I am becoming a little obsessed on the above question 
I'm currently listening to a pair of DIY open baffle speakers in a rather large-ish room(~20'x15'). I took me some time (years) to arrive at a loudspeaker that has a very good tonal balance to my years, low levels of distortion and a controlled (dipole) polar response, at least to my notion. I consists of Eminence Alphas 15A up to about 200Hz, where a 2D array of 8 "progressively" tapered 3" wideband drivers (Visaton FRS 8) takes over up to some 4kHz where they are crossed to a Neo3 PDR (open back).
These speakers perform quite well. The only (kind of) missing piece is the often-quoted three-dimensionaity of the soundstage some systems are able to convey with proper recordings. I am NOT talking about the sensation of diffuse "spaciousness" sensation, which is there, and also the speakers are able to render a wide scene with well-positioned instruments. I am talking about the sense of depth in the scene - which is well defined, but rather 2D.
Am I chasing ghosts here ? Any suggestions to what exactly make a loudspeaker+room capable of conveying depth ?
I'm currently listening to a pair of DIY open baffle speakers in a rather large-ish room(~20'x15'). I took me some time (years) to arrive at a loudspeaker that has a very good tonal balance to my years, low levels of distortion and a controlled (dipole) polar response, at least to my notion. I consists of Eminence Alphas 15A up to about 200Hz, where a 2D array of 8 "progressively" tapered 3" wideband drivers (Visaton FRS 8) takes over up to some 4kHz where they are crossed to a Neo3 PDR (open back).
These speakers perform quite well. The only (kind of) missing piece is the often-quoted three-dimensionaity of the soundstage some systems are able to convey with proper recordings. I am NOT talking about the sensation of diffuse "spaciousness" sensation, which is there, and also the speakers are able to render a wide scene with well-positioned instruments. I am talking about the sense of depth in the scene - which is well defined, but rather 2D.
Am I chasing ghosts here ? Any suggestions to what exactly make a loudspeaker+room capable of conveying depth ?
Stereo recordings can convey depth only via amplitude clues and those clues being properly delayed from the main signal.
In the end it is an illusion that we interpret as "depth".
Among the things that can confound the perception of depth (or forward presence) are included early reflections (room reflections) of too high an amplitude and time/phase issues in the speaker itself. Placement and angle of the speakers and seating location also can make a big difference in the perceived sound stage.
Perhaps certain sorts of distortions in the system will also confound the perception of "depth".
To really see what ur speakers are doing, I suggest using only "acoustic" recordings, preferably ones that were recorded with two mics in true stereo. There are a number of labels that do this, Chesky is one (not for all though), there are numerous others. AND you can always go and record your own these days!
These statements represent the opinions of the poster and are not reflective of the staff or management of DiyAudio.com or any other person or organization on this planet. .
_-_-bear
In the end it is an illusion that we interpret as "depth".
Among the things that can confound the perception of depth (or forward presence) are included early reflections (room reflections) of too high an amplitude and time/phase issues in the speaker itself. Placement and angle of the speakers and seating location also can make a big difference in the perceived sound stage.
Perhaps certain sorts of distortions in the system will also confound the perception of "depth".
To really see what ur speakers are doing, I suggest using only "acoustic" recordings, preferably ones that were recorded with two mics in true stereo. There are a number of labels that do this, Chesky is one (not for all though), there are numerous others. AND you can always go and record your own these days!
These statements represent the opinions of the poster and are not reflective of the staff or management of DiyAudio.com or any other person or organization on this planet. .
_-_-bear
I find depth difficult to do. As Bear says, reflections can kill it, especially reflections from the front wall. Absorption and/or diffusion there can help. Abortion is easiest, but can kill the mid-bass of open baffle speakers. With no wall at all behind an open baffle rig, I've heard depth that was simply amazing, but the low end suffered.
Getting the crossover right is a big help with depth perception. A digital crossover with variable delay and phase can be a big help with depth. It can be a long process, so don't be discouraged.
Getting the crossover right is a big help with depth perception. A digital crossover with variable delay and phase can be a big help with depth. It can be a long process, so don't be discouraged.
You may want to try out some horns. The quality ones are not only smooth but they convey a sense of depth that is lacking in domes and ribbons.
I've always found small sealed boxes with high quality drivers can do the depth thing really well when partnered with excellent amplification (tube or solid state - krell always sounded more 'liquid' to me than any other solid state). I just prefer the sound of open baffle speakers these past few years (with their wider, more ambient and lifelike sound).
i think if you've got a crossover in the midband (~2-3khz) then it needs to be really, really good and phase aligned over a very wide bandwidth. I really like the following "rule" to phase integration:
If your speaker follows the above, it could have that coherency that you're looking for. Your mention of "8 progressively tapered" midrange drivers was a bit of an indicator that the speaker probably lacks driver-to-driver coherency... and that's where the depth goes.
Phil Bamberg said:
If your speaker follows the above, it could have that coherency that you're looking for. Your mention of "8 progressively tapered" midrange drivers was a bit of an indicator that the speaker probably lacks driver-to-driver coherency... and that's where the depth goes.
IMHO a significant part of the sense of "depth" perception comes from the correct balance of low midrange/upper bass frequencies ranging approximately from about 150-300Hz. If there are deep notches measured at the listening position in this range due to floor and front/side wall boundary cancellation effects, (very common) sense of depth can suffer.
These boundary induced notches can NOT be fixed with EQ, but rather must be addressed with speaker design and placement. For example notches due to the floor bounce reflection and/or floor/ceiling modes that occur with a "high" woofer placement (anything above about 0.4m from floor to woofer centre) can be addressed in a number of different ways such as:
1) Lower mounted woofer (<0.3m) with suitable crossover frequency
2) Multiple vertically spaced woofers so the notch frequency of one woofer is filled in by the other(s)
3) Flanking subs with a sub very close to the main speaker and running as high as ~200Hz or so, etc.
For notches due to front and side wall reflections your options are generally limited to flanking subs as above, (for example just behind and to the outside of the main speaker) and/or making sure that the distance from the front of the speaker to front wall and side walls is different so both notches don't stack on top of each other.
If you can get a relatively notch free response at the listening position from 150-300Hz through placement/speaker design its then just a matter of applying a small amount of EQ in the <300Hz region to get the right balance - an overall error of only 1dB either way over the 150-300Hz octave (relative to the rest of the frequency range) I find can affect the sense of depth quite a lot.
I generally adjust this balance by ear as I haven't found a way to get a good correlation between the exact frequency response in this region (as measured in room at the listening position) and the balance sounding right. There are just too many room effects going on in this frequency range for this to be possible.
As pointed out by bear, sense of depth in a recording is an "illusion", and a delicate one at that which is easily upset, but it can be quite convincing with the right speaker/room setup and recording. (Many recordings don't impart any sense of depth, the best to test with are generally minimalist acoustic recordings)
These boundary induced notches can NOT be fixed with EQ, but rather must be addressed with speaker design and placement. For example notches due to the floor bounce reflection and/or floor/ceiling modes that occur with a "high" woofer placement (anything above about 0.4m from floor to woofer centre) can be addressed in a number of different ways such as:
1) Lower mounted woofer (<0.3m) with suitable crossover frequency
2) Multiple vertically spaced woofers so the notch frequency of one woofer is filled in by the other(s)
3) Flanking subs with a sub very close to the main speaker and running as high as ~200Hz or so, etc.
For notches due to front and side wall reflections your options are generally limited to flanking subs as above, (for example just behind and to the outside of the main speaker) and/or making sure that the distance from the front of the speaker to front wall and side walls is different so both notches don't stack on top of each other.
If you can get a relatively notch free response at the listening position from 150-300Hz through placement/speaker design its then just a matter of applying a small amount of EQ in the <300Hz region to get the right balance - an overall error of only 1dB either way over the 150-300Hz octave (relative to the rest of the frequency range) I find can affect the sense of depth quite a lot.
I generally adjust this balance by ear as I haven't found a way to get a good correlation between the exact frequency response in this region (as measured in room at the listening position) and the balance sounding right. There are just too many room effects going on in this frequency range for this to be possible.
As pointed out by bear, sense of depth in a recording is an "illusion", and a delicate one at that which is easily upset, but it can be quite convincing with the right speaker/room setup and recording. (Many recordings don't impart any sense of depth, the best to test with are generally minimalist acoustic recordings)
Last edited:
3D is something i always strive for.
It is important that electronics preserve the information, but the speakers are very important.
Seems to be one of the things that EnABL enhances. In my experience it is almost an expected result transforming a flat soundstage into one that often blows out the wall behind the speaker. I often "measure" how good an untreated speaker is by how much it can portray a 3D stage. The new CSS VRX126 wide band mid (heard only once in a somewhat unfamiliar system) is the latest driver to impress.
3d-ness (with the appropriate recording) is the clue i use to tell an EnABLed speaker from its untreated version in blind tests.
dave
I am becoming a little obsessed on the above question
I'm currently listening to a pair of DIY open baffle speakers..
These speakers perform quite well. The only (kind of) missing piece is the often-quoted three-dimensionaity of the soundstage some systems are able to convey with proper recordings.
Any suggestions to what exactly make a loudspeaker+room capable of conveying depth ?
How close are they to the "front" wall?
Stereo recordings can convey depth only via amplitude clues and those clues being properly delayed from the main signal.
In the end it is an illusion that we interpret as "depth".
Among the things that can confound the perception of depth (or forward presence) are included early reflections (room reflections) of too high an amplitude and time/phase issues in the speaker itself. Placement and angle of the speakers and seating location also can make a big difference in the perceived sound stage.
Perhaps certain sorts of distortions in the system will also confound the perception of "depth".
To really see what ur speakers are doing, I suggest using only "acoustic" recordings, preferably ones that were recorded with two mics in true stereo. There are a number of labels that do this, Chesky is one (not for all though), there are numerous others. AND you can always go and record your own these days!
These statements represent the opinions of the poster and are not reflective of the staff or management of DiyAudio.com or any other person or organization on this planet. .
_-_-bear
hi bear,according to this "Stereo recordings can convey depth only via amplitude clues and those clues being properly delayed from the main signal. " can you refer me to some other reading..or paper..etc.? i am very very interested to read more about those.., thx..thx
Hi Dave,
So, you've (already) heard the VRX126... Can you give us any further impressions and what was the setup; run with the new tweeter or solitary- also was the VRX "Enabld"??? Thanks.
Wow, thanks for the many useful insights in your replies, guys ! It is interesting to see the different takes at the subject from different people
@Godzilla: "I've always found small sealed boxes with high quality drivers can do the depth thing really well"
So that would actually imply that acoustically small (compared to wavelength) sources - i.e. close approximation of point sources - do the depth thing better than larger radiators and polar response alone is not the whole story. That would also be one of the reasons why small fullrange drivers have the reputation of good imaging.
@RockLeeEV: "Your mention of "8 progressively tapered" midrange drivers was a bit of an indicator that the speaker probably lacks driver-to-driver coherency... and that's where the depth goes"
I also had the array un-tapered with all drivers radiating the same signal. It was not better. Actually, the tapered version, which comes closer to a "point source", has better imaging - as I said above, there seems to be some influence from the apparent size of the radiating area. But I take the point on crossover phase. Mine has rather good phase tracking, but probably not in the 5 degree tolerance you quoted. It is also not in the 2-3 kHz, but rather 4kHz.
@DBMandrake: I also remember to have read something about the tonal balance influence on depth, namely that the spectral composition of sounds is different for different distances and the brain relies on this type of information. I'll check again my in-room response, although getting rid of notches can be tough.
@ScottG: speakers are ~1 meter away from a (reflective) wall. I'll do some experimenting, although I will not be able to move them much further away on a permanennt basis
So the things I gathered so far as worth investigation:
1. apparent source size, with point source being better than larger radiators. Unfortunately, though, larger radiating areas sound better, with more authority and I prefer them from this perspective. It seems though that there are large speakers out there that do the depth thing really well, so...
2. electronics
3. narrow radiation pattern like horns: I prefer dipoles in general for their free, airy sound
4. Enabl
5. crossover phase behavior
6. speaker placement and room reflections
I'll probably start with 6.
@Godzilla: "I've always found small sealed boxes with high quality drivers can do the depth thing really well"
So that would actually imply that acoustically small (compared to wavelength) sources - i.e. close approximation of point sources - do the depth thing better than larger radiators and polar response alone is not the whole story. That would also be one of the reasons why small fullrange drivers have the reputation of good imaging.
@RockLeeEV: "Your mention of "8 progressively tapered" midrange drivers was a bit of an indicator that the speaker probably lacks driver-to-driver coherency... and that's where the depth goes"
I also had the array un-tapered with all drivers radiating the same signal. It was not better. Actually, the tapered version, which comes closer to a "point source", has better imaging - as I said above, there seems to be some influence from the apparent size of the radiating area. But I take the point on crossover phase. Mine has rather good phase tracking, but probably not in the 5 degree tolerance you quoted. It is also not in the 2-3 kHz, but rather 4kHz.
@DBMandrake: I also remember to have read something about the tonal balance influence on depth, namely that the spectral composition of sounds is different for different distances and the brain relies on this type of information. I'll check again my in-room response, although getting rid of notches can be tough.
@ScottG: speakers are ~1 meter away from a (reflective) wall. I'll do some experimenting, although I will not be able to move them much further away on a permanennt basis
So the things I gathered so far as worth investigation:
1. apparent source size, with point source being better than larger radiators. Unfortunately, though, larger radiating areas sound better, with more authority and I prefer them from this perspective. It seems though that there are large speakers out there that do the depth thing really well, so...
2. electronics
3. narrow radiation pattern like horns: I prefer dipoles in general for their free, airy sound
4. Enabl
5. crossover phase behavior
6. speaker placement and room reflections
I'll probably start with 6.
I meant the array itself operating over the midrange... i agree with the taper. I wouldn't bother without the taper personally so you're on the right track as it is.
Even if it has good phase tracking on axis, the phase tracking on the vertical axis is probably changing and that means different signals being reflected off the floor and ceiling. Try treating the floor the ceiling for starters, and working on improving phase tracking more and more. I would use 6" or thicker of OC703 or Rockwool 60 to make sure the absorption is balanced.
And a 4khz crossover isn't likely to have good center to center spacing I'd imagine... but i'm sure you have your reasons. Personally i'd go as low as the tweeter will allow. So again, even if the phase tracks well on-axis, it probably falls apart in the vertical off-axis so your ear gets mixed signals. Even a huge 12" coaxial while a larger radiator, has superior center to center spacing if you think about it.
So my suggestion would be to to lots of treatment on the floor and ceiling reflection points and see if you get any improvement. I'd imagine you do but can't say with any sureness.
Last edited:
At Bob's. diyHiFiSupply tube pre/old old Crimson power Amp (the weak link)/Oppo player (don't know which).
PE 1/4 ft^3 box (sealed or porte?). 1st with the XBL dome tweeter cap XOed (too low i think), and then by itself. Does a crediable job as a FR, but this is where it struts its stuff. Getting a tweeter to integrate might be tricky (i have a pair of RT1 and some Foster planars i'll start with) -- Bob said they didn't like what any series inductance did, we'll see. It is designed to be used with woofers. An 8 and a larger one in the same family are likely.
Very smooth and uncoloured, with the best 3D image of any stock driver i've heard.
Very nicely made. Uses a heat sealed woven poly cone -- a cone material developed by Wilson Benesch i understand.
I won't get any to play with until the boat lands...
I can see it is systems similar to the von Swickert (sp?) Unity 3 in place of the FX120 (which is not nearly as good)
dave
In my experience, electronics are always ok enough, though some might sound more open than others. But many speakers sound as if their designers were absolutely unconcerned regarding depth, openness, etc... Especially when designed by guys who consider that freq response flatness is the only thing that matters, and phase issues are considered not audible.
For me a 20$ full ranger with a 20$ Tamp might be ok. Not sure that a phasy egglestongranutopiaslam mess is even acceptable...
Absolute phase is not going to be audible, as long as the order of drivers is correct (that is, tweeters arrive last, bass arrives first etc). There's nothing wrong with for example a 4th order crossover. We just don't hear the 360deg phase shift.
Relative phase is most certainly audible. We can hear differences in phase between drivers.
Last edited:
You concepts about absolute and relativeness seems to be absolutely relative.
Anyway enjoy with anything you might consider relative or absolutely OK...
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.