A thought that could improve the acceptance of active speakers

[pablolie]

My apologies if this has been discussed already. My question is, why don't we have the equivalent of PoE for active loudspeakers? Removing the need for a power cord would simplify setup dramatically, meeting, or conceivably even exceeding what can be done with conventional AVRs and passive loudspeakers now. Thoughts?
@amirm

I think 100W is about the max PoE supports. I'd assume that is plenty, especially if paired with a sub, but some -especially big speaker traditionalists- may think that is not enough. Also - if the power is provided by a PoE switch rather than the DAC/pre-amplifier, it's not like it simplifies cabling all that much, and a new breed of DAC/pre-amp is required?

 

[mcdn]

Thinking about multichannel, does Dolby even allow decoded channels to be sent digitally?

 

[Steve81]

I think 100W is about the max PoE supports. I'd assume that is plenty, especially if paired with a sub, but some -especially big speaker traditionalists- may think that is not enough. Also - if the power ir provided by a PoE switch rather than the DAC/pre-amplifier, it's not like it simplifies cabling all that much, and a new breed of DAC/pre-amp is required?

I was envisioning essentially a new breed of AVR. Some type of PoE/Powerline Ethernet/something-brand-new switch meets pre-pro; advantages vs conventional AVRs as I see it are everything remains in the digital domain till the amp in the speaker (you'd need ADC for analog inputs, but this process happens anyway with modern bass management); digital XOs vs passive (would this ease engineering difficulties for speaker manufacturers?), and the setup is no more complicated than what exists now. As an overall system, I think it beats the heck out of the current paradigm, but I didn't consider cost to develop vs the effective market either.

It now occurs to me...skip the ethernet; just use SPDIF/TosLink or new standard (one with error correction for example) combined with low voltage AC.

As @sigbergaudio mentioned though, an active/passive mix would do just dandy without much additional complexity.

Thinking about multichannel, does Dolby even allow decoded channels to be sent digitally?

Since this would have to be an industry-wide effort for adoption, they'd presumably be consulted in the matter.

 

[Littletycoon]

Thinking about multichannel, does Dolby even allow decoded channels to be sent digitally?

they get their money from the decoding algorithm

 

[lisgotan]

My apologies if this has been discussed already. My question is, why don't we have the equivalent of PoE for active loudspeakers? Removing the need for a power cord would simplify setup dramatically, meeting, or conceivably even exceeding what can be done with conventional AVRs and passive loudspeakers now. Thoughts?
@amirm

If you are are going to combine signal and power into 1 cable, then the best solution is still passive speakers and just 2 wires.

With POE, you still need an external PSU (eg. Poe injector) to supply power and signal to the speaker. if you want to use a centralised power supply (like a Poe switch), then how is that different from AVR?

Apart from added complexity and cost, you have more points of failure and you still need a cable and an external box.

 

[Steve81]

If you are are going to combine signal and power into 1 cable, then the best solution is still passive speakers and just 2 wires.

With POE, you still need an external PSU (eg. Poe injector) to supply power and signal to the speaker. if you want to use a centralised power supply (like a Poe switch), then how is that different from AVR?

Apart from added complexity and cost, you have more points of failure and you still need a cable and an external box.

See this post for further updates. Things are evolving, though still potentially infeasible.

 

[scsim]

Ooh! An audio question where my IT background is relevant! Let's assume we're talking about ethernet cabling and connectors.

To get the cable count to 1 per speaker, the modern answer is wireless powered speakers like the KEF LS60. Still it's interesting to look at why PoE doesn't work very well as a solution.

Unless this thing is to be totally proprietary you need the digital output to be AES50 (https://en.wikipedia.org/wiki/AES50) for audio over ethernet. At which point each endpoint needs a full ethernet stack, which means each endpoint is now a computer. Or you go full AES67 and use IP networking, which requires a full OS on that computer. At that point, why would you not go with DLNA or similar?

And then you need PoE++ for 100W per port. And PoE++ switches with a total power budget over 200W really only exist in 16 port or larger configs, due to the size of the power supply required. So the cost for the power and signal distribution climbs to at least USD500.

Hi! Me is IT too!!

I see POE as convenience/ease of installation esp. when the location does not have any power source. APs and IP cameras (esp outdoors!!) are such examples.

However, it is implemented for very short distances, then I failed to see how a "POE" connector where you combine signal and power into 1 creates ease. Mainly because you still need a physical cable from speaker to source. I see having wire from source to speaker a greater hassle rather than connecting speaker to wall socket. Wireless takes care of this issue.

 

[Steve81]

Hi! Me is IT too!!

I see POE as convenience/ease of installation esp. when the location does not have any power source. APs and IP cameras (esp outdoors!!) are such examples.

However, it is implemented for very short distances, then I failed to see how a "POE" connector where you combine signal and power into 1 creates ease. Mainly because you still need a physical cable from speaker to source. I see having wire from source to speaker a greater hassle rather than connecting speaker to wall socket. Wireless takes care of this issue.

I'm IT too (hence the original PoE thought), but it seems low voltage AC+Toslink or SPDIF would be preferable for a few reasons. Different strokes for different folks on the Wireless+Power Socket vs Single Wire To Switch.

What do you do in the IT field? I spent a chunk of my day mediating a dispute between a dentist's office and SprintRay after they said their product (a 3D printer) was effectively unfixable. The client is super pleased with them at the moment

 

[Galliardist]

The positioning of my stereo setup, and the various rules I have to follow in this building, mean that I have an awkward choice, to say the least, if I tried to go multichannel. One rear would go in/near a corner where I can't run power to it, full stop. I could run a flat cable past a doorway with a small step. The other rear I can't easily run a cable to from where the electronics can go: but I could run wifi signal to a powered speaker.

Possible? I don't know. I could use battery powered wireless speakers as a pair, I suppose, if such a thing is doable: I believe there are a couple of soundbars that have that option.

As for PoE to speakers, I only see that as really an option in a system that is already going to be complex: a high number of surrounds and ceiling mounted speakers, and the ability to mount multiple PoE++ injectors in a central location could be of benefit with active speakers

 

[amirm]

Poe speaker
Another Poe speaker

I jumped on the Genelec 4430A when I saw the announcement and fired an email to them for a sample. I wanted a small high performance speaker for ceiling installation for Atmos. The idea of just finish the cat-5 cable was pretty attractive. As I think was mentioned, figuring out how to feed it from consumer gear became complicated. Genelec was willing to send me a sample so if we can work through the connectivity issue, I can still request and test it.

 

[scsim]

I'm IT too (hence the original PoE thought), but it seems low voltage AC+Toslink or SPDIF would be preferable for a few reasons. Different strokes for different folks on the Wireless+Power Socket vs Single Wire To Switch.

What do you do in the IT field? I spent a chunk of my day mediating a dispute between a dentist's office and SprintRay after they said their product (a 3D printer) was effectively unfixable. The client is super pleased with them at the moment

I am not sure if SPDIF is good idea because SPDIF is rather susceptible to noise. SPDIF do not have error correction and its synchronous communication, so it also susceptible to jitter over long distance as well. I feel that having asysnc mode is better. tcp/ip is something that is already in use and more resilient against noise.

I saw that someone has posted IP based genelec speakers. I think this is something that could work, you just have to plug speakers to a POE switch and able to play music via network. However, its only applicble for low power speakers (saw that its 30W).

High power (say 200w) might be possible via ethernet.

Btw, me is IT manager, mainly just manage my IT team, systems etc etc.... But its mostly administrative...lol.... stuffs like policy, compliance, budgeting, audit, short/mid term plan. Very little hands-on now...sigh... miss those good old days.

 

[Axo1989]

Poe speaker
Another Poe speaker

Thanks, lots of interesting reading on the Genelec PoE model.

 

[Axo1989]

I jumped on the Genelec 4430A when I saw the announcement and fired an email to them for a sample. I wanted a small high performance speaker for ceiling installation for Atmos. The idea of just finish the cat-5 cable was pretty attractive. As I think was mentioned, figuring out how to feed it from consumer gear became complicated. Genelec was willing to send me a sample so if we can work through the connectivity issue, I can still request and test it.

Great, very interesting avenue and product to explore. Especially as a solution to high speaker count systems like Atmos. I'd look forward to reading about it if you go there.

 

[Rja4000]

figuring out how to feed it from consumer gear became complicated.

If you speak about PoE, I suppose you just need a PoE Ethernet switch with 30W or so per speaker, from what I've read in Genelec doc


For audio connection, it seems to be Dante enabled, so all you actually need is a Dante entry point.

It's easy from a PC: you just need the Audinate Dante Virtual Soundcard, which comes with an ASIO driver.

If you need to drive it from hardware, I think the cheapest way is the Dante AVIO AES3, which comes under several brands. It allows 44.1, 48 and 96kHz sampling frequencies. Cost is around $250.
There is also a cheaper AVIO USB, but this one is limited to 48kHz only. Cost is around $200.

Those are similar to the AVIO analog adapter you've measured... except they should be bit-perfect, of course.

So I suppose you know all this.


I use Dante for years, and I've performed measurements with the Dante Virtual Soundcard: it's without any issue (once you get it working).

There are now several thousands of Dante-enabled products.


NB: As you noted in your AVIO review, changing the clock frequency in a Dante network requires to change it for all devices.
You typically don't do that often, for the typical use case (Pro environnement).
But connecting Tidal, Qobuz or the likes as a source would require that you force those softwares to always use 44.1kHz, whatever the source material's sampling rate.

 

[Littletycoon]

Yes, with Dante it is relatively easy, the Dante virtual sound card has a free trial period also. Just connect the speaker to a POE+ switch (note: some rules apply) connected to the pc with the Dante virtual sound card.

Somewhere Q1 next year i hope to go for 12 of these genelec POE speakers for surround & atmos duties via aes67. It's waiting now for some required gear relatively affordable to come onto the market in 2024 for decoding dolby & dts.

 

[krabapple]

You forget the amplifier inbetween (and additionally the dolby/dts decoder, but is not always (mostly not) taken care of by actives). So at worst the same amount of wires, at best half the amount of wires, with those wires being more tidy (mostly electrical sockets).

I didn't forget. When I wrote 'source', that encompasses either a source of the music, or a source of power i.e., an amplifier/AVR. So that is accounted for.

The most parsimonious wiring would be a system using active wireless speakers where the audio source either entirely streamed from online ('cloud') tracks via, say, a tablet connected to a wifi network, or, if such a thing exists, one where that functionality is built into one or more of the powered wireless speakers and can be controlled remotely. A tablet of course still requires intermittent use of a wall socket, to recharge.

My audio library has a great many PCM multichannel files, in addition to numerous raw DD/DTS audio files, and I rely also on Denon/Audyssey system calibration/room correction DSP. I have yet to figure a way to translate that into an analogous powered speaker system. Current system is files on hard drive --> laptop running foobar2k (WASAPI out) --> HDMI to AVR (where DSP and decoding occur) --> AVR wired to 5.1 speakers. Old school.

 

[Steve81]

I am not sure if SPDIF is good idea because SPDIF is rather susceptible to noise. SPDIF do not have error correction and its synchronous communication, so it also susceptible to jitter over long distance as well. I feel that having asysnc mode is better. tcp/ip is something that is already in use and more resilient against noise.

I saw that someone has posted IP based genelec speakers. I think this is something that could work, you just have to plug speakers to a POE switch and able to play music via network. However, its only applicble for low power speakers (saw that its 30W).

High power (say 200w) might be possible via ethernet.

Btw, me is IT manager, mainly just manage my IT team, systems etc etc.... But its mostly administrative...lol.... stuffs like policy, compliance, budgeting, audit, short/mid term plan. Very little hands-on now...sigh... miss those good old days.

Yeah, that thought crossed my mind about SPDIF. Toslink/optical would eliminate that issue, but not sure of the other problems it would raise. Shouldn’t be especially difficult to develop something fit for the purpose either way though (like I know how to do it ).

Can’t say I’d ever want to be a manager. I like people too much (usually) to want to be the one to discipline. I’m pretty close with my boss though, and ping her with my thoughts regularly.

 

[charlielaub]

I'd rather prefer just a power cord and use wireless transfer of audio data. This would make a multichannel system much easier to install.

This is exactly what I do in my current DIY builds, well technically speaking. All of the amplifiers and other components that are required for the wireless streaming are external to the speaker. This is because I build speakers without boxes (nude dipole loudspeakers) so there is no speaker box... Compare this to an active speaker, the Linkwitz/Ear-Food LX521 - there is a dedicated 1U box for each speaker that contains the crossover and the amplification for that side. If wireless streaming could be added (via WISA for example) this would leave only the AC mains input cord.

 

[lisgotan]

Yeah, that thought crossed my mind about SPDIF. Toslink/optical would eliminate that issue, but not sure of the other problems it would raise. Shouldn’t be especially difficult to develop something fit for the purpose either way though (like I know how to do it ).

Can’t say I’d ever want to be a manager. I like people too much (usually) to want to be the one to discipline. I’m pretty close with my boss though, and ping her with my thoughts regularly.

Yeah, that thought crossed my mind about SPDIF. Toslink/optical would eliminate that issue, but not sure of the other problems it would raise. Shouldn’t be especially difficult to develop something fit for the purpose either way though (like I know how to do it ).

Can’t say I’d ever want to be a manager. I like people too much (usually) to want to be the one to discipline. I’m pretty close with my boss though, and ping her with my thoughts regularly.

Toslink is possible but it increase complexity, you will need to integrate optical + copper into a single cable. And then, optical connectors has to be rather precise compared to electrical contact pins, this means the connector will be costly to manufacture.

I don't know how long they can go, I googled and it says average is 10m. So distance may be a concern.

I feel an all electrical connection is a better option rather than mixed.

I can understand. It's not easy to discipline people, even harder when I need to dismiss someone for disciplinary reasons. It's like being the bad guy.

 

[Steve81]

Toslink is possible but it increase complexity, you will need to integrate optical + copper into a single cable. And then, optical connectors has to be rather precise compared to electrical contact pins, this means the connector will be costly to manufacture.

I don't know how long they can go, I googled and it says average is 10m. So distance may be a concern.

I feel an all electrical connection is a better option rather than mixed.

I can understand. It's not easy to discipline people, even harder when I need to dismiss someone for disciplinary reasons. It's like being the bad guy.

Well that would keep the cable manufacturers happy at least

I can't imagine a modern error-correcting protocol running on copper would be that much of an ask if the industry doesn't want to keep essentially the same deal going since multi-driver speakers were invented. Then again, hi-fi as we know it seems to be a dying breed. You can stream a movie on a computer screen and listen on some decent wireless earbuds and get as good or better sound quality than a lot of gear out there now, given that most "hi-fi" speaker manufacturers don't seem to understand what the term means.

Edit: Thought: Why don't we do what we do with bluetooth connections to said wireless earbuds? Transcode the individual channel content to another protocol.