Paradigm Premier 800F Tower Speaker and 500C Center Speaker Review

[Swerd]

Not to poop on Paradigm's parade, but I find this confusing that 2k pair of large towers rated at -2db 50hz and a cheaper large bookshelves (BMRs) claim to have -2db 34hz bass response.
Am I missing something here? (besides the obvious difference in sensitivity)

Where speaker sensitivity is concerned, something called “Hoffman’s Iron Law” takes effect. It offers you any two of the following: small cabinet size, deep bass and high sensitivity. You can't have all three at the same time. See http://www.salksound.com/blogtopic.php?id=3

These paradigm speakers have a 92 db sensitivity and the BMRs only 85 db. So the Paradigm is 7 db more efficient and that is a lot.

I find 89 dB sensitivity for those Paradigm speakers believable, but not 92 dB.

In his review, shadyJ said:

Sensitivity measurements for 2.83v at 1 meter were measured at 88.8 dB for the 800F [tower] and 89.7 dB for the 500C [center speaker]. This is close to Paradigm’s spec for this rating which is 89 dB anechoic for both speakers.​

Is the sensitivity rating of 92 dB a number provided by Paradigm? Is that additional 3 dB of sensitivity an in-room estimation, or is it a number provided by the marketing department?

 

[Swerd]

They are marketed as a full range driver, but they roll off at 170 Hz. From roll off to 20 KHz they have a very uniform nice frequency response. They are only $18 a pop and are probably the widest bandwidth drivers around. You actually don't really need a tweeter. I have a feeling panels of those would compare very favorably to electrostatics.

Those Tectonic Balanced Mode Radiators may be marketed as a full range driver, but as you noted, they roll off below 170 Hz. In fact, they make very poor sounding tweeters. They are best used as a mid range driver in a 3-way speaker.

In the past, Dennis Murphy had tested a 2-way bookshelf speaker, Cambridge Audio Aero 2, which combined a woofer (6½" if I recall) with the Tectonic driver used as a tweeter. His measured frequency response curves (below) were significantly worse looking than the more optimistic curve published by the manufacturer. Based on these initial observations, Dennis decided to try using that Tectonic driver in a 3-way design – a design which he now sells as the BMR monitor.

On axis

Off-axis

 

[shadyJ]

Where speaker sensitivity is concerned, something called “Hoffman’s Iron Law” takes effect. It offers you any two of the following: small cabinet size, deep bass and high sensitivity. You can't have all three at the same time. See http://www.salksound.com/blogtopic.php?id=3
I find 89 dB sensitivity for those Paradigm speakers believable, but not 92 dB.

In his review, shadyJ said:

Sensitivity measurements for 2.83v at 1 meter were measured at 88.8 dB for the 800F [tower] and 89.7 dB for the 500C [center speaker]. This is close to Paradigm’s spec for this rating which is 89 dB anechoic for both speakers.​

Is the sensitivity rating of 92 dB a number provided by Paradigm? Is that additional 3 dB of sensitivity an in-room estimation, or is it a number provided by the marketing department?

Paradigm's specs the sensitivity of these as "Sensitivity: room / anechoic: 92 / 89 dB" so they are pretty honest about this speaker's ability.

 

[AcuDefTechGuy]

It looks like ~73dB at 9kHz and ~ 63dB at 10kHz..

I see what you're talking about - there is a deep, sharp dip in the response from 9 to 10kHz in the off-axis response...

Ah I see now. Yeah, the dip centered around 10 kHz is occurring way off axis....

And I thought I was looking at the ON-AXIS FR!

Is there an on-axis FR graph that we normally see?

 

[shadyJ]

And I thought I was looking at the ON-AXIS FR!

Is there an on-axis FR graph that we normally see?

Lol, I suppose I should have made that more clear. If you want to see the direct axis response, look at the line on the 3D waterfall graph that is '0' on the Z-axis. Trace that line, that is the direct axis response. On the profile view, it is usually the top line, but not always, as sometimes off-axis angles can develop more output, usually from diffraction effects. I'll will try to talk a little bit more about what is happening in those graphs and why I use these particular illustrations sometime later, hopefully today.

 

[Matthew J Poes]

Lol, I suppose I should have made that more clear. If you want to see the direct axis response, look at the line on the 3D waterfall graph that is '0' on the Z-axis. Trace that line, that is the direct axis response. On the profile view, it is usually the top line, but not always, as sometimes off-axis angles can develop more output, usually from diffraction effects. I'll will try to talk a little bit more about what is happening in those graphs and why I use these particular illustrations sometime later, hopefully today.

I just want to add that I use the same approach as James when I report speaker measurements in my reviews. I use different graphics, but we do it for the same reasons. I think we have gotten so used to looking at on-axis that it has given consumers the wrong idea. That the single on-axis measurement is the most important and what happens off-axis is merely nice or incidental. It’s really important to look at everything together.

I might even go so far as to say that the 0 degree is not the most important even if the first ~20 degrees is. In order to break people or the habit of putting so much weight on the 0 degree axis, it’s helpful to use dramatically different ghraphics that provide a better story.

You see speakers all the time that measure great on axis and bad off axis. That’s a bad speaker, it’s on axis response is unimportant. There are speakers which measure bad onaxis but great off axis. That is a good speaker, it’s onaxis response is unimportant. If that’s true, then we need to not focus on a single on axis measurement.

 

[Matthew J Poes]

While I appreciate being conservative in such estimates as much as anyone, I'd reckon for a 12' distance in an actual room, the THX Ultra certification might be a fair place to start for a minimum requirement given that it's specifically for a 3000 cubic foot theater / 12'+ viewing distance (and distance to the front speakers presuming they are adjacent to the screen). Going by that, you could have a 90dB sensitive loudspeaker like the Atlantic Tech 6200e paired with a 150W (2ch driven into 8 ohms) amp like the Integra DTA-70.1 (measurements),and THX would expect that to get the job done.

Regarding the 105dB figure, one should also keep in mind that isn't intended to be a continuous RMS value. 105dB at the seats is expected to correlate with 0dBFS on the digital side of things, i.e. the absolute maximum/peak level available to a channel. Used properly when producing a soundtrack (not always the case),that level should ideally never be reached by the loudest short term peaks in a film to ensure there isn't any clipping of the waveform on the soundtrack.

It is but that isn’t how it’s tested. I have a document that goes through how it’s to be tested in a commercial cinema to ensure that you are achieving the standard. It involved an RTA and pink noise originally and later they adopted a sweep or a stepped sine test. This was actually adopted outside of just THX, Dolby includes it in some of their setup paperwork. That’s an RMS test.

I’d very much like to add this as a test for speakers but have been concerned that manufacturers won’t like me doing that to my loaner review speakers. A standard test procedure exists for distortion limited output and compression. The tests I have done suggest many THX speakers would not actually meet their own standard very well. Being at their absolute limits, needing more than 150 watts as you suggest, and showing significant compression.

Without this kind of testing, consumers can’t know. I really think it’s worth doing. Just a matter of convincing manufacturers to allow me to torture their speakers and add it to reviews.

 

[Steve81]

The tests I have done suggest many THX speakers would not actually meet their own standard very well. Being at their absolute limits, needing more than 150 watts as you suggest, and showing significant compression.

Not surprising. From a manufacturer's standpoint, there's not much incentive to do anything beyond the minimum necessary to achieve THX certification.

Just a matter of convincing manufacturers to allow me to torture their speakers and add it to reviews.

Good luck

 

[Matthew J Poes]

Not surprising. From a manufacturer's standpoint, there's not much incentive to do anything beyond the minimum necessary to achieve THX certification.



Good luck

Yeah it won’t happen. I have a contact with two THX speaker makers and neither was super comfortable. I got “maybe” responses. @shadyJ is also pretty skeptical. My plan would put safeguards in place but it’s not perfect. I may do some tests using this approach and write them up just to make a point, but not be able to do it as routine.

 

[shadyJ]

And I thought I was looking at the ON-AXIS FR!

Is there an on-axis FR graph that we normally see?

To explicate a bit more on this, if you look at these graphs as an example:

The lower graph is simply a 2D profile image of the upper 3D graph. If you want to best see the shape of the direct axis response, look at the top contour of the 2D profile graph. The direct axis response is almost always where the speaker will be producing the highest amplitudes, so it will be at or near the top of that ridge. Every successive line going down from that 'ridge' is just responses of the speaker moving further away from the direct axis angle. The outermost angle is at 100 degrees on these particular graphs, and that is obviously way off axis, in fact it is slightly closer to the rear of the speaker than the front.

So, to answer your question more fully, yes there is a traditional direct response here, but it is packaged in by off-axis responses, so you can see the speaker's behavior not just directly in front of it, but all area in front of it. As Mr. Poes explained, a direct axis response alone is not enough to characterize a speaker's performance. It needs context. Furthermore, many speakers weren't designed to be listened to on direct axis. Some speaker designers expect that the user will angle the speakers straight forward in parallel lines, some expect a slight toe-in, and some expect a severe toe-in. These graphs can indicate what angle will make for a more preferable sound to the listener. For example, take a look at this response:

Obviously that would make for a really bright response on the direct axis. but if you follow the lines, you can find an angle that will make these not quite as bright. Compare the 50 degree angle response with the direct axis response:

The 50 degree angle is much more tolerable- still bright, but a bit easier on the ears. These speakers were not intended to be listened to on direct axis.

 

[Matthew J Poes]

To explicate a bit more on this, if you look at these graphs as an example:

The lower graph is simply a 2D profile image of the upper 3D graph. If you want to best see the shape of the direct axis response, look at the top contour of the 2D profile graph. The direct axis response is almost always where the speaker will be producing the highest amplitudes, so it will be at or near the top of that ridge. Every successive line going down from that 'ridge' is just responses of the speaker moving further away from the direct axis angle. The outermost angle is at 100 degrees on these particular graphs, and that is obviously way off axis, in fact it is slightly closer to the rear of the speaker than the front.

So, to answer your question more fully, yes there is a traditional direct response here, but it is packaged in by off-axis responses, so you can see the speaker's behavior not just directly in front of it, but all area in front of it. As Mr. Poes explained, a direct axis response alone is not enough to characterize a speaker's performance. It needs context. Furthermore, many speakers weren't designed to be listened to on direct axis. Some speaker designers expect that the user will angle the speakers straight forward in parallel lines, some expect a slight toe-in, and some expect a severe toe-in. These graphs can indicate what angle will make for a more preferable sound to the listener. For example, take a look at this response:

Obviously that would make for a really bright response on the direct axis. but if you follow the lines, you can find an angle that will make these not quite as bright. Compare the 50 degree angle response with the direct axis response:

The 50 degree angle is much more tolerable- still bright, but a bit easier on the ears. These speakers were not intended to be listened to on direct axis.

Eek! I don’t think that speaker was designed for listening. Maybe high up on a bookshelf behind a pillow.

 

[Matthew J Poes]

@shadyJ do you think that peak at 3.5khz that seems to turn into a trough at 4khz is a diffraction effect of some kind. It appears to be the same problem. It moves up In frequency and reverses direction on amplitude. That’s an interesting effect.

 

[shadyJ]

@shadyJ do you think that peak at 3.5khz that seems to turn into a trough at 4khz is a diffraction effect of some kind. It appears to be the same problem. It moves up In frequency and reverses direction on amplitude. That’s an interesting effect.

It could be an artifact from cone breakup. If you remember, that woofer is running full range with no high-pass filter. It could also be some kind of integration issue between the woofer and tweeter where a phase summation on axis is turning into a cancellation off axis. Maybe it could be diffraction too, but my first guess is cone breakup. Anyway, that overall response is clearly imperfect, but this speakers didn't sound all that bad, for the $50/pair price. Keep in mind the majority of that excess treble energy occurs above 10 kHz where there is usually not that much content. It's a flawed speaker but I still managed to enjoy music and movies with it. I think audiophiles who would stand aghast at those measurements ought to give them a listen, they might be surprised.

 

[Matthew J Poes]

It could be an artifact from cone breakup. If you remember, that woofer is running full range with no high-pass filter. It could also be some kind of integration issue between the woofer and tweeter where a phase summation on axis is turning into a cancellation off axis. Maybe it could be diffraction too, but my first guess is cone breakup. Anyway, that overall response is clearly imperfect, but this speakers didn't sound all that bad, for the $50/pair price. Keep in mind the majority of that excess treble energy occurs above 10 kHz where there is usually not that much content. It's a flawed speaker but I still managed to enjoy music and movies with it. I think audiophiles who would stand aghast at those measurements ought to give them a listen, they might be surprised.

I would have assumed the same source but the frequency moves. I didn’t think those kinds of artifacts could shift in frequency as the angle changed. I thought that was a unique property of diffraction.

 

[Matthew J Poes]

Actually I just looked up some diffraction on various speakers at different angles and it seems that some of that show changing response with angle. Peaks at one frequency go away off axis, but a big cancelation shows up just above that frequency. Probably is breakup.

That did seem too low in frequency to be diffraction in any meaningful way.

 

[snakeeyes]

Excellent review!

I wonder how the slightly smaller 700F might compare to the 800F if a capable pair of subs are added to the system anyhow at 80hz crossover. Personally I would go 800F anyhow since the price isn’t much more.

That center 500C looks like a nice speaker by the way. That’s what really got my attention.

 

[keepr88]

Great review!

I currently have the monitor 11 v7. Im curious if anyone has experience with both of these speakers and is the premier 800f a large step up from the monitor 11s. Any thoughts would be appreciated.