PSI Audio monitors - do they measure well?

[Jas0_0]

Hi all,
Does anyone have experience of how well the PSI Audio AXXm series measure? Subjective reviews are always favourable, but here measurements of the A21m don't look that great: https://www.resolutionmag.com/wp-content/uploads/2016/03/PSI-Audio-A21-M.pdf. The vertical dispersion doesn't seem good, and the on-axis seems to show a large elevation in mids and low treble. Excuse the low res images - best I could do from the review PDF.
Any thoughts?
James

 

[Purité Audio]

I thought SoundonSound had measured but just a review,
https://www.soundonsound.com/reviews/psi-audio-a17m-a21m

I bought some for here, only the small ones A17 definitely near field only .
PSI make a big thing of being analogue only, although I am not sure that is such a good idea.
Keith

 

[Soniclife]

That shelved down bass looks quite ATC!

 

[Jas0_0]

That shelved down bass looks quite ATC!

That’s exactly what I thought. The overall on-axis shape is not dissimilar.

 

[ernestcarl]

Remembered seeing PSI-a14m measurement here. And you can compare that with their measurement of the KH-120 here.

 

[q3cpma]

They do provide measurements "made in anechoic room at 1 m". For example, the 2-way A17:

So I'd wait for more trustworthy data to discard them completely. Distorsion is already quite high at 90 dB, though.

 

[Soniclife]

Each speaker is individually calibrated in our anechoic chamber and is delivered with its frequency response sheet.
A proof of quality.

Their own measurement looks nothing like the test result.
@Purité Audio do you have a measurement of your's?

 

[Jas0_0]

@Purité Audio do you still have your A17Ms? Would you consider sending to Amir? I guess London to US West Coast might be pricey. Perhaps a whip-round - I’d gladly chip in.

 

[Purité Audio]

I have a photo,


Keith

 

[Purité Audio]

@Purité Audio do you still have your A17Ms? Would you consider sending to Amir? I guess London to US West Coast might be pricey. Perhaps a whip-round - I’d gladly chip in.

Long gone Jas, sorry, they were ok, I had a chance to hear the big A25’s? and matching subs which were also impressive , but to be frank the SBIR ‘type’ of loudspeaker offer real advantages in domestic/ relatively untreated rooms.
Keith

 

[tuga]

Some measurements of the A215-MS:

 

[napilopez]

Sound and recording has measurements of the PSI A-14 and A-17M. However, as far as I can tell, they are only available under the paywalled massive test of 80 studio monitors As they're under a paywall I'm not totally comfortable sharing here, but if you're very curious, the document is about 9 Euro and it's a nice reference: https://www.soundandrecording.de/sh...&utm_campaign=MDDSR188_studiomonitore_special

(it's in German, but graphs and polar plots are universal )

That said...

They do provide measurements "made in anechoic room at 1 m". For example, the 2-way A17:

So I'd wait for more trustworthy data to discard them completely. Distorsion is already quite high at 90 dB, though.

Sound and recording's data looks quite close to PSI's official data, except the dip between 1-2k looks a deeper. It's not quite Neumann flat, but other than the sharp dip at about 1.3khz I think it's about as flat as it needs to be.

Neither the A14 nor the A17M show bass shelf in the A21M measurements in the OP, so either the speakers are tuned differently or those measurements are off. Directivity is mostly good too.

 

[Deleted member 16502]

They do provide measurements "made in anechoic room at 1 m". For example, the 2-way A17:

So I'd wait for more trustworthy data to discard them completely. Distorsion is already quite high at 90 dB, though.

How do they have that phase response flatness? I thought PSI was riding the all analog train (they say no DSP on their general product page) and DSP is needed for that?

 

[napilopez]

How do they have that phase response flatness? I thought PSI was riding the all analog train (they say no DSP on their general product page) and DSP is needed for that?

How it works is a bit over my head but this post talks about the tech some:

https://www.helios.nl/weblog/psi-audio-cpr-technology-

 

[pozz]

How do they have that phase response flatness? I thought PSI was riding the all analog train (they say no DSP on their general product page) and DSP is needed for that?

All-pass filters which allow the signal through but introduce a delay.

 

[jlo]

That shelved down bass looks quite ATC!

I work for a company that sells PSI but also Neumann, Genelec, JBL, PMC,.... and I have measured many PSI loudspeakers in mixing studios, mastering rooms,...
It is clear that all PSI models are engineered for a flat frequency response in anechoic room. Each and every loudspeaker is measured before leaving the factory and the client gets those exact graphs in the box. There is no low frequency shelf in any model but there is also no boost !

 

[Soniclife]

I work for a company that sells PSI but also Neumann, Genelec, JBL, PMC,.... and I have measured many PSI loudspeakers in mixing studios, mastering rooms,...

Are you allowed to share the measurements?

 

[KSTR]

How do they have that phase response flatness? I thought PSI was riding the all analog train (they say no DSP on their general product page) and DSP is needed for that?

You need a cascade of peaking allpass filter to undo the phase rotation of a typical XO. The peaking is in the group delay, not the magnitude function.

The cost is that the implementation is fairly complex, you will need quite a lot of those properly tuned peaking allpass filters to undo the phase response of, say, a 4th order woofer-to-mid XO at 300Hz, two-digit numbers. Noise and distortion is an issue here, and any slight detuning causes magnitude ripple. And the hard part is that one cannot extend the compensation up to very high frequencies, somewhere up high the slope of the eventually falling group delay is even steeper (meaning a very quick phase rotation), it is quite hard to push this edge to above 20kHz.

But even when you do and the passband of the tweeter is higher, the speaker's phase is almost impossible to measure correctly (most people and software will have a hard day here) because the reference point is not the peak of the impulse response (actually, this is the correct reference point only for linear-phase low-pass filters like DAC-filters and such).

The peaking allpass cascade concept is used instrumentation, for example in some oszilloscopes as well (http://cdn.teledynelecroy.com/files/whitepapers/group_delay-designcon2006.pdf), this plot is showing how it is done:

 

[jlo]

Are you allowed to share the measurements?

yes, which one would you like to see ?

 

[Jas0_0]

yes, which one would you like to see ?

Thanks, I’d be interested in both A17m and A21m.
J