WinISD

Just keep in mind that we have found WinISD to be inaccurate.  It's a good tool to learn from, but when building enclosures remember that their enclosure sizes aren't correct.  Although I haven't played with it, our sales guy was saying that Unibox is more accurate.  You can get it here: https://www.pvconsultants.com/audio/frdgroup.htm

The most accurate box software we have found is LspCad.  However it isn't freeware.

Steven Kephart

Adire Audio

Steven, you say that often.  I have recently started using Unibox, still in the infant stage...but I did a comparison with it and WinISD Pro using the Dayton RS150S-8 6" driver in a vented enclosure.  Given the same volume and tuning freq. the port size and length are identical (to within a few hundredths of a centimeter).  Given the same volume and port size the tuning freqs are just as true.  And given the same Fb and port size, the volume was the same within 1/10 liter.

WinISD Pro's graph also proved to be true when I tested optimizing for a 'wanted tuning peak' in response using Unibox.  The WinISD graph displayed the peaked response when I transferred the Unibox change in Fb to WinISD.

Unibox is much more sophisticated, to be sure, but WinISD Pro is very user-friendly and is indeed accurate when calculating volume, at least it was with the test I have done.  Have you done your own testing, by any chance?  What am I missing?

I think the real beauty of WinISD is that it is pre-loaded with many popular drivers.

Anyway, I continue to be uninformed by WinISD naysayers about the _source_ of the error in WinISD.  Is it in the data files or is it in the calculation of the enclosure? 

I suspect that the error is in the data files rather than the calculations used to determine the box parameters.

We have found that the recommended enclosures given by WinISD (using the same parameters in all software programs) provide faulty enclosure specifications and faulty port lengths.  Like I said above, I have personally never played with Unibox but am just passing on what our sales guy said.  I will run your test past him though.  We use LspCad Pro as it has proven to be highly accurate. 

Ok, I just asked our sales guy about it.  Apparently WinISD isn't always wrong, but just sometimes.  I personally have ran into this as well when comparing a sub in LspCad and WinISD.  WinISD said .707 alignment was around 2 cubes, while LspCad said it was around 1.5 cubes.  Apparently it is also wrong at times with port length and vent mach.  Try doing your test with several different drivers and I'm sure eventually you will run into errors.

Steven Kephart

Adire Audio

Steven,

For a point of reference, who is it that you refer to when you say "we"?  You and this forum body;  you and a group of your friends;  you and your company?  I want to separate hearsay from fact.  I have made one comparison with Unibox (which you have recommended as more accurate than WinISD by virtue of someone else's opinion) and I've found WinISD to be accurate enough in the comparison for real life, and you have now said that you made one comparison with it to LspCad and found innacuracy.  One comparison each.

Math is math as you know.  A program will use its calculating formulas to find the result if given the parameters it needs.  If the program uses its built-in calculators to find volume for the desired tuning freq of one box, it will use that same calculator to find it for another.  Why would it change?  It's not on me to do more comparisons until I find a fault.  It's on you to find a concrete reason why WinISD (sometimes) errs in calculating volume....because it is you who paints the picture of a rogue program whenever someone on this forum asks about WinISD. 

I'm not in your league as far as knowledge of things having to do with speakers or the complex calculations in this field of science.  But I don't understand your reasoning for poo-pooing WinISD when the subject comes up.  People aren't investing anything but their time and could certainly benefit from what they get out of it.  In most cases, the actual construction of their box will be slightly off from what they set out to achieve anyway due to many reasons...that is to say, the real life enclosure is not normally going to be to the cu.in. that one sets out to make it.  And we both know that most people are going to use published specs instead of sophisticated individual driver testing because we are all not audio engineers.  So, close is what we are going to get anyway.

I think more people should try it for its relative ease-of-use so that they don't have to cry for help for each little detail of a build.

Sorry, I should have been a little more specific on that.  We (Adire) used to recommend WinISD and even had a link to it from our website.  However with many uses by my boss (Dan Wiggins), he came to the conclusion that it is inaccurate.  It was providing enclosure sizes, vent machs, and vent lengths that were way off.  To be honest, I only have that one instance for my own personal experience of its inaccuracy.  But I have found to not question my boss' word on things.  If he says it is inaccurate, he has plenty reason to believe this. 

I hope I don't come off as poo-pooing the program.  I just try to point out that you can't fully trust its results.  However I do try to point out that it is a great program to learn from, as I even stated in my first post.  I think warning the end user to double check first is fair, and might save them some time and money on an improper enclosure.  A quick e-mail the tech support of the company is a good idea anyway just to be sure.  I know that we offer free tech support on our products to our customers.

Steven Kephart

Adire Audio

Hmmm, the tension perhaps? Anyways...

Can you provide a list of possible drivers that create these such errors in WinISD by chance Stephen? I think that would help better to illustrate the point you are making(Which I believe is showing that there are faults in the WinISD program.)

Thank you for the reply, and my questions were answered.  I will point out that I have used only the latest alpha version and am not aware of bugs in any of the earlier versions....which may be the case in Dan's experience with it.  It is entirely understandable why a company might not make a specific program recommendation but an individual would. 

And I would, too, recommend that it be used in conjuction with other means to find enclosure volume and port sizing...for example, double-checking vent air speed with other calculators like this one.  I just happened to have WinISD opened to the file on that driver while I was working on it in Unibox because WinISD had an easier way to convert units of measurement than I could find in Unibox.  That's when I started doing some comparisons.

Otherwise, in-general comment about box-building to specific tuning:

Although difficult and time-consuming, port lengths may have the capacity to be adjusted after listening tests.  Some pre-made ports are made that way, so lengthening or shortening the length can be accomplished after everything is installed and tested.  The hard part is removing the driver every time you want to make a slight change in port length.  I've personally never done this, but I have popped out a port and replaced it with a longer length before.

......and as related to using a software program like WinISD Pro:

Use several sources of calculators to find answers.  Here is a list of calculators, formulas and general information that I refer to often.  Double-check all the data you've entered.  You will most likely find any mistakes you've made in entering parameter data by checking with other sources.  As with all things electrical where a change on one side of the equation results in a change on the other side (forbidden), a change in the program for tuning frequency will result in a corresponding change in port length.  In WinISD, you must flip back to the next tab from 'box' to 'vents' to find these changes.

All too often I see individuals (on this forum and elsewhere) complain that the program "gave me crazy port lengths and it's just wrong!"  Take the time to read all the Help info and play with the program until you fully understand it.  If you don't want to spend that much of your time - learning - to build one box then you should take your needs to your local custom audio shop.

Hi all,

My experience with WinISD (and it's Pro variant) have been less than impressive.  Many T/S parameters are not correct (I don't think any of our drivers are properly defined).  It requires much data that can be auto-calculated (for instance, given Fs, Vas, and Sd - all typically given on a data sheet - you can calculate Mms and Cms, which typically are not given on datasheets).

WInISD also seems to incorrectly calculate port air speed; for example, a Tempest in a ~193L box tuned to ~20 Hz and powered with 100W will have an airspeed around 18.5 m/sec (based on measurements with a thermal airflow sensor).  LspCAD estimates 19.3 m/sec - pretty close.  WinISD claims 28 m/sec, well out of range of reality.

WInISD does not allow calculations of series/parallel wiring options for DVC drivers.

WinISD does not even attempt nonlinearities.  Oilcanning of spiders - which happens when they are pushed near their limits - can cause a strong jump in actual excursion below the Fs of the driver.  When designing a box for "running near the limits", any software that does not include nonlinearities is woefully lacking.

Note that many of these criticisms also apply to UniBOX, particularly the nonlinear and inductive effects.  However, my experience has been the that the typical UnIBOX user is a bit more familiar with T/S theory and its limitations, and thus typically do not over-do the design nearly as much.

IMHO, WinISD is a very basic tool, and it should be treated as such - do not rely on it for anything more than a few W power input.  The errors in vent sizing alone are reason to avoid it, IMHO.

Dan Wiggins
Adire Audio

1. 

Dan]It wrote:

---

requires much data that can be auto-calculated

---

There is an 'auto-calculate' box at the bottom of the parameters window that is used when entering driver parameters.  Click it to calculate unknown values.  The Help section details the steps in entering the numbers.

• Parameters may be and should be auto-calculated.

2. 

Dan]WI wrote:

---

ISD also seems to incorrectly calculate port air speed

---

WinISD displays vent air speed by using a graph, and by default the output is shown as 'peak'.  Place the cursor on the "peak" and click to change to RMS.  Changes like this are made by using File > Options > Graph.  Changes made will be set as default the next time the program is opened.

In the Tempest example, which I duplicated after entering parameters from the Adire site (and using a port diameter of 4"), WinISD displayed air speed at 27 m/sec.  But, after changing the graph to display RMS, the display showed 19 meters/second.

• Same as LspCAD.

3. 

Dan]WI wrote:

---

ISD does not allow calculations of series/parallel wiring options for DVC drivers.

---

The choice of parallel or series for DVC subs is made in the database editor under the parameters tab.  DVC drivers have a different designation (name) than do their SVC counterparts, so when you enter data for that driver and save the file, the program will remember how you have the coils wired.  If you want to change series to parallel (or vice versa), open the database editor and change it there.  You can save the change and give it a different name, so now you'll have both ways of wiring in your saved database.  And you are not limited to DVC;  any number of voice coils can be entered.

• Multiple voice coil drivers are calculated according to wiring.

4. 

Dan]Wi wrote:

---

ISD does not even attempt nonlinearities.

---

Nonlinearities at max excursion is beyond the scope of my understanding.  And I will bet that is also true with the vast majority of enclosure builders who are looking for a good design help program.

5. 

Dan]Th wrote:

---

errors in vent sizing alone are reason to avoid it, IMHO.

---

Some people expect the program to give an opinion on vent diameter when a new project is opened.  It does not.  Instead, a default value of 102 mm (shown as 4.02 in) appears with the calculated port length for that diameter.  The program just simply calculates port length according to the port diameter, box volume and tuning frequency.  The user must be aware that he can highlight and change the port area to the size  he wants to use,( then input wattage and check the graph display to see if the vent air speed is less than 10% of mach).  If it's too high, enlarge the port opening.  Vent air speed can be verified by using other calculators, such as these two combined.

• I have heard this criticism (errors in vent sizing) more than once, but no one has yet to demonstrate an example where it errs, or has ventured a reason as to why they believe it errs in calculating vent sizes...in any manner that can be substantiated.

There are a number of ways to verify port volume, or diameter vs length calculations, several of which are found in this list.

6. 

Dan]Ma wrote:

---

y T/S parameters are not correct

---

Any part of the overall build can and should be double-checked with other calculators or information, especially the very basic aspect of making sure the driver parameters are correctly entered. 

• One should enter parameters himself and not rely on those entered by someone else at some other time.  I think that would be wise when using any of the design programs.

The more I use this program the more I find how useful it is.  It may be substandard in the eyes of a world-class speaker designer, but for the majority of posters on this forum who want to build their own box it is a good design tool, and becomes very user-friendly with use.  Anyone wanting to get some good help in designing enclosures should give it a fair trial, but put forth the effort to learn how and why it forms its calculations and how to interpret the results.

Here is the link to download:  https://www.linearteam.dk/default.aspx?pageid=winisdpro

It is free to use.