So just how DOES a person slope part of the front baffle backwards 1.5" ? The wood work is not the problem. How in the world does that negatively effect the sound? Having that slope coming up, underneath the tweeter ?
Well ... it was not going to be a 2 box enclosure. I want to build it in one enclosure. I guess I can look at the backwards ramp back up to the bottom of the tweeter. Looks like there was a decent sized 2" or more gap between woofer and tweeter. That may be a gentle enough of a slope.
To what level is this a problem in your case. Are you trying to get your crossover to line up or are you trying to align the physical geometry in one plane or something else?
One way is to just tilt the entire enclosure. This places the drivers slightly off-axis to the listener though, so if you take it too far the higher frequencies of the drivers will start to disappear due to beaming.
A better option (imo) is to delay the tweeter electrically, either using asymmetric crossovers or a ladder delay network. 1.5" is 80degrees phase at 2kHz which works out pretty nicely with either asymmetric crossovers (higher order on tweeter) or an extra 1st order ladder on the tweeter which ideally gives 90degrees extra phase delay without affecting amplitude response. Ladder delay networks are not the easiest to design, but an example is here Zaph|Audio - ZD5 - Scan Speak 15W8530K00 and Vifa XT25
A better option (imo) is to delay the tweeter electrically, either using asymmetric crossovers or a ladder delay network. 1.5" is 80degrees phase at 2kHz which works out pretty nicely with either asymmetric crossovers (higher order on tweeter) or an extra 1st order ladder on the tweeter which ideally gives 90degrees extra phase delay without affecting amplitude response. Ladder delay networks are not the easiest to design, but an example is here Zaph|Audio - ZD5 - Scan Speak 15W8530K00 and Vifa XT25
Last edited:
Putting a step in the baffle always creates reflections which are a much worse problem than the original one. Fix the problem electrically. Tilting the speaker very much creates other problems.
I was hoping to line up the two drivers, to help with time alignment. I was hoping it would not destroy the sound to physically ramp back the front baffle, to the next vertical again where the tweeter mounts .... but it sounds like disaster to do this! The reason I was doing this, is because I am not competent to do the above crossover work. Looks like I should just keep it as it was.
Is time alignment such an important thing? The ideal of it blurs when two drivers are physically offset, and design parameters turn into compromises.
Upsetting the baffle is one way to divert energy out of time, among other issues.
Perhaps look at the audibility threshholds of group delay and see what kind of timing issues are/are not a problem.
8kHz, 2ms
4kHz, 1.5ms
2kHz, 1ms
1kHz, 2ms
500Hz, 3.2ms
Upsetting the baffle is one way to divert energy out of time, among other issues.
Perhaps look at the audibility threshholds of group delay and see what kind of timing issues are/are not a problem.
8kHz, 2ms
4kHz, 1.5ms
2kHz, 1ms
1kHz, 2ms
500Hz, 3.2ms
What you can do, is make a stepped baffle.
Make an extra front plate, covering the midwoofer area, and angling the side that goes toward the tweeter 45 degrees. Cover the entire front with 1/2" acoustic dampening foam. This foam will stop reflections from the baffle step, and you also don’t have to worry about gaskets or worry about flush-mounting the drivers.
Acoustic foam: Sonic Barrier 1/2" Acoustic Sound Damping Foam with PSA 18" x 24"
Johan-Kr
Make an extra front plate, covering the midwoofer area, and angling the side that goes toward the tweeter 45 degrees. Cover the entire front with 1/2" acoustic dampening foam. This foam will stop reflections from the baffle step, and you also don’t have to worry about gaskets or worry about flush-mounting the drivers.
Acoustic foam: Sonic Barrier 1/2" Acoustic Sound Damping Foam with PSA 18" x 24"
Johan-Kr
Unless you are using brickwall crossover filters, you need time (=phase) alignment otherwise the main 'lobe' of your multiway speaker will be pointing somewhere other than on the listening axis. If you don't align phase then you can still achieve flat on axis response but when you stand up from a seated listening position you could get blasted with an earful of the crossover frequency or a huge null could appears. Even if you're ok with having the speaker only deal flat response within a narrow angle, it's not ideal to have a peaking frequency response aimed at the roof, as the reflected sound is now louder and more problematic.
Passive electrical solutions just have to be good enough to get the job done - align the phase of the two speakers at the crossover point. If the phase stays aligned for one octave either side you're doing well, and if you can keep it aligned for 2 octaves you're doing really well. Beyond 2 octaves from the crossover point the time/phase alignment no longer matters as one of the drivers is almost completely rolled off and is contributing minimal sound so even if it were completely out of phase you wouldn't hear a difference.
Last edited:
In what directions do you get worse problems?
My findings tells me horizontal dispersion is close to unaffected to The step. You get some irregularities i vertical domian high up in frequency but not a problem wrt The benefits of ha ung The step.
Phase is just 360 degrees multiplied by the frequency (in Hz) multiplied by time delay (in seconds). They are directly correlated to each other.
e.g. 0.2ms at 1kHz = 360*1000*0.0002 = 72degrees
If two drivers are phase aligned at a particular frequency, they are not necessarily time aligned. If two identical drivers are time aligned then they are also phase aligned at every frequency. Note the caveat 'identical drivers' as two different model drivers will naturally have different phase responses, so it is near impossible to time align (phase align at every frequency) them using only passive components.
You don't need two drivers to be perfectly time aligned in a 2-way speaker because there is only a narrow band of frequencies where they both play audibly. Therefore using passive filters which maintain phase alignment only around the crossover frequency (and drift out of alignment above and below) is audibly acceptable.
When people actually need 'time alignment' (performed with a DSP) as opposed to just phase alignment, its because the drivers are multiple wavelengths apart at the crossover frequency. For a few kHz we are talking like 20 inches or 50cm+ for kHz frequencies. Multiple milliseconds of delay. Typically we only see this in car audio, or any other system where there are severe limitations on where drivers can be physically placed. It's also a bad idea to build a system which needs a lot of time alignment because the delay between driver and listener does not stay constant versus air temperature! I came across this in my own car as I had each driver individually delayed so everything was in perfect alignment in the drivers seat, and between summer and winter everything would need to be readjusted. Because I was dialling out multiple milliseconds of misalignment (hundreds of degrees of phase), if the speed of sound changed by just a few % between a hot day and a cold delay, there would be fractions of a millisecond changes in propagation delay to each speaker which could be a huge phase difference at my tweeter crossover frequency and between left and right. I'd have to move my head a few cm towards the middle of the car for it so sound correct again, or go and edit all the delays in my DSP.
For a typical woofer and tweeter on a flat baffle this is not the case as the misalignment is usually no more than 1/4 wavelength at the intended crossover frequency, so it is more intuitive to look at just phase alignment.
Last edited:
In a 2 way that has a xo within the midrange frequency band a delay of even just 64us can be audible. I'd try my best to get them physically aligned (acoustic centers). Your baffle can be machined to move your round amt back and form a waveguide of sorts in the process.
Last edited:
My speakers are 'upside down' - tweeter at the bottom. Measuring offset from the listening position gives the correct distance for each. No need for a stepped baffle.
Increased consideration of diffraction effects, cost and changes in fashion. Physical time alignment was something of a fad in the mid '70s, which like most fads died out when companies and advertisers found something else to push, mainly to encourage people to buy new products. 'twas ever thus. No doubt it will make a return at some point, as the eternal wheel of fashion revolves.
+1, all wisely said.
This fussing about phase is way out of proportion to the flux and warbling of music and the duration of notes (maybe hundreds of cycles long), not mention two ears, motion of your head, cooking of recordings, and much more.
Barely detectable using a fancy mic and sine waves (and then by eye, not by ear), let alone on music.
This fussing about phase is way out of proportion to the flux and warbling of music and the duration of notes (maybe hundreds of cycles long), not mention two ears, motion of your head, cooking of recordings, and much more.
Barely detectable using a fancy mic and sine waves (and then by eye, not by ear), let alone on music.
I am the OP ... And wish you a happy new year. But this has become confusing. The ONLY reason I was going to move the tweeter back, is because the tech at ESS( Heil), told me that moving the tweeter back to the cone of the woofer, makes a big positive improvement. But I kind of figured the negative effects of stepping the baffle back, would screw up any positives I would gain. I wanted some opinions, when you guys started throwing around math and physics, lol.