Bang & Olufsen BeoLab 18 Reviews



I was part of the development team, and one of the two persons who decided on the final sound design (aka tonal balance) of the B&O BeoLab 18 loudspeakers. So, I’m happy to share some of the blame for some of the comments (at least on the sound quality) from the reviews.


Bernard Dickinson at Live Magazines said:

The sound reproduction is flawless”


Lyd & Billede’s August 2014 review said

“Lydkvaliteten er rigtig god med en åben, distinkt og fyldig gengivelse, som ikke gør højopløste lydformater til skamme.” (The sound quality is very good with an open, clear and detailed reproduction, which do not put high-resolution audio formats to shame.)

and ”Stemmerne er lige klare og tydelige, hvad enten vi sidder lige i smørhullet eller befinder os langt ude i siden. Det er faktisk ret usædvanligt og gør, at BeoLab 18 egner sig lige godt til både baggrundsmusik og aktiv lytning” (The voices are crisp and clear, whether we are sitting right in the sweet spot or far off to the side. It’s actually quite unusual and makes the BeoLab 18 equally suited for both background music and active listening)

  1. Hi Geoff.
    Completely off topic but I might as well ask.
    Once the Transmitter 1 arrives, I am considering adding a BL 19 to the wireless set up..
    Will the T1 send just the correct frequencies to the 18’s and 19?
    What I mean is, will the bass frequencies be handled entirely by the BL 19 and the BL 18’s handle just the mid-high frequencies?
    Or will the BL 18’s perform as they are now but, with the added dimension of the BL 19 lower frequency response?

  2. Hi Ivan,

    The T1 will automatically configure itself to do the bass management, re-routing the low end out of the 18’s and into the 19.


  3. Hi Geoff

    I have a question regarding connecting beolab 18 via TOSLINK (daisy chain),

    According to specifications the beolab 18 only accepts:
    – in: Sample rate 32k, 44.1k, 48k, 96k/PCM16-24 bit
    – out: Sample rate 48k/PCM 24 bit

    Does that mean that the TOSLINK option in not preferable if i want bit perfect sound from CD/stream or high res audio (24/192) etc.

    Many thanks.

  4. Hi Vlad,

    There are multiple aspects of your question to answer…

    The optical input of B&O loudspeakers are limited to a maximum of 96 kHz. This is even true for the BeoLab 90. This is due to the typical unreliability of an optical digital audio connection at sampling rates above 96 kHz. (B&O is not alone in this respect… many manufacturers make a similar limitation to prevent non-obvious errors in the audio signal.) So, a 176.4 or 192 kHz LPCM signal will not work with the BeoLab 18’s optical input.

    As it says in the specifications, digital signals sent to the BeoLab 18’s optical input are sample rate converted to 48 kHz. The reason for fixing the sampling rate is to avoid breaks in the music caused by changes in the incoming sampling rate (a consequence of the time it takes to update all the coefficients in the DSP if its internal sampling rate were to change.). I’ve described some of the issues that have to be considered when making this decision in this posting.

    So, if by “bit perfect” you mean that the LPCM signal that enters the signal processing inside the loudspeaker is a bit-for-bit match with any signal that enters the optical digital input, then the answer is “no – the signal is not bit perfect”.

    However, the question is whether the use of a sampling rate converter is, by default, a bad thing. It is certainly not – as long as the sampling rate converter that is used is not the “bottleneck” in the performance of the entire audio signal path. The advantage that B&O has is that, since the loudspeaker (the BeoLab 18 is just one example) contains all components – the DSP filtering, DAC’s (one per driver), amplifiers (one per driver), and drivers – then we can design the loudspeaker as an optimised system. This give us advantages that we would not have if we made a passive loudspeaker that was connecting to an unknown amplifier, DAC, processor – etc…

    In addition to this, it must be remembered that any multi-way, digitally-implemented, active loudspeaker has many filters in its processing paths to the various drivers. Therefore, even if you created such a loudspeaker that locked to the native (incoming) sampling rate, by the time the signal gets out of the DSP to the DAC’s, it cannot be “bit perfect” (even if it entered the DSP as a bit-for-bit match with the source) – if it were, the speaker wouldn’t work properly.

    This brings us to the last part of the question: is the optical digital input “preferable” to the analogue input? This is something you would have to test to compare, since it is somewhat dependent on the source that you are connecting to the BeoLab 18’s input. As a simple example, if you are connecting a source using the optical digital connection, but the source does not have a gain control, then the output of the loudspeaker will be stuck at full volume. This would probably be bad. :-)
    (Sidebar: if your source does have a volume control and a digital output, then that output will also not be “bit perfect”…)

    Of course, the analogue input is somewhat susceptible to noise, depending on the quality of your source and cabling (proper grounding and shielding are always helpful in optimising the signal to noise ratio).

    Having said all that, I can say that, when I am working on the loudspeakers, I typically use the optical digital input from my sound card. Of course, the Power Link and Line inputs are verified, using measurements and audition – but in a day-to-day situation here at work, I use the digital input. Also, that input is typically running at the sampling rate of my source materials – most of which are 44.1 kHz / 16 bit.


  5. Hi again Vlad,

    I realised afterwards that I was a little vague – so I’m back to clear that up. The sampling rate converter in the BeoLab 18 has a signal to noise ratio of 128 dB. If you had a 16-bit signal that actually used all 16 bits with a TPDF-dithered signal, then the maximum SNR of that input would be approximately 93 dB. If you had a 24-bit signal that actually used all 24 bits with a TPDF-dithered signal then the maximum SNR of that input would be approximately 141 dB – but good luck finding such a recording… :-)

    In other words, the SNR of BeoLab 18’s sampling rate converter is approximately equivalent to that of a 22-bit TPDF-dithered LPCM signal.

    As I mentioned above, this is on-par with the performance of the DAC’s and the amplifiers, although since the signals outputted by those components are band-limited (for and by the tweeter and woofers) then it’s a little complicated (and possibly misleading) to talk about the SNR of the signals in those parts of the signal path.

    Hope this answers any questions that may have arisen as a result of my first answer. If not, please feel free to ask more questions.


  6. Thank you very much for det detailed answer, very insightful :-)

    I should try and compare the analog to the digital inputs or maybe even better – connect via the transmitter 1 :-)


  7. Hey Geoff. I just bought my Beolab 18’s and started thinking about the placement. I will hang them on the wall, but the problem is I have two sofas in front of the speakers. In order not to “hide” the wooden lamelas by the sofas siderests , the most optimal height of placement would be to mount them 30 cm above the floor. In the instructions, though it says that max recommended height from the floor is 20 cm. Would it be an issue to add those extra 10 cm? I could e-mail you a picture of how everything looks.

  8. Hi Bob,

    Sorry for the long delay in responding. I am just getting back from vacation.

    I would not expect that the extra 10 cm would make any difference worth worrying about, based on the description of your configuration. The extra height will change the frequency and impact of the comb filtering caused by a floor reflection, but since you have a sofa in the way, this is less of an issue for you.

    Hope this helps – and it’s not coming too late…