Active Curved Beyma 15MC700nd/AE TD12M/TPL150H

[D!sco]

In addition to the horizontal/vertical lobing issue, a 2-way ribbon will have huge distortion or directivity issues at the crossover anyways. Pairing it with woofer will likely require an unusually high crossover point, both drivers will likely distort at that level, and both drivers will probably be beaming. This would leave you with a narrow listening window on top of whatever other lobing issues occur. It's super tough to make a good 2-way like that.

My favorite designs for a center are a decent two-way with flanked woofers, WT/MW style. Get a nice midrange or midwoofer and cross it so the midrange takes the majority of the vocal range, 300-4000hz, get some fun/price effective drivers to do the rest. This is especially important when using a center for home theater, where most voices come directly from that channel. Having a strong 3-5" mid fill the room, unobstructed by cancellation is the goal, IMO.

 

[morpheusX]

The TPL-150H is not a ribbon, but an AMT, and can have a low XO point (some designs use it at 900Hz).
The L/R designed in this topic, use it from 1530Hz (with a LR24 XO) to 20kHz.

A future center would cross the TPL-150H at the same exact 1530Hz (with a LR24 XO), to guarantee voicing coherence across all speakers.

Unfortunately, i can't do a 3way as i'm limited to 2 outputs in the Active crossover, thus the options i mentioned in the previous posts.

 

[D!sco]

Have you considered using a passive XO on the mid/tweeter and using EQ to fix irregularities? I'm also super biased toward active woofers

 

[morpheusX]

Have you considered using a passive XO on the mid/tweeter and using EQ to fix irregularities? I'm also super biased toward active woofers

Yes, alex-z has suggested the same approach here

The problem with that approach, is that due to the space constrains, the Mid would have to be placed in an individual box, which would make measurements an almost impossible task.

I had to resort to Ground plane measurements for the L/R due to their dimensions, and if going with the above solution, it would be near impossible to place the above boxes in a stand or even in the ground for proper measurements.

So i'll probably go for the easier option, a simple two way, it will make taking proper measurements much easier, which will lead to an accurate result.

 

[3ll3d00d]

I don't understand how one would impact the other (distance to side walls vs lobbing)

I was thinking that if your side wall reflections were way off axis then lobing may not be such a big deal

 

[morpheusX]

Do you think that with the current positioning, that would be the case?

- To MLP: 3m,
- To right wall: 2,5m
- To Left wall: 5,5m

Can this be modeled through VituixCAD or other tool?

 

[D!sco]

Listening position and room can both be set in VituixCAD. No idea how to make a non-rectangular or complex room.

And don't forget to fill out the driver XYZ offsets, especially with a waveguide moving the tweeter back

 

[fluid]

Can this be modeled through VituixCAD or other tool?

You can look at it but it is quite basic and the result will change significantly depending on the surface absorption level. Vituix will let you model that.

Vertical nulls are generally less damaging a problem as you do not tend to change listening height very much unless you like to listen both sitting and standing.

For a single listening position a Horizontal null might be equally benign if it is outside of the normal listening position, but if you sit right in it which would happen if you had a few people on the couch someone is going to wonder why it sounds off.

 

[3ll3d00d]

It's just trigonometry, calculate the angles from the speaker to the extents of the listening area and to the side walls, compare to the position of the lobes in your design

 

[Adi777]

Can such a project compete with, for example, the JBL M2?

 

[morpheusX]

Don't have a clue

These are also not fully optimized, due to the constrains i had to proper measure them.
I cannot say how much better they would be if i would have a full spinorama taken by a NFS or something with the same accuracy, to use for the XO development.

But, the M2 is $6000 per speaker (without amps), and mine were around $1300 per speaker (i was able to get the drivers with a great discount)

 

[morpheusX]

Its a shame the TPL-75 doesn't have a horn.
Being a bit smaller than the TPL-150H, it would have been possible to use it with a 4" mid, allowing to do something like this:

 

[fluid]

Its a shame the TPL-75 doesn't have a horn.

It's only a 3D print away from having one

 

[morpheusX]

It's only a 3D print away from having one

Yes, it has crossed my mind, but i don't have a clue on how to design a Horn.

I'm now toying with the idea of repurposing the cabinet, so i have the available space for a proper center speaker.

The cabinet is modular, so it should be easy to change it from this:

To something like this:

Its just a sketch, but ideally, the center would fill the available space, so it would be aesthetically pleasing (WAF).

I need to talk with my builder, but it doesn't seem to hard to do, and would give me the space needed for a full WTMW, using dual 12" Woofers, and a 10" Medium, with the TPL-150H.

Of all alternatives, this seems the best one

 

[fluid]

Yes, it has crossed my mind, but i don't have a clue on how to design a Horn.

It requires even less effort to do so now Ath can use a rectangular source

https://www.diyaudio.com/community/...-design-the-easy-way-ath4.338806/post-7163666

 

[morpheusX]

Thank you for the link!

Meanwhile, i've spoken to the wife, she didn't like the idea of cutting the cabinet

I'm now reverting to the idea of a simple two way, maybe using the Purifi PTT6.5M08-NFA-01.

I'm not in a hurry, will continue evaluating options until the final decision.

Thank you for all the help

 

[morpheusX]

I'm on vacations, so i had some spare time to learn VituixCAD, so i can improve the crossover a bit more.
VituixCAD has now a new "On-axis linearity' constraint" functionality:

Example how 'On-axis linearity' constraint can be used: Target is to make as linear predicted in-room response (PIR) as possible so we optimize Preference rating with Custom equation and set all weight to SM_PIR and others to 0. That alone could lead to high deviations in on-axis and listening window responses if directivity slope is not linear. Or some unwanted tilt. Now we can set On-axis linearity constraint to keep deviations acceptable and slope close to 0 dB/oct.

It works perfectly, and allows to optimize which target we want to achieve (On Axis, Listening window, PIR, etc).

For example, optimizing for On Axis only i get near perfect "On Axis" response, but Listening window is affected.

If i optimize for "Listening Window", LI is improved, but On Axis is affected:

So i would like to ask a simple question to the experts: considering that the speaker will be corrected by Dirac, which target should i choose to ensure the best final result?

 

[fluid]

So i would like to ask a simple question to the experts: considering that the speaker will be corrected by Dirac, which target should i choose to ensure the best final result?

There is no right or wrong answer, you might prefer it one way someone else another. The on axis response will generally contain the highest number of diffraction related effects that will not be present on any other axis and by making it ruler flat everywhere else will be worse. This might be preferable if you were listening up close dead on axis, but you are a reasonable distance away. In that case I would prefer to follow the idea that any EQ should improve more than just the on axis response and if it doesn't leave it out.

I would not just restrict myself to looking at the main spinorama curves, look to where and why the on axis and listening window end up being different. There is much more information available in the whole set of two plane measurements.

I quite like to use the 3rd preference rating equation with the tick boxes for listening window and PIR slope. If you can hit those without making the on axis wild you should be in good shape for something that sounds balanced. If not then trial and error is the next way to go.

 

[3ll3d00d]

So i would like to ask a simple question to the experts: considering that the speaker will be corrected by Dirac, which target should i choose to ensure the best final result?

from those graphs, the difference between the two looks to be v small (with the marked difference being in a frequency range you may not be able to hear depending on your age/HF sensitivity). I would also look at more than just those 2 curves but I can't say I do one specific approach so follow that.

Another question you may want to ask yourself is whether you should apply full range correction by dirac to speaker that has been subject to this type of quasi anechoic EQ.

 

[morpheusX]

I quite like to use the 3rd preference rating equation with the tick boxes for listening window and PIR slope. If you can hit those without making the on axis wild you should be in good shape for something that sounds balanced. If not then trial and error is the next way to go.

This post, and this post by kimmosto has some tips on how to use the Preference Rating.


He suggest this for a "Decent weighting factors to start with:" (2), and i added "Increase weight of SM_PIR to 2 if typical listening distance is longer than near field." (4):

Which gives me:

I'm a bit "afraid" that, given the lack of proper vertical measurements, and also by the fact i've used GP measurements, it may skew the calculations for directivity, PIR, etc.

This post by Dcibel touches on the above:

Given that you're loading traced data, I would stick to optimize by axial response or listening window at 100% weight. Off-axis information created using diffraction tool has limited accuracy so it is only to provide general idea of driver directivity based on size, but unlikely that it will match reality >1kHz for power & DI charts.

Do you have any formula suggestion to optimize the speaker based on the specific constraints of this design process, as well as speakers/room constraints:
- Measurements using GP, and only to 50º
- No vertical measurements (TPL measurements using traced data from audioxpress for TPL-200H)
- Speakers at 3.5/4m to MLP,
- Low ceiling
- Lively room
- Will be equalized by Dirac