I can build better boxes, with better bracing than I usually do. I've been trying to make a good center channel speaker. Mine are sealed vertical two or three ways typically. Right now I'm working with 2-ways.
I had a leftover box I was reusing. Roughly a 9 x 12 baffle, and about 9 deep. Woofer is 6.5". Box is unbraced 3/4" MDF, and 3/4" baffle. I like little speakers for some things. I Made a new x-over at 3k. The response looked fine. After several days of use, I was going to assemble the x-over, and call it done. I removed the speaker from the system. I wanted to watch some shows on my HT system. I replaced the center with one of my mains. It made a better center than the one I was going to finalize. Now I want to figure out why one sounds better than the other.
The issue is in the low vocal range. A coloration I think. The vocals sounded muddy. I previously thought the problem was the movie soundtrack, but this speaker handled it really well. Other movies seemed good with either speaker. Can lack of box bracing make that big of a difference?
Both speakers are sealed, and the Q is low. Both were crossed to a sub at 60 or 80hz, I think. Their f3s are a little different. Both in the 80 to 100hz range. Drivers are pretty low in distortion, and this problem was not at a loud SPL It was dialog. The movie was Midway, which is on Netflix.
I have REW, and can measure response at the seat, and compare. I don't expect to see any drastic difference though. What else might show the box issue, if that's the cause?
I had a leftover box I was reusing. Roughly a 9 x 12 baffle, and about 9 deep. Woofer is 6.5". Box is unbraced 3/4" MDF, and 3/4" baffle. I like little speakers for some things. I Made a new x-over at 3k. The response looked fine. After several days of use, I was going to assemble the x-over, and call it done. I removed the speaker from the system. I wanted to watch some shows on my HT system. I replaced the center with one of my mains. It made a better center than the one I was going to finalize. Now I want to figure out why one sounds better than the other.
The issue is in the low vocal range. A coloration I think. The vocals sounded muddy. I previously thought the problem was the movie soundtrack, but this speaker handled it really well. Other movies seemed good with either speaker. Can lack of box bracing make that big of a difference?
Both speakers are sealed, and the Q is low. Both were crossed to a sub at 60 or 80hz, I think. Their f3s are a little different. Both in the 80 to 100hz range. Drivers are pretty low in distortion, and this problem was not at a loud SPL It was dialog. The movie was Midway, which is on Netflix.
I have REW, and can measure response at the seat, and compare. I don't expect to see any drastic difference though. What else might show the box issue, if that's the cause?
Without seeing measurements of both speakers, knowing the drivers used, baffle configuration, etc., it's harder to make useful suggestions. But, in a box that small with walls that thick, I doubt what you are hearing is related to bracing.
Do both speakers have similar dimensions? Do both use internal damping/absorption of some kind in useful quantities? Is their polar response similar? Was the speaker location in the room the same for both?
If you are going to measure, it's helpful to collect quasi-anechoic response instead of, or in addition to, in-room response. It involves more details and has limitations but gives you a truer representation of what the speaker is doing on its own. Off-axis measurements may also show you differences between the two.
It can also be revealing to compare a new speaker to a known neutral reference, whether that's a highly regarded speaker or set of headphones. If you are designing for accuracy, you may not really know at this point which speaker is better. Many modern movies and TV shows have awful speech intelligibility. Ultimately, if you are just designing for enjoyment the one with clearer dialog may win out, even if it's less technically accurate.
Do both speakers have similar dimensions? Do both use internal damping/absorption of some kind in useful quantities? Is their polar response similar? Was the speaker location in the room the same for both?
If you are going to measure, it's helpful to collect quasi-anechoic response instead of, or in addition to, in-room response. It involves more details and has limitations but gives you a truer representation of what the speaker is doing on its own. Off-axis measurements may also show you differences between the two.
It can also be revealing to compare a new speaker to a known neutral reference, whether that's a highly regarded speaker or set of headphones. If you are designing for accuracy, you may not really know at this point which speaker is better. Many modern movies and TV shows have awful speech intelligibility. Ultimately, if you are just designing for enjoyment the one with clearer dialog may win out, even if it's less technically accurate.
Last edited:
Both speakers are similar in size. The better sounding one has a 1.5" baffle, and rather large round overs on the sides. The woofer is a Peerless 5.25" HDS fiberglass cone. The other is a Peerless 6.5" HDS poly cone. Both boxes have a lot of firm type fiberglass inside. I think the better sounding one has a layer of thin stick on damping material, but I'm not sure. Both were played in the same location, I think within about 6" anyway. The x-overs are different. The off axis is good on both, but probably better on the 5.25" when measured past 45 degrees. I'm sitting about 15 degrees off axis.
I use gated measurements. I suspect that above 300hz, they measure very close. I think the issue I'm hearing is below 300hz, but that's right in the range of the deeper vocals. If the panels have a resonance there, that could be what I hear. I can add some bracing, or build a better box, but if the woofer cone material, or something else is the reason, I'd like to know before I proceed. The boxes are small, and fitting a x-over inside is a challenge. Adding bracing will make it an even tighter fit. An external x-over is an option if necessary, but I would prefer internal. On the other hand, I could just use the better speaker as my center, but it leaves me with a second matching speaker that goes to waste. Truth is, I'd like to know why they sound different.
I use gated measurements. I suspect that above 300hz, they measure very close. I think the issue I'm hearing is below 300hz, but that's right in the range of the deeper vocals. If the panels have a resonance there, that could be what I hear. I can add some bracing, or build a better box, but if the woofer cone material, or something else is the reason, I'd like to know before I proceed. The boxes are small, and fitting a x-over inside is a challenge. Adding bracing will make it an even tighter fit. An external x-over is an option if necessary, but I would prefer internal. On the other hand, I could just use the better speaker as my center, but it leaves me with a second matching speaker that goes to waste. Truth is, I'd like to know why they sound different.
I was incorrect when I said they measure the same above 300hz. The center with the 6.5" woofer has very minimal BSC. The other 5.25" speaker is pretty much full BSC. It's unexpected, but the full BSC is the one that sounds correct. The speaker sits on a 15" stand, and the woofers are about 21" off the floor.
Again, something of this magnitude should show up in the measured response if the measurements are accurate and quasi-anechoic.
If you have a graphic equalizer at your disposal, you can adjust the response of the center channel speaker to more closely mimic that of the main speakers and see if that resolves the issue. Or just adjust frequency response to taste based on listening tests, and then try to implement that response in your crossover.
From a distance and without seeing more data, we're speculating based on what you say and our general opinions about how to do things. My suspicion is that you are dealing with a voicing issue. It only takes a couple dB difference in the midrange to make certain voices sound subdued. If you're picky about these things voicing a speaker can take significant time. Your ears lose their frame of reference pretty quickly, so you can't have extensive listening sessions that are productive. You have to listen to multiple things, adjust the circuit or equalizer, listen some more, and then walk away for hours or a day. You then come back at it with fresh ears and do it all gain. Lather, rinse, repeat.
If you have a graphic equalizer at your disposal, you can adjust the response of the center channel speaker to more closely mimic that of the main speakers and see if that resolves the issue. Or just adjust frequency response to taste based on listening tests, and then try to implement that response in your crossover.
From a distance and without seeing more data, we're speculating based on what you say and our general opinions about how to do things. My suspicion is that you are dealing with a voicing issue. It only takes a couple dB difference in the midrange to make certain voices sound subdued. If you're picky about these things voicing a speaker can take significant time. Your ears lose their frame of reference pretty quickly, so you can't have extensive listening sessions that are productive. You have to listen to multiple things, adjust the circuit or equalizer, listen some more, and then walk away for hours or a day. You then come back at it with fresh ears and do it all gain. Lather, rinse, repeat.
I swear I looked before, and didn't see any EQ applied. But when I just checked, it shows up on the center. I've been swapping equipment around. This AVR had been adjusted previously. Still, it would seem that both speakers received the EQ. I'll measure anyway.
If two speakers with different response received the same EQ, one will sound better...
Here's a close mic of the speaker I liked, and the larger 6.5" center. I changed the high-pass from 60hz to 150hz., and changed the eq at 250hz from -6dB to -3dB. The 120hz high-pass looked pretty close too, but I'm trying this first. The light green trace was the one I preferred, and the red is the tweaked response of my center.
I'll see how this works out.
I'll see how this works out.
Attachments
Last edited:
OK.. Please pretend I deleted the preceding post. I had a rather large wiring mistake. Maybe I'll get it right next time.
This kind of stuff happens to everyone. I've been guilty of realizing a day or two late that I accidentally had an EQ in circuit on a new speaker I was working on. I've also mis-wired crossovers - which is pretty easy when prototyping something and it's all held together with a pile of jumpers.
My livingroom floor is covered with several x-overs connected with jumpers I tried not to move one that was not in use. I moved the wire going to the woofers, and used it as if it was coming from the amp, so I had two low-pass filters connected, which actually explained a lot. My center has no BSC, but it measured like it did because of the other x-over. This one is just the woofers, and correct x-over.
These are close mic measurements. The actual bass would drop a few feet away.
Here's a comparison of the low end of these two speakers. Both using the same EQ, which was -6dB at 250hz. A few months ago, I was trying to tame the low end. That's were the EQ came from.
What is strange, is that the louder trace is the one that sounded natural in the vocals.
These are close mic measurements. The actual bass would drop a few feet away.
Here's a comparison of the low end of these two speakers. Both using the same EQ, which was -6dB at 250hz. A few months ago, I was trying to tame the low end. That's were the EQ came from.
What is strange, is that the louder trace is the one that sounded natural in the vocals.
Attachments
Last edited:
Here's the center with no EQ. The high-pass for all remains at 60hz. I've not listened yet, but this is what I'll try. I made the x-over for this center with no BSC knowing that previous centers tended to have a room / location resonance peak around 250hz.
Above 1k the speakers should measure pretty close.
Above 1k the speakers should measure pretty close.
Attachments
Last edited:
The floor has an effect. I have about four AVRs that I use in two rooms. The setups vary, and I have more than one center in the works. The EQ available in this AVR is crude. Basically, I have the 250hz band .. The next band is 800hz. One AVR only has a bass knob. I move things around often because I'm always working on projects. It was just a surprise that my 5.25" main, shown in green above, produced the sound I was looking for. I'm hoping the non EQed ( purple ) response of the 6.5" will sound very similar. I suppose that I could make changes to the centers x-over to mimic the low end of the main, and keep the EQ at 250hz. I think it should sound the same. I could do that. It would be nice to not need the EQ though.
I'm also trying to make a center that can handle the loudest of action movies, so a higher x-over point than my mains have would be an advantage. My mains are at 2k, and shallow. I'm shooting for close to 3k on this center.
I'm also trying to make a center that can handle the loudest of action movies, so a higher x-over point than my mains have would be an advantage. My mains are at 2k, and shallow. I'm shooting for close to 3k on this center.
Last edited:
After reassembling the x-over, I decided to make some changes. I've not listened yet, but will tonight.
This is the response at about 20 degrees off axis, which is where my seats typically are. The x-over is at 2.5k, which should be safe for the tweeter with action movies. The x-over is 6 parts, so it should fit in the box without too much hassle. Maybe I can even add a brace between the sides.
This is the response at about 20 degrees off axis, which is where my seats typically are. The x-over is at 2.5k, which should be safe for the tweeter with action movies. The x-over is 6 parts, so it should fit in the box without too much hassle. Maybe I can even add a brace between the sides.
Attachments
Last edited: