I have had a few people ask us why the recent NRC measurements on our Sierra-1 do not look as flat as the measurements on our site. (http://www.soundstageav.com/onhifi/20070901.htm) There are a few reasons for this and I thought it would be a worthwhile thread and something that can be referenced to.
The two curves in question are actually quite close to each other. Attached is a picture overlaying our response curve on top of the response curve taken by the NRC. Our curve is green, the NRC curves are in black. It is important to note that the NRC graph shows on-axis, 15 degrees and 30 degrees off-axis. These off-axis graphs can be identified by the high-frequency roll-off. With my newfound Photoshop skills , I have precisely aligned the graphs.
In the range from 80Hz to 20 kHz, the NRC curve and our curve are within ± 0.5 dB, aside from a small peak at 800Hz that is 1.5dB higher then evidenced on our graph. That is remarkably close. The small 1.5dB difference is due to mic positioning. The NRC (where the measurement was taken) positioned the mic at tweeter level while our graphs are taken with the mic positioned precisely between the tweeter and woofer. It is important to note that most loudspeaker companies keep to a tolerance of ± 3dB (this is the audio industry standard.) Even SEAS themselves, a world renowned high-end tweeter manufacturer (who's tweeters we use exclusively) keep to a ± 3dB standard.
The closeness of these two graphs is a very good indication of just how tight our tolerances are. It would be highly unlikely that your non-Ascend left speaker matched your right speaker as closely as these curves match up and these were taken in different facilities using different equipment and using different Sierra-1. Our tolerances and QC procedures are the tightest in the industry.
I must also highlight that the NRC measurement does not include the output of the rear port, so that from 80Hz and lower, the NRC graph is not accurate (all you are looking at is the woofer output, no port). This is because the Sierra-1 is a rear ported loudspeaker and directly behind the speaker are 54" thick fiberglass foam wedges, designed to absorb all sound waves that hit them (the NRC is a true anechoic chamber).
The rear port of the speaker fires directly into these wedges, rather than firing towards the microphone, thus the output of the port is nearly fully absorbed. Where the rear port output begins to diminish and the front firing woofer takes over once again, this too is evidenced in the NRC response graph, where bass starts to increase in the 30-35Hz range. This is all clearly indicated on the impedance curve of the speaker where the impedance peak at 80Hz is where the port tube output begins and the woofer output begins to diminish, with the woofer output at its lowest point (port tube output is greatest) at the impedance saddle at 50Hz, and then woofer output takes over and port tube output is minimal at the second impedance peak at 30Hz. http://www.ascendacoustics.com/pages.../srm1meas.html
I have attached a picture of a speaker in an anechoic chamber to give you a visual representation of how a rear ported speaker would be "crippled" in an anechoic chamber (that guy is NOT me -- I don't even own a suit )
Sierra-1 NRC Measurements
, I have precisely aligned the graphs.
and these were taken in different facilities using different equipment and using different Sierra-1. Our tolerances and QC procedures are the tightest in the industry.
)
