A very nice midrange is the B&W FST, you can't buy it OEM of course... 
Here is the info
I have a centre that has this driver WITH a normal ring, and my main 804 speakers have the FST. The difference between the 2 is audible, especially in women's voices.
The HTM2 center is not bad at all however, absolutely clear dialog.
Here is the info
I have a centre that has this driver WITH a normal ring, and my main 804 speakers have the FST. The difference between the 2 is audible, especially in women's voices.
The HTM2 center is not bad at all however, absolutely clear dialog.
Rick,
we can add a lot of 2nd and 3rd harmonics to sinewaves without easily hearing it, but when you put some wide bandwith komplex signal into the driver tons of new crap will be created from the non linearities that are responsible for the low order products.
It´s correct as you mention that most of the distortion from a driver is lower order, but there are still some energy higher up.
For example ewven if you where unable to discern a 10% 2nd order harmonic content on a sinewave from a driver, that driver would be totally useless for audio.
Todays typicall better drivers have below 0.5% 2nd and 3rd harmonics at 90dB in the midrange. Still as Dan points out, there will be lots of more distortion when music is playing. If driver technology continue to improve I think we will hear better and more clear sound as we get lower distortion. 1%HD in the bass and 0.1%HD in the 100Hz+ range would be worthwhile when measuring with single sinewaves at 90dB IMO.
Take a look at Dan´s posted links earlier in this thread on driver distortion where you clearly can se the harmonic distortion and intermodulation distortion with one and two sinewaves playing.
/Peter
we can add a lot of 2nd and 3rd harmonics to sinewaves without easily hearing it, but when you put some wide bandwith komplex signal into the driver tons of new crap will be created from the non linearities that are responsible for the low order products.
It´s correct as you mention that most of the distortion from a driver is lower order, but there are still some energy higher up.
For example ewven if you where unable to discern a 10% 2nd order harmonic content on a sinewave from a driver, that driver would be totally useless for audio.
Todays typicall better drivers have below 0.5% 2nd and 3rd harmonics at 90dB in the midrange. Still as Dan points out, there will be lots of more distortion when music is playing. If driver technology continue to improve I think we will hear better and more clear sound as we get lower distortion. 1%HD in the bass and 0.1%HD in the 100Hz+ range would be worthwhile when measuring with single sinewaves at 90dB IMO.
Take a look at Dan´s posted links earlier in this thread on driver distortion where you clearly can se the harmonic distortion and intermodulation distortion with one and two sinewaves playing.
/Peter
Re: Relative and real distortions
Rick,
Every CES, I get the pleasure of spending an evening or two with Earl; this year it was with Earl and his wife, Dr. Lidia Lee (yes, there I was, a lonely BSEE, with Dr. Geddes, Dr. Lee, and Dr. Hyre!).
Distortion from nonlinearities is typically low order; however, the interaction of those distortion products - and from cone breakups - will result in higher order products as well.
Note too that Earl's informative (and IMHO, 100% on the mark) comments about THD relate to JUST perception of distortion; it is assumed that the frequency response will be fairly flat. This is where any higher order products exciting breakup modes will fail - you'll get a spike in the frequency response from the mode itself.
In other words, if the FR is good, then THD/IMD isn't a big problem. If the FR is bad, then don't even worry about THD/IMD - solve the FR first. If the FR is bad because of cone breakup modes, then you have to solve those, either by damping the cone (anethema to most metal cone adherents) or eliminating excitation (requiring proper filtering of the drive signal).
Dan Wiggins
Adire Audio
rick57 said:
Rick,
Every CES, I get the pleasure of spending an evening or two with Earl; this year it was with Earl and his wife, Dr. Lidia Lee (yes, there I was, a lonely BSEE, with Dr. Geddes, Dr. Lee, and Dr. Hyre!).
Distortion from nonlinearities is typically low order; however, the interaction of those distortion products - and from cone breakups - will result in higher order products as well.
Note too that Earl's informative (and IMHO, 100% on the mark) comments about THD relate to JUST perception of distortion; it is assumed that the frequency response will be fairly flat. This is where any higher order products exciting breakup modes will fail - you'll get a spike in the frequency response from the mode itself.
In other words, if the FR is good, then THD/IMD isn't a big problem. If the FR is bad, then don't even worry about THD/IMD - solve the FR first. If the FR is bad because of cone breakup modes, then you have to solve those, either by damping the cone (anethema to most metal cone adherents) or eliminating excitation (requiring proper filtering of the drive signal).
Dan Wiggins
Adire Audio
Scan-Speak vs Seas Excel
I have both speakers. No doubt, the Seas Excel W17EX002 has a brighter midrange over the Scan-Speak 18W/8535. I would guess it would be better for voice frequencies if somewhat aggressive sonics are needed. It is not better for music playback per my opinion.
Both speakers use same best quality aftermarket crossover components (not values).
I have both speakers. No doubt, the Seas Excel W17EX002 has a brighter midrange over the Scan-Speak 18W/8535. I would guess it would be better for voice frequencies if somewhat aggressive sonics are needed. It is not better for music playback per my opinion.
Both speakers use same best quality aftermarket crossover components (not values).
Not sure if I should jump in but...
I agree with Pan. There will be a rise in distortion as the harmonics excite the bell resonance. At least in the example of the w22. This is simply a nonlinear distortion being excited by a linear one. Not too hard to grasp. Dan of course is right that excursion issuses are the major source of nonlinear distortion. In the case of a driver with a sharp peak, any nonlinear distortion falling at this peak will be amplified. For metal cones with severe resonances, both high and low pass frequencies should be chosen and filtered carefully.
As to Dan's comment about breakups being nonlinear, though technically correct, for practical purposes this should be considered a linear phenonema. Linear distortion is reflected in an FR curve, and there is very little change in this with drive level. It would be hard to design a real world speaker, otherwise if the FR curve kept changing with drive level.
As to the GedLee metric-I agree with this kind of a metric. In essence, it just emphasizes higher order distortion, which is clearly more offensive perceptually. One can't infer that lower order distortion is ignorable. If you look carefully at my driver graphs, you can see examples of drivers that do well on higher order and lower order distortions. (Although I've yet to find the perfect driver.)
I agree with Pan. There will be a rise in distortion as the harmonics excite the bell resonance. At least in the example of the w22. This is simply a nonlinear distortion being excited by a linear one. Not too hard to grasp. Dan of course is right that excursion issuses are the major source of nonlinear distortion. In the case of a driver with a sharp peak, any nonlinear distortion falling at this peak will be amplified. For metal cones with severe resonances, both high and low pass frequencies should be chosen and filtered carefully.
As to Dan's comment about breakups being nonlinear, though technically correct, for practical purposes this should be considered a linear phenonema. Linear distortion is reflected in an FR curve, and there is very little change in this with drive level. It would be hard to design a real world speaker, otherwise if the FR curve kept changing with drive level.
As to the GedLee metric-I agree with this kind of a metric. In essence, it just emphasizes higher order distortion, which is clearly more offensive perceptually. One can't infer that lower order distortion is ignorable. If you look carefully at my driver graphs, you can see examples of drivers that do well on higher order and lower order distortions. (Although I've yet to find the perfect driver.)
"I’m sick of missing dialogue in films & TV"
Only a couple people have mentioned that you should be able to hear dialogue fine with your present speakers.
I'd guess on your room having a noise floor too high, or too many hard surfaces. Both things that building a new centre speaker will not help (unless you choose one with higher directivity - horn / dipole, and that will only help the 'bright room' scenario )
Cheers,
Rob
Only a couple people have mentioned that you should be able to hear dialogue fine with your present speakers.
I'd guess on your room having a noise floor too high, or too many hard surfaces. Both things that building a new centre speaker will not help (unless you choose one with higher directivity - horn / dipole, and that will only help the 'bright room' scenario )
Cheers,
Rob
Rob
The acoustics aren’t brilliant, correct. My wife has control of surfaces (which are somewhat on the hard side) so there’s little or no scope there.
I agree that more directive speakers are better. (I’m building horns & dipoles for two other rooms).
But why do you feel that a new centre speaker will not help ~ at all?
Whatever speaker is there will be degraded by the room. But surely a higher resolution speaker eg a metal cone like a Jordan will start from a higher base and sound better than a ‘mid-fi’ entry level (polyX?) Mission?
Cheers,
Richard
The acoustics aren’t brilliant, correct. My wife has control of surfaces (which are somewhat on the hard side) so there’s little or no scope there.
I agree that more directive speakers are better. (I’m building horns & dipoles for two other rooms).
But why do you feel that a new centre speaker will not help ~ at all?
Whatever speaker is there will be degraded by the room. But surely a higher resolution speaker eg a metal cone like a Jordan will start from a higher base and sound better than a ‘mid-fi’ entry level (polyX?) Mission?
Cheers,
Richard
These are only my opinions btw 
If the sound is getting messed up by acoustics so much that you can't understand dialogue, then I can't see how a 'higher resolution' messed up sound will help. ie the extra subtleties / nuances of a better midrange will be masked by bad acoustics.
Have you measured your noisefloor?
The reason I posted is I've got a cheap pair of mission bookshelf speakers, and they give pefectly adequate dialogue for movies/TV. Back in my 'poor apprentice' days I used loads of really really cheap rubbish, and could still watch films without problems.
Maybe a different question could be 'why can't my room produce good dialogue?'
Cheers
Rob
If the sound is getting messed up by acoustics so much that you can't understand dialogue, then I can't see how a 'higher resolution' messed up sound will help. ie the extra subtleties / nuances of a better midrange will be masked by bad acoustics.
Have you measured your noisefloor?
The reason I posted is I've got a cheap pair of mission bookshelf speakers, and they give pefectly adequate dialogue for movies/TV. Back in my 'poor apprentice' days I used loads of really really cheap rubbish, and could still watch films without problems.
Maybe a different question could be 'why can't my room produce good dialogue?'
Cheers
Rob
Greets!
Spot on, Rob.
Rick, study human hearing perception. We are amplitude based creatures, i.e. we react to what's loudest, not clearest. IOW which is the nearest 'threat'. When you hear thunder, do you stop to gauge how clear it sounds to you or how close the lightning threat is?
According to the folks who research this for a living, the key parameters for max intelligibility are:
signal-to-noise (S/N) ratio
reverb time (Rt60)
distance from source
source misalignments
reflections <1ft of path length difference
Since I gather you have little/no control over the room's acoustics, you must control the speaker's output to offset these issues as much as practical. Until this happens, the best midrange driver in the world isn't going to help you much, if any.
Obviously then, the easy 'fix' is to increase the CC's level WRT the rest of the channels, but then the presentation won't be tonally balanced, so about all you can do AFAIK is design/build a large constant directivity (CD) system with 'x' horizontal and 'y' vertical response for all three front channels based on the listening distance(s) and required coverage angles, which isn't a viable option in your (or most folk's) case. Or if you sit back far enough, possibly a ~CD L-R system and use the phantom center mode. Done right, this has a higher intelligibility than a L-C-R system, but for other than a single seating the room needs to be quite large, with a very large screen if a cinema-like presentation is desired.
The downside to the latter approach is that producers are increasingly going back to essentially a mono presentation due to intelligibility problems in the theater, with the L-R acting as pan/surrounds like was done originally. With a phantom CC, the inverse intelligibility losses due to synthesizing stereo from mono occurs. Seems the early multi-channel researchers/designers basically got it 'right' the first time around so once again the old adage, "the more things change, the more they remain the same" is apt.
On a more personal note, the little info available implies that you may have some age, etc. related hearing issues. Try cupping your ears when watching a movie. If this helps and the others aren't having any trouble with the dialog, then consider fabbing some ear baffles you can wear, or do as I do and hunker down in the sofa/chair with my head propped on a throw pillow to 'EQ' what I hear. Growing old sucks!
GM
Spot on, Rob.
Rick, study human hearing perception. We are amplitude based creatures, i.e. we react to what's loudest, not clearest. IOW which is the nearest 'threat'. When you hear thunder, do you stop to gauge how clear it sounds to you or how close the lightning threat is?
According to the folks who research this for a living, the key parameters for max intelligibility are:
signal-to-noise (S/N) ratio
reverb time (Rt60)
distance from source
source misalignments
reflections <1ft of path length difference
Since I gather you have little/no control over the room's acoustics, you must control the speaker's output to offset these issues as much as practical. Until this happens, the best midrange driver in the world isn't going to help you much, if any.
Obviously then, the easy 'fix' is to increase the CC's level WRT the rest of the channels, but then the presentation won't be tonally balanced, so about all you can do AFAIK is design/build a large constant directivity (CD) system with 'x' horizontal and 'y' vertical response for all three front channels based on the listening distance(s) and required coverage angles, which isn't a viable option in your (or most folk's) case. Or if you sit back far enough, possibly a ~CD L-R system and use the phantom center mode. Done right, this has a higher intelligibility than a L-C-R system, but for other than a single seating the room needs to be quite large, with a very large screen if a cinema-like presentation is desired.
The downside to the latter approach is that producers are increasingly going back to essentially a mono presentation due to intelligibility problems in the theater, with the L-R acting as pan/surrounds like was done originally. With a phantom CC, the inverse intelligibility losses due to synthesizing stereo from mono occurs. Seems the early multi-channel researchers/designers basically got it 'right' the first time around so once again the old adage, "the more things change, the more they remain the same" is apt.
On a more personal note, the little info available implies that you may have some age, etc. related hearing issues. Try cupping your ears when watching a movie. If this helps and the others aren't having any trouble with the dialog, then consider fabbing some ear baffles you can wear, or do as I do and hunker down in the sofa/chair with my head propped on a throw pillow to 'EQ' what I hear. Growing old sucks!
GM
GM
That’s probably the most informative response I’ve yet read. Thank you!
After googling on this a fair bit, I was very interested to hear the key parameters for max intelligibility. Do you have a link that elaborates on this?
Maybe I should have said that generally I’m the only one (of up to five pp) that has this problem, and I’ve always been like this, either watching a movie or in conversation. Hearing tests show me within normal ranges; but to correctly discern dialog I want to use all controllable factors to my advantage.
Interesting comments re the evolution of channel balancing of the mix.
You refer to large CD speakers. I know something about dipoles and a little about horns. Are all CD speakers large?
A dipole is not constant directivity, but may be directive enough.
What about a dipole horizontal on the screen, going down to say 180 Hz, about half the area of the Phoenix baffle say 20 inch by 9 inch (50 cm * 23 cm) with a 6-8 inch high Xmax driver (eg the Excel) for the mids.
As you say, increase the CC's level WRT the rest until the most effective is reached.
(In this room we sit between 5-15 feet, though balancing can be optimised for the deaf guy (me).
(I’ll try cupping my ears to see the effect)
Cheers
That’s probably the most informative response I’ve yet read. Thank you!
After googling on this a fair bit, I was very interested to hear the key parameters for max intelligibility. Do you have a link that elaborates on this?
Maybe I should have said that generally I’m the only one (of up to five pp) that has this problem, and I’ve always been like this, either watching a movie or in conversation. Hearing tests show me within normal ranges; but to correctly discern dialog I want to use all controllable factors to my advantage.
Interesting comments re the evolution of channel balancing of the mix.
You refer to large CD speakers. I know something about dipoles and a little about horns. Are all CD speakers large?
A dipole is not constant directivity, but may be directive enough.
What about a dipole horizontal on the screen, going down to say 180 Hz, about half the area of the Phoenix baffle say 20 inch by 9 inch (50 cm * 23 cm) with a 6-8 inch high Xmax driver (eg the Excel) for the mids.
As you say, increase the CC's level WRT the rest until the most effective is reached.
(In this room we sit between 5-15 feet, though balancing can be optimised for the deaf guy (me).
(I’ll try cupping my ears to see the effect)
Cheers
I have to disagree a little bit with this notion that the room acoustics are to blame for voice intelligibility problems.
The typical home room is not markedly reverberant, and the distance from the CC/screen is not so far that the reverberant sound field overwhelms the direct sound. Unless you have an unusually reverberant room and are sitting 20 feet from your CC.
This is really not the same situation as a medium or large sized theater, auditorium etc.
Nonetheless, this is a very common complaint, that CC dialogue is poor.
While a dipole or other constant and limited directivity design would work well, a standard box two way with-
good dynamic range(i.e. good nonlinear distortion)
attention to smooth off axis performance
smooth FR(i.e. linear distortion) in the 300-3k range
should be fine.
When I switched to my sealed Usher 2 ways, which were intentionally designed to take advantage of the main lobe tilt, the dialogue clarity went up significantly.
The other isssue is some of the more budget HT processors, or at least older ones, may not have decoded correctly. I'm not sure if that's a problem now.
The typical home room is not markedly reverberant, and the distance from the CC/screen is not so far that the reverberant sound field overwhelms the direct sound. Unless you have an unusually reverberant room and are sitting 20 feet from your CC.
This is really not the same situation as a medium or large sized theater, auditorium etc.
Nonetheless, this is a very common complaint, that CC dialogue is poor.
While a dipole or other constant and limited directivity design would work well, a standard box two way with-
good dynamic range(i.e. good nonlinear distortion)
attention to smooth off axis performance
smooth FR(i.e. linear distortion) in the 300-3k range
should be fine.
When I switched to my sealed Usher 2 ways, which were intentionally designed to take advantage of the main lobe tilt, the dialogue clarity went up significantly.
The other isssue is some of the more budget HT processors, or at least older ones, may not have decoded correctly. I'm not sure if that's a problem now.
ucla88 said:
We're kind of in agreement here - If you notice I said that I have a pair of 2-way missions, and that there is no probs with dialogue. I suggested that maybe the room was causing problems as, like you stated above, a standard 2 way should do fine.
What i didn't realise was that it was only Rick57 who was having trouble with dialogue, and nobody else in the room
That leads us back to making the dialogue clearer and I believe that a horn / dipole type speaker will help rick, giving him more direct sound versus reflected sound. More signal than noise so to speak. I've recently tried horn tweeters, and they certainly make the vocals more 'direct'.
Cheers,
Rob
Hi all
The relative importance of acoustics would depend I guess on eg
sitting in the centre of the room ~ high proportion of direct sound vs
sitting close to a wall ~ probably much lower ratio of direct sound.
Ucla can you explain “designed to take advantage of the main lobe tilt”?
I though of headphones – best SQ yes (I have some Stax electrostatics but don’t use them), but don’t want the cables, not a big fan of cans.
> budget HT processors
A bit of an unknown quality . .
> When I switched to my . .
(My two bookshelves are of similar quality). Who else has swapped speakers and got a difference?
> I've recently tried horn tweeters, and they certainly make the vocals more 'direct'.
The most important range for intelligibility is 500 – 3000 Hz.
You could get 500 Hz from a horn with mouth about 18 * 12”, length about 12” (45 * 30 * 30 cm). By itself that’s not too big, but by the time you add a woofer etc it’s getting bulky for many rooms.
If a 800 Hz was used, the size could come by about 40%, and then we’d need a midwoofer that would go up that high, an 8 incher would be good and not too big, so it’s sounding quite feasible.
A horn would help the person it’s pointed at (our room is very spread out) but not all people, but I’m the one with the problem & doing the work, so that’s ok.
The relative importance of acoustics would depend I guess on eg
sitting in the centre of the room ~ high proportion of direct sound vs
sitting close to a wall ~ probably much lower ratio of direct sound.
Ucla can you explain “designed to take advantage of the main lobe tilt”?
I though of headphones – best SQ yes (I have some Stax electrostatics but don’t use them), but don’t want the cables, not a big fan of cans.
> budget HT processors
A bit of an unknown quality . .
> When I switched to my . .
(My two bookshelves are of similar quality). Who else has swapped speakers and got a difference?
> I've recently tried horn tweeters, and they certainly make the vocals more 'direct'.
The most important range for intelligibility is 500 – 3000 Hz.
You could get 500 Hz from a horn with mouth about 18 * 12”, length about 12” (45 * 30 * 30 cm). By itself that’s not too big, but by the time you add a woofer etc it’s getting bulky for many rooms.
If a 800 Hz was used, the size could come by about 40%, and then we’d need a midwoofer that would go up that high, an 8 incher would be good and not too big, so it’s sounding quite feasible.
A horn would help the person it’s pointed at (our room is very spread out) but not all people, but I’m the one with the problem & doing the work, so that’s ok.
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.