Keith_W
Major Contributor
A lot of time and anger would be saved if instead of posting people would first search and read what has been written about the same topics in the past from persons who really know what they are talking about and take off their black and white glasses as most truths are somewhere in the middle, but it seems writing in forums is also a hobby for some.
Here are some quotes of Floyd Toole who even owned and enjoyed loudspeakers with close to omni radiation pattern:
Thanks for those quotes. Before I start, I will say that I have no skin in this game. I have heard a few omnis, and they are not as terrible as conventional wisdom makes them out to be. Whenever that happens, my interest is piqued.
I have highlighted the relevant parts of what Dr. Toole said:
Here I will point out an important scientific principle: absence of evidence is not evidence of absence*. Just because there is no evidence might mean that the studies have not been done yet, and in fact they haven't. I have been looking for studies of reflections over the past few months (and started a thread on frontal reflections) and I was surprised by the paucity of evidence. From a scientific perspective, I do not consider Dr. Toole's take on this to be definitive. He is arguing from first principles, and he may even be right. But until there is some evidence, such as a blind preference study comparing omnidirectional speakers to conventional speakers in the same room, then I am sure that even he would agree that his statement is a tentative and not a definitive one.
* Yes, there is something called the Negative Evidence Principle (NEP), where absence of evidence is indeed evidence of absence. But in this case, there needs to be so much evidence collected that the likelihood of assertion X is likely to be false. One example is that the Loch Ness Monster likely does not exist, because of the NEP.
He says as much in this quote, that the notion that omnidirectionality is ideal awaits proof.
He made some remarks about "recordings not being made with omni monitors", so let us talk about that. First, we need to make a distinction between "scientist" type arguments and "engineer" type arguments. A "scientist" type argument goes like this: "we don't know the answer, so we will conduct an experiment and collect evidence". An "engineer" type argument goes like this: "we don't know the answer, so we will derive it from engineering principles". Here is an example of an "engineer" type argument: if we take a perfect recording and play it back on a perfect loudspeaker, perfect reproduction will be achieved.
So let us see if it holds up. First, we will leave out all recordings that are not recorded with a microphone (all electronic music) or recordings where spatiality effects have been added (all recordings with Q-Sound) and only talk about acoustic recordings. What is relevant is not the monitor, but the type of microphone, how the mics were placed, and how they were mixed. To me this point seems to obvious that I am surprised that Dr. Toole uses this argument about monitors. The studios are mastering for tonality, mix, balance, etc. They have no way to manipulate the mix of reflections as captured by the microphone on the mixing desk, so the monitor is irrelevant as far as spatial reproduction is concerned! What is important is the type of microphone and where the recording engineer has placed them. That is what captures spatiality, not the type of monitor that was being used.
Recordings are commonly made with cardioid mics to reject reflections, but some reflections still get into the recording. I wonder what he would say if the recording was made with an omnidirectional microphone.
I mean, what do you think we are trying to recreate? What the mixing engineer heard on his monitors? Or what you would hear if you were sitting in the recording studio? Even the mixing engineer is trying to recreate what is heard in the recording studio or concert hall, which is why the recording engineer takes so much care in placement of the microphones. Maybe you would be trying to recreate what the engineer heard on his monitors if you are listening to music which is 100% artificial, such as electronic. But for any acoustic recording, you want what you would hear in the recording studio if you were there.
And finally, let us talk about what Linkwitz said. Unlike Toole, Linkwitz preferred dipoles and omnis. He wrote an article on ideal directivity that you can find on his website. This is the relevant quote:
He makes the important point that typical box speakers of the type favoured by Toole and Harman have an uneven power response with a 10dB drop in acoustic power between low and high frequencies and asserts that "this can not be the avenue to sound reproduction that is true to the original".
I will criticise him too. This is another example of an "engineer type argument". Note what is missing in this argument: humans. This IMO is the weakness of "engineer type arguments". No consideration is made for preference, psychoacoustics, etc. As an example, an engineer will try to recreate a perfect square wave from a speaker. A scientist knows that it is not audible. Engineers know that there is no point trying to recreate a perfect square wave, because they know it is not audible (thanks to psychoacoustic studies*). But if there were no scientific studies showing inaudibility, you can guarantee for sure engineers would be trying to make square waves come out of speakers.
* I know that modern thinking says that it is probably audible. But I am using this as an example, it is not my intention to digress from this topic. For that discussion, see the last few pages in this thread.
So we need to ask ourselves, are there studies that shows that this type of uneven power response is preferred? Well, there is Toole's study with four speakers, one of which was a dipole (a Quad ESL-63). And there are Harman's studies showing that a falling frequency response (uneven sound power) is preferred. So my take on it is this: Linkwitz is wrong here. Evidence suggests that an uneven power response is preferred by listeners.
This does not mean that you can not engineer an omni with a deliberate uneven power response. Measurements of the MBL 101e that I have seen indeed do show an uneven power response with a Harman-type falling curve. I even have a friend with a pair of MBL 101e's which he DSP'ed to produce a Harman-type falling curve. So the only argument left is whether the directivity is preferred.
Both Dr. Toole and Linkwitz agree on one point - Toole said that the listener's perception will be dominated by reflections, Linkwitz said that the "perceived sound is dominated by the room's acoustic signature". But they come to different conclusions - Linkwitz said that the speakers disappear into the sound field, which he considered a positive. He acknowledged that the direct sound is masked by reflections and affects precise imaging unless one sits closer to the loudspeaker. Toole said that this was an "embellishment".
In the end, there are a few take-aways:
1. There is no scientific evidence in the form of a preference study that shows the superiority of omnis over conventional speakers, or vice-versa.
2. In its absence, we are left with making "engineer type arguments". To me, this is a second rate argument because it ignores the human factor.
3. Scientific evidence, especially studies involving human preference, is usually fuzzy with wide standard deviations and there may still be people out there who actually prefer omnis. And as JJ said, do not argue with someone's preference.
Last edited:
