Here's the picture you ALL have been waiting for

H

Helge A. Bentsen

Senior
Jan 19, 2011
1,017
64
48
44
Oslo, Norway, Norway
drbentsen.no
Or maybe not ;)

I bought 4 used Labsubs from Silas Pradetto a while back; they arrived here two days ago. They were slightly delayed in a hurricane, but made it through unharmed and well. Initial test suggest that their performance is pretty impressive compared to 4xsb1000 or 4x650-r2 which I've previously used with my SH-50s. They outrun two SH-50s pretty easily, and my whole warehouse is shaking in its foundation, I like that.

The picture is from our industry standard test procedure FTLC6P, and I ended up with a 12dB Butterworth hi-pass@25hz, an LR24dB LP@80hz on the Labs, and a 24dB LR@100hz on the SH-50s with 5ms of delay. I currently set my limiter at 57V(800W/4ohm) and they seem to cope with that amount of power just fine. If anybody has some settings they wish to recommend for the Labs, I'm open for suggestions.


I do believe that this is the first operational Labs in Norway, and they are available for rent if you're around in these parts of the world ;)

Many thanks to Silas for the deal :)
 

Attachments

  • DSC00917.jpg
    DSC00917.jpg
    162.3 KB · Views: 0
Re: Here's the picture you ALL have been waiting for

Hey Helge, it looks like they arrived in pretty good shape :D~:-D~:grin:

I am not sure that the high pass you are using is going to be steep enough. I've plotted the magnitude response of the LABs and they start to roll off steeply at 30Hz. Others have told me that they blow a lot of woofers without the HP filter. Your filter falls in the gray area between the 'safe' steep HP at 30Hz, and the 'deadly' total lack of HP. In my years of using 16 LAB subs, I have never blown a woofer, even with 4,000 watts into one box. I suggest some experimentation to see how safe you're going to be...you can't really hear when the LABs start to run out of gas. You'll only know when it's too late.

I do have some various EQ settings for various configurations, when I have time I'll post the list of filters and a screenshot of the target response.

Enjoy!
 
I would be more worried about slope than frequency, a second order filter keeps excursion constant as frequency drops. The penalty for a higher order filter is paid at frequencies that you are cutting anyway, in a band you will be introducing no additional loudspeakers into, so I have no compunction about using 4th order Butterworth.
 
Re: Here's the picture you ALL have been waiting for

Bennett Prescott said:
I would be more worried about slope than frequency, a second order filter keeps excursion constant as frequency drops.
Click to expand...

I never really thought about that before. It makes sense to me, but why then do so many loudspeaker manufacturers recommend a second order filter to high pass subwoofers? Also, I wrote down "compunction". :)
 
Re: Here's the picture you ALL have been waiting for

Bennett Prescott said:
I would be more worried about slope than frequency, a second order filter keeps excursion constant as frequency drops. The penalty for a higher order filter is paid at frequencies that you are cutting anyway, in a band you will be introducing no additional loudspeakers into, so I have no compunction about using 4th order Butterworth.
Click to expand...

I'm slow and from the South - please 'splain this to me nice and slow. Maybe some pics would help?
 
Re: Here's the picture you ALL have been waiting for

Helge Dr. Bentsen said:
the SH-50s with 5ms of delay.
Click to expand...

Hi Helge,

While I have never used Labs or SH50's, 5ms seems low if your goal is a proper alignment, considering the path length alone of a Lab sub would account for more than 5ms. Then factor in group delay for your IIR filters..... Did you measure the phase response at the Lab to SH50 acoustic crossover to come up with 5ms or was it a guess?
 
Re: Here's the picture you ALL have been waiting for

Jeff Babcock said:
Hi Helge,

While I have never used Labs or SH50's, 5ms seems low if your goal is a proper alignment, considering the path length alone of a Lab sub would account for more than 5ms. Then factor in group delay for your IIR filters..... Did you measure the phase response at the Lab to SH50 acoustic crossover to come up with 5ms or was it a guess?
Click to expand...

5ms was all that I needed to match up their phase response at 80 hz. I often have to use shorter delay times with the SH-50s compared to other speakers I've tested, I think this is due to the fact that the SH-50 is a horn and the other speakers I've tested is mostly front loaded in the LF.

5ms was on axis at 5m. Different delay times gave me different results both on axis and off axis, but on average i got what I belive is the best summation with the least side effects in that experiment. I don't know if it's "right", research is ongoing.
 
Re: Here's the picture you ALL have been waiting for

Gene Hardage said:
I'm slow and from the South - please 'splain this to me nice and slow. Maybe some pics would help?
Click to expand...
LEts say you want to use a 30hz highpass. so you are wanting to say cut 20Hz.

Look at the attached graphs. At 20Hz, with a 24dB filter, it is much lower in level than with the 12dB slope.

I agree-a 24dB Butterworth is my highpass (for subs) of choice.

Other bands depends on what I am trying to do with the adjacent band. But the bottom end of subs have no adjacent band-so kill as much as you can.

BUT using higer order filters will result in additional problems-so 24 dB is the max I usually use.
 

Attachments

  • filters.jpg
    filters.jpg
    84.5 KB · Views: 0
Re: Here's the picture you ALL have been waiting for

Bennett Prescott said:
I would be more worried about slope than frequency, a second order filter keeps excursion constant as frequency drops. The penalty for a higher order filter is paid at frequencies that you are cutting anyway, in a band you will be introducing no additional loudspeakers into, so I have no compunction about using 4th order Butterworth.
Click to expand...

The tradeoff in HPF is at the knee. Most of the time a 2nd order is completely sufficient in the stopband, but not in the transition to the stopband. This is where having very good models comes into play, and building the filter TFs into the models.

Often it takes a 2nd order filter and a parametric combined to get the high pass filter correct.
 
Re: Here's the picture you ALL have been waiting for

Phil Graham said:
Often it takes a 2nd order filter and a parametric combined to get the high pass filter correct.
Click to expand...

How would one go about measuring this? Would I simply look at Smaart and use whatever combination of filters gets me to an 'ideal' rolloff? How do I know what an 'ideal' rolloff is? A highpass is primarily designed to (1) eliminate overexcursion below cutoff and (2) reduce distortion below cutoff, correct? Measuring excursion in a LAB is going to be pretty difficult, distortion should be easier. What's an acceptable amount of distortion? Is distortion directly related to excursion (I think it might be...).

Sorry Phil, I think I got carried away with questions. :lol:
 
Re: Here's the picture you ALL have been waiting for

Helge Dr. Bentsen said:
Or maybe not ;)

I bought 4 used Labsubs from Silas Pradetto a while back; they arrived here two days ago. They were slightly delayed in a hurricane, but made it through unharmed and well. Initial test suggest that their performance is pretty impressive compared to 4xsb1000 or 4x650-r2 which I've previously used with my SH-50s. They outrun two SH-50s pretty easily, and my whole warehouse is shaking in its foundation, I like that.

The picture is from our industry standard test procedure FTLC6P, and I ended up with a 12dB Butterworth hi-pass@25hz, an LR24dB LP@80hz on the Labs, and a 24dB LR@100hz on the SH-50s with 5ms of delay. I currently set my limiter at 57V(800W/4ohm) and they seem to cope with that amount of power just fine. If anybody has some settings they wish to recommend for the Labs, I'm open for suggestions.


I do believe that this is the first operational Labs in Norway, and they are available for rent if you're around in these parts of the world ;)

Many thanks to Silas for the deal :)
Click to expand...
très chi[COLOR=#0]c, chapeaux!
Once I'll have some time a stop by at your place and listen to these beats - will let you know way before... [/COLOR]
 
Re: Here's the picture you ALL have been waiting for

I'm testing a Dynacord PowerH 5000 right now, I'ts a pretty impressive amp. It runs all four labs flawlessy in 2.1 ohms according to it's load monitor. I've confirmend my initial delay setting of 5ms, I have to use 7 ms with the Dynacord due to it's internal dsp. I also did some testing of different lo-cut filters, it seems like a 30 hz 24 db butterworth is a pretty good compromise between performance and protection.

Silas: I know that you were using Itechs on your subs, and the limiter section in the Dynacord is pretty similar to that of an Itech. What did you use as limiter settings? I currently limit them at 58v RMS, 82v peak in 2 ohm.
 
Re: Here's the picture you ALL have been waiting for

Silas Pradetto said:
How would one go about measuring this? Would I simply look at Smaart and use whatever combination of filters gets me to an 'ideal' rolloff? How do I know what an 'ideal' rolloff is?

Sorry Phil, I think I got carried away with questions. :lol:
Click to expand...


Out of band EQ has changed my world, and method of setting up x-overs. My method for determining where the out of band EQ is needed goes something like this:

1. I measure the output of my DSP with the HP/LF engaged, and save the trace in smaart.
2. Then I apply the filter to the speaker and measure the response with the HP/LP.
3. I'll then use a few key EQ cuts to make the acoustic slope of the speaker match the electrical slope of the DSP more closely.

Normally it only takes one or 2 very narrow, and very steep cuts to make it work. The most noticeable difference for me was adding out of band EQ to the HPF on subs. You can easily smooth out nasty peaks in the response of subs in their upper band while still using a much shallower HPF.


Evan
 
Re: Here's the picture you ALL have been waiting for

Silas Pradetto said:
Hey Helge, it looks like they arrived in pretty good shape :D~:-D~:grin:

I am not sure that the high pass you are using is going to be steep enough. I've plotted the magnitude response of the LABs and they start to roll off steeply at 30Hz. Others have told me that they blow a lot of woofers without the HP filter. Your filter falls in the gray area between the 'safe' steep HP at 30Hz, and the 'deadly' total lack of HP. In my years of using 16 LAB subs, I have never blown a woofer, even with 4,000 watts into one box. I suggest some experimentation to see how safe you're going to be...you can't really hear when the LABs start to run out of gas. You'll only know when it's too late.

I do have some various EQ settings for various configurations, when I have time I'll post the list of filters and a screenshot of the target response.

Enjoy!
Click to expand...

Eminence has a bit to say about the LABsub and how to not blow the drivers:

From their FAQ page at:
http://www.eminence.com/support/faq/

5. You must use a high pass filter set to 35 Hz and that has a slope of at least 24dB per octave to realize the real potential of the design. Many people are using huge power on these cabinets day in and day out, but they are the ones who run steep high pass filters on them.
 
Re: Here's the picture you ALL have been waiting for

Silas Pradetto said:
How would one go about measuring this? Would I simply look at Smaart and use whatever combination of filters gets me to an 'ideal' rolloff?
Click to expand...

Silas, Its not so much an ideal rolloff, as one that balances the excursion behavior of the cabinet. There's no sense in "over" rolling off the cabinet to the extent that the excursion below Fb is less tha an above Fb. Conversely, you don't want to "under" roll off the speaker below Fb have have a greater excursion maxima below Fb than above. You can simulate such things decently with a free program like WinISD, or you can use several more advanced pieces of software (e.g. SoundEasy, LSPCad, AkAbak, LEAP).

How do I know what an 'ideal' rolloff is? A highpass is primarily designed to (1) eliminate overexcursion below cutoff and (2) reduce distortion below cutoff, correct? Measuring excursion in a LAB is going to be pretty difficult, distortion should be easier.
Click to expand...

Excursion of a horn cabinet like the LAB is pretty readily observable by simulation.

Let me explain my thought process for high pass filtering on a hypothetical vented speaker. Lets say that Fb is 35Hz, and the -3dB point is 37Hz. For such a speaker, the highest driver excursion will be somewhere in the mid 40s or low 50s. Lets say that our hypothetical driver starts to sound bad after 10mm one way excursion. If we hit 10mm one way in the mid 40s, we don't want to go much above 10mm one way below the 35Hz Fb. For simplicity's sake, lets assume that the calculated excursion of the driver is 20mm one way at 25Hz. To cut the excursion in half to 10mm one-way, we need to reduce the drive level 6dB at 25Hz. That means our HPF must be 6dB down at 25Hz, and how we get there depends on the excursion rise between 35Hz and 25Hz.

What's an acceptable amount of distortion? Is distortion directly related to excursion (I think it might be...).

Sorry Phil, I think I got carried away with questions.
Click to expand...

Distortion is related to the declining BL product, which in turn depends on the amount of excursion. The relationship is direct, but not linear. The Bl product is pretty consistent as long as the coil is fully in the magnetic gap, and then starts to taper off more rapidly as the coil leaves the gap.

As for the amount of distortion that is acceptable, our ears are very insensitive to low order order harmonic distortion at low frequencies. Less than 10% THD is not readily audible, and nothing to worry about. Driver air venting and suspensions can produce a lot of anharmonic content, and that is often more audible and annoying than the drivers' THD.
 
Evan,

My method is to measure the sub with no HPF/LPF and then use EQ to try and get it flat for at least two octave above my desired LPF and as far as possible below my HPF. Then I apply filters and sometimes add another peaking EQ to make the filter roll off more like my goal.
 
Top Bottom