LHC Posted November 2, 2014 Posted November 2, 2014 On 31/10/2014 at 11:49 AM, davewantsmoore said: A flat frequency response is good. It should be understood that this means not necessarily dead flat from low bass to high treble, as this can sound unnatural...... but what it means is that you don't want unintended wiggles in the frequency response. For speaker with a "falling power response" .... one which (for example) radiates lower frequencies everywhere .... middle frequencies over some forward area.... and high frequencies over a narrow area ......... we have a speaker which, when viewed from an angle, radiates more bass than mids than treble. This isn't "desirable" .... in so much as it is just what happens when you make a speaker of typical size..... the size of the sound waves VS the size of the speaker .... dictate this behavior. What we say is that the power response should "fall smoothly" ..... this means that at angle we look at the frequency response ..... even though the frequency response is falling.... it should fall smoothly .... we don't want any large wiggles in the frequency response ...... because wiggles in the frequency response ---- are wiggles in the TIME response. They are RESONANCES. "Constant Directivity" ..... is very impractical. It refers to speaker which radiates equal levels of bass, mid and treble in all directions (or all forward directions, or whatever)..... It doesn't necessarily mean that a speaker radiates the same overall SPL in all directions .... but it means that when we measure the speaker from some angle (so 60deg off to the size) .... the bass, mid, and treble level is equal (or at least not distorted from what we want it to be ---- see opening comment that "flat" may not be the ultimately desirable response in a room). Getting more towards constant directivity is a good thing .... to the extent that we cannot achieve constant directivity, we should have an even, smoothly declining power response. I wonder did I help ?!? Try some charts? Constant Directivity Quite constant Directivity above 1Khz Smooth but fast decline in directivity (power response) above 1k Very Bad - Power response does not change smoothly Constant Directivity down to low frequency - Nice Smoohtly declining power response Ok Dave, I think I can understand where you are coming from. So this has to do with the radiation pattern from the speakers and how they may interact with the room before reaching the listener's ears. With constant directivity one aims to have off-axis output that are similar to these on the axis so that any reflections from the room would combine evenly across frequency. I think this is the philosophy behind Mike Lenehan's speakers right? I recalled Rawl explaining this in another thread that allows Mike's speakers to be less dependent on room interaction. So if this can be achieved without severely compromising other areas, surely it can only be a good thing, Now the other philosophy is to accept that speakers can't radiate evenly off axis, and therefore design within that constraint to keep the fall off as smooth as possible. I can appreciate that too. But the unsolved question is which philosophy would result in the most realistic sounding speakers? I read the posts here and I don't think this has been resolved. If all else being equal, for a given room, they can't both be correct.
Newman Posted November 2, 2014 Posted November 2, 2014 On 02/11/2014 at 3:25 AM, LHC said: which philosophy would result in the most realistic sounding speakers? I read the posts here and I don't think this has been resolved. If all else being equal, for a given room, they can't both be correct. What are the two philosophies you want to compare?
LHC Posted November 2, 2014 Posted November 2, 2014 (edited) On 02/11/2014 at 3:36 AM, Newman said: What are the two philosophies you want to compare? As have been discussed before, whether we should have even radiation off-axis across frequency, or one where the levels fall off smoothly depending on the frequency. In a given room, say small and untreated, I can imagine they won't sound the same. Edit: I think Johnmath had claimed that the constant directivity speakers would sound too bright in a room. Edited November 2, 2014 by LHC
davewantsmoore Posted November 2, 2014 Posted November 2, 2014 On 02/11/2014 at 3:25 AM, LHC said: I think this is the philosophy behind Mike Lenehan's speakers right? I recalled Rawl explaining this in another thread that allows Mike's speakers to be less dependent on room interaction. So if this can be achieved without severely compromising other areas, surely it can only be a good thing, Yes, that's right. ML speakers (like all good speakers) pay attention to the off-axis sound. The crossover/EQ is designed to shape the response and blend the drivers together so that the direct sound and the off axis sound don't have any large wiggles and are as much like each other as possible. ... but because of the size/shape/type of speaker .... it will be more omnidirectional at low / mid frequencies .... and it will beam (narrow it's coverage pattern) at higher frequencies ..... so it is not possible for him to have the off axis sound be exactly the same as the direct sound. This is where Toole says that if you are going to have the speaker transition from omni directional at low frequencies to a narrower pattern in the higher frequencies, then you should transition smoothly --- which he does. To make the off axis sound the same (or closer to the same) as the direct sound ..... we need to prevent the speaker from becoming omnidirectional at frequencies we care about. The typical shape/size speaker will do this in the midrange .... and to prevent that you can either use the dipole effect (to cancel the offaxis sound) ... or a horn (to prevent the sound going where you don't want it).
davewantsmoore Posted November 2, 2014 Posted November 2, 2014 On 02/11/2014 at 3:25 AM, LHC said: But the unsolved question is which philosophy would result in the most realistic sounding speakers? I read the posts here and I don't think this has been resolved. If all else being equal, for a given room, they can't both be correct. The are not really two competing philosophies in my way of thinking. The philosophy (Toole, Geddes, Danley, everyone really) says that you should not have wiggles in the frequency response (at any angle of measurement from the speaker) .... and that the frequency response at one angle, should be a close as you can to the frequency response at other angles of radiation. Without using a horn or a dipole .... the best you can do is to have the radiation pattern change gradually from wide to narrow. Toole confirmed this via listener tests.
davewantsmoore Posted November 2, 2014 Posted November 2, 2014 On 02/11/2014 at 3:49 AM, LHC said: Edit: I think Johnmath had claimed that the constant directivity speakers would sound too bright in a room. Depends on the room (size). The speaker (whether it is "constant directivity" or not) needs to be equalised to keep the sound balance correct..... A constant directivity speaker might need it's balance shifted towards the bass some when compared directly with a (more typical) speaker which changes its coverage pattern from wide to narrow. The real difference (benefit) for CD in small room playback systems, is that through the lower mid frequencies (200-1000hz) the constant directivity speaker (unless it is an omnidirectional) will have a narrower coverage pattern than a normal speaker. The sound is being focussed over a more narrow area, rather than being radiated everywhere. This sets the correct balance of direct to reflected sound --- which is essential for clarity.
Recommended Posts