ProSoundWeb Community
Sound Reinforcement - Forums for Live Sound Professionals - Your Displayed Name Must Be Your Real Full Name To Post In The Live Sound Forums => LAB Subwoofer Forum => Topic started by: Benjamin Goulart on November 12, 2013, 06:27:21 PM
-
What kind of power (in volts, please) can each lab sub handle with, say...
no HPF?
HPF set at 25hz 48dB/oct butterworth?
HPF set at 27hz 48dB/oct butterworth?
HPF set at 30hz?
HPF set at 35hz?
What HPF setting would be required for it to handle 3000 watt peaks, or is 2000 watts the limit regardless of roll-off? Not sure if the 2000 watt figure I've seen thrown around is continuous, program, or peak, but I assume it was meant at the default 35hz LR 24db/oct HPF.
Is thermal handling a bigger problem than over-excursion, and would new Class D designs with their PWM supplies be less likely to cause thermal issues with the drivers since they have a tendency to crap-out (or tapper off, to use a euphemism) on extended sin-wave-like signals?
Conservative estimates based on simulation and/or experience, accounting for manufacturing variance is cool. I don't want to necessarily assume the drivers in mine are the best examples Eminence ever produced, though they've apparently run without a problem for 20 years so far.
Thanks for your time.
-
Conservative estimates based on simulation and/or experience, accounting for manufacturing variance is cool. I don't want to necessarily assume the drivers in mine are the best examples Eminence ever produced, though they've apparently run without a problem for 20 years so far.
Thanks for your time.
The drivers are rated at 400 watts each. I doubt you have had them for 20yrs-since the lab 12 has not been made that long-unless you are using a different model number
How much power they can "handle" depends a lot on what the source material is-freq wise-dynamic range wise and so forth.
If you are doing "dance" type material-with long sustained low freq tones-the power "handling" will be less than with a dynamic type source that has short notes.
There are no "exact" ratings-only ratings based on a particular set of test tones or application.
The nice thing about standards is that there are so many to choose from.
-
The drivers are rated at 400 watts each. I doubt you have had them for 20yrs-since the lab 12 has not been made that long-unless you are using a different model number
How much power they can "handle" depends a lot on what the source material is-freq wise-dynamic range wise and so forth.
If you are doing "dance" type material-with long sustained low freq tones-the power "handling" will be less than with a dynamic type source that has short notes.
There are no "exact" ratings-only ratings based on a particular set of test tones or application.
The nice thing about standards is that there are so many to choose from.
Don't know why I said 20 years. I think it said 2003 or 4 inside. Hah. Sorry. Not sure which image of the date I was thinking of in my head when I typed that in.
Do you know of anyone with an already-defined model (and some tutorial info) on the labhorn for hornresp so I could at least see what the simulation suggests the excursion will be at the various settings? I assume that is not taking into account thermal issues. I'm not sure where to even begin with the sim values. I saw one listing of values someone used, but several of them appeared blank from the hornresp version I tired.
Anyone done a simple open air excursion test with a 10hz sin wave to see at what point these drivers reach xmax?
-
What kind of power (in volts, please) can each lab sub handle with, say...
no HPF?
HPF set at 25hz 48dB/oct butterworth?
HPF set at 27hz 48dB/oct butterworth?
HPF set at 30hz?
HPF set at 35hz?
What HPF setting would be required for it to handle 3000 watt peaks, or is 2000 watts the limit regardless of roll-off?
Is thermal handling a bigger problem than over-excursion, and would new Class D designs with their PWM supplies be less likely to cause thermal issues with the drivers since they have a tendency to crap-out (or tapper off, to use a euphemism) on extended sin-wave-like signals?
As Ivan said, "it depends".
If the compression chamber is well sealed (no air leaks) the usual failure mode is thermal, the Lab 12s driver can only handle 200 watts sine wave continuous.
The horn cut off is around 35 Hz, a 24 or 48 BW at 30 Hz will protect the driver from excursion below without affecting the LF much at all.
The highest excursion for most program music will be around 60 Hz, the HP filter won't offer any protection there. However, the impedance is pretty high at 60 Hz, so less power is dissipated in that region.
The way to cook the drivers is hit it hard at 40 Hz, where the cone hardly moves at all and the impedance is about the same as the DCR, about 2.2 ohms for the two drivers in parallel.
The Lab 12 are not very good at loosing heat to start with, and in the small compression chamber the surrounding air does not offer much cooling
An amp's current limiting will make a big difference in that regard, there are some Crown amps that put out less power at 2 ohms than at 4, for instance.
The Hornresp simulations below are from:
http://audioroundtable.com/PiSpeakers/messages/17389.html
Hornresp has gone through many updates since then, it now has a filter simulation as well.
That said, below Fc or Fb, Hornresp is not all that accurate predicting excursion, the Lab 12 suspension limits excursion past Xmax more than the simulation would indicate.
I have used Lab 12s in bass reflex, band pass, front loaded and tapped horns cabinets, they can take 49 volts sine wave (at any frequency) for short duration tests.
They start to smell after about 10 seconds at that level even in open air, vent hole out in a tapped horn.
2000 watt peaks are not a problem, but with a dynamic range of 12 dB (the crest factor for normal pink noise), the drivers will get hot enough that the impedance rises in just a few minutes, and level will drop by 2-3 dB.
If the level is raised another 3 dB to compensate for the thermal compression, smoke may ensue.
Heavy limiting could reduce the dynamic range to a point where 1000 watt "peaks" could be too much average power for a pair of Lab 12.
Art
-
50-60 Volts appears to be the maximum safe thermal limit of the Labhorn sub if there are no restrictions on the type of signals, sweeps, or tests.
Is there a frequency at xmax v. volts graph anywhere for it?
-
50-60 Volts appears to be the maximum safe thermal limit of the Labhorn sub if there are no restrictions on the type of signals, sweeps, or tests.
Is there a frequency at xmax v. volts graph anywhere for it?
That would have to be a 3D graph-as the excursion will vary with freq
-
That would have to be a 3D graph-as the excursion will vary with freq
I'm thinking excursion being 1 at frequency, so it'd be frequency @ reaching xmax versus voltage in 2D. In 3D, only the 1 region would interest me, anyway.
-
50-60 Volts appears to be the maximum safe thermal limit of the Labhorn sub if there are no restrictions on the type of signals, sweeps, or tests.
Is there a frequency at xmax v. volts graph anywhere for it?
60 volts works out to around 1600 watts at the low impedance frequencies of the LabHorn.
From a thermal standpoint the drivers can only withstand 60 volts for peaks, the long term average should not exceed 30 volts.
Doubling voltage is a six dB increase in power, and doubling of excursion.
Using the 45 volt excursion plot in post #3 you can extrapolate what the excursion would be at any given frequency if you increased voltage.
The Lab 12 has 13mm Xmax, around 49V will hit Xmax in the 60 Hz range.
Hornresp also has a function to show thermal and excursion limited levels on the same screen, though I find that to be pretty useless, since I know exceeding Xmax sounds bad, and power level simulations don't have any time constants- there is a huge difference in average power between a punchy kick drum and a droning sine wave like synth bass line.
-
What voltage would xmax be at 30hz?
20hz?
10hz?
I don't think the class d amps with pwm power supplies can provide full power at lower frequencies for more than a few seconds at a time into very low impedances without voltage being time-limited and dropping to a fraction of that rated power with test signals or some dub step synth tone. I would bet it's probably impossible to destroy these labhorns with the amp I bought since it's like that and has dsp & all sorts of limiting regardless. I'll usually further limit it at 25-50 volts depending on how I have the HPF set, but even leaving it at the default 82 volts would probably be safe.
60 volts works out to around 1600 watts at the low impedance frequencies of the LabHorn.
From a thermal standpoint the drivers can only withstand 60 volts for peaks, the long term average should not exceed 30 volts.
Doubling voltage is a six dB increase in power, and doubling of excursion.
Using the 45 volt excursion plot in post #3 you can extrapolate what the excursion would be at any given frequency if you increased voltage.
The Lab 12 has 13mm Xmax, around 49V will hit Xmax in the 60 Hz range.
Hornresp also has a function to show thermal and excursion limited levels on the same screen, though I find that to be pretty useless, since I know exceeding Xmax sounds bad, and power level simulations don't have any time constants- there is a huge difference in average power between a punchy kick drum and a droning sine wave like synth bass line.
-
I don't think the class d amps with pwm power supplies can provide full power at lower frequencies for more than a few seconds at a time into very low impedances without voltage being time-limited and dropping to a fraction of that rated power with test signals or some dub step synth tone.
Most class D amps would LOVE to be able to produce sine waves for a couple of seconds. If you look at the specs-you will see full power for "typically" 0.08 seconds (80ms) before dropping down.
Yes it makes a HUGE difference when doing dance type stuff. Not as big a deal with "kick drum rock and roll".
There are a few out there that can do it a good bit longer-but that is one spec that the manufacturers don't like to talk about.
-
Most class D amps would LOVE to be able to produce sine waves for a couple of seconds. If you look at the specs-you will see full power for "typically" 0.08 seconds (80ms) before dropping down..
Something good to be said about the old Crown VZ5000 "welders" ;)
-
What voltage would xmax be at 30hz?
20hz?
10hz?
I don't think the class d amps with pwm power supplies can provide full power at lower frequencies for more than a few seconds at a time into very low impedances without voltage being time-limited and dropping to a fraction of that rated power with test signals or some dub step synth tone.
Xmax of the Lab 12 is 13mm regardless of frequency.
The voltage required to reach Xmax varies with frequency as you can see in the simulation in post # 3.
I wouldn't put huge faith in the excursion sim below Fc (around 40 Hz) but at 10-20 Hz looks like 46 volts or so would hit Xmax, around 44 at 30, and only 30 or so at 35 Hz.
The SpeakerPower class D amps can deliver full power for around 4 seconds.
The SP1-4000 (4000W/2 ohm, 2400W/4, 1300W/8) weighs only 7 lbs.
It is a nice choice for a pair of Lab 12 (one LabHorn), and will handle four Lab 12, though at that low of an impedance (near 1 ohm at the impedance minima) the breaker will trip with extended full power LF.
It does have plenty of power to smoke the Labs if sine wave type material is played.
Even with pink noise, the SP1-4000 just below clip/limit will induce a couple dB of thermal compression in the Lab 12.
-
Xmax of the Lab 12 is 13mm regardless of frequency.
The voltage required to reach Xmax varies with frequency as you can see in the simulation in post # 3.
I wouldn't put huge faith in the excursion sim below Fc (around 40 Hz) but at 10-20 Hz looks like 46 volts or so would hit Xmax, around 44 at 30, and only 30 or so at 35 Hz.
So you think the excursion is probably more below 30hz than the simulation is predicting?
I was not expecting that sudden peak in excursion at 34hz. Is that normal for a driver in a horn?
-
So you think the excursion is probably more below 30hz than the simulation is predicting?
I was not expecting that sudden peak in excursion at 34hz. Is that normal for a driver in a horn?
The excursion is probably less below 30Hz than the sim, both the small chamber and the Lab 12 suspension are putting the "brakes" on excursion.
FLH typically have the greatest excursion around Fc, so yes, normal, but again, may be a bit exaggerated.
That said, the sim may be a few Hz high from what I recall the measured low corner of the LabHorn is.
Hornresp is a great predictive software (really great considering it is free- thanks David McBean!) but the dips and peaks in impedance, frequency response, and excursion are all a bit sharper than in real speakers.
Of course, short of putting a little video camera in the throat or a Lexan window and measuring excursion, about all you have for to go on for FLH ;).
Art
-
So then, the greatest xmax points would be around 35hz and 60hz? I heard below the Fc the Lab12 drivers behave like in free air. And I always thought xmax skyrocketed in the deep bass on all woofers if allowed to.
Does the xmax graph matter that it's RMS voltage? If I'm using this to set limiters in the amp's DSP, should I be using sqrt 2 * RMSvolts? If I do that, 58.5 Vrms (the predicted xmax it looks like for 60hz, assuming 1.3 * 10mm =13mm and 1.3 * 45V = 58.5V is correct methodology) comes out to about 82V peak value... which is exactly the maximum default stereo output of my nu3000 DSP's hard cap (82Vp). And since they are current-limited to only a few seconds at low impedances & low frequencies with sine-wave-like signals, maybe I don't even need a lower limited voltage for thermal reasons... assuming I have a 48dB Butterworth at 35hz, which will halve the average voltage the Labhorns see at 35hz with program content (-3dB at the corner freq, and assuming half the voltage actually halves the excursion), right? A 35hz test signal cranked way the hell up could still exceed xmax if the gain compensated for the filters, but assuming I or no one else do that, are these rational limits?
[Edit: I misspoke about the -3dB corner freq halving the voltage. -6dB would halve the voltage. Linkwitz-Riley or a parametric would do the trick.]
I've run the pair in stereo this way a few times (assumed the amps were underpowered from prior Labhorn user posts everywhere), with or without parametric EQ to flatten it, and even tried the Dynamic EQ to further reduce the lowest range when it's very cranked. DEG: -15dB, 30 or 35hz, LP12, 1:10, and a variety of threshold settings. I think the Dynamic EQ thing did make it sound cleaner at loud volumes and gave me peace of mind. But if you're right about below 30hz not really being the worst xmax range, then it may have only sounded cleaner because the room was shaking a little less or there was more available power (or power supply swing) for the rest of the range. I'm mostly worried about using these in a live situation either with bands or with crazy-ass live dubstep producers having use of my system. I'm wondering what the proper safety voltage limit is per sub in such situations.
[Edit: Mechanical safety concerns would be xlim, which doesn't account for overheating. But xmax is the linear excursion range that sounds good before nonlinearity sets in. 55Vrms produces xlim at 35hz according to the graph, or 77.8V peak, if a lone single-band signal made it through. The below 45Vp calculation is the xmax voltage limit at 35hz.]
And I was also curious when I do have these plugged in at home, what voltage limit is safe to use when having no HPF on it at all... or if that's even safe at any voltage. If the graphs are either correct or conservative, and my math is right above, wouldn't that mean a Labhorn could handle about 45Vp ((from 13mm/18mm)*45Vrms * sqrt 2) without any filters and not be at too much threat of getting damaged?
So maybe if I'm using them unfiltered at home, or letting a band or young synth head play through them, then 45Vp limiting is safe. But if I'm babysitting them the whole time or DJing through them, then the default 82Vp hard limit is alright per Labhorn? You mentioned some Crests incidentally prevent thermal limits, which I assume is also due to the Class D amp design & switching/PWM power supply.
[Edit: Meant Crown here, not Crest.]
If my thinking/math or whatever is wrong, please let me know. Assuming halving the voltage linearly halves the excursion compared to the graphs is a pretty big assumption, not to mention this live PA thing is very new territory for me.
Oh, and thanks for everyone's input so far.
The excursion is probably less below 30Hz than the sim, both the small chamber and the Lab 12 suspension are putting the "brakes" on excursion.
FLH typically have the greatest excursion around Fc, so yes, normal, but again, may be a bit exaggerated.
That said, the sim may be a few Hz high from what I recall the measured low corner of the LabHorn is.
Hornresp is a great predictive software (really great considering it is free- thanks David McBean!) but the dips and peaks in impedance, frequency response, and excursion are all a bit sharper than in real speakers.
Of course, short of putting a little video camera in the throat or a Lexan window and measuring excursion, about all you have for to go on for FLH ;).
Art
-
1)So then, the greatest xmax points would be around 35hz and 60hz?
2)I heard below the Fc the Lab12 drivers behave like in free air. And I always thought xmax skyrocketed in the deep bass on all woofers if allowed to.
3)Does the xmax graph matter that it's RMS voltage? If I'm using this to set limiters in the amp's DSP, should I be using sqrt 2 * RMSvolts? A 35hz test signal cranked way the hell up could still exceed xmax if the gain compensated for the filters, but assuming I or no one else do that, are these rational limits?
3)I'm mostly worried about using these in a live situation either with bands or with crazy-ass live dubstep producers having use of my system. I'm wondering what the proper safety voltage limit is per sub in such situations.
4) If the graphs are either correct or conservative, and my math is right above, wouldn't that mean a Labhorn could handle about 45Vp ((from 13mm/18mm)*45Vrms * sqrt 2) without any filters and not be at too much threat of getting damaged?
5)But if I'm babysitting them the whole time or DJing through them, then the default 82Vp hard limit is alright per Labhorn?
6)You mentioned some Crests incidentally prevent thermal limits, which I assume is also due to the Class D amp design & switching/PWM power supply.
7)Assuming halving the voltage linearly halves the excursion compared to the graphs is a pretty big assumption, not to mention this live PA thing is very new territory for me.
1) The Lab 12 Xmax is 13 mm, excursion vs.voltage is greatest near 35 Hz and around 60 Hz.
The simulation shows what excursion will be for a given voltage, the voltage could be continuous, as in a sine wave, or peak, as in a kick drum or a dropped microphone.
2) Below Fc the output of the LabHorn is similar to Lab 12 speakers in a small sealed cabinet, free air the drivers don't do squat. Given the same voltage, excursion goes up as frequency goes down in a normal sealed box, but the LabHorn is a sealed compression chamber coupled to a horn with a long air column, different from a sealed or bass reflex box.
3) The safe continuous sine wave voltage is 30, safe peak voltage around 60, depending on frequency.
4) Yes.
5) Depends on the average signal level, 82v peak limiting would not limit a continuous 40v sine wave, which could burn the coils. That said, I don't know if your amp could deliver 30-40 volts for a long enough duration to burn the Lab 12s.
6) That does not make sense, and I have not mentioned Crest amps in this thread.
7) Halving the voltage will halve the excursion at any given frequency, but excursion is not linear with voltage input across the frequency range of the LabHorn.
When the excursion exceeds Xmax, the speakers start to distort. 60 Hz tones will put out 120 Hz (second harmonic), 180Hz (third harmonic) etc. 35 Hz will have harmonics at 70, 105, 140 Hz etc.
Second harmonic distortion is generally highest, but because it is an octave higher, is not all that easy to detect, as octaves sound the same- 40 Hz is the low "E" on a four string bass, 80 Hz the low E on a guitar.
Push the subs hard around 60 Hz with no tops to mask the harmonics and you will become familiar with the sound of distortion, and how loud the subs get before distorting. Use a dB meter (or better, an RTA and a dB meter) and you will become familiar with how loud the subs get before they hit excursion and thermal limits.
Unless you have separate peak and RMS limiters, listening for signs of distress, and backing off level when you hear it is the best way to protect your speakers.
-
Sorry, meant Crown class D amps, not Crest. Reading too much stuff about Crest and the guy who sold me the Labhorns recommending them, I think. Given how many times I re-edited that post for grammar and such, I'm annoyed I didn't catch that!
6) That does not make sense, and I have not mentioned Crest amps in this thread.
When I said 82Vp would be with me babysitting it, I meant going out of my way not to send test signals and trying to not intentionally push them to the limit. That would also be with pre-recorded & mastered music, not live dub step producers or live bands. The latter ones worry me.
From what I can tell, while excursion is not linear across frequency, it is linear at a given frequency as voltage varies up to xmax. Readings I did last night on the subject seemed to indicate that's the whole point of xmax -- the limit where excursion of the driver is no longer linear with voltage at a given frequency, hence distortion increases. And then xlim is the point at which it fails; xmech being xlim * 2. So that makes sense.
By the way, your 30Vrms recommendation for maximum safe sine waves (and I assume live bands and live dub step producer wizards)... is that with HPF or without?
Finally, where are you getting 30Vrms and 60Vp? 30 * sqrt 2 = 42. Is the conversion for a sine wave different than the formula on Wikipedia or something else I'm missing? Or are you just considering thermal limits? Are my excursion calculations wrong when figuring the maximum safe voltage limit of 45Vp WITHOUT a HPF?
-
1)And then xlim is the point at which it fails; xmech being xlim * 2. So that makes sense.
2)By the way, your 30Vrms recommendation for maximum safe sine waves (and I assume live bands and live dub step producer wizards)... is that with HPF or without?
3)Finally, where are you getting 30Vrms and 60Vp? 30 * sqrt 2 = 42. Is the conversion for a sine wave different than the formula on Wikipedia or something else I'm missing? Or are you just considering thermal limits? Are my excursion calculations wrong when figuring the maximum safe voltage limit of 45Vp WITHOUT a HPF?
1) Xlim and Xmech are two terms for the same thing, the excursion limit where damage will occur if exceeded.
Xlim may be around double Xmax (the Lab 12 has an Xmax of 13mm and Xlim of 22mm) or more, or less- there are some speakers with Xmax of 8mm that the coil will hit the back plate and be destroyed at only 12mm Xlim. A B&C 18TBW100 has an Xmax of 12 mm(Xvar of 14) and Xlim of 28.5.
2) At 30v the speaker won't run in to thermal or excursion limits. With a slow time constant the RMS limiter could allow peaks through that could exceed Xlim, so use of a HPF is still advisable.
3) EIA 426 compressed pink noise has a crest factor of 6 dB, twice the average power of "normal" pink noise, sine waves 3 dB, twice the average power of the EIA 426, so the pair of Lab 12s take 400 watts sine wave, 800 watts EIA 426, and 1600 watts peak, double EIA 426, called "music program" on the spec sheet.
I picked an impedance of 2.25 ohms (the approximate impedance minima) for a pair for my calculation.
-
1) Xlim and Xmech are two terms for the same thing, the excursion limit where damage will occur if exceeded.
Xlim may be around double Xmax (the Lab 12 has an Xmax of 13mm and Xlim of 22mm) or more, or less- there are some speakers with Xmax of 8mm that the coil will hit the back plate and be destroyed at only 12mm Xlim. A B&C 18TBW100 has an Xmax of 12 mm(Xvar of 14) and Xlim of 28.5.
2) At 30v the speaker won't run in to thermal or excursion limits. With a slow time constant the RMS limiter could allow peaks through that could exceed Xlim, so use of a HPF is still advisable.
3) EIA 426 compressed pink noise has a crest factor of 6 dB, twice the average power of "normal" pink noise, sine waves 3 dB, twice the average power of the EIA 426, so the pair of Lab 12s take 400 watts sine wave, 800 watts EIA 426, and 1600 watts peak, double EIA 426, called "music program" on the spec sheet.
I picked an impedance of 2.25 ohms (the approximate impedance minima) for a pair for my calculation.
Isn't a limiter that you set max voltage in Vp going to limit peaks?
The nu 3000 DSP in stereo mode is already hard capped out and limited at 82Vp or so, which is just under 1700 Watts peak at 2 ohms. So isn't that an appropriate default setting to use if the HPF is on and I'm only playing properly mastered music myself (no live bands or live synth producers)?
I also use the parametric EQ which is reducing at 35hz (little bleed off below that), boosting at that 45hz-ish dip, few more slight dips and boosts, dip at 90-something, and then boost a little higher up so that the 90hz EQ isn't reducing the filter corner frequency range at 120hz.
Here's what the standalone Behringer software looks like. You can mess with it even if the amp isn't hooked up.
http://www.behringerdownload.de/iNuke/iNuke_Remote_Connect_V01-1.zip
What Vp setting would you use if there was no HPF and they were just reproducing a movie through them when they're not out?
-
Isn't a limiter that you set max voltage in Vp going to limit peaks?
H
THere are a number of different limiters-for different purposes.
A peak limiter is a fast attack limiter that stops/reduces mechanical damage (over excursion), a heat limiter (or other names) is a slow attack for long term heating.
A RMS or music limiter is more of an "average" limiter that is between the others. Its voltage is between them and the attack time is between them.
ANd then there are other that use side chain limiting for different limits at different freq and so forth.
There is no "one size fits all"
-
1)Isn't a limiter that you set max voltage in Vp going to limit peaks?
2)The nu 3000 DSP in stereo mode is already hard capped out and limited at 82Vp or so, which is just under 1700 Watts peak at 2 ohms. So isn't that an appropriate default setting to use if the HPF is on and I'm only playing properly mastered music myself (no live bands or live synth producers)?
3)What Vp setting would you use if there was no HPF and they were just reproducing a movie through them when they're not out?
1) Yes, peak limiters limit peaks, not average. The average level could be near the peak level until voltage limiting kicks in, which it will fairly quickly with an iNuke, since it must voltage limit to keep from blowing a 15/20 amp circuit.
2) 82v x 82v=6724/2.25=2988 watts, about 1500 watts per driver.
3) I would not use LabHorns without a HPF.
I have not used an iNuke 3000 DSP, nor am I familiar with it's output or DSP capabilities, so have no further advice for you.
Read the manual and use the advice from prior posts.
Don't worry, be happy :).
Art
-
3) I would not use LabHorns without a HPF.
I would argue that EVERY speaker needs a high pass filter.
-
I would argue that EVERY speaker needs a high pass filter.
I thought I read Danley state that using the Labhorn for very deep bass was perfectly safe at low volumes and didn't produce more distortion in the very low end than a reflex design. He said you just need to worry about boosting the very deep bass and remembering that any EQ settings are boosting the voltage going to the drivers in that band. I believe someone was talking about a crazy home theater setup. He didn't give a specific voltage limit, though. I was just messing around with them and didn't find 5-10V limits with no HPF was particularly messed up or distorted sounding. Drivers sounded well-behaved, I just didn't think it was incredibly clean sounding and got nervous about sending more into them.
-
I thought I read Danley state that using the Labhorn for very deep bass was perfectly safe at low volumes and didn't produce more distortion in the very low end than a reflex design. He said you just need to worry about boosting the very deep bass and remembering that any EQ settings are boosting the voltage going to the drivers in that band. I believe someone was talking about a crazy home theater setup. He didn't give a specific voltage limit, though. I was just messing around with them and didn't find 5-10V limits with no HPF was particularly messed up or distorted sounding. Drivers sounded well-behaved, I just didn't think it was incredibly clean sounding and got nervous about sending more into them.
You can run any speaker below its rated low freq response-AS LONG AS the levels are below max output. The lower the level-the low you can run it or add boosting down low. So it is a "sliding scale".
I would however make the argument that this is not the "intended usage" of the cabinet-and that you should use a highpass for the INTENDED usage of a cabinet. I guess I did not make that clean
-
You can run any speaker below its rated low freq response-AS LONG AS the levels are below max output. The lower the level-the low you can run it or add boosting down low. So it is a "sliding scale".
I would however make the argument that this is not the "intended usage" of the cabinet-and that you should use a highpass for the INTENDED usage of a cabinet. I guess I did not make that clean
Sure. In its intended usage, I run it with a HPF.
Still, when it's EQed in order to flatten the sound, it's pretty much making it so the excursion will occur, on average, at about the same point given similar dB and at any frequency. The PEQ settings are fairly similar to an inverted version of the excursion graph, but scaled differently to account for 6dB being double/half instead of the scaling in mm. I was thinking the HPF with a Butterworth was halving the voltage at 35hz with its 3dB reduction at the corner frequency. But I think it's 6dB that does that. Linkwitz-Riley will give me 6dB, though. But when I went back to the PEQ settings I used from REW for the sub (I put them into the Behringer DSP software that loads realtime into the nu3000), the 35hz notch is already doing that. So I suspect with the convolving and even no HPF, I'm not at much risk of causing over excursion at 35hz or overheating at 60hz. I don't convolve the deep bass, of course. So that's probably also good. Considering the excursion does not continue to increase the lower you go past 35hz, a HPF would seem to be more useful for utilizing power more efficiently from the amp than really protecting the driver.
-
I don't convolve the deep bass, of course. So that's probably also good. Considering the excursion does not continue to increase the lower you go past 35hz, a HPF would seem to be more useful for utilizing power more efficiently from the amp than really protecting the driver.
What do you mean "convolve" the deep bass?
A HPF does help protect the driver-not only from overexcursion-but also from rejecting freq that the driver can't reproduce-and therefore less heating going on.
Drivers fail from either overexcursion or heat or both.
-
What do you mean "convolve" the deep bass?
A HPF does help protect the driver-not only from overexcursion-but also from rejecting freq that the driver can't reproduce-and therefore less heating going on.
Drivers fail from either overexcursion or heat or both.
By convolving I meant EQing it flat using settings given by REW, which I don't do below 35hz anyway. Interestingly, convolving the rest of the spectrum would be evening-out excursion if the excursion graph is to be trusted.
Right, I realize that over-excursion and heat now are both issues. Would you say very low frequencies in general produce more heat?
-
By convolving I meant EQing it flat using settings given by REW, which I don't do below 35hz anyway. Interestingly, convolving the rest of the spectrum would be evening-out excursion if the excursion graph is to be trusted.
Right, I realize that over-excursion and heat now are both issues. Would you say very low frequencies in general produce more heat?
Then why not say you just eqed instead of using a term that means something different.
I would not be concerned with trying to eq a system to even out excursion. I have never heard of anybody doing or trying to do this.
The only time one freq produces more heat than another depends on the voltage level applied and the impedance at that freq.
By reducing the levels at frequencies that the sub cannot easily reproduce will reduce the heat.
Does it matter? It depends on how hard the loudspeaker is being pushed overall and how much energy is present in the freq that would be cut by the HP filter.
-
Then why not say you just eqed instead of using a term that means something different.
I would not be concerned with trying to eq a system to even out excursion. I have never heard of anybody doing or trying to do this.
The only time one freq produces more heat than another depends on the voltage level applied and the impedance at that freq.
By reducing the levels at frequencies that the sub cannot easily reproduce will reduce the heat.
Does it matter? It depends on how hard the loudspeaker is being pushed overall and how much energy is present in the freq that would be cut by the HP filter.
It's just when I used the REW convolving settings in the nu3000's DSP Parametric EQ, and then compared that to the excursion graph, they seemed to be pretty similar if you considered one in mm of excursion and one in dB with 6bB being half or double the voltage, right? It was actually unintentional but an interesting result.
-
It's just when I used the REW convolving settings in the nu3000's DSP Parametric EQ, and then compared that to the excursion graph, they seemed to be pretty similar if you considered one in mm of excursion and one in dB with 6bB being half or double the voltage, right? It was actually unintentional but an interesting result.
Where did you place the mic to get the measurement?
-
Where did you place the mic to get the measurement?
Right where I stand behind my djing gear, which is about 2/3 into the room. The subs are up against the wall (maybe 4in from it), facing outward, with maybe a foot or two space on the sides. A modified & repaired reflex 18in powered sub is in the middle between them, which I appreciate a little more now at low volumes than I did before now that I have very loud but horn monsters to compare it to. The Labhorns will be great out for volume, but reflex designs do sound more to me like bass I'm acquainted with. I had issues even hearing the single 18in sub unless it was in a corner when at lounges. Some couches and a coffee table are variously arranged in between the big PA system and the DJing gear, and I just pack it all up for gigging but leave it down there the rest of the time for practicing, movies, and background tunes blasting.
And when I say similar graphs, I mean similar but inverted. And not exact flipped versions, but surprisingly similar.
-
Right where I stand behind my djing gear, which is about 2/3 into the room. The subs are up against the wall (maybe 4in from it), facing outward, with maybe a foot or two space on the sides. A modified & repaired reflex 18in powered sub is in the middle between them, which I appreciate a little more now at low volumes than I did before now that I have very loud but horn monsters to compare it to. The Labhorns will be great out for volume, but reflex designs do sound more to me like bass I'm acquainted with. I had issues even hearing the single 18in sub unless it was in a corner when at lounges. Some couches and a coffee table are variously arranged in between the big PA system and the DJing gear, and I just pack it all up for gigging but leave it down there the rest of the time for practicing, movies, and background tunes blasting.
And when I say similar graphs, I mean similar but inverted. And not exact flipped versions, but surprisingly similar.
So you have 2 different types of subs-just a couple of feet from boundaries- in a kinda small room and used a single mic placement.
I HIGHLY bet that if you moved the mic several feet (say 4 or 5 or more) the response will change. Maybe a WHOLE LOT! So NOW what do you do?
Try this for some fun. Put a single tone (say 60Hz) into a single speaker (it doesn't matter which one). Now walk around the room and listen for the peaks and dips in response. Note where they are.
Now change freq-say 80Hz or 50hz. Now walk around and see how the locations of the peaks and dips change. So NOW what do you do to "even it out"?
Do you chase the different responses? or do you "eq for excursion"?
It all depends on what you are trying to do-in what location.
What you should realize is that attempting to eq bass in a small (or even kinda large) room is often an excercise in futility.
The most generally accepted way is to place the mic right at the grill (as close as you can get without touching it (like 1/4" or so). This will give you the best direct response.
That is the ONLY thing you can change. You cannot change the response of the room (the modal response- which is based on the physical dimension of the room) with EQ.
You may "think" you are-but when you move the mic you will find that other positions are now WORSE than they were before.
-
why dont you stack labhorn 2 on top of labhorn 1, with the bottom of the mouth againsts the wall, pointed at the back wall, about 18" out. and remove the 18" sub from the equation. and forget eq, other than a hipass 24 db/oct at around 32 hz. said setup should be Plus/minus 1 db from 35 to 100 hz... and of course room nodes will mess that up, but theres not much you can do about that.
-
why dont you stack labhorn 2 on top of labhorn 1, with the bottom of the mouth againsts the wall, pointed at the back wall, about 18" out. and remove the 18" sub from the equation. and forget eq, other than a hipass 24 db/oct at around 32 hz. said setup should be Plus/minus 1 db from 35 to 100 hz... and of course room nodes will mess that up, but theres not much you can do about that.
Sean,
Pointing the pair at a wall makes the response more peaked, the horn was not designed to reflect off a boundary at 18".
Even measured outdoors, a pair of LabHorns are not flat.
What gave you the idea they are flat from 35 to 100 Hz?
-
I did not EQ for excursion, but coincidentally that seems to have been the result to some extent.
The 18" sub was not being run when REW did the sweep. I let REW do its thing and EQed for flatness from 35hz to 120hz. Then I put those settings into the nu3000 DSP. I sometimes use convolving in Ableton generated directly from REW using either an external, secondary Emu 0204 or JACK ASIO intermediary utility for the 18" and the tops.
Usually I do not have the 18" sub actually running. I most often use it when I'm quickly hooking up *a* system to watch a movie and don't want to be bothered with the Labs.
When the 18" sub is run at the same time with the Labs, it is usually handling very deep bass with everything above about 40hz rolled off.
If I occasionally *do* run the 18" sub at the same time as the Labs, it's not without the 18" and the tops getting delayed to be put in phase with at least a portion of the response of the Labs. I'm using Ableton for the delays and have verified it with some loop back and some SPL meter tests.
That length of the Labhorn path is the biggest reason for a DJ not to use the Labs. The delay is a major hassle to deal with and adding any delay to a system if you can help it tends to be counterproductive to live DJ mixing.
I'm really starting to appreciate reflex designs more as a result. Horns also sound very different. Obviously the 18" sub can't get anywhere near as loud, the back amp plate will begin rattling a little at loud volumes, and the 18" driver doesn't respond as well to being pushed in general, but there's a nice tone to it at low to moderate volumes that I took for granted before. I sort of regret not buying another pair of them for $650 and modifying them, too. Screws, glue, some time and effort paid off.
I have more recently tried running the 18" at the same time and at about the same frequency range as the Labs. Doing this adds a little bit of punch in and out of room (standing in the kitchen). It's obviously going to change the frequency response of the whole system, but is less objectionable than running the Labs alone and without any EQ.
Not EQing the Labs at all results in a ridiculous amount of bass in about the 90hz region. That's how I used them the first few weeks, though.
I have not ever boosted anything below 35hz on them. I might at some point if I also accompany it with some very strict limiting and lower overall volume.
Thanks for the other recommendations. preferably, I would probably put the Labs on the ground with the mouth sides joined up like the bottom of one of those big 4 or 8 stacks, and then treat it as a little DJ stage or actually put it under a stage or something. But it's hard to justify for space-hogging reasons. Even corner-loading them at an angle presents convenience issues. I don't want to annoy my housemates too much beyond what they are already tolerating. They've been good sports so far.
-
why dont you stack labhorn 2 on top of labhorn 1, with the bottom of the mouth againsts the wall, pointed at the back wall, about 18" out. and remove the 18" sub from the equation. and forget eq, other than a hipass 24 db/oct at around 32 hz. said setup should be Plus/minus 1 db from 35 to 100 hz... and of course room nodes will mess that up, but theres not much you can do about that.
What specific frequency would you recommend putting the HPF at on the Labhorn if using a 48dB/oct Butterworth if it were available? I've heard all sorts of recommendations from 25-45hz, and of course the default recommended LR 24 setting of 35hz.
-
Sean,
Pointing the pair at a wall makes the response more peaked, the horn was not designed to reflect off a boundary at 18".
What about pointing them at each other instead of at the walls, and what would be the appropriate distance for that if so?
-
That length of the Labhorn path is the biggest reason for a DJ not to use the Labs. The delay is a major hassle to deal with and adding any delay to a system if you can help it tends to be counterproductive to live DJ mixing.
WHY? I agree the setup/alignment requires more delay time-but the actual "aspect" of "doing it" is the same amount of work.
I strongly disagree about the "counterproductive" part-especially when talking about DJ vs a live source.
With a DJ there is no "initial sound source" that there is a reference to.
With a live band you have the acoustic sound of a snare- or kick drum or even with a person talking you can hear the live sound from their mouth.
With a DJ there is none of that-so what if the sound is a couple of ms late? ALL of the signals will be a couple of ms late-so they are still arriving at the same time. You are not mixing acoustic and amplified signals.
Heck I would argue that if you delayed the whole DJ signal by any amount of time (even a couple of minutes or days) nobody would even notice-except the timing of the lights-but in most cases the lights are not synced to the music anyway and are just flashing/moving around-so who cares?
What is so "special" about a DJ signal" that delaying it will make a difference?
-
What specific frequency would you recommend putting the HPF at on the Labhorn if using a 48dB/oct Butterworth if it were available? I've heard all sorts of recommendations from 25-45hz, and of course the default recommended LR 24 setting of 35hz.
So at one point you are talking about the additional delay being a problem-then you want to add a 48dB/oct filter which will add more time-----------
I would stick with 24dB and use a Butterworth-rather than a L/R so that you can get more low freq before the rolloff.
-
What about pointing them at each other instead of at the walls, and what would be the appropriate distance for that if so?
One of the good things about a large bass horn is it has some directivity towards the upper frequency range (the sound goes where you point it), part of the reason you noted the upper response needed attenuation.
The LabHorn mouth is undersized, but in multiples the mouth area adds, filling in the low frequency rolloff.
Putting the two mouths together, rather than apart, will result in a smoother response as well as more low frequency response.
-
What do you mean "convolve"
That's a perfectly cromulent word. He should be offered the most enthusiastic contrafribularities as it is very compunctuous of you to have caused him such pericombobulation in questioning his choice of word.
Steve..
-
I recomended pointing at the wall vs pointing away pused against the back wall because that would put the mouth about 4 feet away from at least one boundary (1/4 wavelength reflection cancellation at 73 hzish).
Attempting to place the in 1/4 or 1/8 space should help flatten the peaky 1/2 space response of a pair of labhorns. I understand facing them at the corner will not ACTUALLY put it in 1/8 space, but at leastthere will not be the cancellation where d (distance from mouth to wall) is equal to 1/4 wavelength
-
I recomended pointing at the wall vs pointing away pused against the back wall because that would put the mouth about 4 feet away from at least one boundary (1/4 wavelength reflection cancellation at 73 hzish).
Labhorns are 45 inches deep (3 inches short of four feet), placed against a wall, the mouth is basically four feet from one boundary.
Their radiation pattern at 73 Hz is nearly omnidirectional.
Please explain why pointing them at a wall spaced four feet away from the mouth is of any advantage.
-
That's a perfectly cromulent word. He should be offered the most enthusiastic contrafribularities as it is very compunctuous of you to have caused him such pericombobulation in questioning his choice of word.
Steve..
I have no idea what you just said, but I'm strangely aroused by it.
http://convolver.sourceforge.net/
http://www.hometheatershack.com/roomeq/
So at one point you are talking about the additional delay being a problem-then you want to add a 48dB/oct filter which will add more time-----------
I would stick with 24dB and use a Butterworth-rather than a L/R so that you can get more low freq before the rolloff.
You lost me. A 48dB/oct Butterworth filter adds delay? Why and how much?
WHY? I agree the setup/alignment requires more delay time-but the actual "aspect" of "doing it" is the same amount of work.
I strongly disagree about the "counterproductive" part-especially when talking about DJ vs a live source.
With a DJ there is no "initial sound source" that there is a reference to.
With a live band you have the acoustic sound of a snare- or kick drum or even with a person talking you can hear the live sound from their mouth.
With a DJ there is none of that-so what if the sound is a couple of ms late? ALL of the signals will be a couple of ms late-so they are still arriving at the same time. You are not mixing acoustic and amplified signals.
Heck I would argue that if you delayed the whole DJ signal by any amount of time (even a couple of minutes or days) nobody would even notice-except the timing of the lights-but in most cases the lights are not synced to the music anyway and are just flashing/moving around-so who cares?
What is so "special" about a DJ signal" that delaying it will make a difference?
Not talking radio DJ or people who take requests and hit play, while yelling into a microphone "the back bar is open, the back bar is open" between tracks. Talking a DJ with a pair of headphones and hands on the faders and platters. A 12ms delay between the headphone and the sound system is incredibly annoying and prevents the ability to do quick single-ear syncs without double-ear check and repositioning of the phase of the blend.
Adding delay to the headphone helps, but you're still dealing with any additional delay between what you do and what happens to the sound. Delaying the headphones to match the sound system tends to also be more of a hassle than even just the burden of having to delay the tops separate from the subs, as usually you're using the headphone jack strait from a mixer, controller, or interface. I have yet to find a headphone amp with a built-in delay. If anyone knows of any, let me know.
I've tried some various methods of using JACK routing, Ableton, a mono sub, and a mono headphone (for a 4/4 interface), but right now the method of going out from the USB players into a laptop using a separate interface box has been most simple. It allows both DVS from the laptop on another interface/controller besides the Emu used for the delay in/out and the USB media players to be used standalone. And it doesn't require any single DVS software to run 2 interfaces at once, which is prone to glitches eventually. Running DVS on one interface and Ableton on another interface seems fine, though. This, again, also allows me the use of the USB players... though the way everything's routed sort of defeats the purpose of having standalone as a backup.
I have a DSP Crossover on the way that I hope will make things a little more simple, but that still won't resolve the lack of headphone delay when not using the laptop.
-
You lost me. A 48dB/oct Butterworth filter adds delay? Why and how much?
Not talking radio DJ or people who take requests and hit play, while yelling into a microphone "the back bar is open, the back bar is open" between tracks. Talking a DJ with a pair of headphones and hands on the faders and platters. A 12ms delay between the headphone and the sound system is incredibly annoying and prevents the ability to do quick single-ear syncs without double-ear check and repositioning of the phase of the blend.
Adding delay to the headphone helps, but you're still dealing with any additional delay between what you do and what happens to the sound. Delaying the headphones to match the sound system tends to also be more of a hassle than even just the burden of having to delay the tops separate from the subs, as usually you're using the headphone jack strait from a mixer, controller, or interface. I have yet to find a headphone amp with a built-in delay. If anyone knows of any, let me know.
I've tried some various methods of using JACK routing and Ableton, but right now the method of going out from the USB players into a laptop using a separate interface box has been most simple. It allows both DVS from the laptop on another interface/controller besides the Emu used for the delay in/out and the USB media players to be used standalone.
The amount of delay that a filter adds depends on freq. The lower the freq the longer the delay.
And when you say "how much"-as compared to what? NO delay- 12dB/oct or something else? You have to have a reference to tell what a difference is.
I would bet serious money that if you added even 100ms to a DJs mic-NOBODY would know the difference-since there is not a decently loud acoustic source. OK yeah-maybe several seconds might be noticable-----------------
And before you get to "hung up" on delay-consider the delay that exists between the speakers to the side and the DJ. You have already introduced delay just by physical location.
If you ask me-you are getting a bit "hung up" on little things and not looking at the whole big picture.
-
The amount of delay that a filter adds depends on freq. The lower the freq the longer the delay.
And when you say "how much"-as compared to what? NO delay- 12dB/oct or something else? You have to have a reference to tell what a difference is.
I would bet serious money that if you added even 100ms to a DJs mic-NOBODY would know the difference-since there is not a decently loud acoustic source. OK yeah-maybe several seconds might be noticable-----------------
And before you get to "hung up" on delay-consider the delay that exists between the speakers to the side and the DJ. You have already introduced delay just by physical location.
If you ask me-you are getting a bit "hung up" on little things and not looking at the whole big picture.
Is there a calculator for that somewhere on the net?
Well, let's say No Filter...
1) Versus the recommended 35hz L-R 24dB/oct?
2) Versus 25hz Butterworth 48dB/oct?
3) Versus 30hz Butterworth 48dB/oct?
Someone might notice phase issues of even 10ms difference between the tops and subs. But my primary concern is not about what the audience notices. I can already resolve the delay issues the audience would be noticing. The problem left is what affects my ability to actually mix. If I'm single-ear mixing and I have a 12ms delay between the sound system (or more likely: the Labhorn subs next to the monitors facing me) and my headphones, then that affects my ability to quickly mix in. And even if I do some voodoo to delay the headphone signal, now everything I do at the decks is more laggy. I never had such problems with front driver subs. Add to this the fact if you're using DVS, controllers, or some other computer-related DJ stuff, and you've already got a little latency to worry about. Best case: you've essentially doubled or tripled your latency with big horns. Worst case: the tops, the subs, the headphones are all delayed different amounts AND everything you do at the decks is slower to respond.
But I'm trying to make it work.
-
Is there a calculator for that somewhere on the net?
Well, let's say No Filter...
1) Versus the recommended 35hz L-R 24dB/oct?
2) Versus 25hz Butterworth 48dB/oct?
3) Versus 30hz Butterworth 48dB/oct?
Someone might notice phase issues of even 10ms difference between the tops and subs. But my primary concern is not about what the audience notices. I can already resolve the delay issues the audience would be noticing. The problem left is what affects my ability to actually mix. If I'm single-ear mixing and I have a 12ms delay between the sound system (or more likely: the Labhorn subs next to the monitors facing me) and my headphones, then that affects my ability to quickly mix in. And even if I do some voodoo to delay the headphone signal, now everything I do at the decks is more laggy. I never had such problems with front driver subs. Add to this the fact if you're using DVS, controllers, or some other computer-related DJ stuff, and you've already got a little latency to worry about. Best case: you've essentially doubled or tripled your latency with big horns. Worst case: the tops, the subs, the headphones are all delayed different amounts AND everything you do at the decks is slower to respond.
But I'm trying to make it work.
The total delay is not just the physical distance-or the length of the horn-or the type of filter used-but also the latency of the DSP unit. If you are using FIR filters-it gets even worse.
I have done gigs using subs that have longer path lengths than the Lab subs with some very well know DJs and none of them said anything about the delay being an issue.
-
The total delay is not just the physical distance-or the length of the horn-or the type of filter used-but also the latency of the DSP unit.
Yes. The DSP in the Behringer is apparently 0.6ms, so it's negligible.
Using loop back, I'm able to get precise latency measurements on the two interfaces, as well.
Oddly, the headphones need about 40ms delay when I have DVS running and two sessions of Ableton: one 32bit with JACK for delaying the headphones v. the sound system direct to the first interface running in MIDI internal mixer mode, and one 64bit session for delaying the tops using the Emu 0204 looping the DVS interface. The tops and subs seem to be in-phase and I've accounted for distance between the sound system and me. So apparently I'm missing something else or maybe something is 180 degrees out-of-absolute-phase. Obviously the laptop double Ableton delay solution won't help the USB players when I don't use DVS, but it was worth a try last night.
If anyone knows of a cheap stereo line-level or headphone DSP delay solution, please let me know.
-
If anyone knows of a cheap stereo line-level or headphone DSP delay solution, please let me know.
If you are using headphones for beat matching between songs on two decks, delaying their output will make that even more difficult, not a "solution".
If the delay from the loudspeakers (each foot of distance away is roughly one millisecond) is throwing off your ability to hear the phones, get some isolation phones like GK Ultraphones, they cut down the outside world SPL by about 30 dB.
-
If you are using headphones for beat matching between songs on two decks, delaying their output will make that even more difficult, not a "solution".
If the delay from the loudspeakers (each foot of distance away is roughly one millisecond) is throwing off your ability to hear the phones, get some isolation phones like GK Ultraphones, they cut down the outside world SPL by about 30 dB.
Thanks. I already have several different pairs of isolation headphones.
In my experience, delaying the headphones to match the sound system arrival time works better. There is a total lag between what you do and what you hear, but I find it's better to have consistency than otherwise. Obviously no lag would be optimal, though.
-
Is there a calculator for that somewhere on the net?
Well, let's say No Filter...
1) Versus the recommended 35hz L-R 24dB/oct?
2) Versus 25hz Butterworth 48dB/oct?
3) Versus 30hz Butterworth 48dB/oct?
Someone might notice phase issues of even 10ms difference between the tops and subs. But my primary concern is not about what the audience notices.
I think one thing you need to realize is that IIR filters (typical high pass, low pass, and EQ at all) cause a change in the phase response of the circuit (system). This alternation can be thought of as a different delay per frequency. So one filter may cause a very small delay at 45hz, the delay will be longer at 44, longer at 43, longer and 42 and so on. There is no way to get around this without going to FIR filters, which would then cause consistant delay (latency) accross the enire spectrum.
-
I think one thing you need to realize is that IIR filters (typical high pass, low pass, and EQ at all) cause a change in the phase response of the circuit (system). This alternation can be thought of as a different delay per frequency. So one filter may cause a very small delay at 45hz, the delay will be longer at 44, longer at 43, longer and 42 and so on. There is no way to get around this without going to FIR filters, which would then cause consistant delay (latency) accross the enire spectrum.
How much latency would FIR filters cause that have a 48dB/oct at 30hz HPF Butterworth and a 24dB/oct at 120hz LPF L-R?
-
How much latency would FIR filters cause that have a 48dB/oct at 30hz HPF Butterworth and a 24dB/oct at 120hz LPF L-R?
From the answers I got in this post # 535:
http://www.diyaudio.com/forums/multi-way/221434-rephase-loudspeaker-phase-linearization-eq-fir-filtering-tool-55.html
You could be looking at as much as 170ms if you wanted flat phase response from 16 kHz down to 30 Hz using FIR filters.
-
From the answers I got in this post # 535:
http://www.diyaudio.com/forums/multi-way/221434-rephase-loudspeaker-phase-linearization-eq-fir-filtering-tool-55.html
You could be looking at as much as 170ms if you wanted flat phase response from 16 kHz down to 30 Hz using FIR filters.
I was going to plug it into rephase but you beat me too it. The largest latency an ItechHD will allow is about 30ms. It looks like 62.5Hz is the lowest corner frequency it will allow with that penalty.
-
I did some tests with various tones and found the distortion levels below 25hz on the Labhorn to essentially render the sound towards 20hz, let alone below, useless. I'd say the bass is fairly pure in tone down to about 25hz and loud enough at my current parametric settings as previously described (the accidental almost inverted excursion graph EQ derived from REW). You can barely hear clean tone in the 23-24hz range, and anything below was just faint higher frequency distortion harmonics and no significant reproduction of the intended waveform. Anyway, I think right now I have a -6dB 42Vp limit and a 20hz 24dB/oct L-R HPF. The “happy accident” quasi inverted excursion graph parametric EQ is already protecting the drivers pretty well.
I transferred all my elaborate PC delays and EQs from loop-backs into the standalone delay DSP when it arrived. Interestingly, additional filtering of the infrasonic range beyond the parametric EQ, while it adds slightly to available power, seems to change the quality of the rest of the bass. Available power is not a problem, and I'm wondering if the sound change I'm getting, while tighter to some degree, is showing changes in the timing of the deep and mid bass with any HPF that you guys had mentioned. Now that I have all the delays set for every type of arrangement & backup (PC, direct media players, headphones, etc), the 48dB/oct HPF in particular seems most "off" to me now. It might be a placebo effect now that I know it's *supposed* to be worse, but it could also be me having been spoiled by the lack of time distortion from running without a HPF for the last week, with a low voltage limit between 10-20Vp, and all the proper delays.
The only thing I think might be left is finding the exact delay of the LabHorn. It seemed when I was doing summing and cancelation tests of the crossover frequency with the tops, I didn’t get to the right point until like 19-20ms. Is it possibly the parametric EQ on the subs that’s causing that? Anyway, I need to get a dedicated measurement microphone to be sure. The RS dB meter and RCA jack is too much hassle to work out the delays.
-
I did some tests with various tones and found the distortion levels below 25hz on the Labhorn to essentially render the sound towards 20hz, let alone below, useless. I'd say the bass is fairly pure in tone down to about 25hz and loud enough at my current parametric settings as previously described (the accidental almost inverted excursion graph EQ derived from REW). You can barely hear clean tone in the 23-24hz range, and anything below was just faint higher frequency distortion harmonics and no significant reproduction of the intended waveform. Anyway, I think right now I have a -6dB 42Vp limit and a 20hz 24dB/oct L-R HPF. The “happy accident” quasi inverted excursion graph parametric EQ is already protecting the drivers pretty well.
The only thing I think might be left is finding the exact delay of the LabHorn. It seemed when I was doing summing and cancelation tests of the crossover frequency with the tops, I didn’t get to the right point until like 19-20ms. Is it possibly the parametric EQ on the subs that’s causing that? Anyway, I need to get a dedicated measurement microphone to be sure. The RS dB meter and RCA jack is too much hassle to work out the delays.
The Labsub was never intended to go down to 20 or 25Hz. You hear distortion-DUH------------
As has been stated earlier-it it not just the length of the horn-but also the bandpass filters and to a small extent eq (unless using FIR filters) that determine what delay should be used in the processing to align with the tops.
A RS SPL meter is not even close to the right tool.
You need something that will show you the phase-to get a good alignment.
And THEN you need the skills to get a good phase alignment. Delay alone may not do it. Often you will need to use different bandpass filters/slopes/types etc and possibly some out of band eq to get it right.
But that does not mean you can't get "close enough" to be usable.
-
The Labsub was never intended to go down to 20 or 25Hz. You hear distortion-DUH------------
As has been stated earlier-it it not just the length of the horn-but also the bandpass filters and to a small extent eq (unless using FIR filters) that determine what delay should be used in the processing to align with the tops.
A RS SPL meter is not even close to the right tool.
You need something that will show you the phase-to get a good alignment.
And THEN you need the skills to get a good phase alignment. Delay alone may not do it. Often you will need to use different bandpass filters/slopes/types etc and possibly some out of band eq to get it right.
But that does not mean you can't get "close enough" to be usable.
I read Tom Danley saying below the 30-35hz range there was some useable bass if you keep the power lower to the speakers. I also think he said something about that deep range not having any more distortion than like a reflex (or some other design they were talking about) down low. In reality, the labhorn cannot go much past that 27hz horn frequency calculation thing, though. And the whole 35hz L-R 24db/oct HPF I guess is a very simplified analog solution that seems mostly just intended to get the most out of the amp and protect the drivers from the 35hz excursion spike. To get the most out of these, a lot more can be done. And maybe the other design he was talking about as similar deep bass distortion is just so high compared to the labhorns useable range's distortion levels that it's night and day.
-
I read Tom Danley saying below the 30-35hz range there was some useable bass if you keep the power lower to the speakers. I also think he said something about that deep range not having any more distortion than like a reflex (or some other design they were talking about) down low. In reality, the labhorn cannot go much past that 27hz horn frequency calculation thing, though. And the whole 35hz L-R 24db/oct HPF I guess is a very simplified analog solution that seems mostly just intended to get the most out of the amp and protect the drivers from the 35hz excursion spike. To get the most out of these, a lot more can be done. And maybe the other design he was talking about as similar deep bass distortion is just so high compared to the labhorns useable range's distortion levels that it's night and day.
What happens when you go low can get interesting-and especially if you ignore distortion.
There is one manufacturer that claims really low bass reproduction-yet when you listen even almost 2 octave above the claimed low freq spec-the distortion is more than 100% (is that possible?) The harmonic is way louder than the fundamental-HOWEVER the fundamental is still there-so it is still "usable"-you are just not hearing it as you "think" you would.
But you won't find that on anybodies spec sheet and people don't like to talk about it-but it DOES exsit with all loudspeakers.
-
Assuming that reduced amplifier output is not a problem, is there any disadvantage to wiring either each Labhorn internally as a series (driving a pair of the subs in stereo) or putting a pair of them that are already wired parallel internally into an external series (bridging the amp in mono)?
Since I have class D amps, they're better at driving high impedance loads, and I'm voltage limiting to about 42V peak (half the amp's peak capability at this impedance), is there any reason *not* to do this?
Would this adversely change damping or frequency response or something else other than just available power?
Is this simply pointless and increasing the impedance will not help the amps produced sustained dubstep type tones better?
I was just curious. The amps have the ability to bridge. I'm only even using one of them right now. I do have the ability to either rewire the subs internally or keep them that way and turn them into a series externally. Just not sure if there's any point. Heck, I never even opened up the second amp's box.
-
Assuming that reduced amplifier output is not a problem, is there any disadvantage to wiring either each Labhorn internally as a series (driving a pair of the subs in stereo) or putting a pair of them that are already wired parallel internally into an external series (bridging the amp in mono)?
Since I have class D amps, they're better at driving high impedance loads, and I'm voltage limiting to about 42V peak (half the amp's peak capability at this impedance), is there any reason *not* to do this?
Would this adversely change damping or frequency response or something else other than just available power?
Is this simply pointless and increasing the impedance will not help the amps produced sustained dubstep type tones better?
I was just curious. The amps have the ability to bridge. I'm only even using one of them right now. I do have the ability to either rewire the subs internally or keep them that way and turn them into a series externally. Just not sure if there's any point. Heck, I never even opened up the second amp's box.
Since you have multiple cabinets and multiple amplifiers-why not do the experiment yourself?
There are advantages and disadvantages to both series and parallel wiring. It depends on particular situations.
Since this is for YOUR use-then YOU need to decide.
But no, you will not do any damage to anything by putting them in series-assuming everything is run in the linear range.
Also remember that when you bridge you double the voltage-which is 4x the power (into the same impedance.
You also need aware of what "peak" may mean-especially in the amp settings. In many amps the PEAK- is 1.414 times the RMS of the maximum voltage waveform.
Lets look at some simple math. The lab12 is rated for 400 watts and is 4 ohms. Inside the labhorn it is likely higher-but lets ignore that for now.
That is 40V. So the "peak" or max voltage would be double that, or 80V. But if you wanted to set the "peak" on an amp like a lab gruppen-Crown-Powersoft etc you would enter 800x1.414 or 113volts.
But if you measure the voltage (with a typical voltmeter) it would read 80V NOT 113 volts. The peak on an oscope would read 113V.
So when discussing numbers-it is important to be specific what is meant by "peak". Is it "peak" or "max". THey can be very different.
-
Since you have multiple cabinets and multiple amplifiers-why not do the experiment yourself?
There are advantages and disadvantages to both series and parallel wiring. It depends on particular situations.
Since this is for YOUR use-then YOU need to decide.
But no, you will not do any damage to anything by putting them in series-assuming everything is run in the linear range.
Also remember that when you bridge you double the voltage-which is 4x the power (into the same impedance.
You also need aware of what "peak" may mean-especially in the amp settings. In many amps the PEAK- is 1.414 times the RMS of the maximum voltage waveform.
Lets look at some simple math. The lab12 is rated for 400 watts and is 4 ohms. Inside the labhorn it is likely higher-but lets ignore that for now.
That is 40V. So the "peak" or max voltage would be double that, or 80V. But if you wanted to set the "peak" on an amp like a lab gruppen-Crown-Powersoft etc you would enter 800x1.414 or 113volts.
But if you measure the voltage (with a typical voltmeter) it would read 80V NOT 113 volts. The peak on an oscope would read 113V.
So when discussing numbers-it is important to be specific what is meant by "peak". Is it "peak" or "max". THey can be very different.
I know it's not going to hurt anything as long as the amp's still set to 42Vp limit per Labhorn in the DSP section, but I thought I'd read somewhere that sub drivers in general will behave different tonally based on the impedance they're wired. I assumed the amp will have an easier time, but then again since it's also lowering the available watts, maybe it wouldn't make any difference.
I sort of don't want to unseal the other amp. I was thinking in terms of still using just the one, either rewiring each sub internally to series and using the amp in stereo, or leaving the subs internally wired to parallel and linking their cables in series with an adapter of some sort to utilize the single amp in bridge mode. I already have adapters for the bridge pin wiring, and another adapter could be done up to link them, or get one of those little boxes.
So I guess the real question is a two part:
1) will the Labhorns sound different (tonally, dynamically, transients, etc) wired as 12ohms instead of 3ohms, assuming any given amp in question can get them the same real volume in either case?
2) will a class D amp have an easier time with 12ohms than 3ohms using the same model & quantity of drivers but in a different arrangement, or is it completely moot since the available watts is decreased anyway?
I suppose I could also make it easier on the power supplies inside by only using one amp for each sub, but I still would rather keep the extra one sealed until I either have an emergency or decide I'm definitely not selling/returning it. The bridge adapters were made for that, but considering the Labhorns are already wired parallel, the amps are not supposed to go below 4ohms in bridge, and it's way more power than needed, I haven't bothered.
And I think we've already determined that the conservative linear performance cap for the Labhorn was about 45V peak assuming either no EQing or using live production or a band, 85V peak with some EQing with normal mastered music or babysitting, and about 65V peak from a conservative thermal standpoint. 77Vp is the theoretical xlim mechanical damage zone for a 34-36hz crazy peak without any HPF, EQ, or limiter. 45V peak can already get plenty loud to keep up with my tops and allow EQing. That's not even really with the subs coupled, but rather in separate corners, and my third modified reflex powered sub totally turned off rather than adding additional midbass punch.
So I'm comfortable with just 45V peak limiting kept on per Labhorn regardless, just curious about impedance, bridging, and power supply performance as other variables affecting the reproduction of continuous dubstep style tones, since I don't need to worry about mechanically or thermally damaging anything. I'm mostly wondering if I'm shortchanging the continuous tone capability of the amp by presenting a 2-3ohm load to it instead of 6 or 12ohms.
-
Is the dB a given driver produces based on the voltage or the watts?
Would a given arrangement of drivers wired as 12ohms instead of 3ohms require nearly 2X the volts to reach the same volume?
-
I know it's not going to hurt anything as long as the amp's still set to 42Vp limit per Labhorn in the DSP section, but I thought I'd read somewhere that sub drivers in general will behave different tonally based on the impedance they're wired. I assumed the amp will have an easier time, but then again since it's also lowering the available watts, maybe it wouldn't make any difference.
I sort of don't want to unseal the other amp. I was thinking in terms of still using just the one, either rewiring each sub internally to series and using the amp in stereo, or leaving the subs internally wired to parallel and linking their cables in series with an adapter of some sort to utilize the single amp in bridge mode. I already have adapters for the bridge pin wiring, and another adapter could be done up to link them, or get one of those little boxes.
So I guess the real question is a two part:
1) will the Labhorns sound different (tonally, dynamically, transients, etc) wired as 12ohms instead of 3ohms, assuming any given amp in question can get them the same real volume in either case?
2) will a class D amp have an easier time with 12ohms than 3ohms using the same model & quantity of drivers but in a different arrangement, or is it completely moot since the available watts is decreased anyway?
I suppose I could also make it easier on the power supplies inside by only using one amp for each sub, but I still would rather keep the extra one sealed until I either have an emergency or decide I'm definitely not selling/returning it. The bridge adapters were made for that, but considering the Labhorns are already wired parallel, the amps are not supposed to go below 4ohms in bridge, and it's way more power than needed, I haven't bothered.
And I think we've already determined that the conservative linear performance cap for the Labhorn was about 45V peak assuming either no EQing or using live production or a band, 85V peak with some EQing with normal mastered music or babysitting, and about 65V peak from a conservative thermal standpoint. 77Vp is the theoretical xlim mechanical damage zone for a 34-36hz crazy peak without any HPF, EQ, or limiter. 45V peak can already get plenty loud to keep up with my tops and allow EQing. That's not even really with the subs coupled, but rather in separate corners, and my third modified reflex powered sub totally turned off rather than adding additional midbass punch.
So I'm comfortable with just 45V peak limiting kept on per Labhorn regardless, just curious about impedance, bridging, and power supply performance as other variables affecting the reproduction of continuous dubstep style tones, since I don't need to worry about mechanically or thermally damaging anything. I'm mostly wondering if I'm shortchanging the continuous tone capability of the amp by presenting a 2-3ohm load to it instead of 6 or 12ohms.
As I said there are MANY factors involved-NOT just one or two.
For example most modern amplifiers cannot produce full output for very long. Typically it is on the order of 0.08 seconds or less. Longer than that and the power drops-often by 6dB or 1/4 power. Some amps are much less than 0.08 sec. What about your amps? Amplifiers can deliver power into higher impedances for longer periods of time.
When you put speakers in series the damping factor goes up by a factor of 4 (due to the higher impedance as oppossed to them being in parallel).
With a higher load impedance there is less loss across the cable and the damping factor again goes up.
When an amp is in bridge mode the damping factor drops in half-due to the 2 output sections being in series.
However the "control" as seen from each of the speakers is not as high as it would be with a single speaker-since the other speaker is in series-causing a much higher impedance or less control from the amplifier.
So as you can see there are pluses and minuses. YOU need to decide what is most important to you.