That's not my project- just that the measures around 1khz make a pretty clear cut case of a smoother response from the unit with the edge treatment, and that's consistent with use of edge treatment and roundovers, particularly off-axis. With good edge treatment, you get smoother polars. I've experienced this with materials on flat baffles, but just as much so with horns and foam treatments. Different materials and profiles do get different results, but so long as you're thoughtful in your use, you can make some pretty significant improvements that help eliminate localization cues for the loudspeaker enclosure.
@homebuilder:
The larger cone of the 15" provides a directivity advantage as well as meaning the horn (cone) is moving less for the treble for a given SPL. The measured responses are superb on the DIYSG polars, equivalent to some very good horn/waveguide 2 ways (only the top octave gets very slightly wonky).
In my opinion, for what I'd expect from straight sound quality, that 15" is a straight upgrade to the JBL solution, which was previously one of the best for coaxial continuity and directivity.
Last edited:
Sure would be simple for you to post a measurement or two. How about it?
Gotta say, 90% of the time I post that at DIYaudio, I get a "dog ate my homework" response.*
BTW, you may be confusing the benefits of toning down a horn (often a good move) with diffraction effects.
B.
* sometimes "my grandmother was sick" and so I lost the data
Here are 2 measurements taken during the optimization process for a foam collar on the tweeter mentioned. 2 variants of collar.
The only difference was a collar that was applied around the periphery of the coverage pattern.
Last edited:
And, the base case. Notable measureable differences from simply applying and adjusting absorption around the mouth of a horn.
This is a selenium HM17-25 horn that was used in the testing with a soft dome tweeter applied. The small horn starts losing pattern control relatively high in frequency, and so absorption around the mouth does a lot.
These are as consistent a polar testing as I've produced, but I'm not a measurement master. Little tweaks to the collar were easy to see and replicate.
This is a selenium HM17-25 horn that was used in the testing with a soft dome tweeter applied. The small horn starts losing pattern control relatively high in frequency, and so absorption around the mouth does a lot.
These are as consistent a polar testing as I've produced, but I'm not a measurement master. Little tweaks to the collar were easy to see and replicate.
Last edited:
As I said in my initial comments, from which the conversation steadily devolved, diffraction is highly audible but you can't deal with it in isolation from the rest of the speaker.
Baffles have resonances that tend to fall in areas that are highly audible like 1-2khz. Yes that's audible especially in a small room or up close. A small baffle will sound sharper and more shrill, there is no way around this except maybe spherical shapes.
One of the most interesting ways of getting rid of both baffle and cone resonances is to have the tweeter or small driver facing backwards. This also helps deal with the problem of early reflections since the it increases distance before you hear highs.
Some might say you will hurt imaging or mess up phase but these are nothing compared to baffle and cone resonances which are like a "rock in the shoe" and get very annoying.
Baffles have resonances that tend to fall in areas that are highly audible like 1-2khz. Yes that's audible especially in a small room or up close. A small baffle will sound sharper and more shrill, there is no way around this except maybe spherical shapes.
One of the most interesting ways of getting rid of both baffle and cone resonances is to have the tweeter or small driver facing backwards. This also helps deal with the problem of early reflections since the it increases distance before you hear highs.
Some might say you will hurt imaging or mess up phase but these are nothing compared to baffle and cone resonances which are like a "rock in the shoe" and get very annoying.
I must say, that I strongly disagree with 33Polkhigh about the audibility of edge diffractions, in post #566
Diffractions are unavoidable unless we use infinite baffle, ball enclosure or true dipole (A ball will suffer from baffle loss in low freg, but smoothly without narrow band diffractions.) With box speakers diffractions easily get distributed and this way also attenuated.
I have done and measured normal boxes, a dipole and a ball enclosure. Only the dipole sounds different and identifieable in a room or even outdoors (supposing they have identical response axially and don't have gross directivity problems).
A wide curved baffle is something that I would like to make, measure and listen to some day...
Diffractions are unavoidable unless we use infinite baffle, ball enclosure or true dipole (A ball will suffer from baffle loss in low freg, but smoothly without narrow band diffractions.) With box speakers diffractions easily get distributed and this way also attenuated.
I have done and measured normal boxes, a dipole and a ball enclosure. Only the dipole sounds different and identifieable in a room or even outdoors (supposing they have identical response axially and don't have gross directivity problems).
A wide curved baffle is something that I would like to make, measure and listen to some day...
Last edited:
I must say, that I strongly disagree with 33Polkhigh about the audibility of edge diffractions, in post #566
Diffractions are unavoidable unless we use infinite baffle, ball enclosure or true dipole (A ball will suffer from baffle loss in low freg, but smoothly without narrow band diffractions.) With box speakers diffractions easily get distributed and this way also attenuated.
I have done and measured normal boxes, a dipole and a ball enclosure. Only the dipole sounds different and identifieable in a room or even outdoors (supposing they have identical response axially and don't have gross directivity problems).
A wide curved baffle is something that I would like to make, measure and listen to some day...
Let's just start with a simple full range driver in a box. Full rangers are great for learning. Then try it in a larger box.
The size and shape of the box changes the tone of the speaker. I use computer based dsp, and you can't equalize out the tonal difference. The reason is that baffle step shifts the ratio of sound to being more directional.
The issue isn't so much that diffraction is the end of the world. But that if diffraction and cone break up all fall in typical baffle step range of 1-4 khz it creates a messy FR. A good designer can deal with it.
Last edited:
Why would you expect otherwise?
What you're really doing is varying the size/shape of the waveguide and then wondering why DSP EQ (presumably of the axial response) doesn't affect directivity.
Its not entirely about FR. Because a bigger speaker will start becoming directional at a lower frequency. When this happens the sound is emphasized and we hear it differently
This goes back to my initial point that one doesn't want to have the baffle step shift in an area of cone break up. Big baffles help with this but they are better from a distance.
Baffle diffraction
33Polkhigh:
According to your theory concerning audible diffraction, would a driver mounted in a very large baffle, ie a wall, exhibit considerably better performance in a large room? Would it make sense that a large driver would realize more benefit from this type of mounting than a small driver? Say, 15" versus 6"-8" units?
Sorry if some of us have high-jacked parts of your thread, but the reality is that many posts on this forum have value to others in ways that may not seem readily apparent. Especially when one is trying to learn from others, and then apply that to their own situation. With that said:
Badman: "The larger cone of the 15" provides a directivity advantage as well as meaning the horn (cone) is moving less for the treble for a given SPL. The measured responses are superb on the DIYSG polars, equivalent to some very good horn/waveguide 2 ways (only the top octave gets very slightly wonky).
In my opinion, for what I'd expect from straight sound quality, that 15" is a straight upgrade to the JBL solution, which was previously one of the best for coaxial continuity and directivity."
Badman: Thank you for your response on the DIYSG coaxial. Based on your comment, would it seem like a reasonable idea to mate that driver up with JBL 2404 bi-radial crossed over at about 7kh? (I have 4 of these new in box) This would be a very easy, and elegant solution for my room. And the high efficiency of the drivers would allow me lots of options for amplifiers. to remind, you, these would be flush mounted in the end wall of a 24' x 36' room, off-center on the shorter wall.
__________________
33Polkhigh:
According to your theory concerning audible diffraction, would a driver mounted in a very large baffle, ie a wall, exhibit considerably better performance in a large room? Would it make sense that a large driver would realize more benefit from this type of mounting than a small driver? Say, 15" versus 6"-8" units?
Sorry if some of us have high-jacked parts of your thread, but the reality is that many posts on this forum have value to others in ways that may not seem readily apparent. Especially when one is trying to learn from others, and then apply that to their own situation. With that said:
Badman: "The larger cone of the 15" provides a directivity advantage as well as meaning the horn (cone) is moving less for the treble for a given SPL. The measured responses are superb on the DIYSG polars, equivalent to some very good horn/waveguide 2 ways (only the top octave gets very slightly wonky).
In my opinion, for what I'd expect from straight sound quality, that 15" is a straight upgrade to the JBL solution, which was previously one of the best for coaxial continuity and directivity."
Badman: Thank you for your response on the DIYSG coaxial. Based on your comment, would it seem like a reasonable idea to mate that driver up with JBL 2404 bi-radial crossed over at about 7kh? (I have 4 of these new in box) This would be a very easy, and elegant solution for my room. And the high efficiency of the drivers would allow me lots of options for amplifiers. to remind, you, these would be flush mounted in the end wall of a 24' x 36' room, off-center on the shorter wall.
__________________
I'm not sure you even need the 2404 with that coax. It's not perfect in the top octave, but it's not bad at all.
I'm not sure if I can convince you of something that is perceptual. But if you were talking to a person in a very reflective room. And they turned their back to you or moved around their voice would change. More direct sound that is closer is more intense than if it is reflected.
Sure, i don't think any of us remember who's thread this is. My initial point about diffraction was that you can use the ripple to your advantage with certain baffle shapes. You put a diffraction "dip" where a cone peak is. This seems almost common sense to me.
As far as infinite baffles. I do like them from what I've heard. They don't have that typical baffle step transition that creates problems with small baffles. That said its hard to say whether what you want to do will sound good. A 15" driver will be directional fairly low in frequency.
Whose thread? The DIYaudio community! The idea of a threadstarter "owning" a thread is silly IMO.
Regarding using the diffraction that way, that's all well and good on a few angles. A few specific angles. The only consistent solution is to create smooth transitions, use absorption, and minimize baffle artifacts, be it broadband ripple or discrete reflections.
Directional low in frequency is generally the goal for many of us. Very few have big rooms to use so we tend to go for directional solutions to limit room interaction in the mid/HF.
Then you'd lose the direct arriving higher frequencies. Not a clear comparison for speakers. Also dependent on response.
Some angles and frequencies matter much more than others. A direct sound peak in the 1-2 khz region can ruin a speaker imo. A simple roundover can eliminate any diffraction in the 8-16 khz, but there isn't much content up there anyways.
I have been away for awhile, but I'd like to comment on the discussion several pages back.
The Toole reference was interesting, but I don't think that it is evidence that baffle diffraction is not audible.
First, while the infinite baffle shows only small difference from the unbaffled case, the other example shows significant differences. In this second case, the aberrations are baffle diffraction as well, so the two examples seem to show contradictory things, but neither is conclusive as regards the audibility of diffraction.
Second, and this is the most important, there is no reason to believe that a small visual aberration means that the perception is equally small. It may well be that small diffraction aberrations are far more audible that the FR differences would lead one to believe (my research shows that.) I think that it is important to note that diffraction effects are always delayed in time and that can make a huge difference in perception.
Lastly I want to clarify my comment about open baffle diffraction. In thinking it through, I believe that an open baffle will have significant diffraction, in general, but because of symmetry, it may not show up as a large FR aberration on axis. But since the diffracted wave emanates from a very large source, the baffle edge, it will have a much different polar response than the un-diffracted wave. This means that the effect could be quite large in some places and much smaller in others. The more I thought about this issue the more complicated it became and I want to go back to basics and search out the exact nature of the diffraction from a disk (it can be solved in closed form.) This will help me to better understand the importance of diffraction from an open baffle as it relates to perception.
I will also add that in my system I have attempted to eliminate diffraction not only from the speakers but from nearby room components as well. IMO this has paid big dividends in perception, especially in "image". "Tone" to a far lessor degree.
The Toole reference was interesting, but I don't think that it is evidence that baffle diffraction is not audible.
First, while the infinite baffle shows only small difference from the unbaffled case, the other example shows significant differences. In this second case, the aberrations are baffle diffraction as well, so the two examples seem to show contradictory things, but neither is conclusive as regards the audibility of diffraction.
Second, and this is the most important, there is no reason to believe that a small visual aberration means that the perception is equally small. It may well be that small diffraction aberrations are far more audible that the FR differences would lead one to believe (my research shows that.) I think that it is important to note that diffraction effects are always delayed in time and that can make a huge difference in perception.
Lastly I want to clarify my comment about open baffle diffraction. In thinking it through, I believe that an open baffle will have significant diffraction, in general, but because of symmetry, it may not show up as a large FR aberration on axis. But since the diffracted wave emanates from a very large source, the baffle edge, it will have a much different polar response than the un-diffracted wave. This means that the effect could be quite large in some places and much smaller in others. The more I thought about this issue the more complicated it became and I want to go back to basics and search out the exact nature of the diffraction from a disk (it can be solved in closed form.) This will help me to better understand the importance of diffraction from an open baffle as it relates to perception.
I will also add that in my system I have attempted to eliminate diffraction not only from the speakers but from nearby room components as well. IMO this has paid big dividends in perception, especially in "image". "Tone" to a far lessor degree.
Last edited:
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.