I had a look at the spec-page for BL5 over at the official site, and puzzled over the panel below.

Have they changed the processor in the latest version from the original PCM 24/96 to a 32bit/44.1/48/88.2kHz ... ?

http://www.bang-olufsen.com/product-archive?ProductName=BeoLab%205

And if so, when did they make the change? From which serial number?

I'm not a techy but I would like to know how this change effects sounds quality.

I'm almost suspecting they've forgotten to add 96 at the end of that string of sample-rates, as it would be puzzling if they've skipped 96kHz. 88.2 is two times 44.1, and 96 is two times 48 (which is the DVD-sample rate, for instance; and what you get from most cable decoders' coax s/pdif).

88.2 and 96 kHz are becoming the most common formats for high-resolution music files available from suppliers such as Linn, 2L and HD-Tracks, and it would be good if the latest spec BL5 accepted both.

But "their having skipped 96" is just speculation on my part. Anyone know? Alex made mention of the latest spec BL5 and liking the sound from it, maybe he can dig up some specifics.

The 32 refers to 32 KHz sampling rate. Why on earth you'd ever want to listen to audio with 32 KHz sample rate through a pair of £14k speakers is beyond me but there you go.

RE the lack of 96KHz, I also always thought that was a bit strange. Maybe a gaffe too? It states 96 KHz in the service manual I'm looking at now...

Well, it all depends upon how one should interpret the spec's they've posted now.

The speakers' DSP is 32-bit floating point.

And previously, the coax s/pdif feed was up to 24/96 - and whatever entered the speaker, both s/pdif and analog, would be bumped to 24/96 before being sent to the DSP for processing.

But now the spec's state 32-bit 32/44.1/48/88.2 -- and have omitted 96. Which is such a strange decision that I'm pretty convinced they've simply forgotten to add 96 on the official page. I was told of the 32-bit in a phone call, and am not certain that this is the case. I would suspect that it should be 24-bit 44.1/48/88.2/96, as it used to be.

The fact that 96 is mentioned in the service manual for the latest version seems to confirm this. But do we know for certain? An acquaintance has been told by his dealer that the newest version does not accept 96 and that an attempt to play that will result in distortion, and that's what got me started down this trail.

I'm pretty certain the 32 refers to a 32 KHz sample rate capability -  I can't see what benefit 32bit float audio would be purely for playback of PCM audio. With the PowerLink inputs, 32 bit DAC/ADC could potentially offer quite a large performance boost especially when listening at lower levels...

Omission of 96 KHz seems rather strange. This would also suggest the speakers aren't up-sampling in the DSP when fed with a 48 KHz signal...

Alex:

I'm pretty certain the 32 refers to a 32 KHz sample rate capability -  I can't see what benefit 32bit float audio would be purely for playback of PCM audio. With the PowerLink inputs, 32 bit DAC/ADC could potentially offer quite a large performance boost especially when listening at lower levels...

Omission of 96 KHz seems rather strange. This would also suggest the speakers aren't up-sampling in the DSP when fed with a 48 KHz signal...

The 32 bit floating point would be for internal arithmetic within the DSP. There would be no 32 bit ADC's or DAC's. The point being that there is minimal impact on the fidelity of the signal when it it's represented in 32 bit FP during the modifcations imposed by the DSP, before being either truncated or dithered back to 24 bit.

Yup. The mystery here is the "omitted" 96kHz in the new spec's ... which I suspect may be a mistake by our spec's-writers in Struer. I doubt they've dumped that.