So what alternative Mid range drivers can I look at?
Good dome drivers are few and far between and expensive, plus it's been said by many "better a good cone than a poor dome".

So which cone driver, the Scanspeak Illuminator Series are very good and measure pretty flat.
The SB Acoustics Satori Series measure just as well at around half the price of the Scanspeak.
There is also the option of using a full range driver to cover just the mid range and have plenty of scope for crossover position. Mmmmm...........

Before I get too carried away, I'm going to re-visit the D-1100 4" Carbon fiber mid/woofer out of my Quad S1 bookshelf speakers (Pictured below). I never did measure the individual driver response and no information is available from quad, I'm curious as to its frequency range and how flat the curve is.



At this point, I'm also curious as to the accuracy of the measurements I'm taking using the Hypex Filter Designer software (HFD). I will measure this Quad driver with HFD and do a similar sweep with REW for comparison. The Mic calibration file will be loaded into each App. and the Focusrite Audio Interface will be calibrated using loop back.







The first thing to say about the two plots above, is that they show exactly the same results, giving me confidence to proceed using the HFD software.
Second thing to notice is that this drivers response isn't particularly flat, it does have a very wide bandwidth, but then it is a Mid/Woofer. When this driver is coupled with the Fountek Neo X1.0 Ribbon Tweeter (crossed at 3,100Hz) in the Quad S1 bookshelf speaker, the results are very good, particularly with vocals, a point which several reviewers have picked up on. I think it's worth spending some time with this driver, and seeing how it sounds after the response has been smoothed out a little. There is certainly no attempt at this in the stock passive crossover from Quad. I may not use this driver, but it could give me an insight as to what a well matched cone mid might achieve.






LEFT SPEAKER

I flattened the response of the Quad Midwoofer and Fountek Ribbon Tweeter.
This measurement taken in REW shows the crossover region at 3100Hz LR24dB/Octave which is the stock frequency, but I used a steeper slope. I placed the mic mid-way between the drivers at 36cm distance, it may have picked up slightly more of the midwoofer output at this location. Final attenuation measurements will be taken from my listening position and also compared by ear.
One annoying thing with both REW and HFD (AFAIK) is that they only measure up to 20kHz.
If you have a supertweeter or a ribbon tweeter like the Fountek, which is flat up to 35kHz, you miss out on what is happening at the top end. It would be nice to see if it continues flat after EQ has been applied.

This modest set up from the Quad S1 bookshelf speaker can be EQ'd impressively flat, but what does the overall system sound like when partnered with the woofers?

I decided not to measure from my listening position, as the previous attempt, proved to be fantastic from the exact sweet spot, but was severely flawed when moving just a foot to either side. My new approach was to optimize each driver at near field and set the attenuation by ear from my listening spot. This didn't measure well, either from 1m or from my spot, as all the room effects where included, but subjectively it sounded superb and seemingly over a larger portion of the room. Maybe this was the approach to take and later on just work on removing/reducing the worst of the room effects. After all, if I had a musician playing in my living room, his output would be subject to the room and its effects.

One thing that has come out of this so far, is that I won't be using the Yamaha midrange.
Even this little 4" Carbon midwoofer by Quad has a more realistic tone across its range and reproduces the lower midrange of piano better. It's not quite as good at the upper keyboard or with brass instruments and doesn't image quite the same, but only by degrees, it's a lot better at the things the Yamaha doesn't do well. Given a straight choice between the two, I would choose the Quad set up, I couldn't live with the Yamaha long term. This seems to have been the case with every variant and modification I have tried. The NS-1000M does some things very well and you have to accept its flaws, or move on. Reluctantly I will be doing the latter.

As for the Peerless HDS Tweeter, I was hoping the waveguide would flatten the rising response and allow its use with minimal, or no EQ. This wasn't the case and when taking into account the complex mounting required, I think there are other, simpler Tweeter options more suited to the task. So I won't be using this either.

Right, this has really thrown the cat amongst the pigeons.
If this relatively humble 4" cone can produce such good results, what would happen if I used a better quality cone midrange?

This means I'm almost re-starting from the beginning with this project.
What started as an update of the Yamaha NS-1000M has become a search for my ultimate 3-Way active speaker solution. I will still keep all the information gathered to-date though, to show how I got to this stage. The only thing that is fixed at this point is the Woofer in its 55L sealed cabinet. It measures well and sounds superb. I will decide on a midrange then look for a suitable Tweeter.
Lots of specification sheets to go through and I need to read up on other peoples builds and experiences.

Which ever drivers I end up with, they have to be able to fit into the same cabinet footprint as I've been using all along and it can't be too much taller. Coned drivers will generally need their own separate sealed chamber, isolating them from the bass driver, so any drivers over about 5" diameter will be out of the question, due to space restrictions. As a bonus, this size limitation will help reduce beaming.

The search begins!