PSB Alpha P3 Review (Bookshelf Speaker)

[ROOSKIE]

Emotiva B1+ is the same price, though the looks are iffy.

Yah JBL A130 is in range as well and JBL 530 which is on and off sale pricing every other week.
I think the 530 looks cool but many do not. Plus it is larger.
Anyway many better options exist. I just though headless might be harsh. I basically think PSB did okay given the catergorie the products fits. I gave "not terrible".
Be interesting to listen for a few tracks. Prolly is okay at low volumes.

Ignore all subjective views, decipher and judge the data yourself. Problem solved.

Since I am a lover of measuring things and a lover of data I get why you want it to be that simple. Of course I find it just isn't that simple. Even deciphering and judging are extremely subjective actions, even if done by a computer as that computer was created by beings who tailored it having to make subjective choices along the way. You can not avoid having subjectivity participate anymore than you can have up without down or dry without wet. Best action in my mind is to increase your relationship with subjectivity, increase your skill, not try to avoid it. Impossible and frankly honestly a bit boreing.

 

[bluefuzz]

Much of Harman's research is into preference, not hard fact.

And the difference is?

A lot of people seem to misunderstand what preference actually means ...

 

[shrimp_dude]

I got a P3 for nearfield use as it was well reviewed for that purpose online - only to find it to be very boomy and muffled. I then found a PSB Imagine Mini as a replacement and it is MUCH better and clearer for nearfield use. Of course at the price point you'd expect it but I personally did not find any redeeming qualities to the P3 other than its size.

 

[ROOSKIE]

And the difference is?

A lot of people seem to misunderstand what preference actually means ...

You seem to be reinforcing what I said.
If you have a preference for a A over B and I have a preference for C over B but no preference between A & B and you never heard C, which one do you buy and which one do I buy? Which one does that guy over there buy?
A hard fact would be saying a board is 10 feet and measuring the board to be 9'6" and realizing the hard fact is that the board is not 10 feet long.

 

[pseudoid]

When I read @amirm' statement about "...i could easily hear a "zing"..." I had to scroll back up to the top of the review to make sure the speaker he was discussing was really a psb.
The price struck me funny and had to do a fact-check on it. Price is correct (@$249) but who really buys a "single" speaker? Nowhere on the amazon page is the item referred to as "Speakers" although the photo depicts 2 speakers (with and w/o grill).
How odd and who exactly would think to purchase a single bookshelf speaker?
So, the cost of a pair jumps up to $500 and may need 2 headless PPs.
Unless you look at the PSB Alpha P3 as a DIY project; where your involvement is to EQ the p.o.s. before you can use it.
AND then do it again, when you are able to afford the 2nd one.

 

[ROOSKIE]

When I read @amirm' statement about "...i could easily hear a "zing"..." I had to scroll back up to the top of the review to make sure the speaker he was discussing was really a psb.
The price struck me funny and had to do a fact-check on it. Price is correct (@$249) but who really buys a "single" speaker? Nowhere on the amazon page is the item referred to as "Speakers" although the photo depicts 2 speakers (with and w/o grill).
How odd and who exactly would think to purchase a single bookshelf speaker?
So, the cost of a pair jumps up to $500 and may need 2 headless PPs.
Unless you look at the PSB Alpha P3 as a DIY project; where your involvement is to EQ the p.o.s. before you can use it.
AND then do it again, when you are able to afford the 2nd one.

It is 2 for $249. I posted that earlier.

PSB Alpha P3 (Black Ash)

Bookshelf speakers

www.crutchfield.com

However many speakers are sold as singles so that would not be odd at all. In any case these are per pair.

 

[pseudoid]

You are right but check out this amazon link where the only reference"for a pair" is to "magnetic grillES".

Otherwise, I'd think twice... as I should have some 20(?) years ago, when I won a HUGE 10MB SCSI HardDiskDrive at an ebay auction for an unreal price of ~$70USD. When it arrived, I realized it indeed was HUGE... the size of a toaster. I am a forever skeptic online.

 

[shrimp_dude]

Uhh how bout just look at the PSB website where it's listed as $249/pair?

Alpha P3 – Compact Bookshelf - PSB Speakers

Compact and versatile without sacrificing performance. The Alpha P3 is a small two-way bookshelf speaker that brings hifi pedigree to your stereo or home theatre system to fill the room with crystal-clear sound. Featuring a custom 4” (102mm) woofer and aluminium dome tweeter that uses a...

www.psbspeakers.com

 

[richard12511]

Because I am trying to up my game designing speakers as a hobby I am definitely trying to figure out what is what.
I deff have my PEQ and DSP involved often.
I can say that I have many speakers here that do not sound quite like what I anticipated.
I also have speakers that measure somewhat differently and yet sound very similar and others that measure quite closely and yet sound very different.
Despite some seeming flaws in the measurements the 4309 so far has presented a superb subjective sound for me.
Would the PSB here do that? I don't think so even though that predicted in room responce looks similar. There is more to the story and parsing it out will take time.
Testing the 4309, along with many other experiences definitely have me reevaluating some things and weighting things differently.

Where do you live? I'd love to include it in some blind tests that I'm planning in the near future. Speakers like this that measure rather poorly, yet still sound superb enough to get a 5/5, those are the most interesting speakers, imo. If there's more to learn with regards to correlating measurements with subjective preference, edge cases like this are where it's at. It's clear that it has quite a few decent sized resonances(peaks/dips in all three of the top curves), and an overall uneven response, yet you, Amir, and @napilopez all really enjoyed the sound. Makes me wonder if having a superb frequency response and a lack of resonances are maybe not as important as we think they are. Or maybe, the locations of the resonances are in strategic locations that don't affect preference much? Would be great to test under blind conditions, to remove any sort of bias. I might end up having to buy a pair for myself, but it'd be great to save a little money . Would be great to pit it against something like the SVS Ultra Bookshelf. That thing measures great, but sounded much worse(3/5 vs 5/5) to Amir. He's not alone either, His criticisms of the SVS are very common(from what I've read).

 

[richard12511]

Testing the 4309, along with many other experiences definitely have me reevaluating some things and weighting things differently.

Same! The only other speaker I know of that measured that poorly and did that well(5/5) was the Revel M55XC. Interestingly, both the Revel and the JBL show SOTA directivity. They also have different FR errors. The Revel has a FR slope problem(too much treble), but resonance control seems fairly good. The JBL has decent slope, but 5 or 6 "resonances". In both cases, it seems that directivity is far more important than FR(even before EQ).

Another great example would be the Cannon omni speaker that was reviewed. It didn't get a golfing panther, but it sounded way better than it measured(maybe the worst measurements we've seen yet).

In the opposite direction(great measurements/poor sound), the SVS would be a great learning example.

This PSB speaker is much less interesting, as it both measures bad and sounds bad.

 

[ROOSKIE]

Where do you live? I'd love to include it in some blind tests that I'm planning in the near future. Speakers like this that measure rather poorly, yet still sound superb enough to get a 5/5, those are the most interesting speakers, imo. If there's more to learn with regards to correlating measurements with subjective preference, edge cases like this are where it's at. It's clear that it has quite a few decent sized resonances(peaks/dips in all three of the top curves), and an overall uneven response, yet you, Amir, and @napilopez all really enjoyed the sound. Makes me wonder if having a superb frequency response and a lack of resonances are maybe not as important as we think they are. Or maybe, the locations of the resonances are in strategic locations that don't affect preference much?

*Warning pretty off topic as I only mention the PSB in this line.*

Minneapolis, MN.
Yes, blind would be great though hard to pull off even with help. It would be worth trying.
I mentioned this in the 4309 thread. I believe JBL/Harman still blind test during the design phase and require the speaker to win. If so then the 4309 passed their in house testing.
No real sign of resonances of any sort in the 4309 yet. The box is very solid feeling and no odd anomalies at all yet.
By the way not sure what you mean by 5-6 resonances, what do you mean? I will eventually measure and see if I can confirm the port resonance. I can also do the woofer and tweeter again to confirm somethings.

The Polk T50 sounds like a box. Really it does. More than most speakers. (it still sounds good for the price) It has exhibited a couple strange, very audible resonances on songs I know well.
The 4309 really sounds close to what I imagine an invisible vibrating membrane suspended in the room would be doing. If you told me it was that I would nearly believe you.
Not a hint of distortion at high volumes (with an HP at 35hrz to cut out excess excursions), actually one needs to be careful as it is playing much louder than it seems.
There is a touch of something odd about some voices from time to time but who knows. It did take the 1st 4-5 songs for my ears to adjust to the speaker the 1st time. The speaker simply sounds so good to me I now lose all interest in critiquing it. I'd say it is my favorite $2k and under speaker thus far. Again I do have to listen much more, I am more of a long term analyzer.
I don't work for Harman for those who might be worried.

Same! The only other speaker I know of that measured that poorly and did that well(5/5) was the Revel M55XC. Interestingly, both the Revel and the JBL show SOTA directivity. They also have different FR errors. The Revel has a FR slope problem(too much treble), but resonance control seems fairly good. The JBL has decent slope, but 5 or 6 "resonances". In both cases, it seems that directivity is far more important than FR(even before EQ).

In association with this, it goes the other way as well. 1 example, I did not care for the ELAC DBR62 which received 5.8. No sense of realism and somewhat lifeless vocals (like most of the ELAC speakers I am familiar with, just a bit polite and not quite fully engaging). Kind of a dainty mid and treble with thick bass that had less detail than I prefer. (that day I compared it with the JBL 530 and Polk L200, both much preferred here)Still I could see why some love the DBR62.
EQ'ing the Polk T50 (to an estimated 5.5 from 3.7) reduced the unevenness and the sort of glaring quality and slight brightness it can have, downside it that it revealed the speaker relies on that unevenness and elevated highs to make up for having no detail. Tuned to "neutral" it sounds like my GF said, "speakers playing through a paper bag". She is also a pianist and really noticed how the EQd speaker killed the notes. Who knows? At any rate we both though it sounded muffled. Bear in mind she was blind tested as I switched the EQ on and off. She basically aced it. I also did it but I knew what was what, my take away was that the both versions of the T50 didn't hold my interest well and I found them equal but different. Neither the 3.7 stock nor the 5.5ish EQd sounded notably better, just different versions of decent. A good enough speaker set for under $300 retail though. (my set was under $200)

 

[bluefuzz]

You seem to be reinforcing what I said.

On the contrary, as Amir has explained several times, in the context of Harman's research preference means what sounds right to people. There are no other 'hard facts' possible. Regardless of all the measurements we can make with microphones and audio-analysers we ultimately have to relate those measurements to what people actually hear. And, since we can't directly measure subjective experience inside people's brains, all we can do is ask them what they hear, i.e their preference ...

 

[tuga]

Same! The only other speaker I know of that measured that poorly and did that well(5/5) was the Revel M55XC. Interestingly, both the Revel and the JBL show SOTA directivity. They also have different FR errors. The Revel has a FR slope problem(too much treble), but resonance control seems fairly good. The JBL has decent slope, but 5 or 6 "resonances". In both cases, it seems that directivity is far more important than FR(even before EQ).

Another great example would be the Cannon omni speaker that was reviewed. It didn't get a golfing panther, but it sounded way better than it measured(maybe the worst measurements we've seen yet).

In the opposite direction(great measurements/poor sound), the SVS would be a great learning example.

This PSB speaker is much less interesting, as it both measures bad and sounds bad.

Interestingly both the 4309 and the M55XC are Harman speakers, and they both produce much lower harmonic distortion than the Alpha P3.
But the SVS Ultra produces even less distortion... Although some audiophiles describe speakers with flat response (and perhaps low distortion) as "bland" sounding.

I still feel that speaker preference should be assessed in stereo though.


The inescapable fact is that speakers must be listened to. Attempts at determining preference through measurements are only valid on a per individual basis.

 

[ROOSKIE]

On the contrary, as Amir has explained several times, in the context of Harman's research preference means what sounds right to people. There are no other 'hard facts' possible. Regardless of all the measurements we can make with microphones and audio-analysers we ultimately have to relate those measurements to what people actually hear. And, since we can't directly measure subjective experience inside people's brains, all we can do is ask them what they hear, i.e their preference ...

Howdy. You really have no idea what I am saying and keep thinking we disagree. Oh well. Think about it and realize we essentially agree and that therefore the comments I made earlier are basically another version of what you are saying.
Originally @tuga mentioned "I wonder if this sudden change is market/preference driven" & I was pointing out that all of this especially the Harman approach, is preference driven.

 

[SMJ]

Any chance of a 'grill on' 'grill off' plot comparison occasionally?

For three reasons

1. The grill will act a a loosely coupled mass which will change the cabinet resonances.
2. It may alter some of the mid to hf reflections, dispersion and diffraction across the baffle.
3. It may have been designed or tested to be used with the grill on (as per BBC LS3/5a).

And yes, it was Mission who inverted the vertical alignment to compensate for the different acoustic centres of the woofer and tweeter. The conventional arrangement has a in-phase wavefront that tilts downward due to the tweeter being acoustically in front of the woofer, the angle dependent on the distance between the drive units and how far the acoustic centre of the drive unit is from the mounting surface and baffle. For a good approximation it could be considered as a perpendicular line drawn from a line joining the two acoustic centres.

To keep domestic harmony, it was thought by Mission that most people prefer lower height stands rather than try and sit inline with the down tilted alignment from taller stands. By turning things around, the in-phase wave front was now tilted upwards and so lower stands could be used. However, it did mean that the distance between the woofer and the floor had changed too, often causing some unwanted reactions with the floor reflection.

Personally, I preferred Meridian's solution for their M2, an adjustable stand with a tilt mechanism.

SMJ

 

[tuga]

Originally @tuga mentioned "I wonder if this sudden change is market/preference driven" & I was pointing out that all of this especially the Harman approach, is preference driven.

Interestingly even the Harman research seems to indicate that different listeners prefer a different balance; a speaker for untrained listeners should have a booom-tizzz balance:

 

[richard12511]

*Warning pretty off topic as I only mention the PSB in this line.*

Minneapolis, MN.
Yes, blind would be great though hard to pull off even with help. It would be worth trying.

I'm in Texas(near Houston) . I've done a few before, but they're definitely a PITA to setup and run/coordinate. I've got a sweet AVA ABX Switcher now that should make things easier, but I haven't had a chance to use it. I think speakers like this where the measurements differ from the subjective impressions are most interesting to test.

I mentioned this in the 4309 thread. I believe JBL/Harman still blind test during the design phase and require the speaker to win. If so then the 4309 passed their in house testing.

A good point. Presumably, this did well enough in blind testing, despite the measurement flaws.

By the way not sure what you mean by 5-6 resonances, what do you mean?

It's a good question, as I do think we all may have slightly different definitions of what a "resonance" is. That no doubt contributes to some of the debate, even when we likely mean the same thing. When I first heard the word, the idea that came to mind was "some part of the box other than the drivers that resonates, and that alone". I'm no longer so sure of that "alone" part, but maybe it's true. I've even seen Toole describe almost any deviation from flat as a resonance. The most common idea I've seen is "a peak or dip in the frequency response that shows up in all of the top curves(on axis, listening window, early reflections, and sound power)". Peak resonances are more harmful than dip resonances. I actually wouldn't call that port error of the 4309 as a resonance, as it only shows up on axis. I'd call it a port cancellation? I would say the cancellation is right in the middle of a fairly large resonance, though. Either that, or it's surrounded by a couple resonances.

I also am not surprised it's not very audible. My 708p has a similar port problem, and I don't notice the issue at all. When looking for resonances I tend to look for:

- peaks or dips that show up in the top half of the CEA-2034, but not the bottom
- blips in the impedance curve. Resonances may or may not show up here, but blips here almost always correspond with a resonance in the FR(from what I've seen)
- look for somewhat high q, but not too narrow. I don't count all the little tiny blips as resonances, nor would I count super wide dips or peaks. For example, I wouldn't call the 3dB peak from 70-250Hz of this PSB speaker as a resonance, nor the rise and fall from 2.5-10kHz.

This PSB speaker is actually an interesting one to see resonances. Too me, it looks like there are a couple resonances going on in that 1-2kHz region, but also some non-resonance errors, as they also show up as directivity errors at the bottom. Not sure what to call those. Amir calls them "acoustic events". I like that . Also, I see a resonance(or even a couple?)between 2.6 and 3.9kHz, and another at 15kHz. The one at 15kHz I pretty much ignore.

Not a hint of distortion at high volumes (with an HP at 35hrz to cut out excess excursions), actually one needs to be careful as it is playing much louder than it seems.

No doubt this is a huge part of why they did so well(especially when testing with just one speaker). Amir's preferences kinda lean towards dynamic capability being more important than better measurements, though bad measurements still usually lead to worse Amir preference. I get the same feeling with my JTR speakers. You don't really realize who loud the volume is until you try to talk to a friend or sing along with the song. It's an awesome experience.

 

[bluefuzz]

I was pointing out that all of this especially the Harman approach, is preference driven

Yes, and I was pointing out that it could be nothing else. You seemed to be implying it could ...

 

[AlexF]

Any chance of a 'grill on' 'grill off' plot comparison occasionally?

For three reasons:

Yes, but, primary reason: “grill on” is how will be used in practice.

 

[Mart68]

If they were half the price then I'd have given them a pass. But there's got to be better than these for that money?

I thought the poll might be brutal and it was.