@staticV3: can you explain the cause why the Cadentia 3 reaches incredibly good 8.92 with perfect (reaching 14 Hz) subwoofer(s) and eq
only tied by the small Genelec 8030C (8.97) or the Genelec 8331 (8.93),
but the Blade 2 Meta reaches only 8.58?
I'm not sure exactly why you are asking but since I enjoy talking about the score.
It has been discussed already the Harman score, while very accurate in the correct context, was never meant to be handled as so granular.
We really should be at
most rounding to the nearest 1/2 number. Even that is really asking more from the data then is actually there.
Mic variations, variations in humidity, temperature, air pressure and normal driver and crossover component tolerances could easily cause scores to vary by 0.5pts or even more.
If a speaker is 8.23 on a given measurement set, we could round down to 8 even, then take any speakers that are in the confidence range from 8, which I think is +/-1 here and include them as near equals at least in terms of actual user preferences when blind tested.
So a calculated 8.23 could be an actual 7-9 and all speakers in this range have strong potential to be the preferred one.
Obviously there are other factors not considered by the score that may be very important (dispersion is a huge one, along with ultimate clean SPL capability) and the score itself doesn't apply to nearfield, tower speakers, or high SPL playback. The score was generated in a clinical setting at I believe a modest 80db.
I listen louder then that often and we do not hear in a linear fashion especially in the bass regions so it has not been tested whether a different model is needed even for simply louder playback.
I imagine that since we hear a 5-6db increase in bass as doubly loud but a 10db increase is required for the same perception of doubly loud in the midrange that the 80db model is not correct for loud playback.
Personally I enjoy the score because it is fun, reading the complete data set is far more rewarding and informative for actually buying and designing.
The score is valuable and overvalued all at once.
There is just no point in analyzing 0.3 and 0.5 score variations as they are more noise then anything else.
Also I want these speakers, the kit that is. What is holding me back is that tweeter. Generally I prefer hard domes and most of the time my favorite speakers have metal domes but that is really hard to parse out as I like some speakers with soft domes very much and my BMR speakers are wonderful as well with the RAAL tweet. That said what a great kit, I must resist! (but can I?)