Open source Waveguides for CNC & 3D printing!

My final measurements, or the QC I did back in the fall?
Ehm just your own measurements compared to these Klippel measurements?
Not really sure what you're asking here? :D :D

Agreed on the top octave. I think that is where you're are going to see a difference between methods (near vs far) no matter what you do.
Even with a fixed setup, it's really tricky to get consistent results.
If you just have a mic on one of those mic-stands, forget about it above 8kHz or so.

Well that sounds very dramatic, with that I mean you will find differences in the ±1dB ballpark.
 
Over the last 2 decades, I have seen plenty of DIY projects that will outperform the vast majority of "professional" speakers easily.

In fact, ever since Erin started testing, it totally proved my suspicion based on a few of my own measurements, which is that the majority is meh at best.
We are not even talking about nuances, but just fundamental basics that are already wrong or a freq resp that looks more like a mountain range, terrible resonances etc.

cough Wilson cough

The response irregularities tend to fill in to the sides, apart from crossover irregularities, so the power response is not as horrid as the on-axis response might indicate. However this indicates considerable error in diffraction control, no surprise when you look at all the reflecting surfaces and hard edges around the drivers.


Which is mind boggling, if you think about proper literature about this, like Floyd Toole's book that has been available since 2008 or so?
Even before that, there was already a pretty good sense about the importance of directivity and freq resp.

Toole's first AES papers on the subject came out almost a half century ago. Supposed professionals have no excuse for ignorance.
 
When I switched to a 4 ohm woofer after the first prototype was sent to Erin, I put all those measurements in a folder called "archive". I think I deleted it because I didn't need it anymore and didn't want a bunch of clutter. I'm trying to find the screenshot I took of the comparison. And now I'm realizing that since the last measurements from Erin may not have had the newest XO settings activated, but I'm not sure, I can't even compare to my current measurements :( And the speakers are already back at my house. I'll look through imgur and see if I uploaded it there.
 
  • Like
Reactions: 1 user
<150hz = same sensitivity but rolloff was a bit different. Makes sense given mine was a combined port+near woofer response. I've never really put a lot stock in this measurement anyway since accurate T/S will generate better box response in software.

150hz - 1000hz = This is where the biggest difference was. But in a way that my measurement is still accurate, but smoothed. Where the Klippel has some little ripple with its high resolution, my response was like a slightly wavy line that touched the tops of the ripples on the Klippel response.

1000hz - 6000hz = dead on to the Klippel

above 6000hz = slight deviation that became more noticeable >10khz. Maybe 1.5dB different around 12khz or so IIRC. So basically the top octave and bottom octave diverged, but with the middle in good agreement.

Some similarities to our comparison: https://www.audiosciencereview.com/...-how-the-klippel-nfs-works.13139/post-1371581

You can set the distance to calculate the far field to in the Klippel software. This can get rid of some of the differences.. for example if you do your normal measurement at the usual 1m distance, but have the NFS set to calculate to 10m distance.

I don't trust the NFS in the low frequencies yet. I am aiming to do a proper tower measurement (~6m height) to be able to compare the NFS to true 4pi.
 
  • Like
Reactions: 1 user
When I switched to a 4 ohm woofer after the first prototype was sent to Erin, I put all those measurements in a folder called "archive". I think I deleted it because I didn't need it anymore and didn't want a bunch of clutter. I'm trying to find the screenshot I took of the comparison. And now I'm realizing that since the last measurements from Erin may not have had the newest XO settings activated, but I'm not sure, I can't even compare to my current measurements :( And the speakers are already back at my house. I'll look through imgur and see if I uploaded it there.
Whoops, happens to all of us :D :D
No worries, just take your time, I was just curious :)
 
I found one comparison, but this one does not have the low frequency, near field response merged with the far field. So disregard response below 220hz on my measurement. This was the Satori MW16TX and Bliesma T25B on waveguide.

klippelvsmine.PNG
 
Last edited:
  • Like
Reactions: 3 users
With optimizing with different types of window functions you can probably get even lower.

Last week I was able to pull of 100Hz = 10ms

That doesn't give much more detail below 300Hz, but at least gives a bit more frequency resolution.

Hopefully soon I can experiment with a sort of damping cage around the setup.
I have done already some preliminary experiments and it looks very promising :)
 
With my new setup I can get about 8.6ms at 1m, or 7.7ms at 1.5m, or 6.8ms at 1m

I mean 6.8ms at 2m.

I use Blackman Harris window function. I've tried the others but haven't seen anything that would make me switch.

BTW, anyone use both the SB26STAC and SB29SDAC? If you have mention if you used Fs compensation. I considered using the SS 8330 with the SS Discovery woofer, but I haven't made a waveguide that I like yet for it.