Steve/BlueWizard :
I went back and re read the entire thread and I think I know why you are having trouble with the responses to your questions. You appear to be looking for the speaker builders checklist. First you do step one, then step 2, etc until you arrive at step 17 and the speaker is done.
Unfortunately speaker design still has as much art as it does science. i.e there is no right answer, only a series of answers each of which violates some part of the desired result.
The actual process is much more iterative. You pick a pair or three drivers that you think might work. You decide on a set of crossover centers and slopes. Now you model this using some software or even a pencil and paper resulting in a list of predicted performance parameters. In this case looking specifically at "overlap" vs response graphs from the driver manufacturer and then decide if this meets your needs.
We all want a good speaker, small with minimal fuss and low cost that sounds great. Nope, can't have that! One or more of the above needs to be forgiven. You can pay more for better drivers or more drivers. You can increase crossover complexity with steeper slopes and/or traps, you can make a more exotic cabinet or choose a more complex bass alignment. Or of course you can settle for less than the best available sound. Demanding a design that is able to fulfill all the requirements will result in a complex speaker that's far from simple or cheap.
You seem to be somewhat frustrated that none of the responses to your post can tell you how to do this in 25 words or less. Whole books have been written on how this done, some are even designed to be useful to a speaker design newbie, Vance Dickason's, Loudspeaker Design Cookbook is probably the most common example.
If you read some of the group speaker design threads elsewhere on this site, you will see how various trade offs were made for a given design. Everyone of the designs on this site has it's own strengths and weaknesses and much discussion within the thread about the design before it's built. Maybe one of those projects is close to what you want, you can copy their work, that's one reason its here. The other reason is that even seasoned builders can use help making the design decisions.
This site is still very useful for picking a set of drivers
http://ldsg.snippets.org/
The site provides background on various designs, a look at what kind of drivers are available, a list of known decent drivers at several price points and some insights about the process in general. It is dated, but mostly still true. A lot of effort went into that site.
I went back and re read the entire thread and I think I know why you are having trouble with the responses to your questions. You appear to be looking for the speaker builders checklist. First you do step one, then step 2, etc until you arrive at step 17 and the speaker is done.
Unfortunately speaker design still has as much art as it does science. i.e there is no right answer, only a series of answers each of which violates some part of the desired result.
The actual process is much more iterative. You pick a pair or three drivers that you think might work. You decide on a set of crossover centers and slopes. Now you model this using some software or even a pencil and paper resulting in a list of predicted performance parameters. In this case looking specifically at "overlap" vs response graphs from the driver manufacturer and then decide if this meets your needs.
We all want a good speaker, small with minimal fuss and low cost that sounds great. Nope, can't have that! One or more of the above needs to be forgiven. You can pay more for better drivers or more drivers. You can increase crossover complexity with steeper slopes and/or traps, you can make a more exotic cabinet or choose a more complex bass alignment. Or of course you can settle for less than the best available sound. Demanding a design that is able to fulfill all the requirements will result in a complex speaker that's far from simple or cheap.
You seem to be somewhat frustrated that none of the responses to your post can tell you how to do this in 25 words or less. Whole books have been written on how this done, some are even designed to be useful to a speaker design newbie, Vance Dickason's, Loudspeaker Design Cookbook is probably the most common example.
If you read some of the group speaker design threads elsewhere on this site, you will see how various trade offs were made for a given design. Everyone of the designs on this site has it's own strengths and weaknesses and much discussion within the thread about the design before it's built. Maybe one of those projects is close to what you want, you can copy their work, that's one reason its here. The other reason is that even seasoned builders can use help making the design decisions.
This site is still very useful for picking a set of drivers
http://ldsg.snippets.org/
The site provides background on various designs, a look at what kind of drivers are available, a list of known decent drivers at several price points and some insights about the process in general. It is dated, but mostly still true. A lot of effort went into that site.
Sy, nice to see you back in the fray.
Isn't it normal for sound to become more directional as the frequency increase? That is, isn't this a characteristic of frequency independent of drivers? Though, I certainly don't deny that speakers compound this effect.
Thanks to PigletsDad for adding some interesting info to the discussion, though a lot of it was covered before.
As to speakers in break up mode, as I said in my previous post, to the extent that it is possible, we want those irregularities and peaks substantially attenuated when we reach them. The only way to do that is to crossover sufficiently below those aberrations or to somehow trap them with additions to the crossover.
As to rated frequency response being meaningless, I've already addressed this. Rated frequency response isn't meaningless, it's just of very limited value. It can be used to make a very preliminary 'possible/impossible' judgment call on a given speaker.
Again, I want to caution people to keep this discussion in the proper context. I'm not saying, and have never said, that 'frequency overlap' is the ultimate and sole bit of information that is ever considered. I'm saying, it is a rule of thumb that allows you to narrow the field to possible speakers early on. To determine which speakers warrant further investigation and which do not.
Once you've eliminated the 'possible' from the 'impossible', then you need to make other considerations. I have never denied that those other considerations needed to be made...eventually. I've only said, that in the context of my question, I'm not at 'eventually' yet.
Still, I do appreciate the contribution you've made to the discussion.
I'm tempted to extend the discussion with a real world example. Where I lay out a basic design concept, and we work through the design process step by step outlining the considerations that need to be made at each step along the way. We wouldn't have to get into the precise details of actually design (I'm not trying to get a free speaker design), only address the considerations that need to be made along the way.
I think it would be very educational for myself and for every new and older DIY'er here.
But, I fear I've already frustrated you all enough.
A genuine thanks to all who have responded.
Steve/bluewizard
Isn't it normal for sound to become more directional as the frequency increase? That is, isn't this a characteristic of frequency independent of drivers? Though, I certainly don't deny that speakers compound this effect.
Thanks to PigletsDad for adding some interesting info to the discussion, though a lot of it was covered before.
As to speakers in break up mode, as I said in my previous post, to the extent that it is possible, we want those irregularities and peaks substantially attenuated when we reach them. The only way to do that is to crossover sufficiently below those aberrations or to somehow trap them with additions to the crossover.
As to rated frequency response being meaningless, I've already addressed this. Rated frequency response isn't meaningless, it's just of very limited value. It can be used to make a very preliminary 'possible/impossible' judgment call on a given speaker.
Again, I want to caution people to keep this discussion in the proper context. I'm not saying, and have never said, that 'frequency overlap' is the ultimate and sole bit of information that is ever considered. I'm saying, it is a rule of thumb that allows you to narrow the field to possible speakers early on. To determine which speakers warrant further investigation and which do not.
Once you've eliminated the 'possible' from the 'impossible', then you need to make other considerations. I have never denied that those other considerations needed to be made...eventually. I've only said, that in the context of my question, I'm not at 'eventually' yet.
Still, I do appreciate the contribution you've made to the discussion.
I'm tempted to extend the discussion with a real world example. Where I lay out a basic design concept, and we work through the design process step by step outlining the considerations that need to be made at each step along the way. We wouldn't have to get into the precise details of actually design (I'm not trying to get a free speaker design), only address the considerations that need to be made along the way.
I think it would be very educational for myself and for every new and older DIY'er here.
But, I fear I've already frustrated you all enough.
A genuine thanks to all who have responded.
Steve/bluewizard
No, it is absolutely a function of driver and baffle geometry. Otherwise, there couldn't be omnis, point sources, line sources, and the rest of the zoo.
The sort of worked examples you're looking for can be found in Dickason. For the price of a moderately good tweeter, you can buy the book and read it. And I really, really would if I were you.
Thanks Sy,
I'm not saying you are wrong, but let's take a purely hypothetical example. Let's say we have an idealized woofer, mid, and tweeter that radiate sound in an absolutely uniform spherical pattern.
And music (or tones) play from these idealize speakers. Isn't there going to be the psycho-acoustic perception that the bass sphere is large, the mid sphere is medium, and the high sphere is small? Isn't it the nature of bass to sound wide and omni-directional, where as high tone tend to sound 'point source'?
Again, I'm not saying, I'm asking.
Appreciate the help.
To Hermanv,
I haven't read the book you recommended, but I have read books on speaker design, as well as studied Physics in college, and have, very much in my wild impetuous youth, designed a few basic speakers.
Two or three modern books seem to keep popping up in discussion. One is the 'Cookbook' by Dickason. Another is 'Introduction to Loudspeaker Design' by Murphy. And, "Designing, Building, and Testing Your Own Speaker System with Projects" by Weems .
Is this a good selection, or are there two or three other books that people would recommend as a starting point for speaker design and testing?
Steve/bluewizard
I'm not saying you are wrong, but let's take a purely hypothetical example. Let's say we have an idealized woofer, mid, and tweeter that radiate sound in an absolutely uniform spherical pattern.
And music (or tones) play from these idealize speakers. Isn't there going to be the psycho-acoustic perception that the bass sphere is large, the mid sphere is medium, and the high sphere is small? Isn't it the nature of bass to sound wide and omni-directional, where as high tone tend to sound 'point source'?
Again, I'm not saying, I'm asking.
Appreciate the help.
To Hermanv,
I haven't read the book you recommended, but I have read books on speaker design, as well as studied Physics in college, and have, very much in my wild impetuous youth, designed a few basic speakers.
Two or three modern books seem to keep popping up in discussion. One is the 'Cookbook' by Dickason. Another is 'Introduction to Loudspeaker Design' by Murphy. And, "Designing, Building, and Testing Your Own Speaker System with Projects" by Weems .
Is this a good selection, or are there two or three other books that people would recommend as a starting point for speaker design and testing?
Steve/bluewizard
A pulsating sphere is only an ideal for one type of speaker, an omni. It is not an ideal for a dipole or a monopole point source. This goes back to the point I was trying to convey earlier: among the first things you have to decide before choosing drivers and crossovers is what you want the radiation pattern to look like and how you want it to vary with frequency. There is no one right way, there are a lot of options, all of which will interact differently with the room and the listener.
But don't start shooting until you're decided what the target is.
Weems's book is more "tell me how to build a specific project" than "tell me how to make and implement design choices." It's a good book, but not what you need here.
But don't start shooting until you're decided what the target is.
Weems's book is more "tell me how to build a specific project" than "tell me how to make and implement design choices." It's a good book, but not what you need here.
Sy,
Again, I appreciate you replying, but you are shifting the context and ignoring the question. The purpose of 'idealized' source of sound has absolutely nothing to do with attempting to build an 'omni-directional' speaker. These hypothetical idealized speaker were chosen to eliminate all complicating variables. Perfect idealized speakers would likely have a absolutely perfect spherical output.
The real question is about perception of sound relative to frequency. I'm asking if sound coming from the idealized source wouldn't have a size perception attached to them.
Bass frequencies are going to sound big because that is the nature of low frequencies. Mids are going to create the perception of a medium sized sphere because that is their nature, and highs are going to create the perception of a small spherical source image.
Again, is has to do with beaming or narrowing of the sound as frequencies go up. You said this was strictly a characteristic of real world speaker. I'm saying if we eliminate real-world and concentrate on idealized to discuss this one small point, wouldn't high frequencies be perceive as coming from a smaller point source?
This gets back to my statement that high frequencies have a natural tendency to beam or narrow.
Again, I'm not saying the above is correct. I'm asking if this perception I have of high frequencies naturally sounding smaller is real, or if I am merely seeing the fallout of natural speaker characteristics.
In a sense, by using an idealize source, we can forget about speakers, and concentrate on perceived sound and point sources relative to frequency alone.
Not sure if that makes it clear or muddier.
Again, I appreciate you replying, but you are shifting the context and ignoring the question. The purpose of 'idealized' source of sound has absolutely nothing to do with attempting to build an 'omni-directional' speaker. These hypothetical idealized speaker were chosen to eliminate all complicating variables. Perfect idealized speakers would likely have a absolutely perfect spherical output.
The real question is about perception of sound relative to frequency. I'm asking if sound coming from the idealized source wouldn't have a size perception attached to them.
Bass frequencies are going to sound big because that is the nature of low frequencies. Mids are going to create the perception of a medium sized sphere because that is their nature, and highs are going to create the perception of a small spherical source image.
Again, is has to do with beaming or narrowing of the sound as frequencies go up. You said this was strictly a characteristic of real world speaker. I'm saying if we eliminate real-world and concentrate on idealized to discuss this one small point, wouldn't high frequencies be perceive as coming from a smaller point source?
This gets back to my statement that high frequencies have a natural tendency to beam or narrow.
Again, I'm not saying the above is correct. I'm asking if this perception I have of high frequencies naturally sounding smaller is real, or if I am merely seeing the fallout of natural speaker characteristics.
In a sense, by using an idealize source, we can forget about speakers, and concentrate on perceived sound and point sources relative to frequency alone.
Not sure if that makes it clear or muddier.
BlueWizard said:
There will be reflections, the early wall reflections especially would contribute but we don't necessarily perceive the entire sphere. Most wayward sound waves will probably fade out by the time they would reach our ears along their compex path. This makes their initial production seem unnecessary to begin with. Other reflections that survive a certain period of time before reaching us cease to become spatial cues in-so-much as they become echoes, and they are then of questionable necessity.
On the other hand, is it our room we're trying to listen to or the one in the recording?
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.