Please evaluate - too much absorption?

[Audiophile53]

Need evaluation of acoustic treatment and whether the dip at 1khz to 3khz may be caused by too much absorption. I am using Klipschorns which require corner placement so not much of an option. My listening room is also rather uncharacteristic. Thanks for any assistance.

 

[hreindl]

your curve is pretty fine and that dip from 1 to 3 kHz is common - elevation in 1-2 kHz results ina a horrible sound and 2000-3000 Hz is where your ears are most sensitive - you don't want a complete flat curve - the only thing i would worry is the falldown starting with 7 kHz which is a bit too early but thats "brilliance" range and not that much important and pretty sure your speakers charateristics

 

[hreindl]

BTW: in case of too much absorption it won't elevate after 3 kHz while kill 1-3 kHz

 

[BenToronto]

You don't have a dip 1-3kHz, you have a boost 4-7kHz. Might not be too noxious for all (of us) who have hearing deficiencies up there. False to aim for a flat FR when a slope softer at the top is more widely preferred (and false to think one mic freq run tells the whole story of perceived loudness).

 

[DanDan]

Commonly the measurement mic is not precisely the same distance from each tweeter. A signal coming from two sources, one slightly delayed will be subject to comb filtering. You need to measure L, R, individually for full range and level consideration, Then L+R to be viewed for LF only.

 

[BenToronto]

Acoustic measurements are messy. For sure, more sensible to do one speaker (or one driver) at a time. But also interesting to have dual-mono which might reveal places where the speakers are interacting destructively. That shows up when the 3 traces are on the same graph. But I'm not sure if that means much in terms of reproducing/bringing music to your ears.

 

[DanDan]

Hi Ben, REW has become complex. But with Trace Arithmetic one can combine the individual L and Rs to approximate the frequency response delivered to the listener. This can be and is done in more sophisticated ways when required. Dummy heads. Jecklin Discs.
Or just very many samples averaged in a realistic listening zone rather than a single point. There is no possible comparison to the human experience of music. But such data is surely just as informative and useful as say Speaker Frequency Response graphs. Or Polar plots.?

 

[BenToronto]

Although it still leaves moot the original post about absorptions, Toole's Spinorama is an effort to characterize the human reception, given available physical measurements such as directivity index, total power into the room, etc.

B.

 

[DanDan]

Yes and no. The OP writes of a FR anomaly which very likely is being caused by comb filtering due to a small delay between the HF paths. Removing the smoothing makes this more visible, albeit messy looking. But the RT60 is remarkably even across most of the spectrum, a nice 280mS or thereabouts. Audiophile53, your decay data suggests a great job done on treatment. Not too much absorption. And if you measure again, L, R, separately, you may see a more gratifying FR graph. Or not, the comb filtering could be coming from other delayed paths, reflections. The ETC graph does show a lot of loud early spikes. Many of these are of course belonged to the other speaker, which is whey we need individual..... But again from your lovely decay data, and the ETC, no, you do not have excessive absorption. In fact more would be very helpful if you could target those Early spikes in the ETC.

 

[DanDan]

The Blue line is individual L and R measurements combined by the software Fuzzmeaure for viewing purposes. . The Red Line shows L+R a sweep with two speakers running simultaneously. The HF destruction is the result of a small difference in the HF paths from tweeters to central mic, perhaps some other low level early reflections from the floor or rack mounted gear. With smoothing they combine to look like a big band or broad notch. This is a well treated room.