OP
- Thread Starter
- #21
Gating was 1.5 ms. I'll have another go at the 60 degree tomorrow and see if I can smooth it out a bit.
Didn't have time after my wiring conundrum. Will have a go at it tomorrow.
-
WANTED: Happy members who like to discuss audio and other topics related to our interest. Desire to learn and share knowledge of science required. There are many reviews of audio hardware and expert members to help answer your questions. Click here to have your audio equipment measured for free!
Need some help with active speaker conversion
- Thread starter gfinlays
- Start date
Didn't have time after my wiring conundrum. Will have a go at it tomorrow.
how high is your ceiling and how far is mic from speaker?
OP
- Thread Starter
- #23
Ceiling height is 2.4 m. Mic was 1m from the speaker.
My experience is that the frequency range between 200 Hz and 3 kHz is critical for sound source localization,
The frequency range of 200 Hz to 3 kHz indeed plays a pivotal role in sound source localisation, particularly in the context of stereo reproduction. This range encompasses not only the fundamental formants of human speech but also the harmonic content of many musical instruments. The following aspects highlight its importance:
1. Spatial Perception (Interaural Time and Level Differences):
At frequencies below 1.5 kHz, interaural time differences (ITD) are dominant. These arise from the time delay between when a sound reaches each ear, allowing the brain to accurately determine the direction of the sound source.
At frequencies above 1.5 kHz, extending to around 3 kHz, interaural level differences (ILD) become significant due to the head-shadowing effect. This provides additional cues for localisation.
2. Perception of Timbre:
Many of the acoustic characteristics of voices and instruments lie within this range, making it easier to distinguish both the direction and the tonal quality of sound sources.
3. Human Hearing Sensitivity:
The human ear is particularly sensitive to frequencies in the range of 2 kHz to 4 kHz, where the highest frequency resolution occurs. This sensitivity is vital for accurately localising and identifying sound sources.
4. Relevance to Music and Speech:
The range between 200 Hz and 3 kHz is crucial for speech intelligibility and musical transparency. Voices and instruments with strong activity in this range contribute significantly to the perception of the stereo soundstage.
The frequency range of 200 Hz to 3 kHz is critical for the localisation and perception of sound sources because it contains the most relevant spatial information (ITD and ILD) while also being the range where voices and instruments predominantly operate. A well-designed stereo system should reproduce this range with exceptional clarity and accuracy to ensure precise spatial imaging.
Placing a crossover frequency within the 200 Hz to 3 kHz range is generally not a good idea, as it can negatively impact stereo imaging and spatial perception. Here’s why:
1. Phase Issues and Group Delay Distortions:
Crossovers inherently introduce phase shifts because the filters (e.g., Butterworth, Linkwitz-Riley) alter both the amplitude and phase of the signal.
These phase shifts and the resulting group delay are particularly problematic in the 200 Hz to 3 kHz range, where the human ear is highly sensitive to timing differences.
Such distortions can compromise the precision of sound localisation and cause the stereo soundstage to lose clarity.
2. Perception of the Midrange:
This frequency range carries the core information for voices and many instruments. Inaccuracies in this range would degrade not only spatial localisation but also the clarity and naturalness of the sound.
Crossovers can also introduce comb filtering effects if the drivers are not perfectly aligned, further muddying the sound.
3. Optimal Placement of Crossover Frequencies:
Below 200 Hz: Lower frequencies are less critical for spatial localisation, as they rely primarily on interaural time differences. A subwoofer or woofer can effectively handle these frequencies.
Above 3 kHz: A crossover to a tweeter in this range is more suitable, as higher frequencies rely on interaural level differences, which are less affected by phase issues.
A crossover frequency within the 200 Hz to 3 kHz range should be avoided to minimise phase shifts and group delay distortions that could disrupt stereo imaging. Instead, transitions between drivers should occur outside of this critical range, ensuring accurate and natural sound reproduction.
The frequency range of 200 Hz to 3 kHz indeed plays a pivotal role in sound source localisation, particularly in the context of stereo reproduction. This range encompasses not only the fundamental formants of human speech but also the harmonic content of many musical instruments. The following aspects highlight its importance:
1. Spatial Perception (Interaural Time and Level Differences):
At frequencies below 1.5 kHz, interaural time differences (ITD) are dominant. These arise from the time delay between when a sound reaches each ear, allowing the brain to accurately determine the direction of the sound source.
At frequencies above 1.5 kHz, extending to around 3 kHz, interaural level differences (ILD) become significant due to the head-shadowing effect. This provides additional cues for localisation.
2. Perception of Timbre:
Many of the acoustic characteristics of voices and instruments lie within this range, making it easier to distinguish both the direction and the tonal quality of sound sources.
3. Human Hearing Sensitivity:
The human ear is particularly sensitive to frequencies in the range of 2 kHz to 4 kHz, where the highest frequency resolution occurs. This sensitivity is vital for accurately localising and identifying sound sources.
4. Relevance to Music and Speech:
The range between 200 Hz and 3 kHz is crucial for speech intelligibility and musical transparency. Voices and instruments with strong activity in this range contribute significantly to the perception of the stereo soundstage.
The frequency range of 200 Hz to 3 kHz is critical for the localisation and perception of sound sources because it contains the most relevant spatial information (ITD and ILD) while also being the range where voices and instruments predominantly operate. A well-designed stereo system should reproduce this range with exceptional clarity and accuracy to ensure precise spatial imaging.
Placing a crossover frequency within the 200 Hz to 3 kHz range is generally not a good idea, as it can negatively impact stereo imaging and spatial perception. Here’s why:
1. Phase Issues and Group Delay Distortions:
Crossovers inherently introduce phase shifts because the filters (e.g., Butterworth, Linkwitz-Riley) alter both the amplitude and phase of the signal.
These phase shifts and the resulting group delay are particularly problematic in the 200 Hz to 3 kHz range, where the human ear is highly sensitive to timing differences.
Such distortions can compromise the precision of sound localisation and cause the stereo soundstage to lose clarity.
2. Perception of the Midrange:
This frequency range carries the core information for voices and many instruments. Inaccuracies in this range would degrade not only spatial localisation but also the clarity and naturalness of the sound.
Crossovers can also introduce comb filtering effects if the drivers are not perfectly aligned, further muddying the sound.
3. Optimal Placement of Crossover Frequencies:
Below 200 Hz: Lower frequencies are less critical for spatial localisation, as they rely primarily on interaural time differences. A subwoofer or woofer can effectively handle these frequencies.
Above 3 kHz: A crossover to a tweeter in this range is more suitable, as higher frequencies rely on interaural level differences, which are less affected by phase issues.
A crossover frequency within the 200 Hz to 3 kHz range should be avoided to minimise phase shifts and group delay distortions that could disrupt stereo imaging. Instead, transitions between drivers should occur outside of this critical range, ensuring accurate and natural sound reproduction.
Your gate should be around 4.6 ms if your speaker is about halfway up (recommended location). This should be observable if you check the IR by following the REW instructions.
OP
- Thread Starter
- #26
I remeasured the 60 degree responses. Turns out, the TV was sitting out a little from the wall. I've swung it back in to its closest to the wall position which seems to have taken the reflections out of the gating window:
PMS with passive crossover at 0, 15, 30, 45 and 60 degrees:
There is some directivity error around the crossover region, but it's nowhere near as bad as the individual driver measurements would indicate. Even 30 degrees off axis is well beyond either end of the sofa. The question is, with very steep crossover slopes, such as Acourate's UBjPol11, is it going to get a lot worse? Or am I overthinking this/worrying too much?
PMS with passive crossover at 0, 15, 30, 45 and 60 degrees:
There is some directivity error around the crossover region, but it's nowhere near as bad as the individual driver measurements would indicate. Even 30 degrees off axis is well beyond either end of the sofa. The question is, with very steep crossover slopes, such as Acourate's UBjPol11, is it going to get a lot worse? Or am I overthinking this/worrying too much?
OP
- Thread Starter
- #27
The speakers are on the floor, so midrange centre is 765 mm from the floor. Assuming my maths is right, the first reflection is going to reach the mic 2.4 ms after the test signal. I'd love to be able to lift these up, but at 71 kg each, it's not really feasible.
2.4 ms windowing:
I really appreciate all of the help with this.
Roland68
Major Contributor
You could mount the Visaton Waveguide WG 148 R, WG 220x150 or Monacor WG-300 on the Seas 27TFFC, that should fit very well.
Not very expensive to try.
OP
- Thread Starter
- #29
I have a handful of Visaton WG 148R waveguides lying around. Also have a spare 27TFFC on order - the faceplate holes don't align with the holes in the waveguide so some modification will be required - I don't want to mod the ones in my speakers for the moment. I'm hoping to put a test baffle together later this week.
The Monacor WG-300 has sadly been discontinued.
The Monacor WG-300 has sadly been discontinued.
Roland68
Major Contributor
If I remember correctly, the screw holes on the Seas tweeters only need to be filed outwards by about 0.5mm. This shouldn't affect normal function.
Alternatively, there are now many templates for 3D printing. Users are often willing to adjust the hole circle for you, just ask.
You paid very much for a speaker plan and a passive crossover. If you change drivers, as fitting a waveguide is a complete and radical change, you ruin the design you bought from Troels. So you may just be build a new speaker.
I understand when you activate the speaker.
***Personally I do not like the huge graveyards of boutique crossover parts the designer always uses. This extreme (theoretical) linearity often results in a somehow lifeless impression. Sometimes seems to me they are only so complicated to sell Jantzen parts. He uses only very good drivers which should be well behaved with a minimum of correction. Just a private oppinion.***
Removing the passive parts could definetly make them better, but don't mess with the drivers and their position on the baffle. I guess you will be very satisfied, only going active.
I understand when you activate the speaker.
***Personally I do not like the huge graveyards of boutique crossover parts the designer always uses. This extreme (theoretical) linearity often results in a somehow lifeless impression. Sometimes seems to me they are only so complicated to sell Jantzen parts. He uses only very good drivers which should be well behaved with a minimum of correction. Just a private oppinion.***
Removing the passive parts could definetly make them better, but don't mess with the drivers and their position on the baffle. I guess you will be very satisfied, only going active.
If Troels' base design does not require overhaul, could still get some improvements via active crossover and some eq. It would be a major crossover redesign effort though.
OP
- Thread Starter
- #33
The only thing I paid for was the parts kit from Jantzen Audio which cost me €312. Drivers were sourced locally.
I don't believe in boutique crossover parts. Measurements shared on ASR show there's no value in in them. Wire coils and Jantzen crosscaps are all that went into it.
The basic cabinet drawing was free - I went my own route with the internal bracing design and constrained layer exterior to achieve a very heavy, inert, resonance-free cabinet.
The cabinet construction was where most of the cost went.
I could just replicate the existing crossovers in Acourate, however, I think there's more that can be achieved.
With steep crossover slopes, I think the directivity mismatch between mid and tweeter will become an issue, which is why I'm exploring options.
I have a spare midrange, so I can measure the mid/tweeter integration and directivity at various crossover points using a test baffle with a replica of the midrange enclosure fitted to it.
I appreciate it's a complete crossover redesign, but every active crossover is effectively a complete redesign.
It's a lot of work, but other than my time, there's little cost involved.
The first step is probably to go ahead and try things as they are with Acourate. If a major directivity issue appears between the mid/tweeter, then I guess I'm on my own with finding a solution.
Last edited:
IEales
Member
I like Bessel filters as they minimize group delay. For more than 50 years, my mantra has been "It's not the Frequency. It's the TIME!" Getting the time right covers a multitude of FR sins.
I used REW and a miniDSP Flex Eight to time align and mate the tri-amped LFT-8b drivers. I found that many iterations of listening & measuring are needed to get a seamless match w sub, cone woofer, planar mid and 'crinkly' tweeter.
As little as 10µs [miniDSP limit] make a tremendous difference as do very small [≈1°] angular adjustments to toe-in and driver tilt. As these are all inter-related, adjustments are necessary to all. Once the XOs & time alignment are right is the time to worry about EQ. Judicious experimentation on XO points in the room vs on the datasheet can obviate some need to EQ.
As a recording engineer, I never heard a control room playback system that matched the sound in the studio. BUT a well set up HiFi gives a close enough facsimile to get the toes tapping...
I used REW and a miniDSP Flex Eight to time align and mate the tri-amped LFT-8b drivers. I found that many iterations of listening & measuring are needed to get a seamless match w sub, cone woofer, planar mid and 'crinkly' tweeter.
As little as 10µs [miniDSP limit] make a tremendous difference as do very small [≈1°] angular adjustments to toe-in and driver tilt. As these are all inter-related, adjustments are necessary to all. Once the XOs & time alignment are right is the time to worry about EQ. Judicious experimentation on XO points in the room vs on the datasheet can obviate some need to EQ.
As a recording engineer, I never heard a control room playback system that matched the sound in the studio. BUT a well set up HiFi gives a close enough facsimile to get the toes tapping...
I would first ask myself is the DI mismatch of the mid to tweeter causing audible issues that bother me, if no then a tweeter swap is likely not worth the cost. The DI mismatch on this speaker looks to be fairly mild, there are certainly much worse speakers in this regard.
A seas DXT would provide a good dispersion match with the mid, and it's dimensions are similar, a difference of about 1mm which may or may not require a bit of widening on the tweeter route depending on how tight your current route is on the stock seas tweeter.
A seas DXT would provide a good dispersion match with the mid, and it's dimensions are similar, a difference of about 1mm which may or may not require a bit of widening on the tweeter route depending on how tight your current route is on the stock seas tweeter.
Last edited:
I didn't get the fact that the PMS is one of Troels free to copy builds. It is one of his designs with the lowest part count in the crossover, too.
I dont see much to improve in driver placement on the baffle. In fact a very nice design if you don't mind the large size.
I would first copy the crossover in the active setup, do nothing complicated with these well behaved drivers.
Then, for a second try, turn it into a complete 18dB/oct design and drop the low crossover point to 200-250 Hz, while the upper should be fine at 3000 Hz. Some A/B testing of both speakers, side by side with mono music will be interesting. The low crossover point in most passive designs is a compromise. Because passive parts get large, expensive and even may interfere with the resonance frequency, it is usually set higher than really desired.
Troels did all neccesary measurements and the only thing that would be worth experimenting is, indeed, the tweeter. The 27TFFC is a very good, basic part, not expensive (for SEAS), maybe a budget decission. There may be better ones around. The midrange is IMO a gem, one of the best he could choose. So there may be some room for improvement with another tweeter. With such a high crossover frequency of 18dB/oct, a 1 1/8" dome with ferrofluid may not be neccesary. A ferro fluid free and maybe even smaller dome may sound more refined. You are lucky the diameter of the tweeter is 104mm, so many others and even 100mm ones can fit to the baffle. There are many options, the SBAcoustics "SB26ADC-000-4", even as it is less expensive than the Seas, comes to my mind. Worth a try.
I know that wave guides are "en vogue" everywhere, but they are 1. nothing new at all and 2. not an improvement for any driver or speaker. We called them "short horn" in the past.
Today some want to fit every tweeter alive with it, even those that degrade in quality using one. The wave guide maniacs often don't even understand what it does to a well designed baffleI and how the crossover has to be changed. In fact, with this design, I see a waveguide very critical.
I dont see much to improve in driver placement on the baffle. In fact a very nice design if you don't mind the large size.
I would first copy the crossover in the active setup, do nothing complicated with these well behaved drivers.
Then, for a second try, turn it into a complete 18dB/oct design and drop the low crossover point to 200-250 Hz, while the upper should be fine at 3000 Hz. Some A/B testing of both speakers, side by side with mono music will be interesting. The low crossover point in most passive designs is a compromise. Because passive parts get large, expensive and even may interfere with the resonance frequency, it is usually set higher than really desired.
Troels did all neccesary measurements and the only thing that would be worth experimenting is, indeed, the tweeter. The 27TFFC is a very good, basic part, not expensive (for SEAS), maybe a budget decission. There may be better ones around. The midrange is IMO a gem, one of the best he could choose. So there may be some room for improvement with another tweeter. With such a high crossover frequency of 18dB/oct, a 1 1/8" dome with ferrofluid may not be neccesary. A ferro fluid free and maybe even smaller dome may sound more refined. You are lucky the diameter of the tweeter is 104mm, so many others and even 100mm ones can fit to the baffle. There are many options, the SBAcoustics "SB26ADC-000-4", even as it is less expensive than the Seas, comes to my mind. Worth a try.
I know that wave guides are "en vogue" everywhere, but they are 1. nothing new at all and 2. not an improvement for any driver or speaker. We called them "short horn" in the past.
Today some want to fit every tweeter alive with it, even those that degrade in quality using one. The wave guide maniacs often don't even understand what it does to a well designed baffleI and how the crossover has to be changed. In fact, with this design, I see a waveguide very critical.
Well it's early in the day but this is probably going to be the dumbest thing I've read today on here.
IEales
Member
A long time ago, I was hanging out in a studio control room and all the glass was covered with hanging packing blankets. I enquired and was told "to reduce reflections". I tried it next time I was mixing and it definitely changes the perspective.
I cover our flat screen with a heavy duvet and heavy blinds for the windows for the same reason and have demonstrated the effect to visitors who ask.
OP
- Thread Starter
- #39
In the current passive setup it isn't, but the crossover slopes are fairly gentle. With steeper crossover slopes on the active crossover, I think the difference will be noticeable.
The DXT is an interesting tweeter. Directivity at 3kHz is quite similar to the midrange. Nothing I've read about it raises any concerns and it's only a whisker wider than the 27TFFC, albeit the recess depth would need increasing to accommodate the thicker faceplate.
OP
- Thread Starter
- #40
I agree that many folks out there don't understand the need to adapt a crossover for a waveguided tweeter. With a ~6 dB lift between 2-3 kHz, failure to tailor the crossover to the new response would make your ears bleed.
