PSB Alpha P5 Speaker Review

[amirm]

Why is there so much variation in the aspect ratio of the contour maps with each review? It seems to me like they should always be identical, both in terms of aspect ratio and in terms of resolution.

Klippel software seems to have kludged up three different graphing systems each with their own features. The visualization tool where the directivity plots come out of, has no export ability and always starts with a giant window. I resize that manually. Then, I have to pull the SPL bar graph on the right to the left in photoshop. Otherwise the graph is too wide causing it be resized and become blurry. So there is no way I can get them to be the same size in general.

In this review, I went way beyond that due to a comment was made in the last review. That the beamwidth is very revealing of directivity errors whereas contour map is not. I realized part of the problem there was that I was compressing the vertical axis a lot, causing it to lose detail in that axis. So I enlarged it to almost square aspect ratio. I am not sure it helped a lot but that is the reason.

Open to feedback on this.

 

[Sancus]

Klippel software seems to have kludged up three different graphing systems each with their own features. The visualization tool where the directivity plots come out of, has no export ability and always starts with a giant window. I resize that manually. Then, I have to pull the SPL bar graph on the right to the left in photoshop. Otherwise the graph is too wide causing it be resized and become blurry. So there is no way I can get them to be the same size in general.

In this review, I went way beyond that due to a comment was made in the last review. That the beamwidth is very revealing of directivity errors whereas contour map is not. I realized part of the problem there was that I was compressing the vertical axis a lot, causing it to lose detail in that axis. So I enlarged it to almost square aspect ratio. I am not sure it helped a lot but that is the reason.

Open to feedback on this.

I see, it is problematic if the only way to get those images is via manual screenshot. You may be able to achieve better results if you can use a utility like Sizer to force the window size to a fixed dimension every time.

I agree in general that a more square aspect ratio is probably better. The variations in these charts are top to bottom, so the more square you make the aspect ratio the more pronounced they will be and the easier it will be to see issues.

A better long term solution might be to build all(or most) review graphs from raw data instead of relying on the Klippel visualizations. I am not sure if there are any limitations on what data or resolution you can export, I know that the data dumps in each review aren't full resolution. But for example if you can export off-axis data to +/- 1 degree it may be a lot easier to do that and run some software script over it to produce review graphs. Beamwidth should be doable that way too, not sure how you are doing them right now. I think Beamwidth is really good btw and makes it much easier to see and compare the actual dispersion width between speakers so thanks for adding it.

Maybe @hardisj can be of some help as the graphs he's been posting on his site are very nice and perfectly consistent.

 

[ruinevil]

When heard back to back, I liked the P3 over the P5.

 

[YSC]

With Barton being such a big name this is really disappointed

 

[riker1384]

Yeah, I was disappointed with the measurements I saw on other sites for these. The previous Alpha B1 looked much better:

https://www.soundstagenetwork.com/measurements/speakers/psb_alpha_b1/
https://www.stereophile.com/content/psb-alpha-b1-loudspeaker-measurements

I hope they're not slipping. My first pair of good speakers were the PSB Image B25. I was astonished at the clarity of them, it sounded like I could hear a hundred times as much detail as with crummy vintage speakers and such. (Or the overly warm Vandersteen 1C I tried beforehand.) The main flaw was that they always sounded slightly cold and metallic to me in the treble. The probably corresponded to the 10Khz bump on its otherwise flate response: https://www.stereophile.com/content/psb-image-b25-loudspeaker-measurements But it didn't sound harsh in the way that a low/mid treble peak typically does.

 

[daftcombo]

Testing for ability to play deep bass was a fail when playing this track from Burak Malçok's album: Sakli Nefes, Toz Ruhu


The very low frequency bass is almost inaudible but it easily caused the P5 woofer to bottom out. I swapped out the P5 for Revel M16 and it too struggles with this track but played at probably 2X the level (still insufficient).

Hi Amir,
Was that the case with all the standmount speakers you tested?

 

[Koeitje]

Ranks similar to the Klipsch RP-600M, also headless, though the Q Acoustics 3020i also is a 4 but didn't get headless, but with a sub score but it a good deal higher than this and the Klipsch.

In the end the panther is Amir's judgement. With just the score you miss out on spl, I don't want a speaker that bottoms out super easily. So if Amir that takes into account for the panther I'm a happy camper.

 

[digicidal]

In the end the panther is Amir's judgement. With just the score you miss out on spl, I don't want a speaker that bottoms out super easily. So if Amir that takes into account for the panther I'm a happy camper.

Considering the apparent definition of "comfortable levels" that @amirm seems to frequent... I'm guessing that SPL plays rather heavily into that panther equation. At the levels I can commonly listen to at home (unless my wife takes a long trip to the store) I'd be amazed if anything ever bottomed out. I need to find some ear plugs that double as jewelry...

 

[LearningToSmile]

I am super interested in testing a Paradigm. Let me see if I can find something reasonable from them.....

It would be interesting to see Paradigm Monitor SE Atom tested - it's extremely affordable(barely over $300 a pair) and people seem to like it. Not sure how representative it would be of Paradigm as a brand, however.

 

[edechamps]

I have added the PSB Alpha P5 to Loudspeaker Explorer where it can be compared to other speakers.

Okay consistency within the listening window:

 

[flipflop]

@QMuse @flipflop @Maiky76 @pierre pretty please

I've got one ready. It's based on the LW with a -0.3 dB/oct. slope. I'll wait for @pierre to run the numbers before posting it.
Side note: People don't get notifications if you tag them while editing a post

 

[uwotm8]

Why so much drama for nothing?
While FR graph looks OHH OMG SO UGLY for those guys possessed by straight lines, its actually shows declared +-3 dB response over range.
YES, there will be audible tonality changes if compared to dead neutral speakers and NO, this is nothing like something extremely bad for $300 "entry level hi-fi" pair. It's rather typical, "not great not terrible".
THD graphs @high levels are dissapointing tho (but it's questionable if typical owner of such speakers pushes it up to limits).

Would be interesting if Amir measure Infinity Alpha/Beta/Primus speakers of same size and retail price which are now dirt cheap used. Those have decent FR graphs (check Stereophile) while sounding rather crappy.

 

[thewas]

Would be interesting if Amir measure Infinity Alpha/Beta/Primus speakers of same size and retail price which are now dirt cheap used. Those have decent FR graphs (check Stereophile) while sounding rather crappy.

https://www.audiosciencereview.com/...inity-il10-speaker-review-measurements.14310/
https://www.audiosciencereview.com/...infinity-r162-bookshelf-speaker-review.12333/

 

[MZKM]

Unfortunately, PSB doesn't mention the intended axis,

https://www.psbspeakers.com/wp-content/uploads/2018/11/PSB_Owners-Manual_Alpha.pdf

It’s a generic copy/paste, but it is what owners will read.

+20° V is indeed more flat. However, I wonder how that will affect the crossover region off-axis.

 

[MZKM]

Why is there so much variation in the aspect ratio of the contour maps with each review? It seems to me like they should always be identical, both in terms of aspect ratio and in terms of resolution.

This one is 1.19(911 x 762), the Salk WOW1 was 1.9(870x457). But every review is a bit different, despite always covering the same frequency range and angles. The way they are right now, to compare them side by side you HAVE to resize the images first otherwise it's very easy to draw mistaken conclusions.

You can always look at mine , though they are not as pretty (it is a spreadsheet that does a color at each data point, whereas the Klippel uses more points to draw nice lines).

However, older speakers measured were at a higher resolution (more data points), thus their graphs are a tad wider, but it’s mostly in the bass.

 

[EchoChamber]

Thank you Amir for showing us the truth about these speakers!

Looks like those were tuned to sound warm and appeal to shoppers masking poor quality electronics. Or to mask poor speaker components and make them more listenable. That’s my guess anyway.

Many, many eons ago, I went to a friend’s home and heard a small PSB bookshelf speaker, I remember them sounding overly warm too. It might be a house sound.

Product of the year is laughable... lol... Now I remember why I haven’t read Stereophile for years... How can they be taken seriously? Publication awards are often fake, companies pay to get them so they can show an award badge on their marketing communications. Or maybe Stereophile was just trying to seduce PSB to get a bigger share of their advertising budget.

But then I wonder... Who buys budget passive speakers today? As a mainstream consumer, I’d probably get a Bluetooth speaker, or a Sonus One.

 

[raindance]

The upside down tweeter designs are normally designed for listening / measuring on midwoofer axis, Mission started this concept years ago. I'm not sure if there's some complexity in making nearfield measurements and if this "effect" is only evident in the mid field.

 

[thewas]

The upside down tweeter designs are normally designed for listening / measuring on midwoofer axis, Mission started this concept years ago. I'm not sure if there's some complexity in making nearfield measurements and if this "effect" is only evident in the mid field.

 

[richard12511]

Why so much drama for nothing?
While FR graph looks OHH OMG SO UGLY for those guys possessed by straight lines, its actually shows declared +-3 dB response over range.
YES, there will be audible tonality changes if compared to dead neutral speakers and NO, this is nothing like something extremely bad for $300 "entry level hi-fi" pair. It's rather typical, "not great not terrible".
THD graphs @high levels are dissapointing tho (but it's questionable if typical owner of such speakers pushes it up to limits).

Would be interesting if Amir measure Infinity Alpha/Beta/Primus speakers of same size and retail price which are now dirt cheap used. Those have decent FR graphs (check Stereophile) while sounding rather crappy.

I've got some Infinity Beta's I offered up for test. I don't agree with them sounding crappy, though. For the price, they sound excellent when compared to my Revel M105s, JBL 308p mkII and Infinity r263.

 

[hardisj]

Klippel software seems to have kludged up three different graphing systems each with their own features. The visualization tool where the directivity plots come out of, has no export ability and always starts with a giant window. I resize that manually. Then, I have to pull the SPL bar graph on the right to the left in photoshop. Otherwise the graph is too wide causing it be resized and become blurry. So there is no way I can get them to be the same size in general.

In this review, I went way beyond that due to a comment was made in the last review. That the beamwidth is very revealing of directivity errors whereas contour map is not. I realized part of the problem there was that I was compressing the vertical axis a lot, causing it to lose detail in that axis. So I enlarged it to almost square aspect ratio. I am not sure it helped a lot but that is the reason.

Open to feedback on this.

This is why I wound up writing my own Matlab scripts and processing the raw data so I can get it in the format I want. I had a lengthy Skype session with them last month about their graphs and why I don't use them. They should be fixing a lot in their next major dB-Lab release. They already had some items on their list so I just added my two cents. Hopefully, the next major update will fix your ails in (which isn't planned for a while). I will be beta-testing it soon so maybe I can help you out there, if you have any specifics you want to bounce off me.

If you would like, you can send me a copy of the kdbx file (I know it is large; if you want to compress it and throw it on my dropbox that will work) and I can look at it and see if there are any suggestions I can come up with or just see how the raw data output can be made easier using their dB-Extract tool.

Anyway, just trying to help. LMK if this interests you.