New thread, reply to Augpro - RE: project files, dead horse beating...

[critofur]

In an effort not to thread crap too severely, I've moved this reply to it's own thread...

Originally posted by augerpro
But that's not what you are doing, and you still don't seem to understand why although myself and others have tried to explain.

I was simply proving that your argument was invalid.

But you don't have SE so how can you "load up" my design or my drivers?

I can borrow SoundEasy, and, if the community [of SoundEasy users] wasn't so belligerent, unhelpful, and unwilling to share, I probably would buy it.

And you have no idea what my measurement conditions are, mic placement, baffle size and layout, gating, etc. What everyone has been trying to tell is because of all these things my measurements are worthless to you.

They are far from worthless. They would be interesting, useful, and fun.

I know what you want, but what you want is useless,

I have given plenty of reasonable explanations as to why that is incorrect, without a doubt.

and I just don't know how to explain that so you will understand. Consider: if there is a thing that a lot of people want, some one will make it for them. But if no one wants this (except you), what does that tell you?

Other people have also said that they hope this would happen, and that there would be a "sticky" thread with project files and driver files.

I wish what you were asking for was possible, because it would be incredibly useful.

It doesn't have to be perfect, to be informative, useful, and interesting. I never suggested that once someone downloaded your "project" files that they: change a couple components, build that, then ship the speakers across the country to somebody without ever measuring again to make adjustments. You act as if I had.

I hope this doesn't offend, I don't mean to. But you keep asking for this and I keep trying to explain why it just isn't useful. If you don't understand why, then you will never understand how to do Jay's method properly.

It is patronizing, and it is very offensive. I understand all the points you have been trying to make perfectly fine, they are just not correct, for the most part. I had been hopeful, I had thought: "oh, somebody is going to share, but then, it just turns out that you're just goading, or flame baiting, or such. Is "troll" the right word?

Maybe I'm putting the "cart before the horse" ~ I want to find some projects using the drivers that I have access to and take a look at them in either Sound Easy or LspCAD, then "play" with the projects for a while [for the sake of learning] to see what the simulated response does after making various changes [assuming, hypothetically, that I would be using the same baffle]. Then, I would build some cabinets, THEN I would use my PC to simulate the crossover. And maybe, at that point, I would have enough impetus to go ahead and buy/build whatever else it is that I need to do measurements.

I'm sorry, I have to admit, I have some kind of mental deficiency which causes me not to want to learn Sound Easy or LspCAD without first having a look at some other people's project [files] using inexpensive decent drivers which are currently available to me.

Ok, consider this example I will give, then tell me why, exactly, it wouldn't make sense, if you're so convinced with yourself that I'm wrong about these notions...:

Let's say that I wanted to find out if I would want to build Zaph's TMM Waveguide project. The criteria used to determine whether or not I want to build it are as follows: If, using the same baffle, same enclosure, same drivers, same driver placement: I could ESTIMATE, using his SoundEasy files + SoundEasy, whether using a serial 1st order crossover instead, could yield decent results. And, if so, then I would want to build it, then, once built, I would want to measure the results, and, listen to the speakers, before deciding if I was happy to keep them with the serial crossover (as a learning experiment, primarily), or maybe, I might go ahead and build Zaph's original crossover too and do some extended listening in both configurations.

But - before I would do any building, or buying of drivers, WHAT IS THE HARM of simply TRYING some simulation of different crossover configurations on my own PC???

Surely, SoundEasy is beyond capable of generating a rough estimation of the affect of a crossover modification? [rhetorical question]

[Dave Bullet]

Critofur,

Have you gone through this process with something like SpeakerWorkshop?

I know Jay posts his SW files for his designs. They contain respected or commonly used drivers such that you could possibly build something (or fall back on existing designs) should your own tweaking not meet your expectations.

My guess is that SE users are evolved over SW users. They invest in a package of more capability, with probably better measurement baffles, better measurement rooms, better mic etc... because they have more knowledge. They know why they measure and have gone past using others measurements for the issues Brandon outlined. (Hey - I am just a SW user at the moment so I put my self in the un-evolved category

Lastly - you can quickly use SPLTrace then HBT the data to then have something (I'm guessing here) you can import into SE and play with. I know that isn't a complete design - but having the raw data is the start... and this would get you started.

I do agree with you that playing with existing designs and files is another way to learn. They method at least provokes "why did that simulation happen?" for you to then seek answers.

David.

[critofur]

I guess the best bet would be to try to learn SW first then, but I commonly see people saying SW is "harder" than SE or LspCAD...

I tried using it a couple times, all three of them see so insanely counter-intuitive, it's like they TRIED to make the UI hard to use...

For instance, in LspCAD, why not just drag a driver onto the grid, then right click on it and say "load .FRD file" ~ I figured out how to get the files loaded, and how to draw a crossover in the schematic editor, but, when I tried to layout a simple 1st order series xover, it didn't seem to work. Plus, it's a little bit frustrating that you have to switch away from the edit mode before you can see the change on the FR graph - or - maybe I'm just missing something, I would think there would be some way to drag the "value" of a cap/ind. up + down and watch the FR change as you do so.

Speaker Workshop does seem to be the only one with a half decent amount of project files available.

It seems so convoluted to try to do something as simple as:

1) load a woofer + tweeter (have it default to say, an I.B.)
2) draw up a simple cap + inductor crossover
3) see the simulated response from the default mic position (1 m on tweeter axis)?

As a community we should agree on a default baffle and measurement position, then we can easily and meaningfully share all our measurements

Just because your driver is on a baffle, does not mean the measurement is completely useless to others, for one, they could just use the same baffle (as included in your project file, that they could download) for informational purposes, at the very least.

No matter how useless somebody might think those files might be, you could still learn SOMETHING more than just looking at the manufacturer's published response graph. Particularly with a company like Tang Band. In their case, your on baffle response is going to be more useful to me than TB's extremely unrealistic charts...

Zaph's measurement charts are CERTAINLY useful, who would deny that??? Since Jay has been able to take them, use a combination of tools, and generate a fairly accurate simulation, then surely, using Zaph's .WFR file could be at least as useful as just the FR image; and, certainly, much more convenient for a person to simply satisfy their curiosity to play "what if" mixing and matching woofers, tweeters, etc. in simulation even if they can't afford to buy "samples" for themselves.

[Dave Bullet]

I would think there would be some way to drag the "value" of a cap/ind. up + down and watch the FR change as you do so.

In SW you can sort of do this. It's called "perturb". You select which components you want to vary (and by how much) then SW will recalc the network with those combinations, combining all the FRs onto one graph.

It seems so convoluted to try to do something as simple as:

1) load a woofer + tweeter (have it default to say, an I.B.)
2) draw up a simple cap + inductor crossover
3) see the simulated response from the default mic position (1 m on tweeter axis)?

I can't comment fro LspCAD - but in SW:

1. You need both FRD and ZMA files for the drivers
2. Generate HBT data on the FRD. Use the FRD consortium FRC spreadsheet to do this
3. Import all files into SW
4. Create 2 drivers
5. Right click driver properties
6. "Connect" the driver to the FRD / ZMA file.
7. Repeat 6. for other drivers
8. Create a new network (= crossover)
9. Place both drivers onto the network
10. Add components
11. right click "calculate response"

That's all there is to it. SW isn't hard. Although it's feature set is basic, I've found my simulations are very very close to actual measurements so it is useful, although not as fully featured as SE.

Also check out roman's site: www.rjbaudio.com. He has a bunch of albeit older driver FRD files and some SW projects you can download.

Cheers,
David.

[augerpro]

critofor if you had EVER said anything about actually having SE I would have sent you some files. But all you ever mentioned was using SW-which obviously won't work, although you didn't seem to get that.

Look, there is no conspiracy of selfish SE users. IF you get SE I'd be happy to send some files to play with. Until then, no forum will host 2mb files, I don't have a website, I'm not going to go through all of the trouble of zipping up and emailing a big file if you don't even have the software to use it.

Until then, I'm out.

[dlr]

I don't from which thread this split off, but I'll add my $0.02 worth. I hope I'm not being redundant.

Originally posted by critofur
As a community we should agree on a default baffle and measurement position, then we can easily and meaningfully share all our measurements.

There is something of a de facto standard, that being 2.83v @1m on-axis. It's not always necessary to have a common measurement position, though it does make it easier. For example, I have a fair amount of quasi-anechoic large baffle measurements at my site taken at 0.5m. I describe fully the conditions under which they were taken. All that is required to adjust to that de facto standard is to subtract 6db from the measurements, easily done in most software, also doable with the FRD tools. Low end data will always a little suspect due to limitations of conditions such as exist for most DIY efforts, but that can be estimated with the box using T/S parameters for mids and woofers.

Just because your driver is on a baffle, does not mean the measurement is completely useless to others, for one, they could just use the same baffle (as included in your project file, that they could download) for informational purposes, at the very least.

This seems to be a reasonable approach. It would allow one learning to verify that they set it up properly in their software. However, there may be data missing such depending on the source files uses, such as relative acoustic offset in the model if minimum-phase files were used rather than direct measurement data.

I have only one full design posted and it's not likely others would want to duplicate it given the box style, but I do provide the raw measurements for download in the even someone cared to play with them to experiment. i see nothing difficult with this, the files aren't large. I've also got a tweaked Insignia design with downloadable files, but again, only one who tweaks their own woofer could make real use of them.

No matter how useless somebody might think those files might be, you could still learn SOMETHING more than just looking at the manufacturer's published response graph. Particularly with a company like Tang Band. In their case, your on baffle response is going to be more useful to me than TB's extremely unrealistic charts...

I'm in full agreement here. I'm still skeptical of the TB data. I'm more accepting of Seas, S-S, Vifa and Peerless with regard to accurate representation of data. TB is advancing rapidly, though, IMO.

Originally posted by augerpro critofor if you had EVER said anything about actually having SE I would have sent you some files. But all you ever mentioned was using SW-which obviously won't work, although you didn't seem to get that.

How so? I've missed the earlier thread discussion. If the measurements and relevant descriptions of their origin are described, why could it not be re-produced in another CAD program? What's special about it?

Look, there is no conspiracy of selfish SE users. IF you get SE I'd be happy to send some files to play with. Until then, no forum will host 2mb files, I don't have a website, I'm not going to go through all of the trouble of zipping up and emailing a big file if you don't even have the software to use it.

Until then, I'm out.

I understand the issue with SE project files. SE can export the raw driver FR and Z files, either with or without excess-phase for FR. These won't be large files. When zipped they're not large. As for the project, simply describing the design ought to be sufficient to re-produce in another package.

What am I missing?

Dave

[Undefinition]

I'm not sure if I want to get in the middle of this discussion, but Critofur, have you considered Jeff Bagby's Passive Crossover Designer? It's very intuitive, and provides excellent results. If you're really jonesing for project files I can send you some PCD projects I've done and you can tweak them.

You might be thinking that PCD is cheesy because it's just an excel spreadsheet and not a standalone application--but that is selling it VERY short. I have Speaker Workshop (Wallin Jig and all) and LSPCad, and I pretty much only use PCD. First of all, because it IS so intuitive and easy to tweak things. Second, because Jeff's Power Response modeling is one of the greatest inventions since sliced bread.

(Yes, the Aethers were done with PCD, as were my last project, the Vaagens. The same will be true for the next few projects I'll be doing this summer)

[critofur]

Originally posted by augerpro
critofor if you had EVER said anything about actually having SE I would have sent you some files. But all you ever mentioned was using SW-which obviously won't work, although you didn't seem to get that.

Look, there is no conspiracy of selfish SE users. IF you get SE I'd be happy to send some files to play with. Until then, no forum will host 2mb files, I don't have a website, I'm not going to go through all of the trouble of zipping up and emailing a big file if you don't even have the software to use it.

Until then, I'm out.

Well, maybe I'm twisting things a bit, because talking on forums/emails you miss that 90% of communication which is non-verbal - but, it feels like you've just switched from: "you want these files?" "haha, no, you're too dumb so I won't share" to: "if you can use them, I don't mind emailing them to you" - Thanks.

Originally posted by Undefinition
I'm not sure if I want to get in the middle of this discussion, but Critofur, have you considered Jeff Bagby's Passive Crossover Designer? It's very intuitive, and provides excellent results. If you're really jonesing for project files I can send you some PCD projects I've done and you can tweak them.

You might be thinking that PCD is cheesy because it's just an excel spreadsheet and not a standalone application--but that is selling it VERY short. I have Speaker Workshop (Wallin Jig and all) and LSPCad, and I pretty much only use PCD. First of all, because it IS so intuitive and easy to tweak things. Second, because Jeff's Power Response modeling is one of the greatest inventions since sliced bread.

(Yes, the Aethers were done with PCD, as were my last project, the Vaagens. The same will be true for the next few projects I'll be doing this summer)

Please, don't hesitate to chime in, this isn't the sort of thread where I'd be bothered at all by off-topic posts.

I'd be happy to use PCD ~ very happy, to have something that's quick and simple, as long as it's reasonably accurate, which does seem to be the case. I tried to use it, but, it doesn't seem to work with Open Office [I don't have Microsoft Office on my main home PC].

I mentioned before, it's a "comedy of errors" ~ I've tried Speaker Workshop, I've tried LspCAD, I've tried Sound Easy, even tried a few other more obscure programs, always run into one problem or another. I guess I'll go ahead and install some version of MS office so I can use those spreadsheets, too bad they won't work w/Open Office.

Would you be willing to share your Aether project? I'd like to build some, but with some of the tweeters I already have on hand (Audax TM20J3 - http://www.hautparleur.fr/_audax/tm020j3.jpg ) And, for the "bass bins" I'll be using these drivers: http://www.partsexpress.com/pe/showd...number=299-830

[djarchow]

Dave,

As I mentioned before measurements taken by others have little benefit to me except for evaluating if I want to actually use a driver or not. Zaph's and Mark K's measurements are one of the resources I use as part of making a decision which drivers to use for a project.

But for actual design work, I want to measure my drivers on my baffle. I know how to take a measurement like Zaphs and apply baffle diffraction, etc to it to match my baffle. I know that Jay and others can use this process pretty effectively to do a design without ever measuring the drivers.

But for me, I want my own measurements. I have to take measurements once I have the XO done anyway so why not just take the first set of measurements too and then I know what the conditions, baffle etc were like.

With my measurements, I know what gating was used, smoothing, is my mic really calibrated or am I using a calibration downloaded off the web, was the fridge in the garage running during the measurements etc. Sure there are people whose measurements I would trust (and you would be one of them), for example, I have gotten multiple measurements from John K. Taking loudspeaker measurements isn't rocket science, but it isn't exactly point and shoot either. Not only do the conditions under which the were measurement taken impact the results but the skill level and experience of the person taking them can impact them as well and I won't know that from looking at the measurement.

The other drawback with using other people's measurements for design is unit to unit variations. For example, I have measured two sets of the Usher 8945 woofer in the past 3 years and both sets of them were different enough in frequency response and t/s parameters that the previous measurements shouldn't be used. Of course Usher has updated the specs on this driver several times so this explains the problem, but it still doesn't help someone who has no idea if their drivers will match my measurements.

So while I do see the benefit of having independent measurements out there (ala Zaph), I still prefer to rely on my own when actually designing a speaker.

As always YMMV.

Regards,

Dennis

[Dave Bullet]

Hi Dennis,

Your points were pretty much what Brandon said in the other post from which this one started - and I agree.

Critofur

I mentioned before, it's a "comedy of errors" ~ I've tried Speaker Workshop, I've tried LspCAD, I've tried Sound Easy, even tried a few other more obscure programs, always run into one problem or another. I guess I'll go ahead and install some version of MS office so I can use those spreadsheets, too bad they won't work w/Open Office.

Assuming you know some theory, do you generally have problems using computer software? Have you read the manuals available? Have you been to the SE user site on (I think) yahoo? to get help?

Maybe after all that - you might lack the ability to understand and design speakers.

Please don't take your frustrations out on people like Brandon. You can't blame others for your own problems. Doing so only makes them more reluctant to help you - defeating your purpose.

David.