I have used HolmImpulse and that is also free.
Actually, if someone is simply wanting a DIY speaker solution without the learning curve, nothing beats a kit or a tested design.
If the goal is to design a good set of speakers from scratch, spending large sums of cash does not insure success any more than leveraging the free software or cheap software.
It seems we will just have a slight disagreement on that point.
Actually, if someone is simply wanting a DIY speaker solution without the learning curve, nothing beats a kit or a tested design.
If the goal is to design a good set of speakers from scratch, spending large sums of cash does not insure success any more than leveraging the free software or cheap software.
It seems we will just have a slight disagreement on that point.
I'm not so sure we have a disagreement. You need a base of knowledge to work from so you can make wise decision in the design process. The design process still takes some intelligence; you don't simply push a button and out pops a perfectly formed speaker.
But on the other hand, there is no reason not to take advantage of the expertise that is made available to you. The design process is complex, and anything that can simplify it, it worth while.
So, it takes a combination of the two. You need a good base of knowledge and understanding of speaker design and all its intricacies and complication. But that the same time, there is no need to complicate any more than it already is. If you can get your hands on Software that helps you focus on the end result rather than the process, so much the better.
But when it comes to software, you have to make a reasonable and responsible decision. It makes little sense to buy $1000 software to build a single pair of $100 speakers. You have to find some compromise in which the expenditure is justified by the final goal.
When it comes to testing, you need a reasonable knowledge base of room acoustics and how they will effect any speaker you test. You need a reasonable understanding of reflections and standing waves, of peak and nulls in a room. You need all this to get your speaker into a circumstance that allows fair and reasonable testing.
There is no "One Button" solution. You need knowledge, a person needs at least a base education on all the factors that come into play. But, there is no need to bog yourself down in hours of math, the software will do the heavy lifting for you.
At its most basic, speaker design is easy. Make the box look like a speaker box for the size drivers you have. Tune the cabinet and port to the appropriate frequency, use generic nominal impedance crossover design. That's no more than some of the lower end commercial speaker makers do.
But, even this most simple and basic method, still requires some knowledge and understanding of speaker design.
At its most complex, speaker design is a multi-disipline PhD level engineering task, filled with cycles of design, prototype, testing, analysis, and redesign.
Most of us are above the generic cut'n'try design stage, but few of us are anywhere near the PhD Engineering Level stage.
So, yes, you need knowledge, but within the context of your needs and budget, some speaker design and testing software can be very helpful.
For what it is worth.
Steve/bluewizard
Last edited:
As an extension of my above post, and regarding room acoustic, as much as reflections, you need some clear sense of Boundary Re-enforcement. The proximity of your speakers to walls and floors can very much effect their output and performance.
As to the original poster, as I said, your expenditure on software has to be consistent with your ultimate goal. In short - how many speakers do you have to built to justify the outlay of $1000 on software, to justify $200 on software, to justify $40 on software.
The OP already has WinISD, that is a start. Too bad the makers of WinISD and WinISD Pro did not finish there product, it is a nice program and a finish product would be worth a little money. Yet, that is their choice, not mine.
There are many other free or trial software programs related to audio that are available. For testing, it is hard to beat REW (Room EQ Wizard). But for design, the best choice is a little unclear. That is down to your priorities and budget.
Steve/bluewizard
As to the original poster, as I said, your expenditure on software has to be consistent with your ultimate goal. In short - how many speakers do you have to built to justify the outlay of $1000 on software, to justify $200 on software, to justify $40 on software.
The OP already has WinISD, that is a start. Too bad the makers of WinISD and WinISD Pro did not finish there product, it is a nice program and a finish product would be worth a little money. Yet, that is their choice, not mine.
There are many other free or trial software programs related to audio that are available. For testing, it is hard to beat REW (Room EQ Wizard). But for design, the best choice is a little unclear. That is down to your priorities and budget.
Steve/bluewizard
Wow what a lot of info on the post I posted? If no one really likes bass box pro what other programs will help me out better? I was looking at gettting this measuring system? ATB 701 Precision I have never heard of the brand but, the guys at meniscusaudio are really helpful. Again thanks everyone for your help. I will have to take my time and sort it all out. Mr. Jeff
@ Mr Jeff
I think you might like WinISD, i've been using it for years & have designed lots of nice sounding etc cabinets with it. I wrote a Help thread which should assist
http://www.diyaudio.com/forums/subwoofers/148729-how-correctly-use-winisd.html
Latest version is here http://www.linearteam.org/download/winisd-07x.exe
I think you might like WinISD, i've been using it for years & have designed lots of nice sounding etc cabinets with it. I wrote a Help thread which should assist
http://www.diyaudio.com/forums/subwoofers/148729-how-correctly-use-winisd.html
Latest version is here http://www.linearteam.org/download/winisd-07x.exe
Not necessary or recommended.
You need two things to design and test speakers. 1 is software. Most of that is free if you take advantage of what is available.
The second is hardware (mic, USB interface, PC, amplifier, mic stand, etc.).
You can buy a calibrated microphone from Parts-Express and use a free software measurement program such as HolmImpulse or REW to capture the data.
For the mic, buy this...
Dayton Audio EMM-6 Electret Measurement Microphone Allows For Accurate Acoustic Measurements At A Fraction Of The Price 390-801
You need a power amp, but if you are building speakers you already should have that (i.e., stereo amp).
You will need some sort of USB interface between the PC and the mic and the amplifier to drive the speakers.. I use the M-Audio MobilePre, but there are others out there that are less expensive. However, whatever you get, it needs to have phantom power already built in for the electric microphone.
There are software packages that offer design and test from start to finish such as Sound Easy. That package is $250.
Soundeasy In Stock At Parts Express With Free Shipping
The last time I evaluated it, it was a good package, but that was a few years ago and I do not know how that competes with current versions of other software. Sound Easy, I am told, has (or had) a steep learning curve, so you should evaluate it before committing that kind of money. I understand that newer versions have addressed that issue, but I don't know to what end
LspCAD (Standard) is a direct competitor that should do almost the same thing for about the same price and is supposed to be much easier to learn, but the Standard package is not as powerful as Sound Easy.
Untitled
Lastly, you may want to invest in a tool to test the raw speaker drivers and get their T/S parameters (if you need it - LspCAD and Sound Easy already support this)). Woofer Tester II or Parts-Express's version (cheaper) will do an excellent job and make the task easier by automating many of the steps.
woofer tester - Parts Express Ships Fast and Ships Free
I do agree with this to some extent.
It's great to know a bit of the theory but for sealed and even ported enclosures WinISD is so easy to use that most people can just get stuck in.
However, be aware that it is just estimated filter responses and doesn't modelling any resonances. So it wont highlight any issues with enclosures that have a long dimension and it won't model transmission lines.
That said, Sealed, Ported, Band-pass and Isobaric all model quite well.
What I like about it is it works and I especially like it gives me outside dimensions for the boxes. Entering your own speaker their small paramters is nice and easy. (I use woofer tester 2)
I have tried some other software and they want too much information. Like how much padding or absorption ratio of materials of room dimensions or they are just not intuitive for a dummy like me
I have used WinISD to make some 18"H 5" W and 7" D boxes to compare to a pair of voigt boxes I made (same size, same driver) and I am now making some long tall skinny boxes for a CTS 4" driver like the old Bose 901's series I & 2 had.
Resonances, always thought those were fixed by solid bracing and acoustical stuffing.
Transmission lines would be cool, I have a set of Heil cabs that take a 10" woofer, amazing the bottom end that comes out of those boxes.
I have tried some other software and they want too much information. Like how much padding or absorption ratio of materials of room dimensions or they are just not intuitive for a dummy like me
I have used WinISD to make some 18"H 5" W and 7" D boxes to compare to a pair of voigt boxes I made (same size, same driver) and I am now making some long tall skinny boxes for a CTS 4" driver like the old Bose 901's series I & 2 had.
Resonances, always thought those were fixed by solid bracing and acoustical stuffing.
Transmission lines would be cool, I have a set of Heil cabs that take a 10" woofer, amazing the bottom end that comes out of those boxes.
Most measures done at home with computers measure at a spot on the T/S curves much different than where LEAP or LMS measure (almost universally used by the manufacturers). For 1st pass on any box i use the factory numbers, Smith & Larson Woofer Tester 2 is a good 3rd to the above 2.
dave
Since T/S parameters are curves and not scalars, i feel it is important to have some tolerance of their changes under dynamic conditions -- a tightly tuned box will, most of the time, be out of tune.
dave
What this guy said!
People get too hung up on things like:
"this program shows a 1.56dB rise in frequency response at 52.8Hz and program b doesn't. Which one is right?"
"how do I get rid of this 3dB ripple in my response".
"Is the Fs of my speaker 23Hz or 24Hz?"
You need to be aware of the tolerance, errors and 'ripple' of the whole system including, source, microphone, driver, cabinet, room, your ears etc. etc.
Don't try and design to such small tolerances, it's not practical. I tend to work to 3dB ish.
I always lean to a larger box. When using winisd I always round the numbers up on box size, then I'll add any angles I want to the design.
We are not making rockets, no 2 speakers will have the same specs. I've measured 100's they are all over the place.
We are not making rockets, no 2 speakers will have the same specs. I've measured 100's they are all over the place.
wow, your WinISD layout is totally different then mine. Intersting.
@ Mr Jeff
I think you might like WinISD, i've been using it for years & have designed lots of nice sounding etc cabinets with it. I wrote a Help thread which should assist
http://www.diyaudio.com/forums/subwoofers/148729-how-correctly-use-winisd.html
Latest version is here http://www.linearteam.org/download/winisd-07x.exe
Also one small fact that is usually missing in modeling is that power level can and does change performance. T/S was designed around small signal levels where many assumptions were made on linear aspects of materials. High power changes a lot of things. You can still read the original article on how the Thiel/Small parameters are derived (mostly for engineer types (I admit it)) really depends on your math and physics skills. My point is most of the software floating around is based on something created in 1973. Yep some of us have been around this even before that
I have a question about box design software...I recently purchased a new sub and in the process of choosing which driver i was trying to model each driver using the specs manufacturers posted...I kept getting a curve with peaks in the 110-130Hz range...i contacted the various companies and asked why this was...I was told that trying to model it was pointless given that the driver would have a totally different response after it is broken in. does this hold any truth?...to what degree
I believe that the drivers specs that the manufacturer supplies are supposed to be after the driver has been 'broken in'. So although it will change, it should start out 'wrong' and become 'right'.
Therefore it is not pointless modelling drivers. Hence why we all do it!
When you say "...peaks in the 110-130Hz range... what do you mean?
Is that flat-peak-flat or is the simply a response that has a general peak in that range, as in rising-peak-falling response.
The "Q" you specified in the design effects how the speaker will respond. A Q of 2 will have a large bass peak, a Q of 1 will have a slight bass peak, a Q of 0.7, most often regarded as the ideal, will have pretty flat response. A Q of less than 0.7 will have a much slower but likely deep roll off.
In certain software packages this "Q" is determined by the answer to a question - Do you want Most Bass, Flattest Response, or Deepest Response? If you choose Most Bass, then you chose to have this artificial peak.
Could you give us a screen shot of the response graph you are talking about, then we would have some framework for our answers?
Steve/bluewizard
Is that flat-peak-flat or is the simply a response that has a general peak in that range, as in rising-peak-falling response.
The "Q" you specified in the design effects how the speaker will respond. A Q of 2 will have a large bass peak, a Q of 1 will have a slight bass peak, a Q of 0.7, most often regarded as the ideal, will have pretty flat response. A Q of less than 0.7 will have a much slower but likely deep roll off.
In certain software packages this "Q" is determined by the answer to a question - Do you want Most Bass, Flattest Response, or Deepest Response? If you choose Most Bass, then you chose to have this artificial peak.
Could you give us a screen shot of the response graph you are talking about, then we would have some framework for our answers?
Steve/bluewizard
Again, precisely what you mean by a peak in the 110hz to 130hz range is vague. In a manner of speaking, all a Subwoofer is, is the peak of an arc. The response is the peak of a inverted parabolic arc. We simply slice off the top of that arc and call that our working range.
Think about how narrow the response is for a Subwoofer - say 20hz to 200hz. That is a very small window, and could be nothing less than an arc peaking around 100hz.
So again, are we talking flat-peak-flat, or is this simply the top of a gradual rising-peak-falling arc?
Steve/bluewizard
Think about how narrow the response is for a Subwoofer - say 20hz to 200hz. That is a very small window, and could be nothing less than an arc peaking around 100hz.
So again, are we talking flat-peak-flat, or is this simply the top of a gradual rising-peak-falling arc?
Steve/bluewizard
Last edited:
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.