Ricci's Skram Subwoofer & Files

[jay michael]

6 hours ago, Alexlel said:

Hello !

@jay michael Your last measurements has been done with 1 cab ?

@Ricci How close is it to the simulation (21SW152-4) ?

I expect to build 2 Othorns next month but the Skram looks like a real contender, what sensitivity should I expect for 1 Skram, 21DS115-4 loaded ?
My aim is to get the highest output between 25hz and 70hz , can't make up my mind between Othorn or Skram...

Yes this is with 1 cab.  I've built both. The Othorn is amazing, its got a low growl to it like nothing else I have heard before.  That said, the skram more or less equals the low performance of the Othorn and to my ears sounds just as clean doing it.  The skram is substantially easier to build, is also smaller and lighter.   Factoring in cnc time, I built 4 skrams for just a little bit more money than what it cost to build 2 othorns. 

[jay michael]

5 hours ago, SME said:

I think proximity of measurement to the port would affect the balance between upper and lower frequencies, not the low frequency corner.  It looks like the rear wall may be effectively lengthening the port more than anticipated.  The tuning looks to be around 26 or 27 Hz maybe?  I'm curious what the simulation looks like.  How much extra length do you add to the port in the sim to account for the rear wall?

If you take a look at my newest measurement it doesn't seem to show as low of a tune as my earlier measurements.   I am still pretty new to measurements so hard to say if I am getting the best data, but I did try to follow your guy's recommendations on the second measurements

[m_ms]

1 hour ago, jay michael said:

Yes this is with 1 cab.  I've built both. The Othorn is amazing, its got a low growl to it like nothing else I have heard before.  That said, the skram more or less equals the low performance of the Othorn and to my ears sounds just as clean doing it.  The skram is substantially easier to build, is also smaller and lighter.   Factoring in cnc time, I built 4 skrams for just a little bit more money than what it cost to build 2 othorns. 

Can you elaborate on the "low growl" aspect of the sound of the Othorn, and how it differentiates the sound of the Othorn compared to other subs you've heard? Is this something that makes the Othorn sound different compared to the Skram, even though you deem them to be "equal" to each other in the low-end performance?  

[Ricci]

7 hours ago, jay michael said:

If you take a look at my newest measurement it doesn't seem to show as low of a tune as my earlier measurements.   I am still pretty new to measurements so hard to say if I am getting the best data, but I did try to follow your guy's recommendations on the second measurements

That looks close to me. 30Hz tuning is what it should be. Real world measurements usually don't show quite as sharply a defined corner, which can make it more difficult to visually pin point the tuning. Looks close to expected though.

The maximum port output is usually a little below the low frequency corner of the combined system response. This is due to the output cancellation below tuning which nullifies some of the port energy. Peak airspeed is a little below tuning usually as well. If the measurement is closer to the port than the upper frequency radiator it can show increased low frequency output and a lower corner. This was probably what was happening in the first measurements.

I've attached the sim'd response with the 21SW152-4 with semi inductance included. There are 2 traces on the graph because one shows the correct port area and apparent functional length (With the notch at 150Hz) and the other shows the actual straight vent length with adjusted vent area to maintain the correct tuning. I figured the real result would be somewhere between the two but wasn't sure which one it would be closer to. I was hoping it would be closer to the actual vent length with the smoother response above 150Hz, but it seems to be closer to the unadjusted one. It is shifted a bit better than that sim shows though. Notice that the sim rolls off much more abruptly above 120Hz towards the notch. Your measurement shows that the notch due to the pipe resonance and the upper end peaks are shifted up in frequency a little and the response is quite flat up to 150Hz before a much more abrupt notch at the pipe resonance. This is what I wasn't entirely sure about before measuring this one. Basically I was trying to trick the tuning of the ports to be a bit lower than expected so that vent area can be maximized, but also trick the pipe resonance up in frequency a little too. It worked to some extent but I was hoping it might come in a bit higher than it did. Anyway those are a few  tricks I tried to use on these. It's the little things that can really be of interest sometimes.

@SME As far as calculating vent length for tuning purposes when loading into a wall it's a little bit of calculation and a little bit of art. I've been pretty close on mine so far but perhaps I'm just lucky. On the Skram the back wall actually forms a 90deg bend for the vent so I add in the chord length from the bend through the center section and then the pipe termination only terminates on one side because the other is the back panel so I add an additional half of the vent width (length from center end plane of vent to center-point of a 45deg line sketched from the end of the terminated pipe section to the back wall. It seems to be very close on this one.

[jay michael]

Thanks for the explanation, I don't understand most of it but super appreciate the people who do!  Thanks a ton for putting another great design out into the public for people to build!  I do have a question, if my measurement was taken at 1 volt @ 1 meter, what is the assumed power shown in the predictions?

 

[SME]

3 hours ago, jay michael said:

Thanks for the explanation, I don't understand most of it but super appreciate the people who do!  Thanks a ton for putting another great design out into the public for people to build!  I do have a question, if my measurement was taken at 1 volt @ 1 meter, what is the assumed power shown in the predictions?

I forgot to ask, did you calibrate the absolute level of the mic using an SPL meter?  Your numbers look a bit high compared to the sim @Ricci posted.  Note that the sim is for response with 2 volt at 1 meter instead of 2 volt at 2 meter (or the equivalent, 1 volt at 1 meter).  I'd expect your numbers to come in ~6 dB lower, but I could also be missing some detail.

Anyway, assuming the absolute levels of your mic and multimeter are calibrated reasonably accurately, your data shows "1 watt nominal" response at *2 meters*.  Because you were measuring at 1 meter, I had you set the output to 1 volt instead of 2 so that your numbers could be directly compared to those made with reference at 2 meter.  Note that manufacturers (and Hornresp) usually report output at 1 meter rather than 2, so if you want to compare to typical manufacturer specs, add 6 dB to your numbers.

Also note that by "nominal" the "1 watt" assumes that the impedance is 4 ohms for all frequencies.  Then, 2 Vrms (as used when measuring at 2 meters) will consume 2^2 / 4 = 1 watt on average (formula is V^2/Z).  In reality, impedance varies a lot with frequency and actual power consumed will depend a lot on frequency content.

So to translate your figures to manufacturer's specs sensitivity, again assuming your levels are accurate, your cabs are about 104 dB/1W/1m sensitive in the deep bass, from 30-50 Hz!  Of course, a manufacturer would look at the average over the bandwidth or even the peak value and publish that, so maybe 110 dB/1W/1 meter if quoting things generously.

In practice, if you stack the 4 all together, the stack  will gain close to 6 dB, giving you 110 dB/1W/1m at 40 Hz.  With a bridged K10 amp on those that's like 150 dB SPL max output at 1 meter, 130 dB at 10 meters, 110 dB at 100 meters.  Even if your numbers are off by 6 dB, these are impressive numbers.

Clearly your next task will be to preemptively make peace with your neighbors. 

[jay michael]

I did not calibrate my mic with a db meter. Looking at the REW website they mention this.

Calibrating the SPL Reading

Calibrating the SPL reading gives REW an absolute SPL reference by entering a reading from your SPL meter while a speaker or subwoofer calibration signal is playing. Alternatively an SPL calibrator may be used. This step is not required if you are using a USB mic with a cal file that contains a sensitivity figure.

If I am reading this correctly my mic should already be calibrated?? I am using the minidsp umik1 which came with a calibration file. 

Thanks for explaining the calculations, its what makes this place really special,  I have been learning a ton!    If I do need to calibrate the spl, ill try and get that done and remeasure asap.  Regardless,  I can say that I have a huge freaking smile on my face now!  That was a tremendous amount of work getting those 4 cabs completed on my own and now I am feeling the rewards   You guys rock!   

Fedex is showing my SH-46's arriving tomorrow, its like Christmas over here!  Can't wait to get some music through the new rig, should be a potent little system.  

 

[SME]

2 hours ago, jay michael said:

I did not calibrate my mic with a db meter. Looking at the REW website they mention this.

Calibrating the SPL Reading

Calibrating the SPL reading gives REW an absolute SPL reference by entering a reading from your SPL meter while a speaker or subwoofer calibration signal is playing. Alternatively an SPL calibrator may be used. This step is not required if you are using a USB mic with a cal file that contains a sensitivity figure.

If I am reading this correctly my mic should already be calibrated?? I am using the minidsp umik1 which came with a calibration file. 

I bought my UMIK-1 several years ago, directly from Cross Spectrum Labs.  The cal file did not contain an absolute SPL reference.  I had to calibrate the absolute SPL using the procedure described above.  Your situation may be different.

2 hours ago, jay michael said:

Thanks for explaining the calculations, its what makes this place really special,  I have been learning a ton!    If I do need to calibrate the spl, ill try and get that done and remeasure asap.  Regardless,  I can say that I have a huge freaking smile on my face now!  That was a tremendous amount of work getting those 4 cabs completed on my own and now I am feeling the rewards   You guys rock!   

Fedex is showing my SH-46's arriving tomorrow, its like Christmas over here!  Can't wait to get some music through the new rig, should be a potent little system.  

This calibration doesn't affect the shape of the curves, only the absolute number.  There's no need to remeasure as long as you can figure out what the SPL difference before and after calibration is.  You can add or subtract dB to a curve in REW on the "all SPL" tab.  Under controls, look at "Measurement Offsets".  Enter the number of dB to add and click "add offset to data" to apply the change.  Don't forget to save.

The SH-46s ought to be very good as tops.  My only concern would be that they are a bit narrow in coverage at "40 x 60 degrees".  Maybe that's typical for PA, and if they can be turned onto their sides to get 60 degrees in the horizontal, that'd be helpful.  A dance floor is not necessarily "long and skinny" like a concert venue.  Though larger events can be like that.  A lot depends on the details of the setup.

I once went to a multi-day outdoor festival with a huge system in the absolute middle of nowhere (rural Zambia to be precise).  A huge array was flown like 15-20m up (rough guesses on dimensions) on temporary scaffolding.  I think it took a sizable crew a week to build it, even with help from a crane.  The whole dance floor was something like 200m deep x 100m wide.  The camping areas were probably 0.5 km behind the stage, and the bass still thumped clearly in our tents all day and all night.

[SME]

10 hours ago, Ricci said:

That looks close to me. 30Hz tuning is what it should be. Real world measurements usually don't show quite as sharply a defined corner, which can make it more difficult to visually pin point the tuning. Looks close to expected though.

The maximum port output is usually a little below the low frequency corner of the combined system response. This is due to the output cancellation below tuning which nullifies some of the port energy. Peak airspeed is a little below tuning usually as well. If the measurement is closer to the port than the upper frequency radiator it can show increased low frequency output and a lower corner. This was probably what was happening in the first measurements.

Great explanation.  Thanks!

I do want to note that when doing simulations of BP6 designs, I noticed that the roll-off tends to start a bit higher up, relative to the Helmholtz tuning frequency, than a similar plain vented design.  Presumably, the horn is still doing a bit of unloading there, contributing to earlier roll-off.  For the data you posted, Hornresp shows 29.5 Hz for the tune (I assume based on Helmholtz).  The impedance minimum comes in slightly lower in the sim, around 29 Hz.  OTOH compared to the plain vented, the vent in a BP6 doesn't seem to unload as rapidly below tune, so they may have a bit more "usable output" down there.

In any case, the sim shows response that's almost around -2.5 dB at 29.5 Hz.  The same point in the measurement looks to be around 27 Hz.  That's really not too far off.

11 hours ago, Ricci said:

@SME As far as calculating vent length for tuning purposes when loading into a wall it's a little bit of calculation and a little bit of art. I've been pretty close on mine so far but perhaps I'm just lucky. On the Skram the back wall actually forms a 90deg bend for the vent so I add in the chord length from the bend through the center section and then the pipe termination only terminates on one side because the other is the back panel so I add an additional half of the vent width (length from center end plane of vent to center-point of a 45deg line sketched from the end of the terminated pipe section to the back wall. It seems to be very close on this one.

If I understand you right here, shouldn't you include this twice, at least for ground plane because the ground has the same effect as the rear wall in terms of limiting expansion at the inlet/outlet?

[jay michael]

7 hours ago, SME said:

I bought my UMIK-1 several years ago, directly from Cross Spectrum Labs.  The cal file did not contain an absolute SPL reference.  I had to calibrate the absolute SPL using the procedure described above.  Your situation may be different.

This calibration doesn't affect the shape of the curves, only the absolute number.  There's no need to remeasure as long as you can figure out what the SPL difference before and after calibration is.  You can add or subtract dB to a curve in REW on the "all SPL" tab.  Under controls, look at "Measurement Offsets".  Enter the number of dB to add and click "add offset to data" to apply the change.  Don't forget to save.

The SH-46s ought to be very good as tops.  My only concern would be that they are a bit narrow in coverage at "40 x 60 degrees".  Maybe that's typical for PA, and if they can be turned onto their sides to get 60 degrees in the horizontal, that'd be helpful.  A dance floor is not necessarily "long and skinny" like a concert venue.  Though larger events can be like that.  A lot depends on the details of the setup.

I once went to a multi-day outdoor festival with a huge system in the absolute middle of nowhere (rural Zambia to be precise).  A huge array was flown like 15-20m up (rough guesses on dimensions) on temporary scaffolding.  I think it took a sizable crew a week to build it, even with help from a crane.  The whole dance floor was something like 200m deep x 100m wide.  The camping areas were probably 0.5 km behind the stage, and the bass still thumped clearly in our tents all day and all night.

Ok thanks for that, Ill try doing the calibration tonight.  Yeah the 46's have a pretty tight pattern, i plan for the time being to run them in landscape config to utilize the 60 deg coverage angle.  To realize my goal of a highly power dense system in the least amount of cabinets and size,  the 46 seemed to be one of best options I could find.  As they are designed to array with more 46's it keeps the door open to expansion down the road.  For the time being I am planning to add a second pair of smaller synergy horns to act as fills when I need the extra coverage and to use when I don't need the power of the 46's. Working on a deal now for a pair of sm60f's.   I think the 4 subs and 4 synergy horns should give me decent flexibility for both our studio space and our festival stage.  Outdoors we deploy in a large geodesic dome, to expand on size this year we are moving the sound and booth tech outside of the dome under an inflatable stage cover that will basically nest up to the arc of the outside of the dome. The dome will cover the dance floor, using shade sails instead of the full cover to provide shade for the daytime.

https://jaymichaelt.imgur.com/all/

https://imgur.com/a/67pWeUh

 

Another point is everything needs to fit inside a 8 foot by 12 foot enclosed trailer, sound system, dome, scaffolding camping gear etc.  We could barely make it work before, the new system is considerably smaller and approx 400 pounds lighter. 

[Ricci]

7 hours ago, SME said:

Great explanation.  Thanks!

If I understand you right here, shouldn't you include this twice, at least for ground plane because the ground has the same effect as the rear wall in terms of limiting expansion at the inlet/outlet?

That's something I haven't considered much. With vents next to the ground this could account for another 2.5 to 3" (7 cm or so) of apparent vent length give or take. That's enough to drop the tune about 1Hz lower. Might be a bit of that going on in Jay's measurement too. That's something that affects any vented or horn cab that radiates next to the ground or a wall. Since these may or may not be arranged with the vents on the floor I wouldn't add it into the sim by default. The internal geometry is always in effect though.

 

I might have some data in the archives that would show this difference in loading the vents but I'd have to dig around. I may not.

[Ricci]

1 hour ago, jay michael said:

 Working on a deal now for a pair of sm60f's.   I think the 4 subs and 4 synergy horns should give me decent flexibility for both our studio space and our festival stage.

https://jaymichaelt.imgur.com/all/

https://imgur.com/a/67pWeUh

 

Another point is everything needs to fit inside a 8 foot by 12 foot enclosed trailer, sound system, dome, scaffolding camping gear etc.  We could barely make it work before, the new system is considerably smaller and approx 400 pounds lighter. 

Where are you located Jay?

[jay michael]

Calgary Alberta Canada 

[SME]

4 hours ago, jay michael said:

https://jaymichaelt.imgur.com/all/

https://imgur.com/a/67pWeUh

 

Another point is everything needs to fit inside a 8 foot by 12 foot enclosed trailer, sound system, dome, scaffolding camping gear etc.  We could barely make it work before, the new system is considerably smaller and approx 400 pounds lighter. 

The first link doesn't work for me.  It's great to hear you're able to both upgrade the system and reduce its size and weight.  Double win!

[jay michael]

Yup! Try this 

[Alexlel]

Based on your fresh hornresp inputs and your measured B&C 21DS115-4 T&S, it looks different of your simulations in page 2, is it normal ?
Which one should we consider for now ?
It's the same for the LaVoce, higher sensitivity with those inputs.

Gonna wait for some more measurements but if this one is around 98dB/99dB for 1W/1m as shown in the new simulation, I'll go for it instead of Othorn !

 

[m_ms]

Maybe this has been brought up at an earlier juncture, but is the driver in a Skram unloaded below tuning frequency similarly compared to a tapped horn or ported enclosure to necessitate a high-pass filter,  and if so what's the proper/sufficient slope and type to use here - Butterworth, 2nd to 4th order?

7th or 8th order slope HPF are oftentimes considered too steep, but why?  It should follow they offer better protection of the driver while "eating" less dB's down to the cut-off; to my ears a steeper HPF (like 8th order) makes the low-end appear slightly more extended for this reason, but perhaps also a bit too distinctive. I'm back to a 3rd order BW to see how that fares..

[Ricci]

23 hours ago, Alexlel said:

Based on your fresh hornresp inputs and your measured B&C 21DS115-4 T&S, it looks different of your simulations in page 2, is it normal ?
Which one should we consider for now ?
It's the same for the LaVoce, higher sensitivity with those inputs.

Gonna wait for some more measurements but if this one is around 98dB/99dB for 1W/1m as shown in the new simulation, I'll go for it instead of Othorn !

 

Use the newest one. The old ones were probably a few iterations back. I've probably got 50+ Skram records with different drivers, vent configs, front chamber configs, alternate cab designs etc. Also when siming cabs you can come up with quite different results based on where you set the section points (S3,S4,etc) and how the throat is simed. The cab hasn't changed the sim inputs have. This is why I usually try to wait until I have measurements to determine the correct HR inputs. Jay's measurements confirmed which way of siming the cab is closest. Even with a lot of previous experience there are times where it's a bit of a guess until it's measured. Like the vent behavior mentioned earlier. This is also why I never assume other peoples cabinets perform like sim'd without seeing measurements to back it up. 

[Ricci]

9 hours ago, m_ms said:

Maybe this has been brought up at an earlier juncture, but is the driver in a Skram unloaded below tuning frequency similarly compared to a tapped horn or ported enclosure to necessitate a high-pass filter,  and if so what's the proper/sufficient slope and type to use here - Butterworth, 2nd to 4th order?

7th or 8th order slope HPF are oftentimes considered too steep, but why?  It should follow they offer better protection of the driver while "eating" less dB's down to the cut-off; to my ears a steeper HPF (like 8th order) makes the low-end appear slightly more extended for this reason, but perhaps also a bit too distinctive. I'm back to a 3rd order BW to see how that fares..

Yes. These unload below tuning. The only cabs that do not are those with a sealed chamber on one side of the driver, or IB. 

I use 18dB BW most of the time. 12dB when I can get away with it. I never go any steeper than 24dB. It typically causes greatly increased group delay at the corner and removes any possibility of useful output below tune. Typically there is some useful output till about 1/3rd octave below tuning. Also extremely sharp cutoffs just seem to sound unnatural to me. 

[Ricci]

On 7/17/2019 at 4:57 PM, jay michael said:

Yup! Try this 

Beautiful setup. 

[Alexlel]

Here is a 3D Sketchup file for the Skram if someone is interested.
Bracing is missing.

It's rounded for metric units but it should not change that much compared to the original one :

Quote

Skram Dimensions: 24"x32"x36" (609.6mm x 812.8mm x 914.4mm)

Became : 610mm x 813mm x 915mm

Skram.skp

[jay michael]

It’s not how I will normally deploy them...  but figured they should hang and and get to know each other

 

https://imgur.com/gallery/r5UkOor

[SME]

On 7/18/2019 at 5:49 AM, m_ms said:

7th or 8th order slope HPF are oftentimes considered too steep, but why?  It should follow they offer better protection of the driver while "eating" less dB's down to the cut-off; to my ears a steeper HPF (like 8th order) makes the low-end appear slightly more extended for this reason, but perhaps also a bit too distinctive. I'm back to a 3rd order BW to see how that fares..

That's a good description.  And on bandwidth limited systems, it may not necessarily be a bad thing to sharpen the cut-off a bit, as long as the ringing doesn't get too excessive.  I think it'd be worthwhile to try to do a systematic study of listener preference on this point.  It's something I intend to do as I migrate toward testing my low frequency optimization methods in other rooms and on other systems.  Not everyone can have 5 Hz extension like I do.

Unfortunately this is greatly complicated by the fact that HPFs get applied to audio content also.  These are often high Q and can contribute a lot of ringing.  This seems to be especially popular with movies these days.  It's understandable.  Everyone wants "more deep bass" from cinema subs that only play to 30 Hz.  Unfortunately though, 30 Hz sub vary quite a bit wrt HPFs used, so the results will be unpredictable and not translate well.  Ideally, soundtracks should avoid using HPFs above like 10 Hz.  This way, playback systems can be optimized, using a high Q HPF if desired, to get the best compromise of "deep bass sound" vs. "ringing".  Alas, this is very wishful thinking on my part.

[dgage]

2 hours ago, jay michael said:

It’s not how I will normally deploy them...  but figured they should hang and and get to know each other

 

https://imgur.com/gallery/r5UkOor

Trying to recreate the old Maxell ad?   Just a little bit of output capability there.  

[jay michael]

The wife can blow dry her hair.