Question: BeoLab 50 external source: DAC or no DAC? (Geoff?)

[Curly]

Hi all,

As a follow-on to the question I posed on Discord, I wondered if there were other educated opinions here–perhaps Geoff’s (crossing fingers))–regarding the following:

When selecting an external streaming source to use with BeoLab 50, what is there to consider between choosing a “streaming transport” with no onboard DAC versus a streaming DAC?

My gut tells me to use a streaming transport so as to keep the number of D/A and A/D conversions as few as possible and so as to not change the sound of the speaker relative to what Geoff and his team have intended. But others have noted improved sound quality (whatever this really means) when implementing an external DAC.

Further context: Improved sound quality (whatever this really means) over Beosound Core would be a goal in using an external source in addition to having access to TIDAL Connect and a nicer user interface.

So, if one is going to use an external, non-B&O streaming source with BL50, what factors are there to consider in choosing to use an external DAC or omitting one?

Many thanks!

 

[Curly]

If there was a technical reason, and thus a formal recommendation, for whether to use a DAC-free transport or a streaming DAC, it would be very helpful in eliminating a handful of products to audition.

 

For example, Lumin has three streaming transports and four streaming DACs. If I could eliminate one category of product, it would be really useful.

[geoffmartin]

Hi,

This question was one of the first ones I ever answered on Beoworld… but I can’t find it in the Archived Forum section.

So, here’s the short version of what I probably said back then…

The Technical answer is:

IF
you are playing a file / signal that is in a format that either cannot be transported digitally OR cannot be decoded by Beolab 50 (I’m specifically thinking of DSD materials, but bitreamed lossy CODECs like MP3, Ogg Vorbis, and Dolby Digital variants also fit the description)

THEN
you need to decode to analogue through a DAC to get into the analogue input of the Beolab 50.

 

The non-Technical answer is:

Any DAC can also be considered as a filter – it does SOMETHING to the audio. This is also true of the ADC at the analogue input of any system (like the Beolab 50). Therefore

IF
you prefer the filter response of that DAC+ADC combination when you connect a given player to the Beolab 50

THEN
you should use the analogue path.

 

A little update to what I probably said back then:

Personally, I sit firmly on both sides of that fence. I (still) use an Oppo 105 Blu-ray player as my source, connected using BOTH S/PDIF and Analogue (in my case, XLR but the RCA is equivalent) to the Beolab 90s. I then have the 90’s set up to give priority to the S/PDIF input.

So, when I play a PCM file (including CD, DVD-Audio, or Blu-ray audio), the Oppo outputs it on both the analogue and digital outputs, and the 90s play the digital signal (because of the priority setting). When I play DSD (either from a file or from an SACD) then the Oppo doesn’t send anything on the digital output, and the 90s automatically switch over to the analogue input.

 

I also have the Power Link inputs on the Beolab 90s connected to the output of a television (and it has highest priority). The Oppo is, of course, connected to the TV using an HDMI input. In addition, I’ve created a Speaker Group that has all loudspeakers set to Speaker Role = NONE. You may wonder why I have a Speaker Group that uses no speakers… This is because, when I’m using the Oppo for 2-channel audio-only, I can use the on-screen menus of the Oppo viewed on the television, but the  audio signal goes directly to the Beolab 90s on either of those two paths I described above.

However, the default Speaker Group for the Oppo is to use all the loudspeakers connected to the TV, which means that, when I put in a Blu-ray, it automatically defaults to the surround output via the television.

 

Finally, you may be asking why I use the Oppo. There are two reasons for this:

• It still behaves really well. I measured it, and so I trust it. I never buy something new to replace something that still works.
• I still play my shiny discs a LOT.

 

One final small correction:

It’s not MY team – but I’m on it. I’m just the goalie.  🙂

 

Hope this helps.

-g

[Curly]

For my forum friends (and for myself), I’d just like to highlight a portion of Geoff’s response:

“[…] OR cannot be decoded by Beolab 50 (I’m specifically thinking of DSD materials, but bitreamed lossy CODECs like MP3, Ogg Vorbis, and Dolby Digital variants also fit the description)[…]”

This would also include MQA.

My research suggests that DAC-free streaming transports from Lumin and Aurender, for example, can perform the first MQA unfolding and output it over a digital connection to BeoLab 50/90 with a maximum resolution of 96kHz.

Alternatively, these companies offer streaming DACs that can perform the full MQA decoding and rendering up to 192kHz (and beyond) and output it over analog.

The question is which would sound best. And I don’t know. A/B testing required. But it seems logical to want to feed high-quality speakers such as BeoLab 50/90 the richest signal so perhaps using a streaming DAC and outputting over analog is the way to go (if streaming MQA over TIDAL).

 

Thank you, Geoff!

[geoffmartin]

You’re welcome.

Some comments:

You can’t really ask “which would sound best?” since this is a matter of personal preference. It’s like asking which dish on the menu tastes best. Also remember that, even though an audio signal measures better, it doesn’t necessarily mean that you’ll prefer it.

There are plenty of examples that show that there’s no correlation between technical performance and preference. Vinyl and tubes/valves are the most obvious targets here… But but there’s no shame in preferring something that is “worse”.

Regarding MQA: I suggest that you read the article on Stereophile.com about MQA. MQA is not lossless – but it’s also not lossy in the same way that MP3 is, for example. Also consider that Tidal (for one) is setting its default to FLAC over MQA (see this article), which might make the question increasingly irrelevant – but the future is notoriously difficult to predict…

Finally, it’s always important to remember that any conversion device – for example a DAC, and ADC, or a Sampling Rate Converter, will have different behaviours at different I/O sampling rates. In other words, if you do a completely valid A/B test comparing a digital signal path to an analogue signal path (remember that “completely valid” means, at the very least, that it’s a blind test (you don’t know which is which) and that the levels are completely aligned to be identical at the listening position) , then the results of your test are ONLY valid for the sampling rate that you tested. In other words, if you do the test with a 192 kHz / 24 bit file, you still know nothing about a 44.1 kHz / 16 bit file. That might result in the opposite conclusion. I’ve even measured systems that (because of a sampling rate converter in the system) performed worse with a 24-bit file than a 16-bit file of the same sampling rate.

Cheers

-geoff

[Author]

Posts