Oh, sorry, looks like I did a sloppy cut 'n' paste.
I have issues with this comparison.
(1) The size and type of anomaly are different:
Rutcho's worst case MA plot (for the Alpair 10P-A) shows a series of sharp ~10dips followed by equally sharp ~10dB spikes.
Troels' worst case tweeter plot shows fairly wide undulations which add ~2dB variance to the FR plot. This would make little difference to the spikes of the 10P-A plot.
(2) 2dB is the worst case scenario. It does not apply. Small tweeters are more sensitive to mounting than larger (more directional) drivers such as the Alpair 10P-A.
Your own source, Troels, rules out the applicability of your "obvious comparison"
The-Loudspeaker
Testing driver routings. Did I tell these cabs are huge?
No need to flush mount drivers - which makes it much easier.
No need to flush mount drivers - which makes it much easier.
Other sources (with repeatable measurements) agree. The ripple Zaph shows when surface mounting a woofer is much smaller than for a worst-case tweeter.
Zaph|Audio
There is "detailed knowledge" for the comparison you chose. Troels and Rutcho both state:
-test gear used
-mic distance
-signal level
-enclosure / baffle types.
-mic distance
-signal level
-enclosure / baffle types.
...so you could reproduce these tests yourself if you wanted to. Reproducibility is a great way to spot the difference between "trusted" knowledge and magical thinking.
Reproducibility - Wikipedia
You gotta admit that measurements will tell you something about the driver.
e.g. when looking at a series of similar drivers, there's a relationship between diaphragm size and where their breakup modes appear. I can look at Rutcho's measurements, and guess that I'd rather listen to the 70A than the 10P-A. The smaller driver has milder peaks, higher up in frequency, so I wouldn't hear them as clearly.
If I measured those two drivers myself, the same peaks would show up, at the same frequencies. That is: FR plots are a reproducible way to compare these two drivers. Reproducibility - Wikipedia
Subjective descriptions are not reproducible, e.g. if one dude is 60, and the other is 20, the latter will have "test equipment" that is about 40dB more sensitive at 8kHz, and could easily reach different conclusions. For anything else in audio, a 40dB difference would be a really, REALLY big deal.
Rutcho's tests show on and off-axis response. That's 3 points in space. Anyone with similar gear can repeat this. *
Anyone with a useful mobile device can plug in a iMM-6 mic, play noise (hiss) and take a 10 second reading while moving the mic in a small circle around their normal listening spot. That's ~infinite points. Cost of mic and app = $20.
IMO, it is fairly easy to get a visualisation of "all the averaging that goes on".
*I do agree about the "one point" thing. Aiming for a decent, consistent FR over an area will usually be a better goal than aiming for a perfect FR when exactly on-axis.
...but single point tests are ideal for doing before and after measurements of a tweak. Because of the reproducibility.
PS: Yes, I put the Reproducibility link in twice. Pun intended.
That is allright, but is was merely an example of how much difference a "tiny detail" can cause. I chose this as an example because the MA drivers I have are intended for flush mounting, and before purchasing them, I did read that avoiding to do so would cause unwanted effects such as for instance ragged response. If flush mounting can cause a measureable difference like this, what about baffle edges, loose cables, microphone stands etc etc. It adds up, and none of these things are mentioned at all AFAIK on Rutcho's page.
Not trying to start an argument here, just saying that measurements are very hard to standardize to such a high degree that you can get the same results in two different locations, even when using anechoic chambers, and when there are two different persons doing the measurements, well. I think maybe you get the point.
I do not have knowledge about all the circumstances surrounding various measurements, so while some measurements tell part of a story, if there where more measurements from other alternative sources, the picture would get increasingly clearer.
Nice to know we can agree on something.
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.