Celestion F12-X200

[MicFarlow]

Looks like they shipped on the slowest boat

i'm not sure how you got this pic of me but it sums up my feelings well! Been waiting on this driver for years and it has taken F.O.R.E.V.E.R to complete. I know Celestion is just making sure it's right but it has taken a while!

 

[Promit]

Celestion told me today that "we have begun shipping the speaker from our factory to our worldwide warehouses." I don't think that's markedly different from two weeks ago, but we are getting there!

RE beaminess: The Emi Beta setup that Mic uses in the XiTone is the same situation of using the woofer as a waveguide and honestly it's fine. Is it ideal compared to something like the Atomic CLR's purpose designed horn, probably not. But the dispersion pattern is so much more symmetrical (and merges so much sooner) than your typical PA box that there's no sense in being picky.

The really interesting question is what happens when you simply drop one of these new integrated Celestions into a guitar cab, which is what I think they had in mind for the design right from the get-go. I'm curious to see what the driver Qts is.

 

[chrisjnyc]

Celestion told me today that "we have begun shipping the speaker from our factory to our worldwide warehouses." I don't think that's markedly different from two weeks ago, but we are getting there!

RE beaminess: The Emi Beta setup that Mic uses in the XiTone is the same situation of using the woofer as a waveguide and honestly it's fine. Is it ideal compared to something like the Atomic CLR's purpose designed horn, probably not. But the dispersion pattern is so much more symmetrical (and merges so much sooner) than your typical PA box that there's no sense in being picky.

The really interesting question is what happens when you simply drop one of these new integrated Celestions into a guitar cab, which is what I think they had in mind for the design right from the get-go. I'm curious to see what the driver Qts is.

So what shipped out 2 weeks ago? ...and where are those speakers headed??

 

[Tom Morris]

Sooooooooooon........

I'm just going to change my name too "soon".

 

[yeky83]

RE beaminess: The Emi Beta setup that Mic uses in the XiTone is the same situation of using the woofer as a waveguide and honestly it's fine. Is it ideal compared to something like the Atomic CLR's purpose designed horn, probably not. But the dispersion pattern is so much more symmetrical (and merges so much sooner) than your typical PA box that there's no sense in being picky.

It's debatable I think. Yes, directivity will be symmetrical, but doesn't necessarily mean good or better than a typical PA.

Celestion FTX1225 is a good example cus there's data available. And it shows the typical behavior for this type of a coaxial speaker.

Jagged HF response:

Irregular off-axis response:

Trying to use EQ/DSP to correct for on-axis response would cause issues for off-axis response, and vice versa.​

Compare that to a typical non-coaxial PA. I found this data on the DXR8.

On-axis and horizontal off-axis is great:

Issue is the vertical off-axis, big cancellation due to lobing in crossover region:

Pick your poison haha

 

[Rex]

It's debatable I think. Yes, directivity will be symmetrical, but doesn't necessarily mean good or better than a typical PA.

Celestion FTX1225 is a good example cus there's data available. And it shows the typical behavior for this type of a coaxial speaker.

Jagged HF response:

Irregular off-axis response:

Trying to use EQ/DSP to correct for on-axis response would cause issues for off-axis response, and vice versa.​

Compare that to a typical non-coaxial PA. I found this data on the DXR8.

On-axis and horizontal off-axis is great:

Issue is the vertical off-axis, big cancellation due to lobing in crossover region:

Pick your poison haha

Those responses aren't all that different from each other. Example: at 3 KHz, the response for both systems is 6 dB down at 45°.

 

[Promit]

Compare that to a typical non-coaxial PA. I found this data on the DXR8.

Don't get me wrong, I agree that the coaxial has some real problems. But they have smoothed the bajeesus out of those Yamaha curves, and in the real world the vertical off-axis is terrible. This is by design and makes perfect sense for a mounted vertical PA. It's actually desirable. But set it up as a horizontal wedge and ugh.

My XiTone wedge is very noticeably non-flat at any angle and the high frequency behavior is, quite frankly, a complete mess. This is not a knock against Mic or his product, but it is ultimately an assembly of off the shelf components and suffers some problems because of this HF-through-woofer arrangement. It works anyway because in a wedge format it behaves much more consistently off axis than any PA box, and because it doesn't need the 6+ feet that a regular PA wants for the high and low images to merge properly. The response, while messy, isn't obviously colored and that's good enough for instrumental use. It's a terrible device for music playback, but that's not the point and so it's quite alright.

The Atomic CLR, on the other hand, is a very deliberate and focused bespoke speaker for this task. The designer of the CLR has been a very staunch and vocal defender of his design over on TGP, and critical of many other things. I understand why, as that's a purpose built design intended to deal with these problems. I also refuse to purchase the product due to said designer, but that's another problem entirely.

 

[yeky83]

Those responses aren't all that different from each other. Example: at 3 KHz, the response for both systems is 6 dB down at 45°.

Disagreed. I don't see how the two responses are similar. Maybe I'm just not getting it, but at 45 degrees & 3.5 kHz (just 0.5 kHz higher than the 3 kHz example you picked), one is 0 dB in response while the other is -6 dB. So I don't see the point you're trying to make.

Don't get me wrong, I agree that the coaxial has some real problems. But they have smoothed the bajeesus out of those Yamaha curves, and in the real world the vertical off-axis is terrible. This is by design and makes perfect sense for a mounted vertical PA. It's actually desirable. But set it up as a horizontal wedge and ugh.

No disagreement there, I wasn't trying to say that the vertical off-axis of the DXR is good. I did say "pick your poison"

My XiTone wedge is very noticeably non-flat at any angle and the high frequency behavior is, quite frankly, a complete mess. This is not a knock against Mic or his product, but it is ultimately an assembly of off the shelf components and suffers some problems because of this HF-through-woofer arrangement. It works anyway because in a wedge format it behaves much more consistently off axis than any PA box, and because it doesn't need the 6+ feet that a regular PA wants for the high and low images to merge properly. The response, while messy, isn't obviously colored and that's good enough for instrumental use. It's a terrible device for music playback, but that's not the point and so it's quite alright.

Got it. And I wasn't trying to knock against Mic or his product either, or other coaxial products. Just saying it is what it is. Glad to hear another case of it working well for someone. And good point on the PA box speaker's far-field condition being further away.

For wedge use, perhaps the messy high frequency behavior isn't as big a factor as you say. There would be significant floor reflections anyway.
One might still make the case that if you stay within ~30 degree window or so with a PA wedge, it'd behave better than a HF-through-woofer coaxial. But that's a small off-axis window to work with for wedge monitoring IMHO

The Atomic CLR, on the other hand, is a very deliberate and focused bespoke speaker for this task. The designer of the CLR has been a very staunch and vocal defender of his design over on TGP, and critical of many other things. I understand why, as that's a purpose built design intended to deal with these problems. I also refuse to purchase the product due to said designer, but that's another problem entirely.

Interesting, I rather like him haha. But I understand how you'd feel that way.

 

[Rex]

Disagreed. I don't see how the two responses are similar. Maybe I'm just not getting it, but at 45 degrees & 3.5 kHz (just 0.5 kHz higher than the 3 kHz example you picked), one is 0 dB in response while the other is -6 dB. So I don't see the point you're trying to make.

I'm reading the graphs differently.

Look at the HF Contour plot for the coaxial speaker. At 3.5 KHz, the border between the dark red area and the light red area is very close to 45°. That border is the -3 dB point, so off-axis response there is -3 dB.

Now look at the horizontal directivity (Directivé horizontale) plot for the non-coaxial speaker. The 45° line hits -6 dB at 3.5 KHz, so the off-axis response is -6 dB — twice as much fall-off as the coaxial speaker. That's the opposite of "horizontal off-axis is great," but not entire unsimilar to the coaxial plot.


Also, as @Promit pointed out, the non-coaxial plots have been smoothed. Sometimes there are valid engineering or statistical reasons for doing that, but usually it's a marketing decision. Smoothed plots don't show the irregularities.

 

[yeky83]

I'm reading the graphs differently.

Look at the HF Contour plot for the coaxial speaker. At 3.5 KHz, the border between the dark red area and the light red area is very close to 45°. That border is the -3 dB point, so off-axis response there is -3 dB.

Ok. How about 6 kHz? Or 9 kHz and beyond?

Now look at the horizontal directivity (Directivé horizontale) plot for the non-coaxial speaker. The 45° line hits -6 dB at 3.5 KHz, so the off-axis response is -6 dB — twice as much fall-off as the coaxial speaker. That's the opposite of "horizontal off-axis is great," but not entire unsimilar to the coaxial plot.

No, it's not the opposite of "horizontal off-axis is great." Your point seems to be "it's louder at 3.5 kHz so it's better," seems a bad/incomplete way of evaluating off-axis response.

Good off-axis response is about having a controlled roll-off of a wide bandwidth, which in this case the non-coaxial does and the coaxial does not.

Also, as @Promit pointed out, the non-coaxial plots have been smoothed. Sometimes there are valid engineering or statistical reasons for doing that, but usually it's a marketing decision. Smoothed plots don't show the irregularities.

@Promit also points out that the HF behavior is indeed a complete mess

Yes, it's been smoothed (so are the Celestion graphs to a lesser degree), but not to the degree that it goes against my points. Were the irregularities present in the DXR that were present in the Celestion (phase cancellation on-axis at 6k & 10k, jagged off-axis response), it seems to me it would still show through the smoothing (just as it shows the lobing dip at crossover).

If you want a more apples to apples comparison, here it is all from the Celestion catalog. Should be same smoothing, same measurement method, samesies. These measurements are typical of these two design types, and I think it's pretty clear that they don't behave similarly.

HF-through-woofer coaxial on-axis (red):

Horn loaded compression driver on-axis:

HF-through-woofer coaxial off-axis:

Horn loaded compression driver off-axis:

 

[Rex]

Ok. How about 6 kHz? Or 9 kHz and beyond?

At 6 KHz and 45° off axis, the coaxial speaker shows less than 3 dB deviation from on-axis response; the non-coaxial design shows about 6 dB deviation.

At 9 KHz, the coaxial graph is about 6 dB below the on-axis response; the non-coaxial graph shows about 7 dB fall-off. This particular coaxial speaker loses its beam-width superiority only when you get well into the last octave of the audio spectrum.

No, it's not the opposite of "horizontal off-axis is great." Your point seems to be "it's louder at 3.5 kHz so it's better," seems a bad/incomplete way of evaluating off-axis response.

In this case, "better" means "closer to on-axis response." On-axis response is 0 dB. At 45°, this coaxial speaker's response is within 3 dB of that reference across most of the spectrum.

Good off-axis response is about having a controlled roll-off of a wide bandwidth, which in this case the non-coaxial does and the coaxial does not.

Good off-axis response is mainly about keeping off-axis response close to on-axis response. Once you move beyond 45°, the coaxial design attenuates the signal a lot. There's no way to tell what the non-coaxial response does, because the graph gives no data for those angles.

Were the irregularities present in the DXR that were present in the Celestion (phase cancellation on-axis at 6k & 10k, jagged off-axis response), it seems to me it would still show through the smoothing (just as it shows the lobing dip at crossover).

It's risky to assume what irregularities would show through the smoothing (which, by definition, minimizes irregularities). Without knowing the data behind the curves and how much smoothing was applied, there's no way to predict what it would have looked like.

If you want a more apples to apples comparison, here it is all from the Celestion catalog. Should be same smoothing, same measurement method, samesies. These measurements are typical of these two design types, and I think it's pretty clear that they don't behave similarly.

HF-through-woofer coaxial on-axis (red):

Horn loaded compression driver on-axis:

HF-through-woofer coaxial off-axis:

Horn loaded compression driver off-axis:

That's apples-to-oranges. The curves compare a two-driver coaxial system against a single horn-loaded driver (note that the compression driver's response becomes unusable below 1KHz because there's no woofer to complete the spectrum).

 

[yeky83]

In this case, "better" means "closer to on-axis response." On-axis response is 0 dB. At 45°, this coaxial speaker's response is within 3 dB of that reference across most of the spectrum.

Good off-axis response is mainly about keeping off-axis response close to on-axis response.

I agree with your wording, but disagree with your interpretation.

If good off-axis response is about "keeping off-axis response close to on-axis response," and the desired on-axis response is flat, then a good off-axis response should be flat-ish as well.

Having narrow bands of cuts & boosts throughout the frequency range (at 45 degrees: -7dB @2k; -3dB @3k; -6dB @4k; -1dB @5k; -6dB @ 8k; -9dB @ 9k; -12 dB @10k) does not make for "keeping off-axis response close to on-axis response." If you prefer the narrow bands of cuts & boosts as being "close to on-axis response," we simply disagree and we'll have to move on I think.

There's no way to tell what the non-coaxial response does, because the graph gives no data for those angles.

The new graph I provided does, and I'd say it's typical of well designed horn loaded compression drivers.

It's risky to assume what irregularities would show through the smoothing (which, by definition, minimizes irregularities). Without knowing the data behind the curves and how much smoothing was applied, there's no way to predict what it would have looked like.

Agreed. That's why I gave new graphs as examples.

That's apples-to-oranges. The curves compare a two-driver coaxial system against a single horn-loaded driver (note that the compression driver's response becomes unusable below 1KHz because there's no woofer to complete the spectrum).

Save for the previously discussed vertical off-axis lobing issue, the single horn-loaded driver behavior will largely be the HF behavior of a two-driver non-coaxial system that it is installed in, because a crossover is used (provided the crossover design is good). If you evaluate the graphs with this in mind, you can do an apples-to-apples comparison of the high frequency response. One is clearly jagged, the other is not, both on and off axis.

 

[Rex]

If good off-axis response is about "keeping off-axis response close to on-axis response," and the desired on-axis response is flat, then a good off-axis response should be flat-ish as well.

I see your point, but I don't think you're looking at the whole picture. Yes, "good off-axis response" means keeping things as flat as possible when you're off axis. But that's only part of it. It also means keeping the signal level up when you're off-axis.

Let's look at the 45° point, which is about the best spread that you can get with current designs. The coaxial graph you published shows the response to be 3 dB down over half of the spectrum (and rarely more than 6 dB down until you approach the last octave). At the same 45° listening angle, the compression driver is so severely attenuated that other sources of sound (reflections, other speakers, etc.) start to dominate, and the frequency response of the driver becomes irrelevant. That's why I say the first driver has better off-axis response: it puts out a usable signal farther off axis.

Save for the previously discussed vertical off-axis lobing issue, the single horn-loaded driver behavior will largely be the HF behavior of a two-driver non-coaxial system that it is installed in, because a crossover is used (provided the crossover design is good). If you evaluate the graphs with this in mind, you can do an apples-to-apples comparison of the high frequency response. One is clearly jagged, the other is not, both on and off axis.

Yes and no. In the critical midrange, the drivers will interfere with each other, no matter how good the crossover is. You can't say "this apple will be similar to this orange" without doing the same measurements on equivalent configurations. That's like evaluating one ingredient against an entire soup.


This is fun! Despite disagreeing strongly with me, you've kept your comments respectful and factual. I can deal with dissent like that all day.

 

[yeky83]

I see your point, but I don't think you're looking at the whole picture. Yes, "good off-axis response" means keeping things as flat as possible when you're off axis. But that's only part of it. It also means keeping the signal level up when you're off-axis.

It only means "keeping the signal level up when you're off-axis" if that is what you desire. A "good off-axis response" does not equal a "wide off-axis response." In fact, a "narrow off-axis response" might be what one desires, and that would be "good." Depends on the application. So to characterize "wide" as "good" seems the wrong criteria to me.

Further, you can have wide off-axis response without resorting to HF-through-woofer speaker and its jagged response, so even if I were to accept "wide" as "good," I find the point moot.

Additionally, you have not commented so far on the jagged on-axis response of a HF-through-woofer design. It's arguable that this is not a worthwhile trade-off to begin with, so perhaps you could address it.

Let's look at the 45° point, which is about the best spread that you can get with current designs. The coaxial graph you published shows the response to be 3 dB down over half of the spectrum (and rarely more than 6 dB down until you approach the last octave). At the same 45° listening angle, the compression driver is so severely attenuated that other sources of sound (reflections, other speakers, etc.) start to dominate, and the frequency response of the driver becomes irrelevant. That's why I say the first driver has better off-axis response: it puts out a usable signal farther off axis.

The examples I provided were chosen to display the typical off-axis response behavior (jagged or smooth) of different designs. They do not display the comparative max-SPL capability of each, nor the resulting max-off-axis-SPL capability of each. (The "3 dB down" of the coaxial might be quieter than the "-6~-9 dB down" of the non-coaxial, the dB doesn't carry over graphs.) That is an unknown. And it should really be considered on a "product-by-product" basis, not "design type-by-design type" basis, there's many other factors on what makes for a louder 45 degree off-axis product to say otherwise.

If you have data that the HF-through-woofer design type generally has a higher 45 degree off-axis SPL than a non-coaxial design type, I'd be interested to see it. But the examples I've provided do not provide a basis for that discussion.
(BTW, HF-through-woofer type I've seen have less output capability in general than separate high output driver elements.)

Yes and no. In the critical midrange, the drivers will interfere with each other, no matter how good the crossover is. You can't say "this apple will be similar to this orange" without doing the same measurements on equivalent configurations. That's like evaluating one ingredient against an entire soup.

This is true with the coaxial as well, yet you haven't asked for its combined response, and I have not provided one.
You're comparing apples-to-apples with the examples I've provided.

And again, I've already remarked how the non-coaxial drivers will interfere with each other in regards to vertical off-axis response. I've mentioned it several times, but perhaps it's gone unacknowledged because you aren't aware? Lobing? The effect of lobing is very predictable, if one knows what it is, one should easily be able to imagine its effect even with just the graphs I've provided.

This is fun! Despite disagreeing strongly with me, you've kept your comments respectful and factual. I can deal with dissent like that all day.

I try to be respectful to everyone on the forum! Especially you, since there's a lot of great insight from you on this forum, I'm a fan Thanks for taking it for what it is, and not reading any offense or whatnot into it as I dissent

 

[Rex]

It only means "keeping the signal level up when you're off-axis" if that is what you desire. A "good off-axis response" does not equal a "wide off-axis response." In fact, a "narrow off-axis response" might be what one desires, and that would be "good." Depends on the application. So to characterize "wide" as "good" seems the wrong criteria to me.

Interesting point. Our definitions of "good" may indeed be quite different. When I hear "good off-axis response," I think of off-axis response similar to on-axis response. If you look at what you originally posted (which was the basis of my disagreement), you presented an example of a non-coaxial design as having better off-axis response than a coaxial design. But even at a narrower 22.5°, the coaxial plot shows response within 3 dB of on-axis, across most of the spectrum; whereas the non-coaxial plot shows a variation from 2 dB to 6 dB across the same spectrum — that's more jagged and uneven, not less.

Further, you can have wide off-axis response without resorting to HF-through-woofer speaker...

Agreed.

Additionally, you have not commented so far on the jagged on-axis response of a HF-through-woofer design. It's arguable that this is not a worthwhile trade-off to begin with, so perhaps you could address it.

Show me the same kinds of plots, with the same axes and the same smoothing applied, for both speakers. Then we'll have something to talk about. Sources would help, too.

The examples I provided were chosen to display the typical off-axis response behavior (jagged or smooth) of different designs. They do not display the comparative max-SPL capability of each, nor the resulting max-off-axis-SPL capability of each.

It's not about max SPL. I never presented it as such.

And it should really be considered on a "product-by-product" basis, not "design type-by-design type" basis...

Yes.

This is true with the coaxial as well, yet you haven't asked for its combined response, and I have not provided one.
You're comparing apples-to-apples with the examples I've provided.

Then I misinterpreted your post.

And again, I've already remarked how the non-coaxial drivers will interfere with each other in regards to vertical off-axis response. I've mentioned it several times, but perhaps it's gone unacknowledged because you aren't aware?

I haven't taken issue with that because we're in complete agreement here.

I try to be respectful to everyone on the forum!

I've noticed.

 

[Rex]

Derailed big time.

Agreed.

Btw , finally had a bowel movement after 3 days.

See? There was a purpose to it after all.

 

[Promit]

Without wading into the deeper technicalities that @Rex and @yeky83 are debating, I have to point out that all of the measurements are in the far field of the PA box. That's what you would want in most cases, but what you don't get is a minimum distance at which the imaging behaves that way and it's that problem which has been a challenge for me with the cheaper units in wedge mode. The appeal of the Celestion driver this thread is about, and coaxials in general, is they can be treated as nearfield monitors or psuedo-cabs on stage, while a conventional PA really wants to be firing at you from kind of far away.

I can stand in front of a guitar cab and listen to what I'm playing; the same is true of coaxial monitors. It's distinctly not true of your standard affordable PA, which leaves you awkwardly trying to center yourself between the two images at close range. So while it's all well and good to debate the far field behaviors of the two types of monitors, it's not what matters to me personally. I would never use a coax at range.

 

[yeky83]

If you look at what you originally posted (which was the basis of my disagreement), you presented an example of a non-coaxial design as having better off-axis response than a coaxial design. But even at a narrower 22.5°, the coaxial plot shows response within 3 dB of on-axis, across most of the spectrum; whereas the non-coaxial plot shows a variation from 2 dB to 6 dB across the same spectrum — that's more jagged and uneven, not less.

Got it, and that seems a fair criticism. I guess it depends on what off-axis angle one chooses.
But! On the whole it still seems fair to me to say that the HF-through-woofer coaxial has a jagged off-axis that the horn compression driver does not have. One is an effect of inefficient horn loading and diffraction that isn't by design but an accepted consequence of it, while the other is more or less controlled and by design.

Show me the same kinds of plots, with the same axes and the same smoothing applied, for both speakers. Then we'll have something to talk about. Sources would help, too.

Tried to do that earlier, but I know I've written too much to wade through

Here it is again, from the Celestion catalog:
https://celestion.com/files/brochures/Pro_Speaker_Catalogue.pdf
Pick from it most any coax vs. horn loaded compression driver.

On-axis behavior of FTX1225, HF-through-woofer coaxial (red for HF):

On-axis behavior of CDX1-1445, horn loaded compression driver:

It's not about max SPL. I never presented it as such.

Hm, OK. I was talking past you then, sorry about that.

I haven't taken issue with that because we're in complete agreement here.

Cool, got it

Derailed big time.

Sorry! Gonna go on for just a bit more with this post haha

Without wading into the deeper technicalities that @Rex and @yeky83 are debating, I have to point out that all of the measurements are in the far field of the PA box. That's what you would want in most cases, but what you don't get is a minimum distance at which the imaging behaves that way and it's that problem which has been a challenge for me with the cheaper units in wedge mode. The appeal of the Celestion driver this thread is about, and coaxials in general, is they can be treated as nearfield monitors or psuedo-cabs on stage, while a conventional PA really wants to be firing at you from kind of far away.

I can stand in front of a guitar cab and listen to what I'm playing; the same is true of coaxial monitors. It's distinctly not true of your standard affordable PA, which leaves you awkwardly trying to center yourself between the two images at close range. So while it's all well and good to debate the far field behaviors of the two types of monitors, it's not what matters to me personally. I would never use a coax at range.

How close are your ears from your wedge monitor? When considering far-field response, ~6 ft should be fine I think, and my ears are always at least 6 ft from my wedge.

 

[Chiguete]

Celestion please bring out this speaker soon before people here go tech crazy on detail jejeje.

 

[Rex]

I regret starting this thread! Goodnight

The derail ended 24 hours ago. The topic is yours.