Hello,
I have been experimenting with open baffle speakers for a bit now, and read a lot here and on other sites.
The one statement that always comes up is that one has to EQ an open baffle to compensate for the 18db theoretical roll off.
Now, I believe I have understood (and experienced) the reasons why, but what is not often mentioned is exactly HOW one compensates/eq's to "level out" the low frequencies.
DSP is most often mentioned, but I don't often see any detailed answers on how this could be done in the "analog domain", which I would prefer.
Hence my question, would it make sense to incorporate a parametric (pro-audio) EQ to do so? For example the Klar Teknik DN410 seems to be quite versatile for this, and is what I am thinking about.
Please not that I am going to use an analog active crossover and bi-amp the open baffle speaker. My idea is to put the parametric EQ in front of the active crossover. I would also like to do some minor frequency "shaping" on the mid-range.
Thanks in advance!
I have been experimenting with open baffle speakers for a bit now, and read a lot here and on other sites.
The one statement that always comes up is that one has to EQ an open baffle to compensate for the 18db theoretical roll off.
Now, I believe I have understood (and experienced) the reasons why, but what is not often mentioned is exactly HOW one compensates/eq's to "level out" the low frequencies.
DSP is most often mentioned, but I don't often see any detailed answers on how this could be done in the "analog domain", which I would prefer.
Hence my question, would it make sense to incorporate a parametric (pro-audio) EQ to do so? For example the Klar Teknik DN410 seems to be quite versatile for this, and is what I am thinking about.
Please not that I am going to use an analog active crossover and bi-amp the open baffle speaker. My idea is to put the parametric EQ in front of the active crossover. I would also like to do some minor frequency "shaping" on the mid-range.
Thanks in advance!
I don't know anything about the device you're interested in but I do know the mini dsp and Behringer dcx2496/deq2496 tend to be pretty popular for this application. Some of the above might be a little more than you need if you already have an active crossover.
I'm going to engage in a debate but if you are going to try to boost any frequency by 18db good luck.
I am all for it 🙂
Just wanted to see if I am on the right path, and/or if anyone else has done the same already.
Thanks, I know about these devices; however, I would prefer to find an analog solution.
I follow ya now. I guess I should read your post a little more closely. My apologies.
You can look at Linkwitz Labs has an analog signal processor design for several of his designs. I still prefer the analog version although most switched to minidsp without complaint. It might depend on the quality of your preamp/source.... You can use a combo of boost of low frequencies and cut of other frequencies. Do realize this will effect phase... You could also biamp and use different voltage gain amps for low frequencies if your xo point is low enough....
Trying to compensate for that low frequency short circuit that exists with all practical open baffle designs requires a very large SD times Xmax air displacement from one's woofers, along with the substantial power requirements to meet any transients.
Most people trim off the response below resonance to prevent over-excursion of the drivers which can be a little disconcerting to watch at those very low frequencies.
I have built an open baffle with 8 Peerless 830668 10 inch woofers wired in series parallel to get the efficiency up, the baffle was apprximately 8 feet high by 3 feet wide with wings, not something one would have in their lounge room.
If I was doing it again I would have used the Peerless 830669 12 inch woofers for the extra SD they provide.
The amplifier was a home made amplifier and was a 200 watts RMS/ channel power amplifier, the thing is if one is going to do this exercise one needs to raise the efficiency of the woofer driver combination rather than amplifier power, and that means a great number of drivers to gain the necessary SPL for it to be meaningful.
C.M
Most people trim off the response below resonance to prevent over-excursion of the drivers which can be a little disconcerting to watch at those very low frequencies.
I have built an open baffle with 8 Peerless 830668 10 inch woofers wired in series parallel to get the efficiency up, the baffle was apprximately 8 feet high by 3 feet wide with wings, not something one would have in their lounge room.
If I was doing it again I would have used the Peerless 830669 12 inch woofers for the extra SD they provide.
The amplifier was a home made amplifier and was a 200 watts RMS/ channel power amplifier, the thing is if one is going to do this exercise one needs to raise the efficiency of the woofer driver combination rather than amplifier power, and that means a great number of drivers to gain the necessary SPL for it to be meaningful.
C.M
Last edited:
That is the "ultimate" roll off as you approach DC. The shape/slope(s) of your rolloff are very likely a bit more complex that that.
You probably know that an OB type system has a lot of losses because at low frequencies the front and rear sound waves, which are out of phase by 180 degrees, meet and greet and to some degree cancel each other out. The gives rise to the well known "dipole peak", below which the response rolls off. This rolloff is at 6dB per octave. The frequency where the dipole peak happens is determined by the pathlength from the front of the OB to the rear. You can calculate the frequency in Hertz as F = 0.5*v/D, where v=the speed of sound, or 344 m/sec. For example, if the baffle width is 20", the peak is at 345 Hz. The longer the pathlength, the lower in frequency the dipole peak and vice versa. So, if you use a 20" wide baffle with a woofer, there is quite a span between 345Hz and the Fs of the driver in free air (approx. the resonance of the driver on the OB). Over most of this span, the rolloff will be abotu 6dB per octave. As you near the resonance frequency of the woofer, the woofer's Qts influences the shape. Unless Qts is relatively high, like 1.0 or higher, the slope response will become steeper above Fs and after passing thru Fs and going below will eventually reach 18 dB per octave. When the Qts is higher the driver's response around resonance can help to shore up the sagging OB response to a degree and this is why some people prefer to use high Qts woofers in OB systems.
In any case, you have to take all of this into account when designing some kind of EQ scheme for your OB speaker. Unfortunately it is difficult to measure the far-field OB response directly (for a woofer) because it is happening at relatively low frequencies and the room interactions are difficult to separate from the driver's response. You can measure a nearfield response and model the contribution from the OB, and that is a relatively good approximation to a quasi-free-field response. Some people just measure the room response and try to work with that. It's up to you.
Thank you for all your help so far.
As far as I understand, there are several factors that impact the bass performance of an open baffle speaker.
Ignoring the room, this would be:
- baffle width
- driver Qt
- driver fs
Those make perfect sense to me. The following also influence the bass performance, although I am not yet 100% certain why:
- driver SD
- driver Xmax
Taking the room into consideration, the following further influences the bass performance:
- driver distance from floor
- OB distance from walls/ceiling
- listener distance
With this in mind, my thoughts were as follows:
Use the Beyma 18PWB1000/Fe which has a low Fs of 27 Hz, a high Xmax of 12.5 mm, a Vas of 317 l, and the Sd is 0.1255 m².
Mount this monster on a 24" wide baffle, 44" tall, about 3" from the floor.
I am waiting for a 2×170 Watt RMS amp that I want to use exclusively for these (e.g. bi-amping).
According to the formula that CharlieLaub provided, the dipole peak would be at 282 Hz. So, if I understand correctly, the bass response will now roll-off, starting at 282 Hz, with a 6 db slope, until it reaches Fs, which is 27 Hz. After 27 Hz it will roll-off about 18 db. All this is a bit more complex due to speaker placement, floor reflections etc.
I would set the crossover somewhere around 200 Hz. (I am planning to use some rather large full-range drivers for the mids).
Now the problem would be, how to compensate for the 6 db roll-off? Frankly, I don't think I worry too much about what happens after 27 Hz (drivers Fs).
The only "electronics" I ever built were crossovers for which I had the exact schematics (and a lot of help from the DIY Audio community). So I am not too confident about building my own line-eq and/or active crossover (unless you convince me otherwise). Hence my thinking of using a parametric EQ to solve the problem (I already have a nice vintage active crossover).
Please tell me if all this makes sense, of if I am not thinking quite straight here...
As far as I understand, there are several factors that impact the bass performance of an open baffle speaker.
Ignoring the room, this would be:
- baffle width
- driver Qt
- driver fs
Those make perfect sense to me. The following also influence the bass performance, although I am not yet 100% certain why:
- driver SD
- driver Xmax
Taking the room into consideration, the following further influences the bass performance:
- driver distance from floor
- OB distance from walls/ceiling
- listener distance
With this in mind, my thoughts were as follows:
Use the Beyma 18PWB1000/Fe which has a low Fs of 27 Hz, a high Xmax of 12.5 mm, a Vas of 317 l, and the Sd is 0.1255 m².
Mount this monster on a 24" wide baffle, 44" tall, about 3" from the floor.
I am waiting for a 2×170 Watt RMS amp that I want to use exclusively for these (e.g. bi-amping).
According to the formula that CharlieLaub provided, the dipole peak would be at 282 Hz. So, if I understand correctly, the bass response will now roll-off, starting at 282 Hz, with a 6 db slope, until it reaches Fs, which is 27 Hz. After 27 Hz it will roll-off about 18 db. All this is a bit more complex due to speaker placement, floor reflections etc.
I would set the crossover somewhere around 200 Hz. (I am planning to use some rather large full-range drivers for the mids).
Now the problem would be, how to compensate for the 6 db roll-off? Frankly, I don't think I worry too much about what happens after 27 Hz (drivers Fs).
The only "electronics" I ever built were crossovers for which I had the exact schematics (and a lot of help from the DIY Audio community). So I am not too confident about building my own line-eq and/or active crossover (unless you convince me otherwise). Hence my thinking of using a parametric EQ to solve the problem (I already have a nice vintage active crossover).
Please tell me if all this makes sense, of if I am not thinking quite straight here...
If you do decide to roll your own opamp based active EQ, I've got an EQ design that I use for transmission line speakers that could be easily tweaked to give you what yo need.
Here's a sim to do 20dB of bass boost, starting from ~250Hz and finishing boosting at ~80Hz. It's probably a bit faster than you need, but you can easily tweak the rolloff slope by changing R3.
That's a lot of boost though.
Here's a sim to do 20dB of bass boost, starting from ~250Hz and finishing boosting at ~80Hz. It's probably a bit faster than you need, but you can easily tweak the rolloff slope by changing R3.
That's a lot of boost though.
Attachments
I suppose making the baffle smaller would help? I cannot go much smaller though, it's an 18" woofer, so I suppose I should have at least around 2" of wood on it's edges.
Thank you for that. When I feel even more adventurous, I will look into that.
Coming back to my original question:
Is an analog parametric EQ capable of "leveling" this kind of setup?
I don't think what you are trying to do is off base. If you want better chances of getting it right I'd buy the woofer and a way to get some measurements. Packages like soundeasy will help modeling but there is no substitute for measurements. Measuring dipoles woofers at low frequencies is tricky as it's tough to get enough bottom end to get SNR below 50HZ outside. Far field outdoor measurements yield the best results in my experience if you can get enough output. Just dont use ground plane technique for outdoor dipole measurements.... The good news is dipoles have less room interactions so you can get a decent idea in room. Once you know the shape of the driver response in the baffle you can determine how much gain you really need. Then you can decide what method will work and sound best......
That is probably the best idea. I already have a calibrated mic and REW installed, and have used it for some crudely built baffle that I am testing now.
Regarding EQ'ing, is it correct that more than one band on an parametric EQ will work additively? I am playing around with an EQ plugin in Ardour right now, just to see whether I can get a smooth boost with a parametric EQ at all. It seems quite possible.
BTW, the max boost on the EQ I mentioned earlier is +15 db and -24 db, it seems that that is sufficient.
Regarding EQ'ing, is it correct that more than one band on an parametric EQ will work additively? I am playing around with an EQ plugin in Ardour right now, just to see whether I can get a smooth boost with a parametric EQ at all. It seems quite possible.
BTW, the max boost on the EQ I mentioned earlier is +15 db and -24 db, it seems that that is sufficient.
drtebi,
Here is a link that might be helpful to you where OB is being discussed with some depth and explains it quite well, complicated as it is.
HELP: xmax and open baffles
C.M
Here is a link that might be helpful to you where OB is being discussed with some depth and explains it quite well, complicated as it is.
HELP: xmax and open baffles
C.M
A little food for thought....I would favour driver area and a frame of some description over Xmas any day of the week
I came across that thread just minutes before you suggested it! It is very useful and interesting indeed, more to learn.
I am not exactly sure what you are trying to say?
More Xmas means more distortion and more power, it is better to keep it to a minimum by using more cone area and/or an h/u frame for example rather than excessive eq
- Status
- Not open for further replies.