Flat on-axis (quasi)anechoic frequency response is almost meaningless in a normal living room (whither these SF speakers are designed), if we not know the directivity of the speaker at different frequencies and the room acoustics (including listening and speaker positions). Assuming the speaker has no other major faults. IMHO.
I have shown a technical and reasonable argument of why there should be an audible benefit to the "BBC dip" - whether it is the result of purposeful design or discovered by serendipitous chance. To reiterate...
When laterally recorded information is reproduced from a frontal direction (as it is necessarily in stereo reproduction), the head shadowing that would have been apparent were the listener in the original recording environment is no longer apparent - at least to the same extent. A dip over this frequency range can then ameliorate the overly bright lateral information to a limited extent where the source material is recorded over a sufficiently wide angle.
Such gross compensation is, however, only "correct" for one particular source angle and therefore a compromise for sound sources at other angles. It is only applicable to stereo and there are better ways of providing that compensation - and certainly better ways than targetting a non-linear magnitude response in the loudspeaker design process.
But surely most stereo information is recorded on the axis of the mics, minimising the lateral I/P to the them as a result of their polar responses.
Therefore in the recording process the lateral information is, relative to the on axis information, much attenuated because of the polar responses of the mics. (Lobes).
The recorded sound is then replayed from speakers whose axis is again pointing directly at the listener.
I accept that in the replay process the recorded on axis information is, because of the horizontal dispersion of the speakers, directed to the listening room boundaries, and reflected back to the listener.
But as you assert, the ear is less sensitive to lateral energy, and head shadowing of each 'opposite' speaker to a given ear, still occurs.
Therefore in the recording process the lateral information is, relative to the on axis information, much attenuated because of the polar responses of the mics. (Lobes).
The recorded sound is then replayed from speakers whose axis is again pointing directly at the listener.
I accept that in the replay process the recorded on axis information is, because of the horizontal dispersion of the speakers, directed to the listening room boundaries, and reflected back to the listener.
But as you assert, the ear is less sensitive to lateral energy, and head shadowing of each 'opposite' speaker to a given ear, still occurs.
Last edited:
That is likely but not necessarily so. More importantly, ALL stereo encoded information beyond that in the mono (M) channel is the lateral (S) component.
Reproducing the S-channel correctly at the listener's ears would require laterally positioned speakers or at least one extra speaker at the rear of the listener (as in Ambisonics). Making such arrangements work well is stereo is not trivial, however, and two or more speakers over a relatively narrow angle in front of the listener is (in most cases) the best solution.
It has nothing to do with the directivity of the loudspeakers - barring the "serendipitous" outcome apparent in many conventional two-way designs. There is lateral energy in the recording environment that (if high fidelity reproduction is the target) requires reproducing laterally at the listener. So there is an inevitable compromise to be made.
However, it is a further serendipitous outcome that (particularly lateral) reflected energy in the listening environment can serve in compensating for some of the other detrimental factors in stereo reproduction, such as those arising from the inherent comb filtering.
"Off the Wall" can only refer to Michael Jackson!
Michael Jackson said:
If we are talking about the BBC Dip it's about creating an illusion of reality.
Homepage | PMC Loudspeakers
Graham Audio - British manufacturers of high quality loudspeaker systems
Harbeth Loudspeakers
All three very successful companies are in the BBC tradition. If I could pick out something surprising but doubtless true about all of their best engineers, it is that they all know what real music, instruments and voices in a real venue sounds like. I'll bet you they all have been to a Prom in the Albert Hall, London, and heard the real thing.
For sure, Alan Shaw of Harbeth listens to his daughter's voice over his speakers to assess them: YouTube
I think we all want a takeaway from all these musings. That's what a good conversation gives you.
All three companies use SEAS Tweeters. If you aren't too cash loaded, you could probably do worse than these Peerless 830874 6" woofers.
Even Joachim Gerhard likes this idea with the Sonics Oumnia 2.0, even if he can be a bit "off the wall" sometimes:
But who we kidding, it's 3 way or no-way if cost is no object!
And this is 3 way improved, IMO!
The Stirling Broadcast LS3/6 has a slope on frequency response:
Stirling Broadcast BBC LS3/6 loudspeaker Measurements | Stereophile.com
As does the Harbeth Super HL5+ :
Harbeth Super HL5plus loudspeaker Measurements | Stereophile.com
PMC, I don't know much about. They use transmission line and paper cones AFAIK. But if they are good for mixing Miles Davis "Kind of Blue" into Dolby Atmos, they can't be rubbish:
YouTube
I thought their new line of wall-mounted reflex boxes was fascinating.
The original Spendor BC1 seems to have more of a monitor (flat) balance.
Spendor BC1
I have a feeling a monitor needs a well-damped room. One with thick walls, drapes and carpets. A studio, in effect. A domestic speaker works better in a more average lively sort of room.
I am a great fan of Troels Gravesen, who has built more boxes for his lively Danish rooms than most of us have had hot dinners: DIY-Loudspeakers
He takes the view that you should beware anything above 2dB peaks, and this applies particularly to sibilance. They INTRUDE and ANNOY and spoil the musical experience.
Stirling Broadcast BBC LS3/6 loudspeaker Measurements | Stereophile.com
As does the Harbeth Super HL5+ :
Harbeth Super HL5plus loudspeaker Measurements | Stereophile.com
PMC, I don't know much about. They use transmission line and paper cones AFAIK. But if they are good for mixing Miles Davis "Kind of Blue" into Dolby Atmos, they can't be rubbish:
YouTube
I thought their new line of wall-mounted reflex boxes was fascinating.
The original Spendor BC1 seems to have more of a monitor (flat) balance.
Spendor BC1
I have a feeling a monitor needs a well-damped room. One with thick walls, drapes and carpets. A studio, in effect. A domestic speaker works better in a more average lively sort of room.
I am a great fan of Troels Gravesen, who has built more boxes for his lively Danish rooms than most of us have had hot dinners: DIY-Loudspeakers
He takes the view that you should beware anything above 2dB peaks, and this applies particularly to sibilance. They INTRUDE and ANNOY and spoil the musical experience.
How true! Troels just keeps cranking them out. His web site is a trove of DIY information. It's too bad he had to quit sharing his crossover details because people were selling his designs commercially.
FWIW. Here are measured responses (listening window) of a 1978 Rogers LS3/5a and a 1996 KEF LS3/5a. Jeff Bagby used the Rogers’ response as a target FR when he designed his LS3/5a clone called the Continuum. The second is the Rogers response compared to the Continuum response.
Attachments
Slopes are another philosophical idea; different from the dip idea.
PMC does seems to follow flat philosophy.
IMO the home is generally better damped than many studios, especially for HF.
There is evidence, (Toole) that peaks are far more intrusive than troughs, perhaps unsurprisingly.
I am little late to the party, but...
In the first graph (post #174 by system7), on the lower black line (energy response) there is a dip from 1.5 kHz to 3 kHz, and suddenly the tweeter is peaking ("shouting") from 3 kHz to 15 kHz. On the second graph, the lower black line is very smooth, with gently downward slope, without tweeter "shouting", indicating very good crossover design.
Last edited:
Yes, if it is an anechoic response. But downward slope power response is essential for neutral sound.
There's more to a good crossover than just a network. Wasn't that around the time of the move to Jantzen products?
Loudspeaker power response can not be measured on-axis as far as i know
In-room response at listening chair should have a slightly downward tilted response, how much depends how you measure and many other things
My comments were in the context of those probably not very expensive speakers, I would expect much better from a Salon2.
That is exactly how the power response is measured, in a reverberation chamber (room).
+1
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.