touch
ii) It directly affects the workload of the Touch
iii) It indirectly affects all the other components in your system

If i) is true, it can only be by putting the WRONG bits in the buffer - the DTS example proves that the touch is already bit perfect.
If ii) is true, pulling the ethernet cable out of the Touch will demonstrate the effect. Once the Touch is no longer communicating with the server, the workload of the Touch that's related to server comms must reduce to zero, and you should be able to hear the benefit. You've now stated a couple of times that you hear no difference when you pull the cable, so the workload can't be the source of the improvement.
John described a mechanism by which iii) might be having an impact. I was asking for some perspective on those possibilities. What's missing from that short discussion is the old question of interference arriving down the ethernet cable. Again, pulling the cable would test this.

Fidelizer may well have a positive benefit for people playing music from their PC's sound card. But we know that the Touch is NOT analogous to a PC sound card. I'm sure we'd all welcome some input from Fidelizer's author to explain it's benefits in the Squeezebox scenario. I wouldn't be at all surprised to find that he/she disavows any claim to improving things in the Squeezebox world.

Rather than storming off in a huff, why not try and engage with this discussion - what have I missed in my summary?

[garym]

what have I missed in my summary?

For one, you're talking to SBGK, a poster who has demonstrated time and again a complete misunderstanding of how squeezeboxes work.

[garym]

And he referred you to jplay forums for more info. That's like asking Bernie Madoff for investment advice. Search hydrogenaudio.org on jplay.

[lake_eleven]

Soundcheck, in his thread, suggests using TCPOptimizer's 'optimal' settings. Even this does not affect SBT playback?

[chill]

Soundcheck, in his thread, suggests using TCPOptimizer's 'optimal' settings. Even this does not affect SBT playback?

What is the supposed mechanism for any improvement?

We know the server, even a hopelessly underpowered stressed out server, gets all the right bits to the Touch, in the right order.

So assuming the supposed mechanism for TCPOptimizer to have an effect is by improving the 'way' the bits reach the Touch buffer, and hence the workload/duty cycle of the Touch, then the plug pull test will tell you if you have a problem that needs fixing in the first place.

If pulling the ethernet plug out of the Touch gives you 30s of audio nirvana, then your server side setup needs fixing (Fidelizer, TCPOptimizer, different OS, galvanic isolation, etc etc), but if you can't hear a difference when you pull the plug, then don't waste your time.

[JohnSwenson]

John

I can't tell if you're just being mischievous, but if not, are you seriously suggesting that the mere presence of a powered-up computer in the same house as the hifi will have an audible effect? That's a bit of a blow for the whole computer-based audio industry. But moreover, and returning to the OPs point I suppose, are you suggesting that the changes in EMI and mains noise between a computer that's running Fidelizer and one that's not will be audible?

I realise that your post is describing a theoretical possibility only (hence the capitalised 'COULD'), but what is the likelihood of these effects being audible in the real world? How bad would your components have to be for such minuscule things to have an impact? IMO, the plug-pull test is already convincing enough.

I'm being serious. A computer sitting on the same shelf as your stereo system can have a significant affect on said stereo system from airborne EMI and noise on the power line. If it's three rooms away the effects will be much less. This thread never specified any geometrical arrangement of the components. I know several people who have tried to use their laptop to control an SB, the laptop was across the room from the stereo, they could hear noise on the stero system when they moved the mouse on the laptop. What is going on inside the computer can have an affect on the sound without changing bits.

I did some tests on this quite a few years ago testing a bunch of desk tops, laptops, small things such as Mac mini's, embedded devices like FitPCs etc. I did this with my stereo system, and with a few friend's systems. In all these tests it was a computer in the listening room, but NOT right next to the stereo system, usually across the room. Note that the computers being testsed had nothing to do with the audio. The stereo was being fed by an SB conected by wire to a server a long ways away. The computers under test were just doing things like web broswsing etc.

The worst by FAR were the laptops, almost everyone was audible in some way, either directly emitting sound (through the stereo) or changing audio that was playing. Desk tops fared quite a bit better, either not audible at all or not as much affect. Small general purpose computers such as Mac minis did a little better, but still could be heard under some situations, embedded devices such as a FitPC were inaudible no matter what we had them doing.

The method of "contamination" from the laptops seemed to EMI, they did just as bad when run off batteries. We tried wrapping one in aluminum foil (kind of hard to use in that state!) and it's affect went way down. Whether the screen was up or down also had a significant affect.

So yes computers CAN affect sound quality, and something which is changing the underlying behavior of said computer could very well change it's impact on sound quality.

I have no knowledge about the program in question here so I'm not making any comment about it, I was primarily refering to the use of the "pull the ethernet cable" test as being definitive, if the server computer is still running there is the possibility that it can still be affecting the sound quality through means other than direct connection to the SB.

John S.