Paul,
Welcome to a world without boxes, where even small things make audible differences, but luckily they almost always sound good. I'll take a shot at your "exactly". It's a few things plus 2 maybe's.
First, regarding "D" for H baffles (or any baffle with drivers radiating on axis for that matter). D, the extra travel distance for the rear wave, is the distance from the driver to the rear edge of the baffle plus the distance from the rear edge to the plane of the driver. For a simple symmetrical H baffle it is about the same as the total depth of the H. With yours the average D is greater because the distance over the top for the bottom driver is far greater than around the sides. The bottom line is that the Fequal for your H is more like 110hz instead of 200hz. Note that Fequal is a free space concept and most things add bass back, but it is a useful starting point, because with no baffle you wouldn't get much bass at all.
1. Maybe the woofers- If you're only running one voice coil, then the Q would be very high and you'd have some boost as you approach resonance. Have you measured them, and if so how accurate is that .45 Qts?
2. Maybe measuring position- With dipole, your bass is now highly directional and other than on axis reflections are greatly reduced. The increases the impact of on axis bass reflections and boundary gain from the rear wall can greatly affect response.
2. Your room- You're getting some boost from the on axis room mode, which shows up in each of the settings.
3. Boundary gain- The floor is giving you a boundary gain for the front wave. It is also blocking the rear wave from travelling in that direction.
4. Placement- That big cab is acting as a lot more baffle for the rear wave to travel around. Also, remember that woofer output is highly directional on axis, so the bulk of the rear wave bounces off of that back wall at an angle, then the side wall to the listening position. It looks to be about 3-4 meters extra distance vs the front wave. Plug that into your Fequal formula and you'll see that those reflections are reinforcing the front wave output. This means while some of the bass is bending around the baffle to go and partially cancel some of the front radiation, the way some of the rear wave reflections are directed is offsetting some of the dipole cancellation.
To get a feel for the effect of each of this items, just change them one at a time from the setup pictured. eg Move that cab away from the H baffle. Listen to the difference with the H on its side. Get the H up off the floor. Change the angle of the H. Change the placement of the H. Unfortunately the math is too complex to easily predict and quantify these differences.
I use Fequal as a starting point and rely on a very visual approach for how the rear wave will interact with the front wave as far as bass is concerned, and I haven't needed EQ yet.
Welcome to a world without boxes, where even small things make audible differences, but luckily they almost always sound good. I'll take a shot at your "exactly". It's a few things plus 2 maybe's.
First, regarding "D" for H baffles (or any baffle with drivers radiating on axis for that matter). D, the extra travel distance for the rear wave, is the distance from the driver to the rear edge of the baffle plus the distance from the rear edge to the plane of the driver. For a simple symmetrical H baffle it is about the same as the total depth of the H. With yours the average D is greater because the distance over the top for the bottom driver is far greater than around the sides. The bottom line is that the Fequal for your H is more like 110hz instead of 200hz. Note that Fequal is a free space concept and most things add bass back, but it is a useful starting point, because with no baffle you wouldn't get much bass at all.
1. Maybe the woofers- If you're only running one voice coil, then the Q would be very high and you'd have some boost as you approach resonance. Have you measured them, and if so how accurate is that .45 Qts?
2. Maybe measuring position- With dipole, your bass is now highly directional and other than on axis reflections are greatly reduced. The increases the impact of on axis bass reflections and boundary gain from the rear wall can greatly affect response.
2. Your room- You're getting some boost from the on axis room mode, which shows up in each of the settings.
3. Boundary gain- The floor is giving you a boundary gain for the front wave. It is also blocking the rear wave from travelling in that direction.
4. Placement- That big cab is acting as a lot more baffle for the rear wave to travel around. Also, remember that woofer output is highly directional on axis, so the bulk of the rear wave bounces off of that back wall at an angle, then the side wall to the listening position. It looks to be about 3-4 meters extra distance vs the front wave. Plug that into your Fequal formula and you'll see that those reflections are reinforcing the front wave output. This means while some of the bass is bending around the baffle to go and partially cancel some of the front radiation, the way some of the rear wave reflections are directed is offsetting some of the dipole cancellation.
To get a feel for the effect of each of this items, just change them one at a time from the setup pictured. eg Move that cab away from the H baffle. Listen to the difference with the H on its side. Get the H up off the floor. Change the angle of the H. Change the placement of the H. Unfortunately the math is too complex to easily predict and quantify these differences.
I use Fequal as a starting point and rely on a very visual approach for how the rear wave will interact with the front wave as far as bass is concerned, and I haven't needed EQ yet.
For some reason I didn't get email notification of responses! I didn't realise I had these responses.
The mic is a diy mic which is based on an electret mic very similar to the popular panasonic mic. It is not housed in a case, and the response curve in the data sheet is ultra flat down to 20 Hz.
The mic was placed in the listening position about 2.8m from the speakers in a direct line. I use Ultracurve DEQ2496 to play pink noise. It then runs auto EQ to get it flat. It was able to get flat down to about 28 Hz without eq. The screenshot shows EQ applied to get it flat, so 0db corresponds to no EQ, +3db means the response is 3db down since that much boost is required to get it flat, and -3db means the opposite. This is not ideal, but I find if I measure the same way two times, the EQ is usually the same.
I'm not surprised at midrange differences, which I'd expect to relate to a different baffle width (double), dipole roll-off, the rear wave as well as a different room interaction.
John, you could say I tried to not give the OB too much of an easy ride. I knocked them together with particleboard 15mm thick, no glue, just some plasterboard screws. I figured that if OB is as good as some many say it is, and I expected, then it should sound good even with crappy construction. They perform well.
One thing that surprises me is how the perform when pushed. Somehow the sound quality seems to suffer less when the SPL goes up. The drivers work harder with more excursion, the amp reaches its limit and yet the sound appears cleaner, with less distortion and strain.
There must surely be more distortion here, yet perhaps this highlights something often forgotten about distortion - the room actually adds distortion. Try turning up your system with some monopole subs and go outside the room so what you hear is the walls vibrating. The sound is disgusting. It's a secondary form of distortion I rarely hear mentioned.
I haven't measured them. I'm using both VCs wired in series so this gives me 4 ohms, each driver being driven by one channel. The Qts is actually 0.366.
I usually get a room mode peak eq'd out at 35 Hz no matter where I put speakers in the room. This is the front to back wall mode. The side walls have glass doors and some windows which seem to take a lot of the energy out of other modes. Since the room is lightly built, it is fairly well damped and the modes in this room are easy to EQ.
Something I hadn't considered.
UPDATE
Today I cut the H frame in half and moved to baffle to test as a stacked pair of U frames, as well as U frames one on each side. I did a quick run of Auto EQ to get a comparison. See results in the attachment below.
The dotted black line shows the EQ settings on the H frame on one side, hence its actual response is the inverse of those settings. The red line shows the U frame in the same position but it is relative to the response of the H frame. The orange is also relative and shows the U frames where they are placed to each side of the mains as dual mono.
Dual mono U frames have no advantage over the H frame. They have some peaks which have to be removed.
The U frame has about 4db more output from 40 - 80 Hz with 2db less output at 30 Hz towards the bottom end of its useable range.
I could use more eq and get down lower but I prefer to use very little eq to get down lower. These AV12 woofers with nearly 2" p-p excursion are handy. I almost never came close to their xmax. Now one inch excurions are more common but the main amp runs out of steam still before them.
Subjectively the bass is more natural, but at times I feel I could use the extra sense of ooomph for movies and some music. For acoustic music, it is far superior. I didn't like the sound I was getting from a CD by an Australian group (John Butler Trio), but now they sound a lot better. They have acoustic bass and are fairly percussive.
I haven't had a chance to listen to the U frames and decide what I think.
sqlkev said:
The mic is a diy mic which is based on an electret mic very similar to the popular panasonic mic. It is not housed in a case, and the response curve in the data sheet is ultra flat down to 20 Hz.
The mic was placed in the listening position about 2.8m from the speakers in a direct line. I use Ultracurve DEQ2496 to play pink noise. It then runs auto EQ to get it flat. It was able to get flat down to about 28 Hz without eq. The screenshot shows EQ applied to get it flat, so 0db corresponds to no EQ, +3db means the response is 3db down since that much boost is required to get it flat, and -3db means the opposite. This is not ideal, but I find if I measure the same way two times, the EQ is usually the same.
I'm not surprised at midrange differences, which I'd expect to relate to a different baffle width (double), dipole roll-off, the rear wave as well as a different room interaction.
John, you could say I tried to not give the OB too much of an easy ride. I knocked them together with particleboard 15mm thick, no glue, just some plasterboard screws. I figured that if OB is as good as some many say it is, and I expected, then it should sound good even with crappy construction. They perform well.
One thing that surprises me is how the perform when pushed. Somehow the sound quality seems to suffer less when the SPL goes up. The drivers work harder with more excursion, the amp reaches its limit and yet the sound appears cleaner, with less distortion and strain.
There must surely be more distortion here, yet perhaps this highlights something often forgotten about distortion - the room actually adds distortion. Try turning up your system with some monopole subs and go outside the room so what you hear is the walls vibrating. The sound is disgusting. It's a secondary form of distortion I rarely hear mentioned.
I haven't measured them. I'm using both VCs wired in series so this gives me 4 ohms, each driver being driven by one channel. The Qts is actually 0.366.
I usually get a room mode peak eq'd out at 35 Hz no matter where I put speakers in the room. This is the front to back wall mode. The side walls have glass doors and some windows which seem to take a lot of the energy out of other modes. Since the room is lightly built, it is fairly well damped and the modes in this room are easy to EQ.
Something I hadn't considered.
UPDATE
Today I cut the H frame in half and moved to baffle to test as a stacked pair of U frames, as well as U frames one on each side. I did a quick run of Auto EQ to get a comparison. See results in the attachment below.
The dotted black line shows the EQ settings on the H frame on one side, hence its actual response is the inverse of those settings. The red line shows the U frame in the same position but it is relative to the response of the H frame. The orange is also relative and shows the U frames where they are placed to each side of the mains as dual mono.
Dual mono U frames have no advantage over the H frame. They have some peaks which have to be removed.
The U frame has about 4db more output from 40 - 80 Hz with 2db less output at 30 Hz towards the bottom end of its useable range.
I could use more eq and get down lower but I prefer to use very little eq to get down lower. These AV12 woofers with nearly 2" p-p excursion are handy. I almost never came close to their xmax. Now one inch excurions are more common but the main amp runs out of steam still before them.
Subjectively the bass is more natural, but at times I feel I could use the extra sense of ooomph for movies and some music. For acoustic music, it is far superior. I didn't like the sound I was getting from a CD by an Australian group (John Butler Trio), but now they sound a lot better. They have acoustic bass and are fairly percussive.
I haven't had a chance to listen to the U frames and decide what I think.
Attachments
If I decide to go ahead and build a proper version to replace my TLs, it might look something like this.
Compact H frames (push pull loaded, not as shows push push) or a dual H frame stack on one side.
Solid hardwood baffle one layer MDF second layer with small wings. Might use some recycled floorboards.
Waveguide on tweeter (if I manage to come up with one with performance I'm happy with) ... Vifa D25 aluminium tweeter
Tweaked Vifa P17 midbass with a copper phase plug.
Compact H frames (push pull loaded, not as shows push push) or a dual H frame stack on one side.
Solid hardwood baffle one layer MDF second layer with small wings. Might use some recycled floorboards.
Waveguide on tweeter (if I manage to come up with one with performance I'm happy with) ... Vifa D25 aluminium tweeter
Tweaked Vifa P17 midbass with a copper phase plug.
Attachments
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.