WerTicus, measurements can feel mistake in sense that it could make you uneasy in case of issues, but not a mistake in sense that you have gained control over the situation. As long as you can control your emotions over it and solve the issues 😀
Here is how to solve your mystery:
~4kHz period on the individual horn measurement interference indicates there is secondary sound. 4kHz wavelength is ~8.6cm long, delaying a secondary sound 8.6cm would make opposite phase at 2kHz and you've got your pattern and secondary sound source solved: Something is making a reflection, which arrives to mic 8.6cm later than direct sound.
Further, assuming your graph has 2db grid so that the individual L / R measurements have ~3db peak to peak in the interference pattern it means the secondary sound is roughly 14dB down relative to direct sound. This is quite much attenuation already for such short path length, so I think it's from your measurement rig. Perhaps your mic is on the sofa, and it's reflection from the sofa attenuating the sound some, or something like this. It cannot be diffraction inside the horn for example, because it would yield different interference pattern. If one wants to analyze further, the interference amplitude reduces toward top end, which indicates horn directivity affects it, so it's reflection of something between DUT and mic.
This stuff is easy to resolve using vituixcad with ideal drivers, just adjust delays and attenuation until you have similar interference pattern.
Here is how to solve your mystery:
~4kHz period on the individual horn measurement interference indicates there is secondary sound. 4kHz wavelength is ~8.6cm long, delaying a secondary sound 8.6cm would make opposite phase at 2kHz and you've got your pattern and secondary sound source solved: Something is making a reflection, which arrives to mic 8.6cm later than direct sound.
Further, assuming your graph has 2db grid so that the individual L / R measurements have ~3db peak to peak in the interference pattern it means the secondary sound is roughly 14dB down relative to direct sound. This is quite much attenuation already for such short path length, so I think it's from your measurement rig. Perhaps your mic is on the sofa, and it's reflection from the sofa attenuating the sound some, or something like this. It cannot be diffraction inside the horn for example, because it would yield different interference pattern. If one wants to analyze further, the interference amplitude reduces toward top end, which indicates horn directivity affects it, so it's reflection of something between DUT and mic.
This stuff is easy to resolve using vituixcad with ideal drivers, just adjust delays and attenuation until you have similar interference pattern.
Last edited:
I have them crossed over at 800hz now - They sound amazing - which just seems to conflict with what the REW measurements are showing me.
I don't feel like my ears are uneducated either - I went to the store near me and had a listen to the Kii three speakers to make sure i was not going mad and these sound better to me.
Yeah my mic is resting on the sofa for all measurements... the horn itself is 13cm in front of the baffles....
I don't feel like my ears are uneducated either - I went to the store near me and had a listen to the Kii three speakers to make sure i was not going mad and these sound better to me.
Yeah my mic is resting on the sofa for all measurements... the horn itself is 13cm in front of the baffles....
Last edited:
I did a proper time alignment with REW last night. On all drivers and subwoofer too (previous time alignment was done with measuring tape) The imagining is uncanny now. Maybe the 2nd best i've heard in my life. The best being a system that used the full range raven ribbons from 300-40khz
I did a little bit more tuning and tweaking just with my ears - it sounds fantastic. I think im just going to leave it for now - very happy
thanks for all your advice everyone and for the horn design Mabat
I do have plans to make an entirely new speaker at some point from 3d printing and i also have the drivers already so i will certainly make that happen eventually once i get bored again!
thanks for all your advice everyone and for the horn design Mabat
I do have plans to make an entirely new speaker at some point from 3d printing and i also have the drivers already so i will certainly make that happen eventually once i get bored again!
This is the final result i got: I did not spend as much time as i should have with the glue / sanding process but this is just a first draft 😀
BTW, I'm finishing a new horn optimized for the DFM-2535, based on the A460G2.
This time including the stand and everything -
This time including the stand and everything -
Guess I need to buy a new set of plans 😂😬
How much difference do you think there will be with the A460G2?
I don't know, a few days, hopefully. Typically, I get so many ideas right after I "finish" something that I tend to always wait for a while...
It's not finished yet, I already have some further ideas. Actually, there's also no cap yet in the picture, that's just a "placeholder" for the driver. But it will be enclosed.
I think it looks nice being slim, but understand the raw driver could look cheap. The competition is stiff.
Went to a listening bar in Brooklyn and since been flooded with vintage HiFi multicell porn on Instagram. They invest so much into the looks, it’s annoying.
At least they had the daytime music playing from a bluesound streamer...
In Hamburg, Germany, a guy is running a kissa and they were so proud the whole chain is analogue (willingly ignoring the mastering process of the vinyl), showing off their fancy-schmancy crossover, and their horn-loaded beauty had the worst radiation pattern ever. You could not enjoy the music 30 degree lateral.
Complete bogus but it sells the feeling.
Went to a listening bar in Brooklyn and since been flooded with vintage HiFi multicell porn on Instagram. They invest so much into the looks, it’s annoying.
At least they had the daytime music playing from a bluesound streamer...
In Hamburg, Germany, a guy is running a kissa and they were so proud the whole chain is analogue (willingly ignoring the mastering process of the vinyl), showing off their fancy-schmancy crossover, and their horn-loaded beauty had the worst radiation pattern ever. You could not enjoy the music 30 degree lateral.
Complete bogus but it sells the feeling.
Last edited:
I’ve been sanding for what feels like an eternity but am almost ready to put a finish on the horns. I had to rethink the support because they ended up being much heavier than expected - so now the adapter slips through a steel ring which connects directly to the horn stand. I’m not too far away from firing it up!
Attachments
Mabat,
Any specifics on the new optimized horn for DFM-2535? any changes to the radiation pattern of the A460G2 or is it just cosmetic changes?
Any specifics on the new optimized horn for DFM-2535? any changes to the radiation pattern of the A460G2 or is it just cosmetic changes?
It has been re-optimized for that particular throat, without the constraints on the (Gen2) body shape. It is still very close to the existing one, with slightly less rising DI at the top.
- BTW, this is a new Gen2 adapter for the Celestion CDX1-1720:
Raw unsmoothed response (4.5 ms window):
12 ms window, 1/6 oct:
Available soon
- BTW, this is a new Gen2 adapter for the Celestion CDX1-1720:
Raw unsmoothed response (4.5 ms window):
12 ms window, 1/6 oct:
Available soon
I might keep going forwards with the G2 then for versatilities sake.
Will you make that stand design for the A460G2 as well?
Last edited:
I've just installed the latest version of ATH4 on my new Win11 laptop. It's a whole new game compared to the old version I was running (4.3.4).
Everything seems to work OK, but I'm not getting the "Final width x height = " or "Final length =" results come up when I run the program. This makes iterating to a size very clumsy.
I've tried with several different project files including demo1.cfg
Does any one have ideas about what could be wrong?
Everything seems to work OK, but I'm not getting the "Final width x height = " or "Final length =" results come up when I run the program. This makes iterating to a size very clumsy.
I've tried with several different project files including demo1.cfg
Does any one have ideas about what could be wrong?