FR Design Goal? Which is the "best" curve?

Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.
I've been doing the modelling thing for a while now and there's alway one thing that never stops bugging me and that's the target curve of the simulations.

I'm doing a BR (2-way Bass Reflex) with a 15" mid-woofer.
It's possible to get insaine LF response if you tune it for an extended bass shelf.
It looks much neater with a B4 or Be4 style curve but I'm not sure what is the best compromise?

The standard model is equal to an anechoic chamber, we are obviously not sitting in one of those at home.
Add diffraction/baffle step and the curv looks very different, pretty awful really.
Add room gain and we're back in business again.
Percieved loudness isn't making things easier either, do we want equal'ish loudness (rising LF response) or flat FR?

A standard 1A electric bass guitarr has it's E-string at approximately 41Hz.
Some will have a B-string at 30Hz but I can't say for sure how common that is?
A kickdrum is usually tuned above 40Hz so I guess I'm wondering how much emphasis you should put on the bottom octave?
We have electronic/artificial instruments and there's always the pipe organ I guess?

So, keeping in mind box tuning, diffraction, room gain and the low bass beeing firmly in the modal region for the room, what is a good FR curve to aim for?
How low is it necessary to tune it?
How will a bass shelf "sound" compared to a smooth rolloff?

The devil is in the details...
 
Last edited:
In short:

In the bass, it all depends on your room (modes).

In the mid/highs, it depends on your room (RT60) and the dispersion width of your speakers:

- If they exhibit "regular but uniform beaming", then the response can be flat.

- If they are wide dispersion, then they need a slightly falling response.

- If they are omni they need more tuning.

- And if you use a FR driver that beams quite strongly (like a 5" or bigger), you need a rising response. But in my opinion the sound of the latter really $uck$ :smash:
 
Hi 6.283,

I just had a look at your website, which I would like to recommend to others. Your analysis of what makes speakers disappear is right on the mark and fully coincides with my own experiments. It may see a bit ot, but on the other hand, Markus is going to build, and may find some inspiration from it.

The point is, it is not just about getting the FR straight; the way a loudspeaker radiates its sound, as well as the interference caused by the cabinet, are essential issues that will have to be adressed as well.

Edit: just see I cross posted, and the approach you take is a valid one in this respect.
 
Last edited:
Administrator
Joined 2004
Paid Member
A solid 40Hz can be quite satisfying. As mentioned, unless you're doing dubstep, it's plenty. Get a solid 40 and you'll be surprised how deep it sounds.

For the overall curve I like the classic B-K falling response. That is, the response falls to -6dB from ~400Hz up to 20KHz. You can shift the 400Hz knee to your liking, but the -6dB point at the top is a very good target.
 
Sounds like a flat'ish response with a very slight downward slope as the frequency increases is the preferred way to go? 10dB drop over 2 decades looking at the graphs from the "The Subjective and Objective Evaluation of Room Correction Products" by Sean Olive.
Half of that below 1kHz.

Tailoring the response like in the article will probably require a dsp but it would seem making the speakers a little hot in the low end and slightly rolled off hf responce wouldn't be a cardinal sin. :)
40Hz will have to be good enough, that's what I have today with my ML-TQWT.
I'll take your word for it. ;)

My usual daily listening is pretty low, low enough to keep a conversation going without to much of a problem.
When I really listen I probably average 90dB or so.

I'm very partial to a fleshed out LF and a rolled off casual HF.
My head is spinning, time for bed. Hope you can read what I've written.
I can't...
Se you in the morning.
 
Both, the speakers for the B&K target curve and the speakers used by Olive had a wider dispersion than what are you going to build. So for them a slight roll off is indicated. A 1" in a 15" pasta bowl will have very narrow dispersion, so it should stay flatter.
If you roll off the highs it does not sound awful. Just find the slope that you like. But I also would compare it to flat.
No DSP required imho. Use a passive down-shelve.
 
Administrator
Joined 2004
Paid Member
Agree - do compare it to flat. The speakers and the room do make a difference. I use the B&K curve, I think the other would sound too heavy in my room. Your ears and brain don't work exactly like a microphone and chart, so you have to do some judgment by ear.

But either of those targets will give you good subjective results.
 
Well, I doubt there is any such thing as a perfect first trial crossover. It does however appear to affect the response quite a bit.

My latest efforts have been focused on trying to achieve a flat response without the aid of BSC or room contribution. I'm thinking the room gain should make the bottom a little more "full" as long as the baffle step doesn't cause a dip.
It's proving to be a challenge and F3 appears to end up at roughly 50Hz.
Adding a "typical" room gain will llower F3 to 40Hz.

I'll include a screen dump of the last example. I'm using the B&C 15TBX100.
LP crossover and baffle step included, no bsc and no room gain.
80L (20% fill) BR tuned at 50Hz.
 

Attachments

  • FR.jpg
    FR.jpg
    71.7 KB · Views: 224
So, keeping in mind box tuning, diffraction, room gain and the low bass beeing firmly in the modal region for the room, what is a good FR curve to aim for?
How low is it necessary to tune it?
How will a bass shelf "sound" compared to a smooth rolloff?

The devil is in the details...

Given the fact that the end response of any "tuning, diffraction, room gain" will be completely dominated by the room, and that no two rooms are the same, what difference does any of this matter?

If you EQ the modal region - which is the only way to get good LF response - then the free field response of source does not matter at all. Just use several closed box subs, an electronic crossover with EQ and be done with it. Just tune the system the way you want it with the electronics.

Its so much easier that way.

All this talk of sub tuning is such a waste of time.
 
Earl, you are one of my greatest sources of inspiration for this project and one of few who I percieve as really knowing what he's talking about. I very much appreciate your input and from what I've read earlier it sounds like a slightly unorthodox but successful way of solving the issue.

If I had limited time and plenty of resources I would probably just go ahead and do it like that. However, unfortunately my situation is very much the other way around. I have plenty of time and very limited resources.

Once I'm done building the speakers it will probably take me at least 2-3 years to save up enough cash to get a dsp (or electronic crossover) and a couple of cheap subs.
I wish things could be different but some matters are simply out of my control.

So, being unable to implement that solution in stage one, I have to build the speakers as stand alone units with no low end reinforcement from any sub. I'm very much aware that this is not the optimal way of doing things but I try to make the best I can with what I do have.

Any tips regarding target response, cabinet design, crossovers and EQ's are greatly appreciated but I simply can't afford a solution including any additional subs at the moment.
I could perhaps scrounge up the cash for a simple dsp build within a reasonable time frame but the subs will have to wait.

That's why I started this question regarding an extended bass shelf vs. flat response and what a desirable target response actually looks like..
A dsp could flatten the response after the fact since the tuning is lower with an EBS.
If I tune it higher for a flat response the dsp can't do all that much for the LF extension due to the physical limitations from the ported design.

A closed box with a Linkwitz transform would bring the low end extension but limit the peak output in a significant way and require a lot more power from the amp.
The Bass Reflex cabinet appears to be the best compromise although far from perfect.
I guess I'm just trying to find the best compromise and the best bang for the buck.
 
Fair enough. You did not make clear that you were limited in budget. I still think that the HP characteristics of the woofer tuning are not really that important once one gets the speakers in the room. The thing is that the tuning of the system on paper will not mean a thing once the speakers interact with the room modes. But some DSP is still likely to yield a better LF in-room response than without it. I would still use a closed box and EQ the system in-situ (only at the LFs) with say a miniDSP based on some in-room measurements. Don't EQ anything above about 150 Hz though as you are only likely to make things worse rather than better.

There should be, as discussed, some falloff of the power response with frequency. I shoot for typically a -3 dB per decade drop in the listening axis response after about 1 kHz. But this depends very much on the speaker and if it is constant directivity or not. A true CD speaker up to 10 kHz and beyond needs a lot more axial response drop than a speaker with a beaming response.
 
Good solid advice here, thanks for getting me to reevaluate the closed box option and speaking plainly regarding the FR. :D
I'll have to do some more modelling but it doesn't look half bad.
My pair of 15TBX100 will be shipped tomorrow, I'm as giddy as a schoolboy.
 
Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.