Would 7.5ms be enough to make a big improvement? On a scale from 1 to 10, how difficult is it to time adjust?
It should be enough in a normal room.
1 being easy. 0.25. You just type it in with all the other settings. In 0.02ms intervals. All the work will be the measuring to determine the amount. If you leave it hooked up to your computer, do what Waltersys was saying by reversing the phase on the woofer. Adjust the delay (start with 1ms intervals and iterate) until the null is deep. That's the delay you should use. Save it to the minidsp and flip the polarity back. You can even reverse the phase on the mini I think...? Not sure about that though.
1 being easy. 0.25. You just type it in with all the other settings. In 0.02ms intervals. All the work will be the measuring to determine the amount. If you leave it hooked up to your computer, do what Waltersys was saying by reversing the phase on the woofer. Adjust the delay (start with 1ms intervals and iterate) until the null is deep. That's the delay you should use. Save it to the minidsp and flip the polarity back. You can even reverse the phase on the mini I think...? Not sure about that though.
That's cool...
But regarding cabinet options, what do you suggest I use? I'm not sure if I will be able to use the boosted bottom response that the ported box gives compared to a closed one. I mean, my room is about 16sqm. Should I go with a sealed box and maybe use some LT EQ (can miniDSP do that?) and boost bottom end if I feel the need?
But regarding cabinet options, what do you suggest I use? I'm not sure if I will be able to use the boosted bottom response that the ported box gives compared to a closed one. I mean, my room is about 16sqm. Should I go with a sealed box and maybe use some LT EQ (can miniDSP do that?) and boost bottom end if I feel the need?
The low the frequency the more GD you can get away with. The sound of the bass won't change just the time it gets to you and your ears aren't very sensitive to changes in bass distortion wise. Personally I'm a bit of a perfectionist I don't like to see GD rise above 50ms at 20hz. Slopes bass generally means poor quality drivers with low power amps and poorly tuned ported enclosures. If you use a passive crossover it doesn't matter how good you kit is you've already lost any hope what so ever of good bass reproduction. The passive crossover is the worst thing in hifi that has ever existed, along with tubes...
Cabinet decisions should be made with the sound pressure level level desired, BR,TH,and BP type boxes all have a big level advantage over sealed subs.
Since you are limited to around 7.5ms of time correction, cabinet designs showing more group delay at the crossover point would not be a good choice.
The TH 200 sq shown below, a long path tapped horn, requires a bit more than 7.5ms to align a sealed top cabinet.
The Sammi is a ported cabinet using a pair of $5 speakers, it would require little or no delay to align with a sealed top cabinet.
The 7506 is a Sony 7506 headphone, used worldwide in studios and soundstages. It has fairly flat response, where it is flat the phase response is also flat. Unfortunately for the screen shot below, most of its response happens to fall at about -180 degrees, so much of the trace ends up not being visible, it would be right at the top or bottom of the display.
Phase and frequency response are reciprocal, if DSP is applied to flatten the frequency response, it also flattens the phase response. If a PEQ setting were applied to correct the 7506 broad boost centered at 60 Hz, and the dip at 250, it’s phase response would also flatten out.
Same with the Sammi problem at 475 Hz.
The TH response above 250 would require far too many filters to correct, but sounds quite good in it’s intended pass band of 30 to 100 Hz. It would require sealed top cabinets to be delayed by around 9ms optimally.
The difference between 7.5ms and 9 is within 1/4 wavelength at 100 Hz, so your DSP would still work fine for that large tapped horn, but many home theater type TH designet to work down to 16 Hz would require twice that delay time to be integrated perfectly.
That said, there are many proponents using several subs all around the room without any time alignment to even out room response, things can subjectively sound good even though that type of set up is terrible when phase or group delay is considered.
In a small room, getting too picky about phase/group delay is akin to complaining about a kid with a squirtgun shooting at you in a rainstorm.
Art
Attachments
Transmission curve (cross section) is a gaussian curve,
It holds for optics as well, but I don't think it is all that practical for acoustics in air except in limited frequency ranges - like Keele's CBT arrays.
Group delay by itself is inaudible. A recording is a group delay. The issue becomes important when the group delay of two drivers operating at the same frequency at the same time have different group delays. Their differences will cause an interference pattern in space where at some point at some frequencies they will add in phase to combine for a 3db gain in amplitude and at others will cancel completely causing a net amplitude of zero. This is the theory behind the so called "time aligned" speaker system, matching group delays between drivers. However, unless the drivers also happen to be coaxial there will be similar cancellation and reinforcements at various points in space. This is a direct consequence of classical wave mechanics analysis. If you want a speaker system to be phase coherent it must be time aligned at the crossover region and coaxial. Generally the tweeter has a shorter group delay than a midrange, a midrange shorter than a woofer. This is due to differences in inertial mass. If the drivers are coaxial, one way to align them in time is to apply the correct digital time delays to each driver using active crossovers and multiamplification and then sit on the coincident axis of the speakers. This will not cancel out effects due to room reflections however. The good news is that it's all probably very unimportant. There are far more serious issues related to accurate sound reproduction. In the 1970s it was one of the fads du jour, today it's been replaced by newer more lucrative fads in this industry.
This is why it's best to get both tweeter and midrange driver as close together as possible, xover design is very critical , odd order works best.
Coaxial drivers have other serious issues , which eliminates any if not all advantages IMO...
Yes, that right. It means the pressure profile across ths surface is strong in the middle and falls towards the edges. Its actually easy to do with any line array and is called shading. I showed curves of it in my AES line array paper after I saw it in Augspurger's line array paper.
A rectangular aperature always has a sin x/x polar profile. In optics you see that as diffraction rings on a focused star pattern (the optical scientists impulse response). A smooth weighting function will improve the polar curve and reduce lobing for all frequencies, but the dispersion will continue to vary with frequency unless you take steps to scale length with frequency.
The antenna people preceeded us by 50 years. (But they don't have to make things work for 10 Octaves!)
David S.
I always liked this story:
"...we received the Fletcher System at MGM. We set it up for music playback almost the day after we turned it on. Eleanor Powell had a dance mat on the floor of the stage, which she was practicing on. When she came to the studio in the afternoon, we decided, 'well gee, here's something that will really show this system off.' So we asked Eleanor if she would-kindly give us a tap-dance rehearsal and we would monitor it. And lo and behold, we got two taps and this was because the Fletcher system had a re-entrant horn that had somewhere between 11 and 12 feet of air column in it."
"It had a 36-inch deep multicell horn, and they were set so that there was an 8-foot differential between the time the sound arrived at the mouths of both horns, When Eleanor Power did the tap dance, we heard two taps-8ms apart."
"We disconnected the high-frequency unit and listened to her tap from the gut-no problem at all. And then we listened to the high-frequency unit alone, which was excellent; it had good definition."
"We started to move that high-frequency horn back until we got to a point where it jelled and where we had no more echo in the tap, which turned out to be somewhere in the first 2 feet. We gradually got down to 1 millisecond where we thought noblody could detect the difference at that time differential, and that's where we tied it off."
John K Hilliard, long term technical director of Altec Lansing talks of his experience with time allignment of theater speakers in the early 1930s.
http://www.aes.org/aeshc/docs/afternoon_hilliard.pdf
David S.
"...we received the Fletcher System at MGM. We set it up for music playback almost the day after we turned it on. Eleanor Powell had a dance mat on the floor of the stage, which she was practicing on. When she came to the studio in the afternoon, we decided, 'well gee, here's something that will really show this system off.' So we asked Eleanor if she would-kindly give us a tap-dance rehearsal and we would monitor it. And lo and behold, we got two taps and this was because the Fletcher system had a re-entrant horn that had somewhere between 11 and 12 feet of air column in it."
"It had a 36-inch deep multicell horn, and they were set so that there was an 8-foot differential between the time the sound arrived at the mouths of both horns, When Eleanor Power did the tap dance, we heard two taps-8ms apart."
"We disconnected the high-frequency unit and listened to her tap from the gut-no problem at all. And then we listened to the high-frequency unit alone, which was excellent; it had good definition."
"We started to move that high-frequency horn back until we got to a point where it jelled and where we had no more echo in the tap, which turned out to be somewhere in the first 2 feet. We gradually got down to 1 millisecond where we thought noblody could detect the difference at that time differential, and that's where we tied it off."
John K Hilliard, long term technical director of Altec Lansing talks of his experience with time allignment of theater speakers in the early 1930s.
http://www.aes.org/aeshc/docs/afternoon_hilliard.pdf
David S.
The link in an interesting look into the early history of how the horn speaker and much of the technology and standards used in the motion picture industry evolved. It gives insight into the origin of the "West Coast Sound." It's understandable how wealthy people who wanted superior sound for playing records and radios in their homes acquired these monsters that fit into their large rooms and resulted in domesticated models that dressed them up as furniture and scaled them down to more manageable sizes for greater market penetration. At that time the "East Coast Sound" hadn't existed yet, it was still decades away. I always think of horn speakers as relics of a bygone era, more primitive times, equipment intended for an entirely different purpose than high fidelity except in the context of that era.
More on group delay
in your seat , it is the transfer function you hear, the sum of the speaker and the room
it is incorrect to consider the speaker alignment in isolation
smaller higher F3 speakers have more group delay in the sensitive range and filter augmented reflex designs are the worst offenders.
Simply, bass line percussion is late and smeared , subjectively is too loud, and fails to time, to precisely synchronise, with mid-treble transients.
Careful comparison on such programme with I/B alternatives is immediately instructive.
The Meridian DSP 7200 three way system has major fix for group delay , delaying mid and treble the equivalent of three feet to synchronise with the bass , the latter a 25Hz F3 ported, but over damped Bessel alignment.
This is one of the best timing full range speakers made and the absence of group delay results in bass tune playing roughly comparable with the mid range for clarity.
we have put up with a lot of linear lower frequency distortion for a very long time
Martin Colloms HIFICRITIC
in your seat , it is the transfer function you hear, the sum of the speaker and the room
it is incorrect to consider the speaker alignment in isolation
smaller higher F3 speakers have more group delay in the sensitive range and filter augmented reflex designs are the worst offenders.
Simply, bass line percussion is late and smeared , subjectively is too loud, and fails to time, to precisely synchronise, with mid-treble transients.
Careful comparison on such programme with I/B alternatives is immediately instructive.
The Meridian DSP 7200 three way system has major fix for group delay , delaying mid and treble the equivalent of three feet to synchronise with the bass , the latter a 25Hz F3 ported, but over damped Bessel alignment.
This is one of the best timing full range speakers made and the absence of group delay results in bass tune playing roughly comparable with the mid range for clarity.
we have put up with a lot of linear lower frequency distortion for a very long time
Martin Colloms HIFICRITIC
A year ago I sat in on a blind test involving group delay. DEQX was used for the DSP and it is able to dial in zero group delay in the listening position. Two profiles were set up so that you can instantly switch with the remote from one to the other. The purpose of the test wasn't to determine if group delay is audible, but to see what difference this expensive DSP system made to the sound. Two profiles were set up and level matched. The frequency response was also matched. The profile without group delay correction was aligned as much as possible with fixed channel delays. In other words, the time domain was already better than a typical system. It was a 4 way system with dual subs using PHL drivers mostly in a treated room. This is a system about which many say it's the best they have heard.
The difference was obvious. Two things were heard, and everyone could pick it blind, easily.
Firstly, the sound stage was affected. With the group delay correction, the sound stage was bigger and individual sounds were more clearly separated from each other. When the correction was switched off, sounds appeared to bunch together and overall the sound was quite flat in comparison. Normally this system has both precise imaging and a big sound stage. The owner listens to a lot of electronic music where this kind of thing really stands out.
Secondly, the bass was clearly tighter. With the GD correction off, it sounded like there was more bass, as if a loudness control had been turned on. It was not the case - there was no more bass.
This test was in response to other tests in which I found that DEQX was no better than MiniDSP or DCX in terms of sound quality. Blind testing here with instant switching was behind that conclusion, with a small group.
The results were a surprise to me and has piqued my interest in more advanced DSP. Bodzio Ultimate Equalizer (another Aussie product like DEQX) is another way to do it. It's more than likely I'll end up with a PC solution.
The difference was obvious. Two things were heard, and everyone could pick it blind, easily.
Firstly, the sound stage was affected. With the group delay correction, the sound stage was bigger and individual sounds were more clearly separated from each other. When the correction was switched off, sounds appeared to bunch together and overall the sound was quite flat in comparison. Normally this system has both precise imaging and a big sound stage. The owner listens to a lot of electronic music where this kind of thing really stands out.
Secondly, the bass was clearly tighter. With the GD correction off, it sounded like there was more bass, as if a loudness control had been turned on. It was not the case - there was no more bass.
This test was in response to other tests in which I found that DEQX was no better than MiniDSP or DCX in terms of sound quality. Blind testing here with instant switching was behind that conclusion, with a small group.
The results were a surprise to me and has piqued my interest in more advanced DSP. Bodzio Ultimate Equalizer (another Aussie product like DEQX) is another way to do it. It's more than likely I'll end up with a PC solution.
more recent tests looked at noise floor for large (50l) enclosure run ported and sealed which versions were adjusted for a similar F3.
Specifically listening for transparency the complex low level noise from the ported version were rendered obvious, confusing detail in the lower mid range, reducing dynamic contrasts and partially obscuring image depth. We do pay quite a a price for a nominal 3dB efficiency gain.
Martin Colloms
Specifically listening for transparency the complex low level noise from the ported version were rendered obvious, confusing detail in the lower mid range, reducing dynamic contrasts and partially obscuring image depth. We do pay quite a a price for a nominal 3dB efficiency gain.
Martin Colloms
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.