A rough approximation. The actual lne shape will determine the optimum Zd.
This little bit from MJK illustrates the optimum offset for specific TLs, a rsnge which does not include 1/5 = 0.20.
dave
Not sure which part.... Anyway:
My four-year-old Xiaomi phone control panel includes a real hearing test that outputs an EQ curve to be saved under a name and used (or not) -- it's an effective FR chart (inverted) for speaker/positioning/room-effect/hearing.
Frequency Sound Generator APP has three original-iPod-like dials that lets you tone-sweep with your finger and easily/directly compare several frequencies. Of course what you get out is an effective FR combining the above-listed factors. Ears are extremely fine at comparing loudness, subject to the Fletcher-Munson curve: extra sensitive 2-4khz, and depending on SPL (loudness), increasingly insensitive outside 100-10khz range.
The HF drivers should be axially aimed at your two ears from front-L/R (where hearing is most sensitive as well as most stereoscopically effective), otherwise HF will be greatly diminished.
The main point is that HF (5khz in particular) is also responsible for an overall sense of (micro-) dynamics. A dip there will rub the music of articulation/contrast/energy/liveliness -- i.e. too soft-spokenly polite.
Another possibility is too little/too thin midrange (bowl-shaped response) relative to HF. My collection of full-range drivers almost all fall into one of two groups: this (extended, often ear-bleeding highs) or the opposite (rolled off highs), described earlier. The APP will tell you which.
First adjust toe-in to aim the HF driver at your ears. Depending on which needs boosting, midrange or HF, adjust the XO if you have access to it (or to separate binding posts for the two drivers). Try diminishing the other driver with a chain of small resistors (using electrician's twist-caps) until satisfied.
Last edited:
Personally can't 'travel down this road' due to lifelong tinnitus, but I've noticed a preponderance of folks preferring drivers with a lot of sibilance, especially among the 'FR' aficionados, which I assumed was for best speech intelligibility, but if otherwise, then might as well try all the way up to 'pierce' around 12 kHz or even boost the whole ~ 5 kHz - up BW a few dB.
What are the units on this graph? SL? S0? The symbol on the right column?
Too much of a good thing...
Vintage Wharfedale fullrangers all seem to have the same (published) uneven response curve, "house sound". Very dynamic sounding.
Anyway, I first noticed this phenomenon with Hifi-Bird "English drum paper" 4.5" and 5.5" drivers (5 and 6 nominal). Dynamic linearity to kill for -- reminiscent of my Goodmans Axiom 80 (cost 50X). My review here:
https://www.diyaudio.com/community/...e-driver-for-under-eur100.393788/post-7216726
https://www.diyaudio.com/community/threads/newbie-diy-speaker-project-questions.391819/post-7166899
(and earlier in the thread)
Last edited:
I’ve noticed this with my transmission line speakers that use smaller drivers but have good bass extension. I have a set similar to yours, it has two 4” woofers one crossed over at 500hz, and the other crossed over at 5,000hz to a tweeter. They sound really good for jazz (MJQ sounds great!) and, gack, quartets sound great too.
Large symphonic works however are odd. There is plenty of bass but no smackum-jackum. Emulsified music is OK but dynamic music is dulled. There is something about the direct cone radiating area relating to percussive sounds in the recording.
I think a system with an eight inch woofer in a T-line of the same length would provide a better, bigger dynamic picture than two fours, but unfortunately, not the same imaging. Tradeoffs.
Also the room size matters a lot. In a small room your four inch woofers will be OK, but put them in a slightly bigger room and meh… This seams to be a T-line characteristic. It’s different than a small speaker in a big room that would lose bass and schmak. T-lines tend to lose schmak but not bass.
I thought about using a sealed 6 1/2” woofer in the mid-bass region to bring in some boomshakalaka. I don’t know how to integrate it. I’m still pondering…
Just impressions, no math or data to back it up. You simply might need bigger speakers.
Large symphonic works however are odd. There is plenty of bass but no smackum-jackum. Emulsified music is OK but dynamic music is dulled. There is something about the direct cone radiating area relating to percussive sounds in the recording.
I think a system with an eight inch woofer in a T-line of the same length would provide a better, bigger dynamic picture than two fours, but unfortunately, not the same imaging. Tradeoffs.
Also the room size matters a lot. In a small room your four inch woofers will be OK, but put them in a slightly bigger room and meh… This seams to be a T-line characteristic. It’s different than a small speaker in a big room that would lose bass and schmak. T-lines tend to lose schmak but not bass.
I thought about using a sealed 6 1/2” woofer in the mid-bass region to bring in some boomshakalaka. I don’t know how to integrate it. I’m still pondering…
Just impressions, no math or data to back it up. You simply might need bigger speakers.
The TL Design tool in Hornresp uses the method described in Martin's document.
Attachments
Obviously someone listening to a badly designed TL.
Unfortunately, plenty of those out there.
Thanks, my observations are very consistent with yours: different music plays very differently and they are really outstanding on smaller ensembles, vocals, etc where they beat most of the 2-3 way systems I've heard. I used them initially in a bigger room and it wasn't working at all, much better now in a small to mid-size room.
Could you please comment on the <1 tapered TL conjoining mass-loaded/resistive Onken slot, i.e. my TLonken hybrids? (Either here or in your giga-thread, though I think diyaudio needs a sub-thread-with-title feature.) Very much appreciated.
IMO, if your XO right at 5khz is at all phase-incoherent there would be destructive interference leading to less than "full-sounding" dynamic transcients. When I notched down 5.4khz my 5.5" drum-paper fullrange went from world-champion-level dynamics to merely excellent. I suspect my ear-bleeding Lowther PM6A once notched would be the same.
You can easily verify XO phase-alignment by moving the tweeter back-and-forth while playing the XO frequency test tone and listening for wavelength-periodic maxima. Uncanny experience when next played music even mono -- to be suddenly transported to the venue.
Regarding placement of damping material.
The location of harmonic velocity maxima is quite predictable in a straight pipe. But folding or tapering it will move this around from strictly geometrical positions.
So I think that you end up with an imperical itarative process. Unless there is some detailed FEM analysis tool to do it.
The location of harmonic velocity maxima is quite predictable in a straight pipe. But folding or tapering it will move this around from strictly geometrical positions.
So I think that you end up with an imperical itarative process. Unless there is some detailed FEM analysis tool to do it.
I have no experience with the loudspeaker system you refer to, so I am not really in a position to comment on its design or performance.
Interesting. I’ve thought about this. During my design phase I did set the tweeter back 3/4” but maybe that’s not far enough? It’s kind of impossible to re-align it now since it’s built as an MTM. I’ll try this out in future designs.
I mean, does HR simulate <1 tapered ML-TL where the ML is a slot (possibly resistive Onken)?
Like this design schematic:
https://www.diyaudio.com/community/...ir-5-2-in-evansound-3-75l.393909/post-7260419
Last edited:
Regarding offset: The optimal placement seem to be close to one third down the line (at the pressure minima of the third harmonic, The placement should also be close to the 5th and 7th harmonic.
One way is to look at the phase angles of those harmonics and see where the sum of them work out the best along the pipe lenght.
On other way is to get the third harmonic and add damping material from the closed end up to the driver and rest on the fact that all higher harmonics will have velocity maxima at least once in that damping material.
Anyhow the folding geometry is a large factor for how the harmonics turn out (and the classical IMF folding is really bad)
One way is to look at the phase angles of those harmonics and see where the sum of them work out the best along the pipe lenght.
On other way is to get the third harmonic and add damping material from the closed end up to the driver and rest on the fact that all higher harmonics will have velocity maxima at least once in that damping material.
Anyhow the folding geometry is a large factor for how the harmonics turn out (and the classical IMF folding is really bad)
Yes, that causes a pretty big suck-out at 5,000hz which leads me to believe I got the timing setback, maybe not perfect, but pretty close.
I think the ‘politeness effect’ is due to the size of the speaker cone and a lack in the ability to reproduce transients. I’ve noticed people adding 8 or 10” woofers to their systems to provide more output in the lower-mid region.