Having the drivers center lines offset in the baffle will give uneven polar response. Polar response effects the in-room power response, having flat (or slightly tilted down) power response means the speakers will interact with the room less.
Your drivers are different to mine so my XO/EQ will not necessarily work with your setup.
How I derived baffle shape EQ and XO
Steps
1 EQ individual driver on axis for flat frequency response (FR).
2 Measure individual driver on and off axis FR.
3 Adjust baffle shape to give smoothest off axis FR. Ideally off axis FR plot should be identical to on axis but down in level. My setup is good until about 6kHz.
4 Select XO to splice on and off axis FR for flatness. On axis is easy so I concentrated on splicing 45deg off axis for flattest response.
Your drivers are different to mine so my XO/EQ will not necessarily work with your setup.
How I derived baffle shape EQ and XO
Steps
1 EQ individual driver on axis for flat frequency response (FR).
2 Measure individual driver on and off axis FR.
3 Adjust baffle shape to give smoothest off axis FR. Ideally off axis FR plot should be identical to on axis but down in level. My setup is good until about 6kHz.
4 Select XO to splice on and off axis FR for flatness. On axis is easy so I concentrated on splicing 45deg off axis for flattest response.
Thanks for your reply.
I was only focus on the open baffle lower cutoff frequency and diffraction. Diffraction software show that it is better to offset...
As I crossover at 400hz I am not sure if I need such a big baffle...
For the time being I'd like to find a nice shape that would optain nice baffle step and diffraction results.
I might try the linkwith shape or the truncated pyramid...with centered speakers as mentioned.
Thanks again
I was only focus on the open baffle lower cutoff frequency and diffraction. Diffraction software show that it is better to offset...
As I crossover at 400hz I am not sure if I need such a big baffle...
For the time being I'd like to find a nice shape that would optain nice baffle step and diffraction results.
I might try the linkwith shape or the truncated pyramid...with centered speakers as mentioned.
Thanks again
Hi, yeah some asymmetric shape could be utilized to reduce the ripple of diffraction at listening position.
Daily thought experiment:
If you want to balance the diffraction and "baffle cutoff" the simplified thought process is that as you enlarge the baffle the diffraction effects get lower in frequency forcing you to simultaneously move crossover lower in frequency in order to avoid the diffraction, which goes lower and lower in frequency as you enlarge the baffle. If you think about the situation now, with enlarged baffle, you have small driver crossed low, which is bad for SPL capability especially here with open baffle where we have falling lows anyway and in order to get the few octaves (max bandwidth) per driver you would want and need all the transducer surface area possible to reduce excursion. Basically, we want to push the bandwidth as high as possible not as low as possible to retain SPL capability. This is exact opposite what you try achieve with wider baffle and demands no baffle at all. In addition large baffle probably resonates and emits sound, its surface area is now lot bigger than the driver and the contribution might be large. From this simple thought experiment logical conclusion is to use (enough) big drivers and no baffles at all for best performance all around.
At least, you could try with minimal / no baffle at all and then you would find out:
- if the wide baffle is really required to augment lows
- if the diffraction is something to worry about
- if you find out diffraction was audible and not the sound is clean but missing lows, just add more drivers / amplifier / EQ.
Hope it helps
Exactly, after all it is what you want to achieve / optimize. Every system is just bunch of compromises. If there is set of drivers that you need to use then perhaps there is optimal baffle size that would be good compromise for overall performance even if it ment little bit of diffraction or some other small issue. If you want to minimize effect of diffraction with the shape make it so that distance to any edge from center of the cone is varied. Basically almost any shape with offset driver. You could offset the baffle(s) and keep the drivers inline. I'd just use minimal baffle and take the max SPL that the system is capable of. SPL capability reduces as the drivers are asked to do things they are not comfortable with. If SPL capability is not enough then sell the drivers and buy set of better suited ones For sure acoustic design would be optimized with minimal / no baffle and crossover each driver below diffraction effects. Use EQ to boost the low band of each way, instead of wide baffle, and use enough ways with suitable transducer sizes.
This is why buying drivers is the last step for speaker design, not the first like often seen. If buying drivers first you are stuck with design space they allow and this will impose limit to performance (what ever that is). As long as everything is open, nothing purchased, assuming everything is available and affordable, every crossover is doable and amplification is plenty, one is able to pencil in very high performance system and it doesn't have to be very costly or use trendy drivers ("high end") but it certainly can. In this regard the system design is more important than the drivers it ends up using. If starting with a set of drivers you might notice you'd need more or different size drivers, or perhaps very high spec drivers if required to work at limits and if this is not possibility there is no option but choose the best compromise and accept the performance you get, best you can have with the drivers. Or what ever.
I'm sorry I don't know all your requirements, wishes and limitations regarding the system so all I can do is to help with the thought process how to approach a design and how to take it further. I've seen your photos and you try various baffles, this is very good thing you'll learn a ton and I'm sure you end up with very good system eventually, what ever that is. Just have fun with the project!
As disclaimer I've never built open baffle speaker so don't know what would be good set. See this gallery what people have done, dipoluser system on previous page seems nice, enough drivers, no baffle other than what it takes to support the structure. Hope it helps!
For sure acoustic design would be optimized with minimal / no baffle and crossover each driver below diffraction effects. Use EQ to boost the low band of each way, instead of wide baffle, and use enough ways with suitable transducer sizes.This is why buying drivers is the last step for speaker design, not the first like often seen. If buying drivers first you are stuck with design space they allow and this will impose limit to performance (what ever that is). As long as everything is open, nothing purchased, assuming everything is available and affordable, every crossover is doable and amplification is plenty, one is able to pencil in very high performance system and it doesn't have to be very costly or use trendy drivers ("high end") but it certainly can. In this regard the system design is more important than the drivers it ends up using. If starting with a set of drivers you might notice you'd need more or different size drivers, or perhaps very high spec drivers if required to work at limits and if this is not possibility there is no option but choose the best compromise and accept the performance you get, best you can have with the drivers. Or what ever.
I'm sorry I don't know all your requirements, wishes and limitations regarding the system so all I can do is to help with the thought process how to approach a design and how to take it further. I've seen your photos and you try various baffles, this is very good thing you'll learn a ton and I'm sure you end up with very good system eventually, what ever that is. Just have fun with the project!
As disclaimer I've never built open baffle speaker so don't know what would be good set. See this gallery what people have done, dipoluser system on previous page seems nice, enough drivers, no baffle other than what it takes to support the structure. Hope it helps!
Last edited:
vessk0 wrote:
My baffle is almost 125cm wide (including the wings) and the roll-off of my Supravox 165GMF is happening at 200Hz.
Measured nearfield as a full range with no XO:
Sorry, but nearfield measurement doesn't tell anything about open baffle effect. With baffle that wide, mic distance should be roughly 2m.
Open baffles should also have more free space around to let the backside wavefront to meet frontside.
http://gainphile.blogspot.com/2011/01/measuring-open-baffle-loudspeakers.html
My baffle is almost 125cm wide (including the wings) and the roll-off of my Supravox 165GMF is happening at 200Hz.
Measured nearfield as a full range with no XO:
Sorry, but nearfield measurement doesn't tell anything about open baffle effect. With baffle that wide, mic distance should be roughly 2m.
Open baffles should also have more free space around to let the backside wavefront to meet frontside.
http://gainphile.blogspot.com/2011/01/measuring-open-baffle-loudspeakers.html
Last edited:
With everything said this morning I reduced my baffle in a smaller linkwith like shape.
Nice results so far.
I understand what you are saying that rear wave matters. Otherwise we would go sealed and that's it...
But then questions are asked :
-how do you know good placement?
-do you treat rear wall with absorbant, diffraction panels...?
Regards Greg
Nice results so far.
I understand what you are saying that rear wave matters. Otherwise we would go sealed and that's it...
But then questions are asked :
-how do you know good placement?
-do you treat rear wall with absorbant, diffraction panels...?
Regards Greg
Attachments
^nice!
Some thoughts regarding the placement and treatment questions, because you are getting into the nugget of it all, thinking about the whole speaker system including the room
For this you need to study room acoustics, or perhaps first reflections, how hearing system works, or perhaps how studios do it to hear what is happening at the recorded sound. I've got no links with me to the research papers but as you start investigate there are some stuff like precedence effect, Haas effect, all kinds of stuff going on that affect how we perceive the reflections what your question eventually is. From the studies you'll get some numbers and ideas how to think about it, basically the speaker and the room work together, this is what you listen at, sound in room.
I've gathered few simplified numbers and ideas to start, and you can go deeper by doing some research yourself:
Few things one can do:
Basically, controlled directivity on the speakers and positioning of the speakers and listening spot relative to room boundaries would affect all of this, as well as room acoustic treatment. Your OB speakers have strong nulls to sides and up / down, and with juggling positioning and toe-in you can get nice attenuation towards first early reflections and this increases relation of direct sound to reflected sound. Ideal dipole are constant directivity devices but diffraction and real world driver (baskets) and what not would probably change this some, so sound toward all first reflections might not be exactly what the direct sound is but perhaps quite close. You asked about the back wave and for this you might be able to get it's path length, which requires two reflections to reach the listening spot if toed-in assuming square room, so that it can be >15ms contributing just nice reverberation and not altering timbre or other aspects of the sound. Perhaps 10ms is fine, or something.
Sound travels roughly 1ms per foot or 3ms per meter. 15ms delay means about 5 meters or 15 feet path length difference relative to direct sound. This could happen with the back wave if the speakers are well out from the front and side wall, attenuation of this reflection depends on the toe in. Basically mind what the first reflections are as well as proper stereo listening triangle and what the main design listening axis is and its response. You might want to iterate with this, first figure out good positioning and toe-in, figure out the angle of the listening axis and then optimize speaker response to that axis (instead of typical on-axis, which you most probably won't sit at after the juggling). Just making the speaker response nice to all directions (nice power response) and minding about the first reflections (response, attenuation, delay) would be very fine start and then tune if you find out aspects that need tuning.
For room acoustic treatment, I'm not sure what is appropriate, personally I don't have anything else than typical living room furniture and decor + controlled directivity speaker and it feels fine, not too long or obvious reverberation. For OB as controlled and rather narrow directivity speaker I'm not sure if you have to treat anything else than large flat surfaces to kill flutter echo and excessively long reverberation times. If you have dedicated listening space you can go far with the acoustics but for living room I'm not sure if there is too much need as long as the loudspeakers have controlled directivity and good positioning is possible and found. I'm not expert on this, just the stage my knowledge is at and you should do your own research on it. Hope it helps, have fun!
edit. if you haven't checkout Floyd Toole book / studies about the stuff, and checkout dr. Geddes thoughts on the subject too. You might have to figure out what you prefer, more or less early reflections as they affect on "clarity" and perhaps "envelopment" and "sound stage" and other qualities that might be hard if not impossible to achieve all together. Fortunately changing speaker positioning and toe-in is free, just take your time!
Some thoughts regarding the placement and treatment questions, because you are getting into the nugget of it all, thinking about the whole speaker system including the room
For this you need to study room acoustics, or perhaps first reflections, how hearing system works, or perhaps how studios do it to hear what is happening at the recorded sound. I've got no links with me to the research papers but as you start investigate there are some stuff like precedence effect, Haas effect, all kinds of stuff going on that affect how we perceive the reflections what your question eventually is. From the studies you'll get some numbers and ideas how to think about it, basically the speaker and the room work together, this is what you listen at, sound in room.
I've gathered few simplified numbers and ideas to start, and you can go deeper by doing some research yourself:
- early reflections happen roughly 15ms before (typical room) goes into diffuse reverberation, basically this is number from speed of sound and room dimensions. Within this time is also many studio recording space cues if I remember correctly, room sound entering the mic.
- any reflection happening within this time should perhaps sound much the same as direct sound in order not to color the sound.
- any reflection happening within this time more or less makes the "room sound" coloring what is on the recording.
Few things one can do:
- attenuate the first reflections
- while making sure they would sound as similar to direct sound as feasible
- delay the reflections as much as possible, possibly out from the first 5ms and if blessed with big enough dedicated listening space clear out the 15ms altogether (look for Reflection Free Zone studio design stuff).
Basically, controlled directivity on the speakers and positioning of the speakers and listening spot relative to room boundaries would affect all of this, as well as room acoustic treatment. Your OB speakers have strong nulls to sides and up / down, and with juggling positioning and toe-in you can get nice attenuation towards first early reflections and this increases relation of direct sound to reflected sound. Ideal dipole are constant directivity devices but diffraction and real world driver (baskets) and what not would probably change this some, so sound toward all first reflections might not be exactly what the direct sound is but perhaps quite close. You asked about the back wave and for this you might be able to get it's path length, which requires two reflections to reach the listening spot if toed-in assuming square room, so that it can be >15ms contributing just nice reverberation and not altering timbre or other aspects of the sound. Perhaps 10ms is fine, or something.
Sound travels roughly 1ms per foot or 3ms per meter. 15ms delay means about 5 meters or 15 feet path length difference relative to direct sound. This could happen with the back wave if the speakers are well out from the front and side wall, attenuation of this reflection depends on the toe in. Basically mind what the first reflections are as well as proper stereo listening triangle and what the main design listening axis is and its response. You might want to iterate with this, first figure out good positioning and toe-in, figure out the angle of the listening axis and then optimize speaker response to that axis (instead of typical on-axis, which you most probably won't sit at after the juggling). Just making the speaker response nice to all directions (nice power response) and minding about the first reflections (response, attenuation, delay) would be very fine start and then tune if you find out aspects that need tuning.
For room acoustic treatment, I'm not sure what is appropriate, personally I don't have anything else than typical living room furniture and decor + controlled directivity speaker and it feels fine, not too long or obvious reverberation. For OB as controlled and rather narrow directivity speaker I'm not sure if you have to treat anything else than large flat surfaces to kill flutter echo and excessively long reverberation times. If you have dedicated listening space you can go far with the acoustics but for living room I'm not sure if there is too much need as long as the loudspeakers have controlled directivity and good positioning is possible and found. I'm not expert on this, just the stage my knowledge is at and you should do your own research on it. Hope it helps, have fun!
edit. if you haven't checkout Floyd Toole book / studies about the stuff, and checkout dr. Geddes thoughts on the subject too. You might have to figure out what you prefer, more or less early reflections as they affect on "clarity" and perhaps "envelopment" and "sound stage" and other qualities that might be hard if not impossible to achieve all together. Fortunately changing speaker positioning and toe-in is free, just take your time!
Last edited:
The lower mid/woofer the uncorrected plot is 2 pages eariler, I didn't keep the uncorrected plot for the upper mid NEO8.
This is a good read on the design process of a loudspeaker
https://www.linkwitzlab.com/Fitz/acoustics-mechanics.htm
I’m working on a wooden version (originally in cardboard) of my slot loaded open baffle dipole speaker.
More info in thread here:
https://www.diyaudio.com/community/threads/cheap-and-fast-ob-literally.249984/post-6971724
More info in thread here:
https://www.diyaudio.com/community/threads/cheap-and-fast-ob-literally.249984/post-6971724
this is a variant on the Pass design with more H frame approach I like it. Have you tested the final version yet?