So you would prefer an 18" over a 12" for example, correct?
What Xmax, in your opinion, is acceptable?
What Xmax, in your opinion, is acceptable?
Or multiple drivers. It depends how low you want to go. Even Linkwitz advocates sealed subs below 40Hz. Have you looked into H/U frame? http://www.quarter-wave.com/OBs/U_and_H_Frames.pdf
As far as I know parametric eq is additive or subtractive meaning that if you add gain where the filter bandwidths overlap then it sums. I've only used parametric EQ in digital mixing consoles and software though.... I'd test for clipping at the outputs before hooking it up to amp and speaker. Make sure output voltage doesn't exceed amplifier input specs...
I have the Linkwitz LX521 and love the bass. I have a pair of sealed 18's for HT that do not sound nearly as good for music. The Seas drivers used in the LX521 are flattened to the low 30s and I enjoy them without subs. If you stay within Xmax with a decent driver I doubt you will have complaints about distortion. 18's will likely have less distortion for the same output but they also need to be crossed at a lower frequency for acceptable integration. Tradeoffs[emoji16]
Here are some approached to equalizing gradient woofers using shelving and Q boost filters. Gradient woofeer equalization
The mentioned EQ has an overload warning LED, which will be useful for this purpose. Thanks for the tip on output voltage, I need to verify this.
The midrange I am currently testing is actually 12", and has really nice mid-bass. Therefore a low crossover point to the 18" is possible.
That page, and in fact the entire site, is just full of useful info. I only recently discovered it, and need to dive deeper into it. Thanks for the link.
The one advantage I see in using a dedicated parametric EQ is, that I can "play around" with different filter settings, and sort of dial in what's sounds (or measures) best for the room. This is something that would be a lot more difficult to accomplish with an active circuit, where replacing components would be necessary for every tweak...
I suppose one could experiment with things like parametric EQs and variable active crossovers, and once suitable settings have been accomplished, build an active circuit using the desired parameters.
One thing I haven't completely understood yet is the phase shift that occurs when one compensates for the low frequency roll-off. How does boosting the low end affect phase?
Yep, the room, the speakers and the positioning will all effect the required EQ
Don't worry about it, the low order filters will not have an audible effect
No. That will just make the dipole peak move up in frequency and increase the amount of loss you have at low frequency.
An off-the-shelf PEQ unit has the wrong type of filters for EQing an open baffle speaker.
You will need a couple of first order shelving filters and possibly some EQ for adjusting the final (lowest frequency) rolloff "knee" shape. See the link given by JohnK in post 25 - it explains how to use these filters. Some digital pro-crossover units like the venerable Behringer DCX2496 include shelving filters but I am not aware of any analog EQ unit that have them.
The Klark-Teknik PEQ I want to use has 12db low- and high pass filters. The latter would definitely help at the very bottom end. Here a large picture of the unit, where you can see all the possible settings:
Anyway, after having done a few measurements on an experimental baffle (about 22" x 42") with a 12" full-range driver, I have to say I don't think just applying a 6 db boost after the dipole peek plus a "roll-off knee" will result in a perfect low-end response. It is only a starting point (a proper one, I should say).
The measurements looked vastly different from the theory...
From what I have been observing and learning so far, there are many other factors that come into play, especially the room size, baffle position, floor reflections etc., that influence the bass response. Studying these acoustic phenomenons, taking measurements, and experimenting are in my opinion the only way to get closer to that light in the end of the tunnel...
That's also why I like to have the flexibility of "a bunch of knobs" so that I can experiment and learn.
The filters you mention are high pass on the left, low pass on the right, so they won't help with low end boost. You could use the parametrics to experiment
Yes, I do understand that. The reason I mentioned the high-pass filter is because of CharlieLaub's comment:
You will need [...] some EQ for adjusting the final (lowest frequency) rolloff "knee" shape.
For that I suppose a high-pass filter will help.
Gotcha, it might, but it will also roll off everything below at an increasing rate, again, one of the parametrics may be more useful. I would also save the high pass to attenuate sub sonic frequencies. Be very careful if you use parametrics in close proximity, it's easy to end up with dangerously high levels of boost, they are much better for cutting troublesome frequencies
Thanks for that tip. I will definitely be careful.
I found that a good way to get a feel for parametric EQ settings is to simply play around with a software-based one. I have Ardour installed (similar to Cubase, Protools etc.) and a couple of different parametric EQs. When adjusting the EQs, you will get a visual picture of the result.
Here is an example:
I found that a good way to get a feel for parametric EQ settings is to simply play around with a software-based one. I have Ardour installed (similar to Cubase, Protools etc.) and a couple of different parametric EQs. When adjusting the EQs, you will get a visual picture of the result.
Here is an example:
Hmmmmm....
Just came across this Marchand Electronics component, quite interesting, it almost looks as though it was designed for OB speaker's and their bass roll-off:
subwoofer equalizer, bass correction equalizer, bass boost
Any thoughts on that, gentlemen?
Just came across this Marchand Electronics component, quite interesting, it almost looks as though it was designed for OB speaker's and their bass roll-off:
subwoofer equalizer, bass correction equalizer, bass boost
Any thoughts on that, gentlemen?
Member
Joined 2009
Paid Member
It is just so much easier with DSP... I have meanwhile invested into the RME ADI-2 Pro DAC (a AD/DA pre-amp, really), which works wonders with the EQ. I am also now using 18" woofers...
What really made the biggest difference however, was the addition of a more powerful power amp. Now the bass is really kicking 🙂
Still a prototype, but very good indeed. Unfortunately I have too many other things to take care of before I get back into this.
What really made the biggest difference however, was the addition of a more powerful power amp. Now the bass is really kicking 🙂
Still a prototype, but very good indeed. Unfortunately I have too many other things to take care of before I get back into this.
Attachments
That's quite the pyramid you've got going there!
Sorry I missed this thread back over the summer - a lot of good stuff here.
When I started building open baffle speakers about 12 years ago, I took the typical "Boost that Bass!" EQ route. After all, why not? If the OB lacks bass, just use EQ to boost it. It worked, sort of, but never seemed to work as well as I'd hoped. After a lot of coaching from and OB expert (John Busch) I started to understand these weird OB beasts better.
As Charlie pointed out, the bass roll off on an OB can be unusual. I've often seen 9dB per octave, which comes from a combination of baffle size, room placement, woofer Q and other stuff. It's rarely classic slopes. And as Scott points out, you have to beware of clipping in EQ boost.*
So there is another way to do it: Attenuate.
When you are looking for a fairly flat frequency response and good tonal balance, you just have to accept that the bass inefficiencies of your OB woofer set the limit. Every other part of the speaker will be more efficient, need less power, than does the bass. To get a flat response everything else needs to come down to match that inefficient OB bass. You might think "I'll just turn up the bass instead" - but it's the same thing. You are simply pushing the low end as far as it will go with the signal chain and woofer X-max, and everything else will be below that.
How do you deal with that? In the crossover. A woofer on OB will have significant low end loss. When plotted, that will look like a rising woofer response. To flatten out that rise, cross the woofer very low - much lower than your nominal crossover point. Using a first or second order slope at 60-100 Hz low pass often works wonders for flattening the woofer response and serving as your crossover. Seems crazy low, but acoustically it works. You may still need a trap or a slope tweak, but that's the basic approach. Cross low and shallow to flatten the woofer response.
*You can cheat here some, as once you've split up the signal with a crossover, its average level will be much lower - per section. You can use that lower level to your advantage and apply some boost without clipping. Playing around with low pass filters, EQ and musical signals will quickly show you how much boost will be safe.
Sorry I missed this thread back over the summer - a lot of good stuff here.
When I started building open baffle speakers about 12 years ago, I took the typical "Boost that Bass!" EQ route. After all, why not? If the OB lacks bass, just use EQ to boost it. It worked, sort of, but never seemed to work as well as I'd hoped. After a lot of coaching from and OB expert (John Busch) I started to understand these weird OB beasts better.
As Charlie pointed out, the bass roll off on an OB can be unusual. I've often seen 9dB per octave, which comes from a combination of baffle size, room placement, woofer Q and other stuff. It's rarely classic slopes. And as Scott points out, you have to beware of clipping in EQ boost.*
So there is another way to do it: Attenuate.
When you are looking for a fairly flat frequency response and good tonal balance, you just have to accept that the bass inefficiencies of your OB woofer set the limit. Every other part of the speaker will be more efficient, need less power, than does the bass. To get a flat response everything else needs to come down to match that inefficient OB bass. You might think "I'll just turn up the bass instead" - but it's the same thing. You are simply pushing the low end as far as it will go with the signal chain and woofer X-max, and everything else will be below that.
How do you deal with that? In the crossover. A woofer on OB will have significant low end loss. When plotted, that will look like a rising woofer response. To flatten out that rise, cross the woofer very low - much lower than your nominal crossover point. Using a first or second order slope at 60-100 Hz low pass often works wonders for flattening the woofer response and serving as your crossover. Seems crazy low, but acoustically it works. You may still need a trap or a slope tweak, but that's the basic approach. Cross low and shallow to flatten the woofer response.
*You can cheat here some, as once you've split up the signal with a crossover, its average level will be much lower - per section. You can use that lower level to your advantage and apply some boost without clipping. Playing around with low pass filters, EQ and musical signals will quickly show you how much boost will be safe.
I agree with your findings...
My crossover is where I set the proper volumes. There I have the subwoofer at max, and the other three somewhere just above the middle. I was lucky enough to find a Sony TA-D88, which works great for this purpose. The lowest crossover point I can set for the subwoofer is 140 Hz. Lower might be interesting, and there is a module for the TA-D88 that can do 62,5 / 100 / 125Hz, but it's pretty much unobtainable. To my ears, 140 Hz works just right though.
The RME ADI-2 Pro has a 5-band parametric equalizer. The beauty about this device is, that it automatically adjusts its output to avoid distortion. I have not yet seen this method fail... so all in all, I am quite happy with the current setup. Although, as I mentioned, it is currently on hold, since the house I live in is not yet fully remodeled, and I happen to be "the remodeler" 🙂
I have done some measurements, but it's hard to get those right, and even harder to understand what's really going on. In the end I just listened, experimented, and tried a couple of different baffle shapes.
Going from a 2 feet wide, 4 feet tall baffle to this pyramid shape made a huge difference... in my opinion, the wide baffle did colorize the sound quite a bit.
What do yours look like? Tell me more... 😉
My crossover is where I set the proper volumes. There I have the subwoofer at max, and the other three somewhere just above the middle. I was lucky enough to find a Sony TA-D88, which works great for this purpose. The lowest crossover point I can set for the subwoofer is 140 Hz. Lower might be interesting, and there is a module for the TA-D88 that can do 62,5 / 100 / 125Hz, but it's pretty much unobtainable. To my ears, 140 Hz works just right though.
The RME ADI-2 Pro has a 5-band parametric equalizer. The beauty about this device is, that it automatically adjusts its output to avoid distortion. I have not yet seen this method fail... so all in all, I am quite happy with the current setup. Although, as I mentioned, it is currently on hold, since the house I live in is not yet fully remodeled, and I happen to be "the remodeler" 🙂
I have done some measurements, but it's hard to get those right, and even harder to understand what's really going on. In the end I just listened, experimented, and tried a couple of different baffle shapes.
Going from a 2 feet wide, 4 feet tall baffle to this pyramid shape made a huge difference... in my opinion, the wide baffle did colorize the sound quite a bit.
What do yours look like? Tell me more... 😉
There have been a lot. Here's a sample.
Suggestions on OB problems
Fast, fun, Inexpensive OB project
Caveman Speakers. The Troglodytes.
It's all fun stuff.
Next up -21" woofers!
Suggestions on OB problems
Fast, fun, Inexpensive OB project
Caveman Speakers. The Troglodytes.
It's all fun stuff.
Next up -21" woofers!
- Status
- Not open for further replies.