There's been a lot going on so it's not always easy to keep track of what things are the direct comparisons and for which aspect. I understand.
In the bullet points above about what affects frequency response and whether I can control this, I should have put ( of course )
4. The frequency response of each driver, in the 90Hz to 1000Hz region, approx.
Ok so I'm quite familiar with the sound of each of these drivers ( Fostex & B&C 8PE21 ) in that region. Bear in mind this is in various ways - from listening unequalised on the floor, or on OB. There will a certain amount of up-tilt in FR of course , lots for unequalised, but less on OB. I am happy with this and have accustomed my ears to judging drivers in this state . Both drivers sound good. I have also listened to both these drivers on the short tractrix (straight ) horn with the 'teardrop' phase plug in front of the cone. This plug maintains the low-mid tone and improves the upper-mid and treble tone by taking out dustcap effects in HF or averaging the ripple somewhat lower down.
Once fitted to this horn, the driver FR will tend to 'imprint' onto the horn mouth/expansion/folding characteristics. I can make allowances for that and can see the FE208Ez dips on the test I did with the radioshack a few weeks ago.
Ok, next point, I'm not ( now ) making changes to the cabinet, by which I mean the horn expansion. After the Mk.1C mod I'm done, anything better will have to wait for the Mk.2 . See further comments below.
What I've been doing is modifying the back volume ( behind the driver ) from open , down to around 6L. This is having profound effects on the sound quality , either in the upper bass or mids or both. The back volume and its shape is not really capable of doing a lot to colour the lower mids or upper bass it's not big enough for the waves at those frequencies. Most of the time this back-volume has been a rather complex internal shape with triangles, slabs & chunks of wood at odd angles. It does not lend itself to generating clear box modes. It's main effect is to add acoustic loading to the rear of the driver and I believe that's more or less all it's doing. Different acoustic loading WILL affect phase.
So what I'm hearing all the time is either having strong tonal colour, or losing it into a greyness ; or having pianos and drums 'pinging' in a fast, crisp way, or being mushy and slow. These to me are all phase-related effects and cannot be mimicked or 'fixed' by small changes up or down in the amplitude response. For example - I have played around with the FE208Ez a while back with a notch filter for the 3.5/4kHz area and while it stops it from sounding overtly harsh, it does not improve the tonal colour through that region, because it hasn't fixed the phase effects going though the area the cone or dustcap has resonances. I accept that tuning of an amplitude response will improve the exact tonal rendition of voices and instruments to make them more believable, but it won't increase or decrease the amount of tonal colour - what Romy calls 'absolute tone' and I call 'tonal gamut' .
The situation I'm in at the moment is all about validating or otherwise whether the back-volume adjustment can or can't bring this horn design up to a worthwhile level for continuation. I'm realising that it's the biggest thing going on in this system and needs to be fully understood and fixed before it's worth worrying about smaller stuff. At present I have an unacceptable choice between upper-bass speed and attack, or midband tone. There is also the side issue of the driver's basic resonance which comes into the potential for successful adjustment of the back-volume for the upper bass. The Fostex starts at 40Hz, the Eminence at 69Hz . That's the reason for that additional purchase, I need this driver to help me understand the effects better.
Ps. I do not have the option to 'EQ' , other than perhaps one notch filter. The basic LP filter will be in the amp so the Zout from the amp stays low to the speaker. There may be an additional coil at the speaker to steepen the roll-off higher up. This is another ground rule which is different between me and you. I have not heard a convincing DSP-based system to my ears - not yet. If someone wants to loan me a Najda one for a couple of weeks, I might get a bit more interested - otherwise not an option. What you're talking about above, as things being harsh or sibilant or not is mainly in the KHz region, anyway, where I'm now pretty much sorted with the 288/Yuichi and G3 ribbon. One of the advantages of working with horn loading is that these drivers are then operating a couple of octaves lower ( in the velocity-controlled region ) where their responses are smoother and less complicated.chris661 wrote: ↑Thu Mar 16, 2017 12:39 pmI see your point. I'd counter that a dB or two deviation from somewhere near flat is just fine, but what is "detail" to one person is "harsh" and "sibilant" to another, and a measurement mic tells you it's a 6.6kHz peak, Q=5, and 8dB in amplitude (Fostex FE126E, if anyone's wondering). Flattening that out with EQ will make the second person much happier, and the first person might actually like the neutrality so they can listen to a wider range of recordings instead of those that sound nice with that peak.Nick wrote: ↑Tue Mar 14, 2017 4:47 pm I think the very fact that we can live with loudspeakers in a room where the frequency response can shift by dB just by moving out head indicates that while amplitude variation is easer to measure its not that big a deal when it comes to the sound we get. We can see a similar difference in phono stage eq. we can hear and compare amplitude mistakes in the eq, but what seems to make the sound real is being correct in phase.
And yes, it sounded good Mark.
Speakers are often minimum-phase systems (except around bass reflex tuning, etc), so EQing the frequency response flat can often flatten the phase response out, too.
I don't believe that speakers are free of phase problems where they have ripples . In many of these light-coned speakers there is a dip in the region 500 to 1200Hz where the first cone breakup occurs, and I'm damned sure that means phase shifts. The 208 certainly has something around 500-600Hz, but I'll have to live with it , although I have the Eminence coming now to compare. That's the only thing that ultimately might respond to a bit of EQ if it could be done cleanly.
But what I can say from the sounds I've heard from this unit already, is that the sum total of the tone I can get from this, adding up the low end and upper end when either working correctly, is it's very nice and good enough for my purposes . It's a bit coloured at present, but there's a fair bit of improvement to come in the more open, smoother Mk.2 design, This should be fine long term. It just needs to be brought together.
Bear in mind there's more going on in the background than I can get on this thread. I do correspond almost daily with NIck, and have been emailing Bjorn Kolbrek who has been helpful. Just the last couple of days I was quizzing Troels Gravensen on his experience of the Vox Olympians at Munich , and on a ported back-volume applied to a horn & whether the Oracle/Vox had this.
'Phew, long reply'