WinISD/SPL charts

[M0ses]

Two questions: one easy, one that we could probably talk about for years.

1. Is there any sort of database, or at least somewhere to start looking, for cabinet SPL charts? Where does BFM get his?

Now on to WinISD...which I don't know a ton about.

When you're using WinISD, you choose a driver and BAM, it gives you and almost flat SPL chart and box that would get it to you, right? So, what's so hard about building a flat cabinet? WinISD just designed it for me with any driver I choose. Will building a box with those dimensions actually yield a cab with an SPL chart that looks like that?

 

[greenboy]

Q:

When you're using WinISD, you choose a driver and BAM, it gives you and almost flat SPL chart and box that would get it to you, right? So, what's so hard about building a flat cabinet?

A: Unfortunately a laptop equipped to run WinISD is too large to shove inside an enclosure through a round port tube. Some designers have shifted back to shelf-style ports in a desire to accomodate these laptops, but pushing them past that last bend invariably causes Windows to "blue screen".

 

[will33]

If you look at the size of a lot of the cabinets WinISD automatically plugs in you'll find a lot of them to be impractical, as in way big. It may be the best mathematical equation but it's too big to haul around and often tries to put out deep lows that just aren't needed to make a bass sound good. Yes, there is a such thing as too much bass for the bass.

 

[bobcruz]

And the flat curve you see in WinISD is good only up to about 200 Hz or so, if I recall correctly. There are a lot of frequencies above that, around 19,800 more, to deal with if you want a cabinet with a flat response. Sure it's do-able but how would it sound? I have some really high end JBL home stereo speakers that are flat from 35 to 20,000 Hz, and for a lark I hooked my bass amp to them to see how they'd sound. Not good. Too much bass and not enough mids to hear myself in any gigging situation. That got me away from thinking about using a flat PA cab for bass.

 

[Passinwind]

Hints: look at the max power and excursion charts. Be sure to specify the driver's rated wattage when you do the latter one.

Now, what's so easy about designing a reasonably sized cab with no displacement limiting ?

 

[billfitzmaurice]

1. Is there any sort of database, or at least somewhere to start looking, for cabinet SPL charts? Where does BFM get his?

I measure them.

Will building a box with those dimensions actually yield a cab with an SPL chart that looks like that?

No. WinISD is only accurate within the region of purely pistonic function. That's up to perhaps 200Hz. Above that the driver data sheet gives on-axis, but off-axis is more important. On-axis would lead one to believe that most drivers are useful to at least an octave higher than they actually are. That's why my charts are measured on-axis and at 45 degrees off-axis.

 

[will33]

As far as accurate to 200hz. WinISD has adjustments you can make for the whole frequency band but how accurate is it?

In the mids/highs, when you shift the listening position to 30 or 45 degrees off axis it makes a very rounded and even/repetative or whatever "comb" looking pattern. Is that a wildass guess or can it be considered "close enough"?

Increasing coil inductance numbers on a speaker brings the top end down but is that also an estimation?

It is very useful in tayloring a lowend and getting a driver and a box to work well together. As far as the rest of the sound, I go by what an SPL chart says if there is one or most often my ears.

 

[billfitzmaurice]

As far as accurate to 200hz. WinISD has adjustments you can make for the whole frequency band .

Not that I'm aware of. WinISD, like all modeling programs, uses T/S specs, and it's those specs that only work up to roughly 200Hz for the average driver.

 

[will33]

Maybe I'm not clear. If you go into the section where you can adjust input power in watts, humidity, etc., you can adjust the listening position and the farther you put it off axis, the more descending humps it has the higher you go in frequency.

If you go in to edit the speakers parameters, raising coil inductance with show a smooth treble rolloff on the graph, the higher the inductance, the more treble rolloff.

How accurate is that stuff above 200hz?

Edit: General example of what's going on? Specific to that speaker at all? Educated guess?

OP, not trying to derail but gain more info.

 

[M0ses]

The WinISD window goes up to 1kHz. That data beyond 200Hz basically worthless, then? Is there no reasonably accurate way to predict the response of a cab without building it? How do people design cabs, other than stabs in the dark?

I measure them.

I figured.

A: Unfortunately a laptop equipped to run WinISD is too large to shove inside an enclosure through a round port tube. Some designers have shifted back to shelf-style ports in a desire to accomodate these laptops, but pushing them past that last bend invariably causes Windows to "blue screen".

Hints: look at the max power and excursion charts. Be sure to specify the driver's rated wattage when you do the latter one.

Now, what's so easy about designing a reasonably sized cab with no displacement limiting ?

The only charts I have are gain, phase plot, SPL, and group delay. The third of which I only partially understand and the fourth I am clueless.

 

[will33]

I don't understand the group delay thing either, never have. I've read random things on the internet (how much stock would you like to put in that?) about if you keep it below a certain whatever it likely won't cause a problem. I don't know what that whatever is/was but if I remember right, my WinISD reading (whatever that means) was a bit below where whoever said something might interfere with something or whatever.

I would like to know what it is/does. I'd like to know how or if it matters with a bass cab and why and if it matters with anything else, higher frequencies for example.

 

[alexclaber]

The WinISD window goes up to 1kHz. That data beyond 200Hz basically worthless, then?

Yes.

Is there no reasonably accurate way to predict the response of a cab without building it? How do people design cabs, other than stabs in the dark?

You use the manufacturers' measured plots, you figure in what you know about the driver dimensions and construction (and thus off-axis response), you then build it and test it (properly test it with measuring equipment and ears) and see how it compares to what you predicted. Then use that data to tweak the design, test again, and round and round until you get where you'd originally planned (or suitably near) or give up and start the process again with a fresh design.

 

[KJung]

You use the manufacturers' measured plots, you figure in what you know about the driver dimensions and construction (and thus off-axis response), you then build it and test it (properly test it with measuring equipment and ears) and see how it compares to what you predicted. Then use that data to tweak the design, test again, and round and round until you get where you'd originally planned (or suitably near) or give up and start the process again with a fresh design.

This is the real difference between the Barefaced and fEArful cabs versus, for example, the LDS custom cab type thing (of course, Don has some amazing 'tried and true' tweaked designs also, but that's not my point). Sometimes, a pure 'one shot' custom design using the computer program to generate port dimensions, etc., etc. can sound FANTASTIC. However, sometimes, the results, while always at least 'OK', can be just a bit disappointing.

The ideo of a designer like Alex or Greenboy or Jim Bergantino or Duke going through many iterations to fine tune a limited number of box designs and the crossover, etc. to be 'optimimal' guarantees that the box will deliver on the designer's goal (which may or may not work for a given individual's tone goals.. but that is another topic).

 

[billfitzmaurice]

I don't understand the group delay thing... I'd like to know how or if it matters with a bass cab .

It doesn't. For group delay to be bad enough to be audible the speaker would have such poor response as to be unlistenable at all.

 

[bgavin]

Sometimes, a pure 'one shot' custom design using the computer program to generate port dimensions, etc., etc. can sound FANTASTIC.

If you repeat the design/build/test cycle long enough, one can hone their design sheets (XLS for me) to an exact science for weight and tuning.
This lets me get difficult designs like slot ports dialed in nearly perfect the first time out.
The T/S design software is sufficient to get the measured performance very close to the designed performance.

If you do any amount of this, a few test boxes helps with the initial concept.
For example, I have over-size boxes with removable baffles to mount different drivers.

The vent is a series of round tubes cut to different lengths for rapid plug-in tuning changes.
The external vents do not affect the internal volume.
It is easy to change the internal volume using filler blocks of 1.0 cubic feet.

The impedance sweep is the revealing tool in the process.
It shows you real world performance that cannot be duplicated in software.