Klipsch RP-600M Speaker Review

[Shazb0t]

If there is no compression chamber then it isn't a horn. It's pretty simple. This speaker just has a large waveguide.

Actual horn

 

[napilopez]

If those graphs are accurate, they just clearly indicate that the crossover is badly designed. It's nothing new. It's a fact that many speaker manufacturers don't have the qualified personnel to design smooth passive crossovers. The curves clearly show that the woofer response start to slope down at a too low frequency, whereas the tweeter's knee frequency is too high, so their responses fail to properly cross at a specified frequency.

Yeah as @ROOSKIE said I find it extremely hard to believe this response wasn't intentional. Even if it was somehow a cost-saving measure, they chose where to make their compromises.

Part of me wonders whether they chose to prioritize directivity because many listeners will use room EQ via receivers? Seems like a stretch, but one possibility. It also seems weird that on their higher end speaker they do go for a flatter response.

Yah one of the forum members reviewed the active BW DUO and the measurements kinda were crap. All the DSP available and they still don't go for flat. They have a whole other thing going. Not sure what their intent is but is something...

That was me. The Formation Duo measured decently on axis but off axis they did not seem to go for any kind of directivity control, and their soundstage suffers for it. I felt they had a narrow sweet spot.

I still enjoyed them though. Their saving grace was their bass, which is very impressive. They get a ton of mileage out of those 6.5-inch drivers, which for the no-subwoofer crowd(a large portion of audiophiles willing to spend that much money on speakers in the first place) likely makes them more impressive than many alternatives.

 

[Frank Dernie]

PS. I haven`t shared Amirs review at Klipsch forum yet. If someone is brave enough.. Usually it leads to "there must be something wrong about the data.." and bashing.

That is what happens in all reviews of all makes.
If the review is bad there is a plethora of comment from mainly owners but also fans of the brand that the data is bad/operator is an incompetent idiot etc..
This usually goes on and on, and on...
It doesn't usually happen when the measurements are good where the equally bad data from the still equally incompetent measurements are what people wanted to see.
Human nature and, I suspect, the fact that aptitude in things technical is rare.

 

[John Atkinson]

The RP600M's tweeter and woofer are connected in the same polarity - see fig.7 at https://www.stereophile.com/content/klipsch-reference-premiere-rp-600m-loudspeaker-measurements - and while fig.4 at that URL shows the acoustic crossover slopes, 4th-order high-pass on the tweeter, close to 3rd-order low-pass on the woofer, it doesn't show the phase of the drive-units in the crossover region.

I retrieved the MLSSA files for the Klipsch RP-600M from my archive and added the tweeter and woofer responses after inverting the tweeter. You can see from the attached graph that in this condition (dotted trace), the suckout in the crossover region is very much deeper than it is in the factory condition (solid trace). (Note that I haven't corrected for the microphone's departure from flat in this graph, just for the antialiasing filter's passband behavior.)

John Atkinson
Technical Editor, Stereophile

 

[Vladimir Filevski]

With thanks to Mr. Atkinson, this concludes the Klipsch RP-600M frequency response was deliberately tailored with dip between 1 and 3 kHz.

 

[MZKM]

Hi Amir, when you write in this graph “Peaking above 10kHz will sound bright,” that hasn't been my experience. Excess energy in the top octave, if also present off-axis, tends to add "air" or fizziness and lispiness if extreme. If the excess above 10kHz is small and the tweeter's dispersion is limited in this region (as they almost all are), then it can actually result in a neutral top-octave balance in all but small rooms. Brightness, I have found, tends to be due to too much energy between 5kHz and 10kHz.

John Atkinson
Technical Editor, Stereophile

Yep. If we look at the predicted in-room response, this will tell us whether the emphasis >10kHz will cause “fizziness”.

If we look at it for the RP-600M (gray trace):

Dispersion is not that narrow as to forgive the emphasis on-axis, it’s pretty much neutral from ~3kHz-14kHz, where a slope-down is preferred.

I also agree that brightness is most usually caused by emphasis lower than 10kHz.

 

[tuga]

With thanks to Mr. Atkinson, this concludes the Klipsch RP-600M frequency response was deliberately tailored with dip between 1 and 3 kHz.

Or maybe this was the best that they could do with the available drivers.

 

[Shazb0t]

Or maybe this was the best that they could do with the available drivers.

That doesn't make sense.

 

[kejar31]

Is this dip in the 1 to 4khz range not called the BBC dip?

 

[thewas]

The "BBC" presence dip is usually a bit higher, often between 2-3 or 2.5-4 kHz.

 

[Kw6]

That doesn't make sense.

Klipsch sells the floor standing model of this and GR research also found a big dip in it!

 

[richard12511]

Am I confused, or do Amir and Joe seem to have reached opposite conclusions about this speaker?

I trust Amir's review a little bit more

Joe seems like a good guy, though.

 

[richard12511]

A sub can't fix this mess....should be more like 4.1 with a sub.

EQ can fix this. Do you use EQ?

 

[LTig]

If there is no compression chamber then it isn't a horn. It's pretty simple. This speaker just has a large waveguide.

Actual horn

What then do you call those audiophile highend horn speakers which have no compression chamber?

 

[kejar31]

Klipsch sells the floor standing model of this and GR research also found a big dip in it!

I own the RP-8000F speakers and have installed the GR-research crossover (did not install the tube connectors as I absolutely do not believe they make a difference, aka my brain refuses to accept that sh**) I also took a FR measurement from before and after, and the crossover definitely resolved the suck-out issue. Also I very much prefer the speakers post crossover replacement FWIW.

Notice: my measurements are in room 1 meter away so they are not the best but they still show the changes made by the crossover swap.

 

[richard12511]

As are the GR research measurements shown in the video earlier.

Yeah this isn't a "bad sample" issue. Joe and Tell also said that all of his measurements showed the same dip.

 

[napilopez]

Without specifically stating what it means, he basically said that these speakers only sound good when you toe them in and align your ears perfectly in line with a spot between the tweeter and woofer. I don't want speakers that require such a narrowly defined listening position.

However, that conclusion seems to conflict with Amir's review that said "The 'horn' is actually not a horn but a waveguide that is doing its job to provide uniform off-axis response allowing one to EQ the mid-range dip."

Am I confused, or do Amir and Joe seem to have reached opposite conclusions about this speaker?

I actually do not think these are mutually exclusive opinions - perhaps I'd need more detail on what Joe meant. On a narrowish directivity design, moving slightly to the left or right can make the soundstage disappear and the image coalesce around one speaker because the SPL change is relatively drastic. Based on the Klipsch's measurements, I would totally assume a relatively narrow defined listening position, regardless of how smooth the directivity is. The soundstage will seem well defined with sharp imaging, but the physical sweet spot will still be small especially if your LP isn't very far (larger angle changes per equivalent movements).

You can counteract this with extreme toe in (toe-ing in in front of the LP), but few people are likely to do that (it's quite effective imo, but aesthetically it bothers me quite a personally).

 

[MZKM]

The "BBC" presence dip is usually a bit higher, often between 2-3 or 2.5-4 kHz.

The actual BBC dip for the LS3/5A was at 3700Hz.
Also, so many people wrongly use the dip that was added to that speaker as proof that flat on-axis is not preferred. The dip was added as that speaker peaked heavily off-adis at 3700Hz:

Audyssey for instance states that their 3dB dip at 2kHz is inspired by the BBC dip. Besides the wrong frequency, the major issue is that Audyssey is taking a spatially-averaged in-room response, not an anechoic on-axis one.

 

[tuga]

The actual BBC dip for the LS3/5A was at 3700Hz.
Also, so many people wrongly use the dip that was added to that speaker as proof that flat on-axis is not preferred. The dip was added as that speaker peaked heavily off-adis at 3700Hz:

Can you infer that from the LS3/5a white paper?

This is what it says:

(c) Units

The low-frequency unit is a KEF type B110 unit specially selected to BBC specification; the nominal impedance is 8 ohms and the free-air resonance frequency is 35 Hz (with a statistical spread having 95% confidence limits at 33 and 40 Hz).
The high-frequency unit is a KEF tweeter type T27 with a nominal impedance of 8 ohms and a nominal resonance frequency of 1200 Hz (95% confidence limits of 1000 and 1450 Hz). As the diaphragm of this unit is exposed and could therefore be easily damaged in use, it has been protected by a domed perforated metal cover. This has a small effect on the frequency response of the T27, which is wholly beneficial as it raises the output somewhat at high frequencies. The radiating surface of the T27 is small and the radiator is therefore nearly omnidirectional; in order to prevent the acoustic discontinuity presented by the edge of the cabinet from setting up an interference pattern, the tweeter is surrounded by a thick felt strip mounted on the baffle front surface.

(d) Equaliser/crossover network

The circuit diagram of the network used for this purpose is shown in Fig. 2. The inductance L1 and the resistor R1 are employed to equalise the generally rising axial response/frequency characteristic of the bass unit; the group C5 , L 2 , R2 , compensates for a hump in this characteristic and the crossover frequency to the high-frequency unit is at about 3 kHz. Forthehigh-frequencyunit,inductorL3 servessimultaneouslyas a shunt inductor for the crossover network, and as an auto transformer to allow different relative sensitivities of individual l.f. and h.f. units to be matched. When used for this purpose, capacitor C2 is adjusted to keep the crossover frequency constant. This convenient form of network was first used in the design of the LS5/1 loudspeaker3 and has proved to be very useful. R3 serves as a damping resistor to prevent ringing, whilst R4 , and C6 serve to adjust the frequency response at the upper end of the band.
Physically the circuit board is mounted just behind the T27 unit and is prevented from resonating mechanically by means of a thick felt pad, placed between the board and the unit.

 

[Jon AA]

I think this highlights an interesting case where, once EQ comes in to the equation, the speaker actually ends up being better than quite a few speakers that measure better pre-eq. Does that make it a better speaker? I'm not sure how I feel on that. As someone who uses Dirac Live, the answer for me - personally - probably should be yes, but I still don't really like that answer for some reason.

I think the answer to that really has to be yes. From the research it's pretty clear the biggest contributions to sound quality are a flatish on axis response and smooth off axis curves. When EQ is in the equation, this would tend to mean the directivity curve is the most important measurement of a speaker. A speaker with poor directivity control that may have flatter on-axis response out of the box will be stuck with uneven off axis curves, while a speaker with an uneven frequency response out of the box but a smooth directivity curve can be corrected much closer to ideal than the first speaker will ever get.

There are limitations, of course. Decent driver selection and crossover design will increase the chances of a good result. For example, in the case of the RP-600M, I'd worry that EQing it flat would eat up a lot of headroom--possibly making distortion from the tweeter audible at a lower volume level as you try and squeeze more output from the bottom of its range.