GR Research X-LS Encore Kit Speaker Review

[ctrl]

If you really want to demonstrate that baffle edge diffraction is the primary cause of the response irregularity in the off-axis response, the thing you should do with your simulator is double the baffle width, leaving the edges sharp. If the dip-followed-by-peak is in fact caused primarily by baffle edge diffraction, as you certainly seem to think, it will absolutely, positively shift to half-lower frequency.

Yes sir, no problem sir! - you don't need to shout

Let's look at three different baffle widths - 0.22m (original Encore width), 0.44m and 0.15m.
For all three cases the complete loudspeaker is simulated, each with a crossover frequency of 1.8kHz (LR4).

Here are the respective crossovers at 1.8kHz



I show again the frequency responses normalized to on-axis from 0° to 90° in 15° steps. Please keep in mind that there are also other cabinet edges (top and bottom edge) that influence the on-axis frequency response.

First the original Encore (baffle 0.22m width) with a simplified tweeter (no wave-guide) and a roughly similar woofer:

Then the width of 0.44m required by @KaiserSoze

... and finally the ultra slim baffle with 0.15m

The edge diffraction wanders merrily back and forth - more precisely the first dip (which presents itself as an peak in the off-axis frequency responses in the normalized representation) and the following already significantly weakened ripples.

"The edge diffraction is always accompanied by a dip on axis - if no attempt is made to straighten it via the crossover. As a result, the off-axis FR shows usually no or only a small sound pressure increase in the affected frequency range."

Perhaps there is a language difficulty here, or perhaps this indicates an understanding that isn't as complete as might be desired. Theoretically, diffraction produces a ripple in the response,

I meant the interference with the highest amplitude impact on-axis (compared to off-axis). With a usual baffle width of 0.2-0.25m, this interference is the most important, as it's in the sensitive frequency range of 2-4kHz.

... and yes, my understanding is certainly not complete, but I'm aware of ripples and don't like Pringles:
Source (it's not porn ): http://hannover-hardcore.de/infinity_classics/!!!/Richtwirkung erzeugen.pdf

 

[Thomas_A]

Yes sir, no problem sir! - you don't need to shout

Let's look at three different baffle widths - 0.22m (original Encore width), 0.44m and 0.15m.
For all three cases the complete loudspeaker is simulated, each with a crossover frequency of 1.8kHz (LR4).
View attachment 75689 View attachment 75690 View attachment 75691

Here are the respective crossovers at 1.8kHz

View attachment 75692 View attachment 75693 View attachment 75694

I show again the frequency responses normalized to on-axis from 0° to 90° in 15° steps. Please keep in mind that there are also other cabinet edges (top and bottom edge) that influence the on-axis frequency response.

First the original Encore (baffle 0.22m width) with a simplified tweeter (no wave-guide) and a roughly similar woofer:
View attachment 75695

Then the width of 0.44m required by @KaiserSoze
View attachment 75697

... and finally the ultra slim baffle with 0.15m
View attachment 75698

The edge diffraction wanders merrily back and forth - more precisely the first dip (which presents itself as an peak in the off-axis frequency responses in the normalized representation) and the following already significantly weakened ripples.




I meant the interference with the highest amplitude impact on-axis (compared to off-axis). With a usual baffle width of 0.2-0.25m, this interference is the most important, as it's in the sensitive frequency range of 2-4kHz.

... and yes, my understanding is certainly not complete, but I'm aware of ripples and don't like Pringles:
Source (it's not porn ): http://hannover-hardcore.de/infinity_classics/!!!/Richtwirkung erzeugen.pdf
View attachment 75688

Nice simulation. It shows that the 44 cm wide baffle do not have the same type of peaking 2-4 kHz off-axis as the 22 cm baffle. The total width of my DIY speakers is 38 cm, but then is the shape of the cabinet quite different and "heavily" beveled.

 

[KaiserSoze]

Okay, I like it.

While you demonstrated that diffraction accounts for roughly half of the original peak above the crossover point, you also demonstrated that directivity mismatch accounts for roughly half of it. In your wide-baffle simulation the diffraction-associated peak shifted from its original location to half-lower frequency, leaving behind a still-appreciable plateau above the crossover point. The overall shape of the remaining plateau above the crossover point is strongly indicative of the tweeter's greater dispersion just above the crossover point vs. the woofer's weaker dispersion just below the crossover point. I.e., directivity mismatch. It looks as though each causative effect accounts for roughly half of the original peak above the crossover point.

It isn't usually smart to generalize, but I will anyway. I think you have demonstrated that when a speaker exhibits pronounced "tweeter flare", that this is very likely due as much to diffraction as it is due to directivity mismatch.

What you did also proposes a design approach for two-way speakers where there will inherently be a directivity mismatch between the woofer and tweeter. In this design approach the baffle is wide such that the 1st peak in the off-axis diffraction ripple occurs at frequency somewhat below the crossover point. The indicated formula for the baffle width is: W = 2 x V / (.8 x F), where W is the baffle width, V is the velocity of sound, and F is the crossover frequency. The numerical constant .8 is a guesstimate. Certainly there are more practical ways to deal with the directivity mismatch, but probably not diffraction, notwithstanding that an awful lot of builders think that a small roundoff radius such as 3/8" will be effective at frequencies lower than upper treble. In my mind I picture a tower speaker about 30 cm wide using a column of small woofers. The implied crossover frequency will be about 2.8 kHz. Maybe six 4" woofers ... A shallow waveguide would still be needed ... hmmm...

I haven't yet looked at the last link you included, for a PDF that hopefully isn't porn, but one comment about the last graph you included, illustrating the combined baffle step and diffraction ripple (for what I expect is the circular baffle above), is that the magnitude of the peaks (both positive and negative) should probably decrease with increasing frequency, probably following an exponential decay. The reason is simply that the driver (tweeter) dispersion decreases as frequency increases, i.e., SPL at the edge of the baffle decreases as frequency increases, and as such the magnitude of the ripple should correspondingly decrease.

 

[Kustomize]

Would be so interested in hearing how the GR Research XLS stack up against the Zaph ZA5.2!

 

[Rick Sykora]

Would be so interested in hearing how the GR Research XLS stack up against the Zaph ZA5.2!

Comparably, by measurement at least. But Amir felt the Zaph 5.2 lacked bass. See here:

https://www.audiosciencereview.com/...a5-2-diy-kit-speaker-review.12086/post-352041

From my limited time with the X-LS, it is louder and has more bass. I like both for different reasons.

P.S. Both designs are pretty old at this point too. Would consider newer offerings by Bagby and others at Meniscus Audio

 

[Jmitch3]

A roundoff radius of 3/8 inch is barely going to have any effect at all at frequencies below 4 kHz.

Exactly what problem is this supposed to solve? Even if this amount of roundoff where particularly effective at the crossover frequency (it wouldn't be), do you think that the dip in the off-axis frequency response at about 1.6 kHz and the broad plateau about an octave higher is caused by diffraction? If this is what you believe, could you please explain why?

I’m not a speaker designer so I can’t say.

However the actual designer of the speaker calls for a 3/8 roundover on the left and right edges of the front baffle in the speaker design. So. This is effectively a “modded” box not quite to designer spec.

 

[amirm]

However the actual designer of the speaker calls for a 3/8 roundover on the left and right edges of the front baffle in the speaker design. So. This is effectively a “modded” box not quite to designer spec.

We could make this change to the other unit. Along these lines, anyone want to buy this pair? I have no use for it and it is tying up fair amount of cash.

 

[Jmitch3]

We could make this change to the other unit. Along these lines, anyone want to buy this pair? I have no use for it and it is tying up fair amount of cash.

that would be interesting to try given a previous posters analysis and simulation.

Here we also see some areas of quality control when evaluating diy speakers. Without a complete analysis comparing the box construction, bracing, damping, etc, to the original specs it’s hard to know if what we are evaluating is THE speaker.

a YouTuber who does some reviews has actually taken speakers apart and done such an analysis on retail speakers to see the bones of the unit. I find that quite interesting.

 

[amirm]

Here we also see some areas of quality control when evaluating diy speakers. Without a complete analysis comparing the box construction, bracing, damping, etc, to the original specs it’s hard to know if what we are evaluating is THE speaker.

The responsibility for this falls squarely on the shoulders of GR Research. Danny does a lot of youtube reviews but until we tried to build this kit, he had not bothered to make one for this speaker. Nor provide detailed instructions on his website.

 

[xarkkon]

The responsibility for this falls squarely on the shoulders of GR Research. Danny does a lot of youtube reviews but until we tried to build this kit, he had not bothered to make one for this speaker. Nor provide detailed instructions on his website.

Yeah, think he's finally putting up build videos for the xls encore, but even then at a rather slow pace

 

[Deleted member 2944]

I’m not a speaker designer so I can’t say.

However the actual designer of the speaker calls for a 3/8 roundover on the left and right edges of the front baffle in the speaker design. So. This is effectively a “modded” box not quite to designer spec.

The official cabinet plans call for no such round over, nor is any requirement mentioned on the system description page.
http://gr-research.com/pdf/X-LS Encore box.pdf
http://gr-research.com/x-lsencorekit.aspx

I think the build here by Rick is consistent with what a typical novice builder would construct. A typical builder probably would not have a router or 3/8" radius bit to create it.
Calling this a "modded" speaker is stretching.

Dave.

 

[Jmitch3]

The official cabinet plans call for no such round over, nor is any requirement mentioned on the system description page.
http://gr-research.com/pdf/X-LS Encore box.pdf
http://gr-research.com/x-lsencorekit.aspx

I think the build here by Rick is consistent with what a typical novice builder would construct. A typical builder probably would not have a router or 3/8" radius bit to create it.
Calling this a "modded" speaker is stretching.

Dave.

welcome to diy. Check out the last 10+ years of people posting their builds on the forum. Let me know what percentage you see that don’t have an front baffle edge roundovers. https://www.audiocircle.com/index.php?topic=150120.0

 

[Deleted member 2944]

welcome to diy. Check out the last 10+ years of people posting their builds on the forum. Let me know what percentage you see that don’t have an front baffle edge roundovers. https://www.audiocircle.com/index.php?topic=150120.0

Thank you very much. I've been DIYing speakers for the better part of forty years now.

You've missed the point completely. Whether or not creative DIY'ers choose to chamfer/round/whatever the front baffle on speaker projects is fine and dandy. In this particular instance though, it's NOT called out in the plans. So, people that don't frequent the AudioCircleJerk forum would never know about such a "requirement."

Dave.

 

[ctrl]

While you demonstrated that diffraction accounts for roughly half of the original peak above the crossover point, you also demonstrated that directivity mismatch accounts for roughly half of it.

The overall shape of the remaining plateau above the crossover point is strongly indicative of the tweeter's greater dispersion just above the crossover point vs. the woofer's weaker dispersion just below the crossover point.

You are right, for a loudspeaker using a tweeter without any "radiation control" the above would apply. But keep in mind that the Encore tweeter is using a small wave-guide. Nevertheless, since I cannot simulate the encore exactly due to missing data, I have no problem to agree with your analysis.

I agree with your theoretical arguments and considerations too, only in the rating of the effects on the real loudspeaker we have different opinions.
Even if there is, after a baffle optimization, a small remaining plateau above the crossover point it could be dealt by the crossover (by changing the on-axis FR).

To close the circle, here is the optimized loudspeaker with a crossover at 1.8kHz LR4 (like the Encore) to show that good radiation is possible even with simple DIY options. With this "hypothetical" loudspeaker there is in no case a directivity mismatch





To be noted:
Since I'm not paid by GR-Research to optimize their loudspeakers, I used (because I'm lazy) drivers I had already used for simulations to show the basic principle of a possible optimization. The woofer used in the Encore has a slightly larger cone area than the SBAcoustics woofer in the simulation (137/120cm²), but with a crossover of 1.8kHz this makes little difference.
Whether the small wave-guide of the Peerless T26SG is as effective as that of the Seas DXT I can't judge, but both tweeters are about the same size (possible wave-guide area).

 

[KaiserSoze]

I’m not a speaker designer so I can’t say.

However the actual designer of the speaker calls for a 3/8 roundover on the left and right edges of the front baffle in the speaker design. So. This is effectively a “modded” box not quite to designer spec.

The first simulations that ctrl generously provided confirmed that even in computer simulation a 3/8" roundoff has barely any effect except at very high frequency (where the simulation may be exaggerating the effect). Siegfried Linkwitz made a detailed study of baffle edge diffraction using a tweeter and a few different shapes and sizes of baffles. An excerpt:

"Real cabinet edges are usually at 90 degrees or rounded, which reduces the magnitude of the diffracted wave, since the wave does not immediately expand into 4pspace upon reaching the edge. The benefits of edge rounding come into play only when the radius is greater than 1/8th wavelength. Thus a typical 1/2 inch radius begins to diffuse the diffracted wave at frequencies above 3.4 kHz, but will decrease in relevance at higher frequencies, when the driver illuminates less of the edge due to its increasing directivity. With most speaker cabinets the radius or chamfer is acoustically too small and is primarily cosmetic."

Several times I have tried to think of a way to rephrase his rule that the radius must be greater than 1/8th wavelength. As with many things, an effect of concern comes into play gradually, without any well-defined threshold. For a given roundoff radius R, the diffraction-reducing effect is a little bit effective at wavelength 8 x R. In order to be more than a little bit effective at this wavelength, the radius needs to be greater than R. Another way to say this is to say that for roundoff radius R, the diffraction-reducing effect is more than a little bit effective at wavelengths shorter than 8 x R, i.e., higher in frequency. For a 3/8" roundoff, the value of 8 x R is 3", which corresponds to 4.5 kHz. Below this frequency the roundoff has no appreciable effect, so it cannot intervene and prevent the diffraction effect from exacerbating the effect of the directivity mismatch in the vicinity of the crossover. Furthermore, as Linkwitz points out, at the higher frequencies where a 3/8" roundoff would be effective at mitigating diffraction, diffraction is not a concern because at these higher frequencies the tweeter is so highly directional that that the edge of the baffle is only weakly "illuminated".

There are a few simple strategies that are useful at mitigating baffle edge diffraction, but corner rounding is not one of them except for when it is done in the extreme. For it to be effective you would pretty much have to round the corners to the point where the enclosure is a vertical circular cylinder. The strategies that are effective are ubiquitous. The most effective strategy is the use of a baffle that is rectangular rather than circular (although if it is square it will be only slightly better than circular). The rectangular baffle works because the distance from the driver to the edge is not constant. It is meaningful to note that rounding the corner will vary the distance by maybe an inch whereas the variation in distance produced by a typical rectangular baffle is substantially greater than an inch. The same fundamental reason, variation in the distance to the edge, is also the reason that it is effective to locate the tweeter off-center and with the distance to the baffle's top edge not equal to either of the two lateral distances.

Another excerpt:

"It has been my experience that the on-axis and off-axis frequency response of a given driver and baffle combination must be measured to assess overall uniformity of response, and that computer models give pessimistic predictions, because the underlying assumptions do not sufficiently describe the real acoustic behavior of drivers and of cabinet shapes. ... While the "baffle step" cannot be avoided, the additional higher frequency ripples can be easily reduced to a magnitude that is much smaller than the first arriving direct sound, by simply optimizing the proportions of a rectangular baffle. The absolute width of a cabinet is not the critical parameter that many people think it is. Much is hypothesized, little is proven and much is overrated when it comes to diffraction."

http://linkwitzlab.com/diffraction.htm

The last simulations that ctrl did (in response to my suggestion that doubling the baffle width in the simulation should show us the extent to which baffle edge diffraction was contributing to the irregularity in the off-axis response in the vicinity of the crossover frequency) indicated that the rise in response above the crossover point (in the off-axis response) is due partly to directivity mismatch and partly to diffraction. The 1st peak in the diffraction ripple 90-degrees off-axis lands at frequency just above the crossover frequency, exacerbating the directivity mismatch. His simulations suggest that these two distinct causative effects each contribute roughly half of the response rise observed just above the crossover point, in the off-axis response. By doubling the baffle width the frequency for this 1st peak in the diffraction ripple shifted to half-lower frequency. About half of the response rise just above the crossover point remained, and it morphed in shape to a shape consistent with a rise associated with the tweeter's dispersion behavior. Doubling the baffle width had no significant effect at suppressing the dip just below the crossover point, which indicates that this dip is not attributable to any significant extent to diffraction ripple. It is attributable to the strong directivity of the woofer at frequency region where the wavelength is not any greater than the piston diameter.

 

[KaiserSoze]

Thank you very much. I've been DIYing speakers for the better part of forty years now.

You've missed the point completely. Whether or not creative DIY'ers choose to chamfer/round/whatever the front baffle on speaker projects is fine and dandy. In this particular instance though, it's NOT called out in the plans. So, people that don't frequent the AudioCircleJerk forum would never know about such a "requirement."

Dave.

Oh, man. You know why that made me laugh, the "AudioCircleJerk". Nearly everyone participating in the JerkFest has a strong opinion as to what steps must absolutely be taken in order for the finished product to be all that it can be.

As for rounding the baffle edge, the point you make is perfectly valid. But to my way of thinking it is more important for people to understand that rounding the corners is not going to have any effect whatsoever on that off-axis response undulation seen in the vicinity of the crossover. The simulation that ctrl did strongly suggests that diffraction is making the rise seen just above the crossover point about twice as bad as it would be without the diffraction effect. But even so, this does not mean that rounding the baffle edge is going to make it go away. It won't. To have the necessary effect, the rounding would have to be done to such a strong extent that the cabinet would be turned into a vertical circular cylinder.

 

[KaiserSoze]

You are right, for a loudspeaker using a tweeter without any "radiation control" the above would apply. But keep in mind that the Encore tweeter is using a small wave-guide. Nevertheless, since I cannot simulate the encore exactly due to missing data, I have no problem to agree with your analysis.

I agree with your theoretical arguments and considerations too, only in the rating of the effects on the real loudspeaker we have different opinions.
Even if there is, after a baffle optimization, a small remaining plateau above the crossover point it could be dealt by the crossover (by changing the on-axis FR).

To close the circle, here is the optimized loudspeaker with a crossover at 1.8kHz LR4 (like the Encore) to show that good radiation is possible even with simple DIY options. With this "hypothetical" loudspeaker there is in no case a directivity mismatch

View attachment 75786 View attachment 75787

View attachment 75788 View attachment 75789

To be noted:
Since I'm not paid by GR-Research to optimize their loudspeakers, I used (because I'm lazy) drivers I had already used for simulations to show the basic principle of a possible optimization. The woofer used in the Encore has a slightly larger cone area than the SBAcoustics woofer in the simulation (137/120cm²), but with a crossover of 1.8kHz this makes little difference.
Whether the small wave-guide of the Peerless T26SG is as effective as that of the Seas DXT I can't judge, but both tweeters are about the same size (possible wave-guide area).

For my part I found the exercise stimulating and enlightening. To my way of thinking the simulation you did that was most illuminating was the one where you doubled the baffle width (at my suggestion ). This was the simulation that revealed the relative contributions of each of the two distinct causes of the response rise above the crossover point (in the off-axis response).

I don't know whether you've ever read the study that Linkwitz did on baffle edge diffraction. You may find it worth a quick read. You can find the link in one of my posts in this thread, about two posts prior to this one.

When you wrote: "With this "hypothetical" loudspeaker there is in no case a directivity mismatch.", would you agree that it is a bit more accurate or precise to write something like, "With this hypothetical loudspeaker the inherent mismatch in the directivity of the two drivers at the crossover frequency is fully resolved by the edge rounding and by especially the waveguide, such that the off-axis response is very nearly as smooth as the on-axis response."? Maybe I'm nitpicking, but to my way of thinking, the way you word it muddles the distinction between a directivity mismatch that has been laid to rest vs. a directivity mismatch that never lived in the first place. I am admittedly much pickier about this kind of thing than just about everyone else.

It would be interesting to me to be able to directly compare the outcome for three variations of your hypothetical speaker, all three with the waveguide, but different in respect to the corner treatment. As you have it here, vs. no corner rounding at all, vs. consistent 10 mm radius rounding from top to bottom for both side corners.

 

[Thomas_A]

For my part I found the exercise stimulating and enlightening. To my way of thinking the simulation you did that was most illuminating was the one where you doubled the baffle width (at my suggestion ). This was the simulation that revealed the relative contributions of each of the two distinct causes of the response rise above the crossover point (in the off-axis response).

I don't know whether you've ever read the study that Linkwitz did on baffle edge diffraction. You may find it worth a quick read. You can find the link in one of my posts in this thread, about two posts prior to this one.

When you wrote: "With this "hypothetical" loudspeaker there is in no case a directivity mismatch.", would you agree that it is a bit more accurate or precise to write something like, "With this hypothetical loudspeaker the inherent mismatch in the directivity of the two drivers at the crossover frequency is fully resolved by the edge rounding and by especially the waveguide, such that the off-axis response is very nearly as smooth as the on-axis response."? Maybe I'm nitpicking, but to my way of thinking, the way you word it muddles the distinction between a directivity mismatch that has been laid to rest vs. a directivity mismatch that never lived in the first place. I am admittedly much pickier about this kind of thing than just about everyone else.

It would be interesting to me to be able to directly compare the outcome for three variations of your hypothetical speaker, all three with the waveguide, but different in respect to the corner treatment. As you have it here, vs. no corner rounding at all, vs. consistent 10 mm radius rounding from top to bottom for both side corners.

As additional simulation of the 44 cm baffle, I would be happy to see 45 degree bevel left and right , ”all the way possible” + a smaller beveling at the top. No waveguide. This may be off- topic and belongs to the DIY forum but since it is a DIY kit that comes without cabinet it could be useful for any person buying the kit and want to experiment with other cabinet types that are fairly easy to execute.

 

[maty]

Audiocircle Measurements and review of the X-LS Encore kit by www.audiosciencereview.com

https://www.audiocircle.com/index.php?topic=171596.msg1815901#msg1815901

by Hobbsmeerkat

Reading the comments was a trip. I've never actually heard of Danny being called "divisive" or "polarizing" of course i don't have any experience with trying out different caps/connectors. But I have heard the XLS with & without Norez & it definitely makes a difference in the mid bass.

Without having listened to them but having improved a few loudspeakers I have no doubts about it. The last case was my modded KEF Q100 (with closed front bass-reflex), the sound change (improvement) was substantial. Now they weigh more than the KEF LS50.

 

[Kustomize]

Comparably, by measurement at least. But Amir felt the Zaph 5.2 lacked bass. See here:

https://www.audiosciencereview.com/...a5-2-diy-kit-speaker-review.12086/post-352041

From my limited time with the X-LS, it is louder and has more bass. I like both for different reasons.

P.S. Both designs are pretty old at this point too. Would consider newer offerings by Bagby and others at Meniscus Audio

Thankyou!

I ask because I’ve a ZA5.3T which is the 2.5 way using 2 woofers. I don’t think it lacks bass, plus I use subs anyway. I always wondered if going to the X-MTM Encore would be an upgrade or not! I guess your reply tells me probably not an upgrade!