Is absolute phase audible?
YES.
Will you notice it most of the time?
NO.
Good recording for hearing absolute phase:
Mickey Hart - Planet Drum.
Cut one. Flip the phase, it sounds like a different mix in that some instruments go to the back and some come to the front. Subtle though.
Regarding direct to reflected sound - can you say "Bose 901"??
_-_-bear
YES.
Will you notice it most of the time?
NO.
Good recording for hearing absolute phase:
Mickey Hart - Planet Drum.
Cut one. Flip the phase, it sounds like a different mix in that some instruments go to the back and some come to the front. Subtle though.
Regarding direct to reflected sound - can you say "Bose 901"??
_-_-bear
>>> For me a 20$ full ranger with a 20$ Tamp might be ok.
GDO, for much of my listening i agree. I don't like crossovers and i do like T-amps! IME tubes offer the most 'depth' but they are not perfect either.
But i love 'helper' woofers and tweeters to go along with fullrangers to round out their sound.
GDO, for much of my listening i agree. I don't like crossovers and i do like T-amps! IME tubes offer the most 'depth' but they are not perfect either.
But i love 'helper' woofers and tweeters to go along with fullrangers to round out their sound.
Is absolute phase audible ? (As in, reversing the phase of both speakers)
Some will say yes, but from what I have read/heard the answer is no, our hearing mechanism itself is not sensitive to absolute phase - a tone which begins with a positive polarity (pressure increase) sounds exactly like a tone that begins with a negative polarity. (pressure decrease) Likewise an impulse that spikes in a pressure increase direction sounds the same as an inverted impulse which is spiking in a pressure decrease direction - the frequency spectra produced is the same, and that's what matters.
However this assumes that speakers are perfectly linear and they are not. One example sometimes cited as sounding different if the phase is reversed are kick drums. Again, as far as I'm aware if the speaker was perfectly linear a kick drum would sound identical whether the phase of the initial waveform rise was positive or negative.
All woofers have some degree of even order non-linearity (quite a lot in many cases) so reversing the phase would in fact alter the response of the initial waveform rise depending on whether it was causing an outwards or inwards cone excursion. This is a potentially audible difference, but its the non-linearity of the speaker to blame, not absolute polarity being audible per-se.
Also, there is no "right" or "wrong" phase in this case - it may sound slightly different one way to the other, but if even order non-linearity is present both are actually wrong, just in different ways. In that sense attempting to "preserve" waveform polarity right through the record/playback chain doesn't make sense, because the slope and shape of the woofers non-linearity will vary from woofer to woofer anyway. (One phase might sound "better" with a particular woofer, but the opposite phase could sound "better" with a different woofer due to differences in their linearity)
Many recordings will have the phase flipped somewhere between microphone and disc anyway - for either individual instruments or the entire recording, and we would never know if this had occurred. Some amplifiers invert the phase as well - I've seen amplifiers where the output is in phase with the tone defeat control active, and out of phase with the tone controls enabled. (Due to an extra inverting buffer stage being switched in)
Worrying about absolute phasing of speakers is a bit of a wasted effort IMHO, if it does sound different with the phase reversed (and can be proven in some sort of double-blind test, rather than an unreliable sighted test) then I would be looking at the speakers as the culprit, particularly if the woofer had a lot of even order non-linearity.
Last edited:
Obviously woofers and tweeters help to get a better tonal balance. But if the speaker is not able to deliver proper imaging and depth, i am afraid that there is no "helper" around.
Big nonsense multiway monsters freaks with a concern for imaging might get back to basics and find a refreshing inspiration in minimalist full range zen setups...
Perhaps, although maybe somewhere in between electrically in and out of phase is correct, with each frequency being again different.
As to why it's important, with your on-axis reasonable as it is, you'd notice differences between the on-axis response and the power response of the system and room. Your targeting of on-axis measurements for a flat response may be a red herring as to how flat your system is, also since you are using a highly room dependant style of speaker this will be all the more the case.
Crossover phase variations will vary the relationship between on-axis response and sound power contributing to this problem of it measuring one way, but sounding another. The phase itself is not the issue, and while phase matching causes the above variations, the phase matching in and of itself is not the primary concern in your case.
It is important that you can take the next step with your measurements and focus on all sound emitted from the speaker (weigh in all directions).
Taking a look at your measurements, specifically the mids contribution there seems to be a reflection at a distance of (possibly) a little under 4 feet longer than the direct path. With a variance up to +-2dB in places I'd consider that worth improving.
Then looking at the tweeter it seems you could improve it if you could find the cause of the ripples. Maybe some damping material laid over the baffle would isolate this.
Number 2, the electronics is essential. I find an overbuilt power supply has a great effect on speaker imaging. In a 2A3 SE amp, I use 18 pound Power Transformers, and dual tube rectifiers, to start to get the original recoding reproduced intact. Wire in my power trannies and chokes can do 900 mA., and the amp draws about 53 mA. My chokes are all under 10 Ohms DCR. Then you will have what you always sought. The amps, not the speakers, are the BIGGEST TURKEYS of all. With an amp, in the end, you are hearing the Power Supply. Just how well designed is it ??? Its always been that way in my experience over 50 years.
Jeff Medwin
Jeff Medwin
All good advice, thanks !
I'm also worried by the tweeter ripples. I do have some 4mm think damping material all over the baffle and the baffle has roundings, but that did not get rid of the ripples. I did not see any measurements of Neo3 PDR in dipole on a balffle, but I wouls assume it should be much smoother.
I think it's due to diffraction effects at the currently not very well done boundary between Neo3 and the baffle (the tweeter has no face plate)
Last night, I changed the crossover from 4 to 3kHz. Definitely an improvement in that instruments positions are more stable in the scene, still the depth stuff is not quite there. More experiments needed and the backwards response hole not fixed yet.
An externally hosted image should be here but it was not working when we last tested it.
Big nonsense multiway monsters freaks with a concern for imaging might get back to basics and find a refreshing inspiration in minimalist full range zen setups...
..which in turn would have some other limitations that might bring you back to other "basics" and have you try a multiway monster that overcome those limitations
There are compromises in loudspeaker design, I'm learning that the hard way. And there is not really a one-size-fits-all, different people prefer different compromises.
Yes, but you simply have to choose which compromises and limitations you are ready to live with.
For me big systems with conventional analogue xovers all suck, and their big slam for slam sake simply does not impress me at all.
If you strieve for something big, I believe that linear phase xover dsp based systems is the one and only way to go. You have to accept to put a dsp in you life, but then you might get the big sound, with a big image, and last but not least...the depthhhhh.....
For me big systems with conventional analogue xovers all suck, and their big slam for slam sake simply does not impress me at all.
If you strieve for something big, I believe that linear phase xover dsp based systems is the one and only way to go. You have to accept to put a dsp in you life, but then you might get the big sound, with a big image, and last but not least...the depthhhhh.....
Your on-axis response looks OK. What gating and/or smoothing have you used?
If you feel like trying something and have the means, EQ your on-axis response to sit on the -30dB mark (according to your last posted graph) between 2k and 6k. If nothing else it might encourage your mid/bass to be all it can be.
If you feel like trying something and have the means, EQ your on-axis response to sit on the -30dB mark (according to your last posted graph) between 2k and 6k. If nothing else it might encourage your mid/bass to be all it can be.
The green curve is 1/3 octave smoothed. The red curve (which is the same, but louder) is 1/24 octave smoothed. I don't have the exact gating any longer, but it's usually ~5ms (hope I am not wrong, I usually cut before the first big reflection in the impulse response)
I'll try your suggestion - if I got it right, you suggest to EQ the green curve down about 2dB in the 2-6 kHz area. For the current tweeter filter it might be as simple as using a smaller capacitor for the tweeter, but I need to get the phase right (currently, phase tracking in crossover is quite good, with a broad deep reverse notch)
OK. You might get a hint of the overall response if you extend the gating to include the major reflections and take it from your listening position. Smoothing at 12th octave should give reasonable results. Compare it to your gated plot.
Only if it's convenient as knowing only what I know about your system this is a shot in the dark, but generally speaking, the curve I've suggested trends well.
You may be able to go easy on the phase by simply lowering the level of the tweeter. You'll lose a little of the top end sparkle this way but it will still let you feel out the mids this way.
OT posts removed.
in my experience, room and placement are very important. the speakers that throw an impressive depth at the show room can sound like crap in your home.
I once listened to a pair of speakers that used a wrong replacement woofer (not even impedance was right). it's obvious that the magnitude and phase responses were messed up. but they still conveyed the sense of depth.
based on that, my conclusion was that sound staging capabilities have to do with, well, a lot of things.
I once listened to a pair of speakers that used a wrong replacement woofer (not even impedance was right). it's obvious that the magnitude and phase responses were messed up. but they still conveyed the sense of depth.
based on that, my conclusion was that sound staging capabilities have to do with, well, a lot of things.
Last edited:
Happy new year !
You mean, how much baffle edge diffraction negatively impacts depth, right ?
In my case, roundovers are about 1-1.25" (the wings of the baffle). I was not able to measure differences with and without them (except in the bottom end due to dipole cancelation) but, as said, I will need to fix tweeter to baffle diffraction first, that looks bad.
I changed the response curve as you previously suggested. Not much listening impressions so far (dit it last night late after the children went to bed), but it sounds at least as good as before. The response at the listening spot would be close to the second curve (~30 degrees off-axis)
I realized that the top end is a bit too lively for my taste. I have a fairly reflective room and the graphs also show a lot energy and high dispersion above 7kHz. It's nice to hear all top end details, but it sounds a bit unnatural. I'll try lowering the overall tweeter level a bit. Might also affect depth
), but it sounds at least as good as before. The response at the listening spot would be close to the second curve (~30 degrees off-axis)I realized that the top end is a bit too lively for my taste. I have a fairly reflective room and the graphs also show a lot energy and high dispersion above 7kHz. It's nice to hear all top end details, but it sounds a bit unnatural. I'll try lowering the overall tweeter level a bit. Might also affect depth
An externally hosted image should be here but it was not working when we last tested it.
Last edited:
Imo, you have a problem of power response exagerate deep and tweeter bloom. This is caused by a LR12 xover used in a MTM config at a much too high xover freq.
LR are no a good match with MTM especially for lr12, and if you cross too high... A more directional tweeter might help but this one has a too wide dispersion... for this design.
Probably a fix would be lower the xover to around 2 khz with an LR24 or but18.
Maybe this is worth reading:
Power
Though in this study, LR12 MTM is not even taken into consideration...
LR are no a good match with MTM especially for lr12, and if you cross too high... A more directional tweeter might help but this one has a too wide dispersion... for this design.
Probably a fix would be lower the xover to around 2 khz with an LR24 or but18.
Maybe this is worth reading:
Power
Though in this study, LR12 MTM is not even taken into consideration...
Imo, you have a problem of power response exagerate deep and tweeter bloom. This is caused by a LR12 xover used in a MTM config at a much too high xover freq.
LR are no a good match with MTM especially for lr12, and if you cross too high... A more directional tweeter might help but this one has a too wide dispersion... for this design.
Probably a fix would be lower the xover to around 2 khz with an LR24 or but18.
Maybe this is worth reading:
Power
Though in this study, LR12 MTM is not even taken into consideration...
Thanks for the hints.
Before taking power response into consideration, I realized I overlooked an aspect of this tweeter that does not show very clearly in the graphs above, but it does on a tweeter measuement. In dipole mode, this tweeter has a rising response that peaks at about 10 kHz - see http://meniscusaudio.com/images/NEO3pdr.pdf
It can - and probably should - be addressed with a series LRC as in this design (which uses a LR36 filter !) rather than "incorporating" it in the overall response.
Another option would be to use a shallow waveguide on the tweeter(loading both front and back). A proper waveguide will lift the response before cca 7kHz and will also increase directivity.
Last edited:
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.