About midrange driver choice in a 3-ways speaker

[diyiggy]

Thanks Adason
I thougth about two reed paper cone from Vifa...to reach my spl needs but it is far from the two vifa that are flat till 3k, sota thd, titanium voice coil , high qms and low no number...
The ne149we is far to be flat which is making the passive filter more complex...lcr, etc..

 

[chargedcapacitor]

My DSP weapon of choice is the RME ADI2 pro fsr. You get 5 bands of parametric EQ and possibly the best sounding DAC under 5k. Paired with a philips CDM1 based transport, it plays everything I put through it perfectly from my 4000+ CD collection.

I was a big fan of the RME Adi 2 as well; I used it for room correction for over a year. However, the MiniDSP Flex has made the RME obsolete, as it has 4k tap FIR capability. I'm currently using a hybrid crossover approach for my DIY speakers: MiniDSP Flex x4 balanced, FIR DSP crossing the MT, and a 3rd order Butterworth for the MB at 400hz, with the IIR filtering for the low end room correction. It's a much cleaner and livelier sound than the passive / RME setup I had before.

If you have the money and space, you can get the MiniDSP flex x8, and use it for 3 way crossovers plus two independently tuned subs. There is only an RCA version, which performs slightly worse than the balanced x4 version.

 

[diyiggy]

OK, let's go for a 6ND430 from 18THSounds....

 

[profiguy]

Let me look and see if someone has leftover stock on NE123s. Don't give up yet. Its disgusting Peerless did what they have with the availability of most of their drivers to diy. Typical chinese...

 

[diyiggy]

Hi, thanks profiguy,

there are 4 ohms unit here and there in Europe, but 8 ohms seem NLA... Seems also this vendor putted this driver choice without any checking !

 

[profiguy]

@diyiggy You have a PM

 

[Bryguy]

Pick drivers with flat smooth responses and little to no breakup if possible. Set the midrange for as wide a bandwidth as possible so that it contains the majority of the content similar to a fullrange driver so that spatial and imaging cues contained in the music are coherent (aka, the magic of the fullrange driver speaker - circa the decade from 500Hz to 5kHz, sometimes called the “telephone band”). This can for example end up with a speaker that has 400Hz and 4kHz or 500Hz and 5kHz or even 600Hz and 6kHz crossovers.

Avoid crossovers at the traditional 1.5kHz to 2.5kHz range as that is one of the most audible regions to the ear so mistakes or imperfect transitions are more noticeable, plus you lose the phase coherence of the fullrange driver in this critical part where stereo imaging and soundstage are important.

Hey X! I've lately been thinking a lot about this idea you often bring up about the important imaging and spatial content of the signal existing in the telephone band. Intuitively, this makes sense to me. I'm sure that through your exhaustive experimentation with speakers, you have simply heard this idea prove itself to you through listening, but I was wondering if you have any sources that suggest the same idea? I'd like to read up on this subject if there is content available

 

[Juhazi]

I support same idea as xrk971. I can't remember where it came, but after making several multiway speakers I am sure that it pays for. Info is from bits and peaces...

Range from 500 to 5kHz brings many difficult challenges to loudspeakers because the basic dilemma is wavelength and dimensions of drivers and boxes/baffles coincide and make interferences.and introduce rapid changes in directivity. Then add diaphragm resonance of woofer... then add phase shift because of crossover!

There are many ways to overcome these issues - fullrange driver, wide baffle, no baffle, line source, MEH, 1st order xo etc. but every concept has caveats or bring other kind of problems.

 

[xrk971]

Most of what I have learned is from reading threads in the Fullrange forum. But a little searching around in publications turned this paper up. It seems to show that frequencies below 500Hz don’t work well because the head diameter is similar to the wave length of sound and diffraction and shadowing effects make it hard. The graph in Fig 4 shows the acoustic transfer function of a typical human head/torso. It seems to have maximum sensitivity from 500Hz to 10kHz but 5kHz will capture the peak.

I got that paper from looking at the references of this paper.

I’m sure more references can be found in the above paper.

One more thing, the spatial info is buried in the phase info of the sound. We hear frequencies up to 20kHz but we can discern phase down to 2-3 usec. This is in the hundreds of kHz frequency in amplitude. This is how we can know where sound direction is with one ear because the folds are asymmetric and impose phase shifts of order 2-3 usec.

There is a paper that measured this phase difference. It was done by US Army at the Aberdeen Proving Ground. I can’t seem to locate it at the moment.

But here is a great paper by same group.
https://apps.dtic.mil/sti/pdfs/ADA431963.pdf

Hope this helps.

 

[Juhazi]

Troels Gravesen is fanboy of low order and high MT xo http://www.troelsgravesen.dk/Diy_Loudspeaker_Projects.htm#PAGE_INDEX
- sadly he doesn't show xo slopes of his latest designs

on-axis resposes of some of my latest actives. Basically acoustic LR2(elliptic) or LR4



Psychoacoustics of stereo
http://www.hovirinta.fi/audio/psykoakustiikka/psychoacoustics.htm
https://mutor-2.github.io/ScienceOfMusic/units/03/
http://www.davidgriesinger.com/
http://www.davidgriesinger.com/Localization, Loudness and Proximity 8.pptx (check your browser's safety settings and skip warning)

 

[Bryguy]

Thanks X and Juhazi that should keep me busy for awhile!

 

[Bryguy]

Most of what I have learned is from reading threads in the Fullrange forum. But a little searching around in publications turned this paper up.

For me, this is a dead link. Which of the references is it from the second link you gave? I can probably find it online if I know the name

edit: I found it. I have to click it from this link to get it to work https://pubs.aip.org/physicstoday/a...We-Localize-SoundRelying-on-a-variety-of-cues

 

[Bryguy]

There is a paper that measured this phase difference. It was done by US Army at the Aberdeen Proving Ground. I can’t seem to locate it at the moment.

if you ever come across this paper again, I'd definitely be interested in reading it. I'll try to find it, and if I do I will post it

 

[xrk971]

I found the link but it doesn't seem to work anymore. Fortunately, I had saved a copy of the paper, attached.

My mistake, it was by Batteau of Tufts Univ sponsored by China Lake Naval Ordnance Test Station. I think the dtic server was located at Aberdeen.

 

[Bryguy]

I found the link but it doesn't seem to work anymore. Fortunately, I had saved a copy of the paper, attached.

My mistake, it was by Batteau of Tufts Univ sponsored by China Lake Naval Ordnance Test Station. I think the dtic server was located at Aberdeen.

No wonder I couldn't find it! Lol

Thanks for sharing it!

 

[airvoid]

I think the hifi portion of the audio community in general, and the diy in particular, would do well in obsessing less with the ‘where’ aspect and focus more on the ‘what’.
Musical communication is achieved through the language of rythm, pitch and harmony not location. Personally I see location mainly as a nice feature, helpful in enabling my brain to follow certain instruments or voices in the mix, but hearing what they are playing or the musical expression in terms of timbre, modulation, inflection and pace take precedence of where they are in the space between the speakers.
For a more current summary on the subject at hand I would like to recommend
An Introduction to the Psychology of Hearing by Brian C.J. Moore the leading textbook of auditory perception. It’s in its 5th edition from 2004 and describes first principles, state of current research, with interpretation and evaluation, and includes extensive references for more reading for those interested.

 

[Bryguy]

Musical communication is achieved through the language of rythm, pitch and harmony not location.

I don't see why these need to be mutually exclusive. Some people, myself included, really enjoy the "space" in which the recordings we listen to exist. Why tell us we're listening the wrong way? By paying attention to my system's portrayal of space, I'm not consequently ignoring other aspects of musical playback. It's part of a whole, and at least for my musical enjoyment, it is an integral piece of the puzzle. Not that I can't jam out through my monophonic phone speaker every now and then

This is not even to mention the fact that many musicians use the soundstage as a big part of their musical expression. Baba O'Riley by The Who comes to mind. That track is an entirely different listening experience played in stereo vs mono. I love it either way, but the band put a lot of time into creating the stereo effect, which I don't think should be dismissed. Here's a more extreme example

Try listening to this song in stereo, on a good system, then in mono. So much of the magic is lost when you take away the spacial cues and immersive sound stage. If we're talking "musical communication", I think this track conveys so much just through its use of space and location. It's not the whole pie, but it's certainly a big piece

 

[stv]

for my musical enjoyment

I second that.
The immersive experience of "being there" can make the magic happen.

 

[Bryguy]

The immersive experience of "being there" can make the magic happen.

The song as a whole I could take or leave, but the opening to this track gives me goosebumps on a good system. If I'm in the right mood, I can almost place myself in the wilderness. It's pretty spooky, in such a fun way

 

[Bryguy]

For a more current summary on the subject at hand I would like to recommend
An Introduction to the Psychology of Hearing by Brian C.J. Moore the leading textbook of auditory perception. It’s in its 5th edition from 2004 and describes first principles, state of current research, with interpretation and evaluation, and includes extensive references for more reading for those interested.

Thanks by the way. I'll check this out!