Non-PC processors

[Funkstar]

Something slightly different to get thinking about.

We all know that modern high performance processors cut back on their power requirements etc. when idling.

What about other processors though?

There is an on going argument on the SlimDevices forum that FLAC audio decompressed to WAV on the server produces a better sound than when the same FLAC file is streamed directly to a SqueezeBox. The argument to support this is: "more CPU cycles to decompress might cause some power supply interferences with the digital or analog output stage".

The squeezeBox is built round a Ubicom IP3000 processor.

Do older or more simple porcessors scale back their demands while not under load like an AM2 or Core2 would?

Anyone got any thoughts on this?

[shaithis]

The only way to get an answer to that would be to ask the ppl who make the Squeezebox "Is the decompression algorithm lossy when there isn't enough CPU time available?"

I fail to see why any CPU could not be throttled, the chipset would control it......We all know from overclocking that CPUs are not bound to a frequency or voltage and have a range that varies, nothing stops a manufacturer from implementing power-saving technologies.

[schmunk]

Whilst not knowing the answer to your direct question, Funkstar, I can answer the more general question re the 'sound quality', i.e. that it won't make even the slightest bit of noticeable difference...

[Funkstar]

The only way to get an answer to that would be to ask the ppl who make the Squeezebox "Is the decompression algorithm lossy when there isn't enough CPU time available?"

That isn't an issue. The audio would break up rather than be degraded in quality. And there is more than enough processor power in there for 16bit 44.1KHz decoding. It even manages to downsample 24bit 96KHz audio to 48KHz as well as decode the FLAC.

i.e. that it won't make even the slightest bit of noticeable difference...

Agree completely. I would just like to have some amunition against the audiophiles on their forums that believe that unrelated scenarios make a difference to sound quality.

Their argument is that FLAC decoding uses more processing power than streaming WAV that is just passed to the DAC. This processing power puts more strain on the power supply causing the analogue outputs to give a sligtly different signal. Or the other one is that the increased processing power increases EM interference at the analogue stange.

I guess Ubicom would be the best people to answer this though

[KowShak]

The fact is that the CPU will consume more power decoding FLAC vs streaming a WAV. On the other hand, the streamed WAV will have a higher data rate than the FLAC so the streaming will need more power. Which case needs more power is open for debate.

Whether the additional power requirements either way will make the PSU produce a more noisy output depends on the design of the power supply.

[Agent]

Its because they are using non gold plated, oxygen free cables.
Fools!

[mroz]

Snake oil. IMO they're not worth arguing with. But, if you want an argument, offering to set up a double blind trial is an option.

[DanceswithUnix]

It sounds like they have already made their mind up, but to answer your question:

Most of the current PC throttling ideas came from the battery powered embedded processor world in the first place, so just about any modern CPU can throttle. For example, mobile phones must have good battery life to survive as a product. That doesn't mean you have to turn those options on, so even if it has power saving there is a good chance the designers won't bother making their design more complex by using it.

I would just point out to them that if they want superb sound quality they should get a record deck and listen to LPs

[Fraz]

Its because they are using non gold plated, oxygen free cables.
Fools!

And don't they know these cables are directional? They should check they are the right way round!

Audiophiles make me laugh with their dubious physics.