Well, that depends on how you simulate them. For hornresp and AJHorn the Le is an elementary for the simulation. I can only repeat, you have to have very accurarate values of these measurements. In such situations BassCAD and WinISD are much better off because they simulate on partly different parameters. You will find that these simulations are much so closer to the actual box if something like Le doesn't have the right value (or is even missing completely).
Self measured parameters are great and very useful but on some of these the measurement error margins are huge, just think about a multimeter which becomes inaccurate at higher frequencies. But that's not the only thing, there are tolerances in the measurements, in the speaker and even different results of the same driver and measurement setup, depending on humidity, temperature and air pressure changes. That's exactly what I was talking about better results for extreme constructions with other simulation software and why you should not trust every simulation (or at least check/verify with other software and measurements), especally if experience and measurements of others say contrary.
I don't see how any simulation tool can be MORE accurate than Hornresp if it does not take the effect of voice coil inductance into consideration. At least in Hornresp it uses a simple fixed value for inductance (Le). The only way I can see a simulation tool being better here is if it can better model how Le varies with frequency and takes that into consideration in the sim. But to not take inductance into consideration at all in the sim? Sorry, I'm not buying that.
It happened several times, and these were not my own projects, I help a lot of folks in different forums or while visiting someone. I had a suspicion and it would not turn out like simulated, I compared the simulations with my own because they looked too good. My concerns were shot down (ofcourse, like always). And, ofcourse, it did not work. You don't get an appreciation or just a 'thank you', you become hated because it was correct.
I did not document it (do you documentation on other's projects?), and without enough information there IS nothing to contribute, that's more confusing and time wasting which I try to avoid, especally if it's not mine. What would you think if I'd come up to you with 'that does not work' and don't support the claim with data?
On domestic use you practically always got much closer walls and smaller rooms. To take that into account is the right way to deal with it - you have to plan for it, you can't rely on having quarterspace and then realizing your setup lacks 6 dB or more. The opposite is actually much better, it can't hurt if you have 'too much' at the venue. That saves a lot of gigs.
Well, if you do not have/remember references on that. Things happen, and memory is inaccurate at best. Etc.
Please consider informing David in his thread regarding "Hornresp" the next time, it is incredibly much more helpful than just keeping it to yourself.
Regarding 1/pi, 2/pi, 4/pi sims; I am of the very firm and personal opinion that it is better to plan for 2/pi in-house, because not every frequency see a bit of cardboard (or closet doors) as an absolute boundary, and where there is room gain, there is room loss. I find 2/pi give me more realistic expectations in regards of final performance when the build is complete.
Please consider informing David in his thread regarding "Hornresp" the next time, it is incredibly much more helpful than just keeping it to yourself.
Regarding 1/pi, 2/pi, 4/pi sims; I am of the very firm and personal opinion that it is better to plan for 2/pi in-house, because not every frequency see a bit of cardboard (or closet doors) as an absolute boundary, and where there is room gain, there is room loss. I find 2/pi give me more realistic expectations in regards of final performance when the build is complete.
Last edited:
That's because the port is close to the enclosure wall, you have to shorten it by 20-30%.
Because they use less critical values. Don't you understand there are more ways than just the one? Let me give you an simple example. Let's assume you want to measure the power of some electrical unit. Easy, is it? Yes, ofcourse. You measure the voltage and put an ampere meter/multimeter in series. Oh, that multimeter introduces a serial resistance because it has to measure the voltage drop over it. That gives you an error margin. What's better? Well, you could measure the resistance and then calculate with the voltage. Oh, the power consumption isn't constant or you can't measure relyable the resistance? Well, you could use a clamp-on ammeter, that does not have the disadvantages of the former ones. Well, it does not take into account if there's a phase shift. Do you see how each of these methods would work but all of them got also disadvantages? There isn't just one way to do it.
The T/S parameters are tied together and with some you can calculate others. You can measure the difference if you add something (enclosure volume or weight to the cone), measure the stiffness of the suspension, measure the impedance and inductance etc, with some of these values you fill in the missing ones. Some of these parameters are cruicial and if you calculate with already off-values, you'll get bigger and bigger error margins but a simulation is only as accurate as the underlaying data. You still don't buy it? Fine by me but keep in mind, the physics doesn't give a crap if you believe in it or not.
Um, no. None of the other quoted t/s parameters "fill in" for Le. It cannot be calculated from the other t/s parameters.
Yes, some box design programs allow you to sim without a value for Le, but that makes them LESS accurate, not MORE accurate.
You can easily see the impact of Le on a sim by emulating it in Hornresp and comparing the results obtained when Le is set to 0 rather than the value quoted by the manufacturer or measured by a tool like DATS. In some sims it will make more difference than others, but it will make a difference.
And finally I think that you're making some pretty big assumptions about calculations being made with values being "off" to back up a position that IMO really doesn't have any merit.
Thank you, I guess.
But I wanted to indirectly put emphasis on wording and phrasing. It is so difficult sometimes, to try and remain neutral.
Here comes a guy with a blunt argument, there is someone with a sharp one. And then you have it, page up, page down of irrelevant crap. In many parts of this forum it seems quality of threads is quite good, but still, it is so easy to creep down a trench and take defensive positions. It's not about "can't everyone just get along" because that is never going to happen. It is more about "please have a look at your post and try to see it from someone else's point of view" before clicking *post*.
Not gonna pretend I am better at this than the next guy, I have my mostly blunt and some few sharp moments, it's just trying to be aware.
For me at least, it is work in progress.
But I wanted to indirectly put emphasis on wording and phrasing. It is so difficult sometimes, to try and remain neutral.
Here comes a guy with a blunt argument, there is someone with a sharp one. And then you have it, page up, page down of irrelevant crap. In many parts of this forum it seems quality of threads is quite good, but still, it is so easy to creep down a trench and take defensive positions. It's not about "can't everyone just get along" because that is never going to happen. It is more about "please have a look at your post and try to see it from someone else's point of view" before clicking *post*.
Not gonna pretend I am better at this than the next guy, I have my mostly blunt and some few sharp moments, it's just trying to be aware.
For me at least, it is work in progress.
Last edited:
What does that say about the accuracy of the software?
LOL, well technically that's not an issue with the software but with how the topology of the build is being sim'd with the software. A PEBCAK problem
What does that say about the accuracy of the software?
Maybe you should have read the full post.^
Slot ports adjacent to the wall will virtually lengthen the port due to additional losses.
Therefore one needs to shorten the port.
No fault on the software.
I said..
..in the context of not precise (or off-) parameters.
I can only repeat: If you're using BR simulation on hornresp or AJ Horn, verify it with a different simulation software, even if it's in maybe 95% of the cases not necessary. Maybe you can even drop that verification on standard tuning. But this sub isn't anything like that, small, deep, loud, extreme in every way and that means it's the prime example you have to verify.
And I want to note that "Lossy Le" is not an issue on BassCADe or WinISD, that's not an issue with the driver either, it's an issue with hornresp and AJHorn because of their special simulation method.
Hi Brian,
You are correct, of course - my bad.
I need to be more careful in framing my comments in the future
.Kind regards,
David
Then you didn't read the thread I linked. Any bass reflex simulator can simulate drivers where "Lossy Le" should apply, and they'll show a flat response just like Hornresp does. That isn't what the finished cabinet will produce, though. What the cabinet will produce is simulated well by Hornresp, with Lossy Le applied.
We're looking at drivers with long coils and heavy cones - home theatre or car subs, usually.
Chris
I did read it. They only simulated with hornresp. Nothing else. No WinISD or other Qt based simulation program. Are you not capable of grasping they use different parameters to simulate the response? I've just done the simulation of 2 of the drivers in said thread and the simulation of how the low drop slope with the same start of decline. If you still claim the same, you didn't compare the hornresp and WinISD simulations. And you did not understand that the other simulation programs work differently.
For the last time: They. Do. Not. Use. Le.
That means, you do not get a wrong simulation because of 'lossy Le'. That in turn means: It's absolutely impossible to get a false simulation because of high Le because that got no influence whatsoever on the simulation. And that also means it's a hornresp problem ONLY. No need for any adjustments, you don't tweak the simulation until it matches the measurement, the goal is the other way around, simulate it and get an idea if the cabinet you'd like to build performs on the low end.
Yes, that got a downside, it does not simulate the upper part (which is influenced by much more stuff btw). But the HUGE advantage is, you have a very precise simulation of the low end that's reliable. I repeat it again: Hornresp is a great tool but if you are building extreme vented speaker, other programs are far better and reliable. I get it, you love hornresp and noone may dare to say anything against it but everything got its advantages aswell as disadvantages, but nothing gets 'better' by ignoring the downsides.
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.