A "foray" is defined as an initial attempt, so it was an incorrect use of that word. It would be comparable to saying, "Elon Musk introduced the Model 3 today as he launched Tesla's foray into electric vehicles."
Not in my dictionary. A foray is "a sudden attack or incursion into enemy territory, especially to obtain something; a raid." It doesn't have to be an initial attempt. The word "foray" seems to trace back to an Anglo-French term for "raider" or "forager". The word probably has the wrong implication when applied to a commercial offering to the public, like the HomePod.
The ongoing debate here is a prime example of how irrelevant comments can derail the discussion, particularly when the comment in question is one of the very first ones on the article (and thus one of the first ones everyone reads).
I'm deleting the comment and all subsequent talk of it, not because I'm particularly sensitive, but because it's not relevant to the story at hand. Readers who want to point out typos or have grammatical debates are able (and encouraged!) to contact us via email, or twitter, or elsewhere outside of the comments section.
Also, feel free to check out our handy dandy commenting guidelines while you're at it.
For a small, 5lb, single unit, it’s great, but it’s definitely not as loud as I expected from the reviews. In-store you ger a square paper bag made for HomePod.
Setup was a breeze, like with AirPods. I don’t think that BT process can be streamlined any more than it already is.
It’s definitely not creating any stereo-like sound. Maybe beamingforming audio is coming in a future update, but at this point if the device to one side all the sound will clearly be coming from that one side. It pretty much has to be directly behind or in front of me to not be able to determine the side the device was placed relative to my head in my simple tests.
If you already have a good home entertainment center (which could have an echo or Google Home attached) then HomePod won’t sound better.
The far-field microphones are impressive. The Echo was impressive 3 years ago in this regard, which was always the first reason Siri would come across as incompetent. The HomePod outpaces even the new Echo Plus insofar that you have have the volume maxed out and Siri will still be able to hear you, even over spoken audio, at a normal speaking voice from across the room or from a different room. The Echo Plus will require you to raise your voice a little when the volume is maxed. (Caveat: My Echo Plus is connected to other, better, louder, and more speakers that aren’t centreally locates which even the HomePod’s A-series chip probably couldn’t process as well, either.)
Sirius XM spoken audio does sound more crisp on the HomePod. And while that’s an EQ setting on the HEC, it’s nice that it’s automatic.
I am disappointed that it can tell me what’s on my Calender next. I assume that was covered in a review, so I must have missed it or it needs some additional setup.
Beamforming audio is operational.
Could you give us some details on how the HomePod is located in your room/space? A diagram would be great.
The room dimensions ans setup seem pretty much identical to what Apple demoed, except flipped to the left side of the room in relation to the Tv and couch.
As previously stated, I did had it moved the different locations so I could try to discern where the sound was coming from when on the couch, but on he side table would be ideal in terms of where it would look best.
Just curious; do you have a wall behind the couch?
Yes. An exterior, load bearing wall with a window, like the other walls. No sliding glass doors or anything like that on any wall. The couch isn’t up against it. You can walk behind it.
I'm not versed in the HomePod's placement obviously, but does the HomePod have a orientation that we would consider "front"? There would definitely be a "back" orientation where the cord connects. Otherwise, all I see is the HomePod optimizing for the room, whatever its position, not the position of the "sweet" spot of listening.
In my view, the best position would be placement a few feet away from the wall opposite of the couch, if you were maximizing for separation, but as you state, that isn't ideal for you.
1) Nope, but that’s by design.
2) The big feature fanatics have been jacking each other off over—despite not have ever tested it—is the ability to place it anywhere (reasonably) in a room and have it intelligently adjust how the sound leaves the teeeters to idealize itself for an environment.
Looking forward to getting one soon. I do have a question, though, and maybe it will be covered in the full review.
How does/will it respond to multiple voices? By design, this is not a personal device. It sits in a home and should be able to be used by more than one person. Can Siri be trained to recognize more than one voice? And if it recognizes different people, can it access different music profiles, different reminders or text messages based on who is making the request?
Siri does not distinguish between people nor does it allow multiple accounts. If you enable access to messages, notes and reminders, and your iPhone is within range, anyone can access them by voice.
If there are two people in the same room, in range, which does it go to? It doesn't distinguish voices, but does it respond to all voices? Currently, my wife's devices ignore my "Hey Siri" and vice-versa.
I'm not really sure what you're asking here. Are you both saying Hey Siri at exactly the same time? It listens to whoever prompted it. I don't see where this would be a problem.
It responds to all voices.
OK. Gotcha regarding the voices. To clarify my other question. If there are two phones (sorry I typed people earlier) within range and you ask it to set a reminder or read back a text, which phone does it grab it from? Thanks for the responses, Neil.
It can only be set up with one phone and one account, that's it. So if you ask it to pull text messages, it will only do so from the one authorized phone that was used to set up the account.
Alternatively, as part of the initial setup, you can disable HomePod access to messages, reminders and notes. This means that not you nor anyone else will be able to read messages from your phone (or whatever one it is paired with).
If you have access to messages enabled, there is no security feature. If your phone is within range, anyone can ask Siri to read your unread messages.
How does the “Hey Siri” canceling work with multiple people in the room with multiple devices? I mean you initially set up HP with some voice samples, but that doesn’t matter as far as account recognition as it’ll respond to anyone, right? So what if someone else says Hey Siri but they’re talking to their watch or phone, both on a different Apple ID, and the HomePod on the other account is in the room and “takes” the request? Honestly confused how this thing works with potentially several people using Siri in proximity of the HP and how having only a single user setup for your personal information/etc makes any sense at all unless you live alone and have no visitors ever.
How does Hey Siri work with multiple devices in the same room? Is it still buggy? As it is now for me on the latest software, if I raise my wrist and say Hey Siri to my Apple Watch, my iPhone also activates Siri at the same time. Will I now have three devices activating Siri every time?
Just tried asking Siri for the time from way across the great room, whilst playing really loud music and holding my iPhone close to me. The iPhone Siri motion graphic displayed but nothing happened on the phone, it didn't answer, the HomePod quietened down and told the time and went back to the music. My gosh it's powerful.
Looking forward to getting one soon. I do have a question, though, and maybe it will be covered in the full review.
How does/will it respond to multiple voices? By design, this is not a personal device. It sits in a home and should be able to be used by more than one person. Can Siri be trained to recognize more than one voice? And if it recognizes different people, can it access different music profiles, different reminders or text messages based on who is making the request?
Siri does not distinguish between people nor does it allow multiple accounts. If you enable access to messages, notes and reminders, and your iPhone is within range, anyone can access them by voice.
If there are two people in the same room, in range, which does it go to? It doesn't distinguish voices, but does it respond to all voices? Currently, my wife's devices ignore my "Hey Siri" and vice-versa.
I'm not really sure what you're asking here. Are you both saying Hey Siri at exactly the same time? It listens to whoever prompted it. I don't see where this would be a problem.
It responds to all voices.
OK. Gotcha regarding the voices. To clarify my other question. If there are two phones (sorry I typed people earlier) within range and you ask it to set a reminder or read back a text, which phone does it grab it from? Thanks for the responses, Neil.
It can only be set up with one phone and one account, that's it. So if you ask it to pull text messages, it will only do so from the one authorized phone that was used to set up the account.
Alternatively, as part of the initial setup, you can disable HomePod access to messages, reminders and notes. This means that not you nor anyone else will be able to read messages from your phone (or whatever one it is paired with).
If you have access to messages enabled, there is no security feature. If your phone is within range, anyone can ask Siri to read your unread messages.
Looking forward to getting one soon. I do have a question, though, and maybe it will be covered in the full review.
How does/will it respond to multiple voices? By design, this is not a personal device. It sits in a home and should be able to be used by more than one person. Can Siri be trained to recognize more than one voice? And if it recognizes different people, can it access different music profiles, different reminders or text messages based on who is making the request?
Siri does not distinguish between people nor does it allow multiple accounts. If you enable access to messages, notes and reminders, and your iPhone is within range, anyone can access them by voice.
If there are two people in the same room, in range, which does it go to? It doesn't distinguish voices, but does it respond to all voices? Currently, my wife's devices ignore my "Hey Siri" and vice-versa.
I'm not really sure what you're asking here. Are you both saying Hey Siri at exactly the same time? It listens to whoever prompted it. I don't see where this would be a problem.
It responds to all voices.
OK. Gotcha regarding the voices. To clarify my other question. If there are two phones (sorry I typed people earlier) within range and you ask it to set a reminder or read back a text, which phone does it grab it from? Thanks for the responses, Neil.
It can only be set up with one phone and one account, that's it. So if you ask it to pull text messages, it will only do so from the one authorized phone that was used to set up the account.
Alternatively, as part of the initial setup, you can disable HomePod access to messages, reminders and notes. This means that not you nor anyone else will be able to read messages from your phone (or whatever one it is paired with).
If you have access to messages enabled, there is no security feature. If your phone is within range, anyone can ask Siri to read your unread messages.
How does the “Hey Siri” canceling work with multiple people in the room with multiple devices? I mean you initially set up HP with some voice samples, but that doesn’t matter as far as account recognition as it’ll respond to anyone, right? So what if someone else says Hey Siri but they’re talking to their watch or phone, both on a different Apple ID, and the HomePod on the other account is in the room and “takes” the request? Honestly confused how this thing works with potentially several people using Siri in proximity of the HP and how having only a single user setup for your personal information/etc makes any sense at all unless you live alone and have no visitors ever.
HomePod gets priority over all other devices as long as they are not active (ie, iPhone is locked, wrist is not raised). If you want to do an iPhone-specific command without HomePod, while HomePod is within earshot, grab your phone and unlock it first.
I couldn't get on with the way the Sonos system required everything to be played through its own app.
One of our popular activities with friends is to do am music quiz in which everyone searches using their own iOS device for a song, and then plays it either using the headphone cord or Bluetooth. I'm concerned we won't be able to do this with HomePod.
For a small, 5lb, single unit, it’s great, but it’s definitely not as loud as I expected from the reviews. In-store you ger a square paper bag made for HomePod.
Setup was a breeze, like with AirPods. I don’t think that BT process can be streamlined any more than it already is.
It’s definitely not creating any stereo-like sound. Maybe beamingforming audio is coming in a future update, but at this point if the device to one side all the sound will clearly be coming from that one side. It pretty much has to be directly behind or in front of me to not be able to determine the side the device was placed relative to my head in my simple tests.
If you already have a good home entertainment center (which could have an echo or Google Home attached) then HomePod won’t sound better.
The far-field microphones are impressive. The Echo was impressive 3 years ago in this regard, which was always the first reason Siri would come across as incompetent. The HomePod outpaces even the new Echo Plus insofar that you have have the volume maxed out and Siri will still be able to hear you, even over spoken audio, at a normal speaking voice from across the room or from a different room. The Echo Plus will require you to raise your voice a little when the volume is maxed. (Caveat: My Echo Plus is connected to other, better, louder, and more speakers that aren’t centreally locates which even the HomePod’s A-series chip probably couldn’t process as well, either.)
Sirius XM spoken audio does sound more crisp on the HomePod. And while that’s an EQ setting on the HEC, it’s nice that it’s automatic.
I am disappointed that it can tell me what’s on my Calender next. I assume that was covered in a review, so I must have missed it or it needs some additional setup.
Beamforming audio is operational.
Could you give us some details on how the HomePod is located in your room/space? A diagram would be great.
Beamforming performance is still limited to 1) fundamental sound propagation principles, 2) the physical characteristics of the speaker array and how it relates to beamforming, 3) the speaker enclosure design, and 4) the physical geometry of the room.
1) Sound propagation is frequency dependent. Low frequencies have less directionality and tend to flood the room with sound rather than having a discernible source location. Woofer and midrange speaker cones improve the directionality of low frequencies somewhat but at the extremes, like subwoofers, the speaker placement is much less critical other than to avoid echos and standing waves/resonance. High frequencies on the other hand are very directional and tweeter projectors/cones/openings are designed to help spread the high frequencies sound to reduce overly extreme directionality. A lot of popular music is heavily biased towards the lower frequencies so the composite sound will be naturally less directional.
2) Transmit beamforming (speaker array sound projection) is basically the superposition of sound from multiple individual speakers. An individual unbaffled sound projector will produce sound omnidirectionally, i.e., no beams. When you add in a second sound projector next to the first one the effect of beamforming will produce a single donut shaped beam, but a beam that's lacking left-vs-right and up-vs-down directionality. Adding additional sound projectors in the same linear plane will produce additional donut shaped beams all still lacking left-vs-right and up-vs-down directionality To obtain left-vs-right directionality you have to wrap the sound projectors into a cylindrical or spherical shape and baffle (block) the parts of the beams you don't want, like the up and down and backside parts of the donut, which is what Apple is talking about when they refer to "ambient" beamforming.
3) While the active beamforming done by the onboard DSP creates the donut shaped beams the ambient beamforming effects are the result of the enclosure design, i.e., baffles and openings (apertures). The net result are a series of beams that project out like the petals on a flower when viewed from above. Most but not all of the undesired parts of the beams are suppressed but less so for lower frequencies because of the general lack of directionality of lower frequencies. About the only thing Apple can practically do to achieve a stereo effect is to manipulate the sound levels and source channel data (stereo left vs right) that are directed to specific beams. But with so few sound projectors they are only producing a limited number of beams and the beam widths are fairly wide and overlap quite a bit. Couple that with the preponderance of lower frequencies and there's only so much they can do for stereo effect and directionality. If you look into sonar and radar systems that rely heavily on beamforming the number of projectors is often in the hundreds and sometimes over a thousand and they are arranged in 2D planar, cylindrical, or spherical arrays so they can form highly directional and steerable pencil shaped beams.
4) The room interferes with the projected beams because echos further superimpose themselves on top of the projected beams. If you're standing in the room you're hearing a mix of both projected sound and reflected sound. Apple does apply compensation to try to minimize the negative effects of echos and standing waves but they can't prevent echos and standing waves, only deemphasize their effects. And again, the lower frequencies tend to flood the room with nondirectional sound and the reflection of low frequencies act like additional low directionality sound sources and reduce the stereo separation sound effect.
I'd bet that in a lab or anechoic chamber with pure sound tones the stereo sound effect produced by the HomePod is quite spectacular for what it is. But in a real listening space and with composite sound sources its going to be less precise but still better than if they were not performing any of the advanced sound processing that they are doing to make the listening experience better.
Neil, you said: Upon unwrapping HomePod, the top of the box slides off to reveal the speaker sitting upright.
IMO, this is the best packaging Apple has ever done - no wasted space or weight... And you can remove the clear plastic covering from the box without any knife, scissors, etc. If you want to return it (fat chance), you could place all the contents in the original box as if it had never been opened
Neil, you said: Upon unwrapping HomePod, the top of the box slides off to reveal the speaker sitting upright.
IMO, this is the best packaging Apple has ever done - no wasted space or weight... And you can remove the clear plastic covering from the box without any knife, scissors, etc. If you want to return it (fat chance), you could place all the contents in the original box as if it had never been opened
I agree. I just packaged the empty box all up again to store but I just could get that amazing clear plastic wrap that seemed only held by static back on again lol.
For a small, 5lb, single unit, it’s great, but it’s definitely not as loud as I expected from the reviews. In-store you ger a square paper bag made for HomePod.
Setup was a breeze, like with AirPods. I don’t think that BT process can be streamlined any more than it already is.
It’s definitely not creating any stereo-like sound. Maybe beamingforming audio is coming in a future update, but at this point if the device to one side all the sound will clearly be coming from that one side. It pretty much has to be directly behind or in front of me to not be able to determine the side the device was placed relative to my head in my simple tests.
If you already have a good home entertainment center (which could have an echo or Google Home attached) then HomePod won’t sound better.
The far-field microphones are impressive. The Echo was impressive 3 years ago in this regard, which was always the first reason Siri would come across as incompetent. The HomePod outpaces even the new Echo Plus insofar that you have have the volume maxed out and Siri will still be able to hear you, even over spoken audio, at a normal speaking voice from across the room or from a different room. The Echo Plus will require you to raise your voice a little when the volume is maxed. (Caveat: My Echo Plus is connected to other, better, louder, and more speakers that aren’t centreally locates which even the HomePod’s A-series chip probably couldn’t process as well, either.)
Sirius XM spoken audio does sound more crisp on the HomePod. And while that’s an EQ setting on the HEC, it’s nice that it’s automatic.
I am disappointed that it can tell me what’s on my Calender next. I assume that was covered in a review, so I must have missed it or it needs some additional setup.
Beamforming audio is operational.
Could you give us some details on how the HomePod is located in your room/space? A diagram would be great.
Beamforming performance is still limited to 1) fundamental sound propagation principles, 2) the physical characteristics of the speaker array and how it relates to beamforming, 3) the speaker enclosure design, and 4) the physical geometry of the room.
1) Sound propagation is frequency dependent. Low frequencies have less directionality and tend to flood the room with sound rather than having a discernible source location. Woofer and midrange speaker cones improve the directionality of low frequencies somewhat but at the extremes, like subwoofers, the speaker placement is much less critical other than to avoid echos and standing waves/resonance. High frequencies on the other hand are very directional and tweeter projectors/cones/openings are designed to help spread the high frequencies sound to reduce overly extreme directionality. A lot of popular music is heavily biased towards the lower frequencies so the composite sound will be naturally less directional.
2) Transmit beamforming (speaker array sound projection) is basically the superposition of sound from multiple individual speakers. A individual unbaffled sound projector will produce sound omnidirectionally, i.e., no beams. When you add in a second sound projector next to the first one the effect of beamforming will produce a single donut shaped beam, but a beam that's lacking left-vs-right and up-vs-down directionality. Adding additional sound projectors in the same linear plane will produce additional donut shaped beams still lacking left-vs-right and up-vs-down directionality To obtain left-vs-right directionality you have to wrap the sound projectors into a cylindrical or spherical shape and baffle (block) the parts of the beams you don't want, like the up and down and backside parts of the donut, which is what Apple is talking about when they refer to "ambient" beamforming.
3) While the active beamforming done by the onboard DSP creates the donut shaped beams the ambient beamforming effects are the result of the enclosure design, i.e., baffles and openings (apertures). The net result are a series of beams that project out like the petals on a flower when viewed from above. Most but not all of the undesired parts of the beams are suppressed less so for lower frequencies because of the general lack of directionality of lower frequencies. About the only thing Apple can practically do to achieve a stereo effect is to manipulate the sound levels and source channel data (stereo left vs right) that are directed to specific beams. But with so few projectors they are only producing a limited number of beams and the beam widths are fairly wide and overlap quite a bit. Couple that with the preponderance of lower frequencies and there's only so much they can do. If you look into sonar and radar systems that rely heavily on beamforming the number of projectors is often in the hundreds and sometimes over a thousand and they are arranged in 2D planar, cylindrical, or spherical arrays so they can form highly directional and steerable pencil shaped beams.
4) The room interferes with the projected beams because echos further superimpose themselves on top of the projected beams. If you're standing in the room you're hearing a mix of both projected sound and reflected sound. Apple does apply compensation to try to minimize the negative effects of echos and standing waves but they can't prevent echos and standing waves, only deemphasize their effects. And again, the lower frequencies tend to flood the room with nondirectional sound and the reflection of low frequencies act like additional low directionality sound sources and reduce the stereo separation sound effect.
I'd bet that in a lab or anechoic chamber with pure sound tones the stereo sound effect produced by the HomePod is quite spectacular for what it is. But in a real listening space and with composite sound sources its going to be less precise but still better than if they were not performing any of the advanced sound processing that they are doing to make the listening experience better.
So, would I be out of line to suspect that best separation for a stereo input source would be achieved at the center of symmetry of a room with symmetric placement of the HomePod?
Soli noted he was configuring a mirror image of what Apple was demoing, and was achieving poor separation results, which I think would have been expected by that configuration and position of the listener.
How does Siri work with tasks the HomePod can’t handle? Do they get pushed to your iPhone or iPad or does Siri say sorry I can’t do X on the HomePod?
If I ask "Hey Siri, what's on my calendar?" I get the response: "I can't access your calendar here. Sorry about that!"
The lack of Calander data despite connecting to my iCloud account is less than ideal. I've also been unable to connect the HomePod to my Mac either through my Mac or through the HomePod, yet with the Echo I can have it connect with a simple voice command. HomePod doesn't even show up under AirPlay when on the same network as my Mac and with the same iCloud account. This is less convenient than AirPods which does this seamlessly.
For a small, 5lb, single unit, it’s great, but it’s definitely not as loud as I expected from the reviews. In-store you ger a square paper bag made for HomePod.
Setup was a breeze, like with AirPods. I don’t think that BT process can be streamlined any more than it already is.
It’s definitely not creating any stereo-like sound. Maybe beamingforming audio is coming in a future update, but at this point if the device to one side all the sound will clearly be coming from that one side. It pretty much has to be directly behind or in front of me to not be able to determine the side the device was placed relative to my head in my simple tests.
If you already have a good home entertainment center (which could have an echo or Google Home attached) then HomePod won’t sound better.
The far-field microphones are impressive. The Echo was impressive 3 years ago in this regard, which was always the first reason Siri would come across as incompetent. The HomePod outpaces even the new Echo Plus insofar that you have have the volume maxed out and Siri will still be able to hear you, even over spoken audio, at a normal speaking voice from across the room or from a different room. The Echo Plus will require you to raise your voice a little when the volume is maxed. (Caveat: My Echo Plus is connected to other, better, louder, and more speakers that aren’t centreally locates which even the HomePod’s A-series chip probably couldn’t process as well, either.)
Sirius XM spoken audio does sound more crisp on the HomePod. And while that’s an EQ setting on the HEC, it’s nice that it’s automatic.
I am disappointed that it can tell me what’s on my Calender next. I assume that was covered in a review, so I must have missed it or it needs some additional setup.
Beamforming audio is operational.
Could you give us some details on how the HomePod is located in your room/space? A diagram would be great.
Beamforming performance is still limited to 1) fundamental sound propagation principles, 2) the physical characteristics of the speaker array and how it relates to beamforming, 3) the speaker enclosure design, and 4) the physical geometry of the room.
1) Sound propagation is frequency dependent. Low frequencies have less directionality and tend to flood the room with sound rather than having a discernible source location. Woofer and midrange speaker cones improve the directionality of low frequencies somewhat but at the extremes, like subwoofers, the speaker placement is much less critical other than to avoid echos and standing waves/resonance. High frequencies on the other hand are very directional and tweeter projectors/cones/openings are designed to help spread the high frequencies sound to reduce overly extreme directionality. A lot of popular music is heavily biased towards the lower frequencies so the composite sound will be naturally less directional.
2) Transmit beamforming (speaker array sound projection) is basically the superposition of sound from multiple individual speakers. A individual unbaffled sound projector will produce sound omnidirectionally, i.e., no beams. When you add in a second sound projector next to the first one the effect of beamforming will produce a single donut shaped beam, but a beam that's lacking left-vs-right and up-vs-down directionality. Adding additional sound projectors in the same linear plane will produce additional donut shaped beams still lacking left-vs-right and up-vs-down directionality To obtain left-vs-right directionality you have to wrap the sound projectors into a cylindrical or spherical shape and baffle (block) the parts of the beams you don't want, like the up and down and backside parts of the donut, which is what Apple is talking about when they refer to "ambient" beamforming.
3) While the active beamforming done by the onboard DSP creates the donut shaped beams the ambient beamforming effects are the result of the enclosure design, i.e., baffles and openings (apertures). The net result are a series of beams that project out like the petals on a flower when viewed from above. Most but not all of the undesired parts of the beams are suppressed less so for lower frequencies because of the general lack of directionality of lower frequencies. About the only thing Apple can practically do to achieve a stereo effect is to manipulate the sound levels and source channel data (stereo left vs right) that are directed to specific beams. But with so few projectors they are only producing a limited number of beams and the beam widths are fairly wide and overlap quite a bit. Couple that with the preponderance of lower frequencies and there's only so much they can do. If you look into sonar and radar systems that rely heavily on beamforming the number of projectors is often in the hundreds and sometimes over a thousand and they are arranged in 2D planar, cylindrical, or spherical arrays so they can form highly directional and steerable pencil shaped beams.
4) The room interferes with the projected beams because echos further superimpose themselves on top of the projected beams. If you're standing in the room you're hearing a mix of both projected sound and reflected sound. Apple does apply compensation to try to minimize the negative effects of echos and standing waves but they can't prevent echos and standing waves, only deemphasize their effects. And again, the lower frequencies tend to flood the room with nondirectional sound and the reflection of low frequencies act like additional low directionality sound sources and reduce the stereo separation sound effect.
I'd bet that in a lab or anechoic chamber with pure sound tones the stereo sound effect produced by the HomePod is quite spectacular for what it is. But in a real listening space and with composite sound sources its going to be less precise but still better than if they were not performing any of the advanced sound processing that they are doing to make the listening experience better.
So, would I be out of line to suspect that best separation for a stereo input source would be achieved at the center of symmetry of a room with symmetric placement of the HomePod?
Soli noted he was configuring a mirror image of what Apple was demoing, and was achieving poor separation results, which I think would have been expected by that configuration and position of the listener.
I also noted that when I placed in the center line of the room (under the wall mounted TV) it was still obvious when the location moved a little to the left or right.
Comments
I'm deleting the comment and all subsequent talk of it, not because I'm particularly sensitive, but because it's not relevant to the story at hand. Readers who want to point out typos or have grammatical debates are able (and encouraged!) to contact us via email, or twitter, or elsewhere outside of the comments section.
Also, feel free to check out our handy dandy commenting guidelines while you're at it.
https://forums.appleinsider.com/discussion/184333/appleinsiders-updated-commenting-guidelines/p1
2) The big feature fanatics have been jacking each other off over—despite not have ever tested it—is the ability to place it anywhere (reasonably) in a room and have it intelligently adjust how the sound leaves the teeeters to idealize itself for an environment.
One of our popular activities with friends is to do am music quiz in which everyone searches using their own iOS device for a song, and then plays it either using the headphone cord or Bluetooth. I'm concerned we won't be able to do this with HomePod.
1) Sound propagation is frequency dependent. Low frequencies have less directionality and tend to flood the room with sound rather than having a discernible source location. Woofer and midrange speaker cones improve the directionality of low frequencies somewhat but at the extremes, like subwoofers, the speaker placement is much less critical other than to avoid echos and standing waves/resonance. High frequencies on the other hand are very directional and tweeter projectors/cones/openings are designed to help spread the high frequencies sound to reduce overly extreme directionality. A lot of popular music is heavily biased towards the lower frequencies so the composite sound will be naturally less directional.
2) Transmit beamforming (speaker array sound projection) is basically the superposition of sound from multiple individual speakers. An individual unbaffled sound projector will produce sound omnidirectionally, i.e., no beams. When you add in a second sound projector next to the first one the effect of beamforming will produce a single donut shaped beam, but a beam that's lacking left-vs-right and up-vs-down directionality. Adding additional sound projectors in the same linear plane will produce additional donut shaped beams all still lacking left-vs-right and up-vs-down directionality To obtain left-vs-right directionality you have to wrap the sound projectors into a cylindrical or spherical shape and baffle (block) the parts of the beams you don't want, like the up and down and backside parts of the donut, which is what Apple is talking about when they refer to "ambient" beamforming.
3) While the active beamforming done by the onboard DSP creates the donut shaped beams the ambient beamforming effects are the result of the enclosure design, i.e., baffles and openings (apertures). The net result are a series of beams that project out like the petals on a flower when viewed from above. Most but not all of the undesired parts of the beams are suppressed but less so for lower frequencies because of the general lack of directionality of lower frequencies. About the only thing Apple can practically do to achieve a stereo effect is to manipulate the sound levels and source channel data (stereo left vs right) that are directed to specific beams. But with so few sound projectors they are only producing a limited number of beams and the beam widths are fairly wide and overlap quite a bit. Couple that with the preponderance of lower frequencies and there's only so much they can do for stereo effect and directionality. If you look into sonar and radar systems that rely heavily on beamforming the number of projectors is often in the hundreds and sometimes over a thousand and they are arranged in 2D planar, cylindrical, or spherical arrays so they can form highly directional and steerable pencil shaped beams.
4) The room interferes with the projected beams because echos further superimpose themselves on top of the projected beams. If you're standing in the room you're hearing a mix of both projected sound and reflected sound. Apple does apply compensation to try to minimize the negative effects of echos and standing waves but they can't prevent echos and standing waves, only deemphasize their effects. And again, the lower frequencies tend to flood the room with nondirectional sound and the reflection of low frequencies act like additional low directionality sound sources and reduce the stereo separation sound effect.
I'd bet that in a lab or anechoic chamber with pure sound tones the stereo sound effect produced by the HomePod is quite spectacular for what it is. But in a real listening space and with composite sound sources its going to be less precise but still better than if they were not performing any of the advanced sound processing that they are doing to make the listening experience better.
IMO, this is the best packaging Apple has ever done - no wasted space or weight... And you can remove the clear plastic covering from the box without any knife, scissors, etc. If you want to return it (fat chance), you could place all the contents in the original box as if it had never been opened
Soli noted he was configuring a mirror image of what Apple was demoing, and was achieving poor separation results, which I think would have been expected by that configuration and position of the listener.
This is torture.