Hi!
TLDR: Why is Madisound (ultra-high-end speakers) recommending an enclosure with a Qtc of .8 for my mids?
https://www.madisoundspeakerstore.com/approx-5-woofers/scanspeak-15w/8530k-01-5-revelator-woofer/ (.3 cuft + 20% fill)
LONGER VERSION:
I am building a 4-way system using active crossovers (MiniDSP Flex8). I was big into hi-fi car audio when I was young. I am fanatical about sound quality but I also want the music to have... impact. Read: I will be using a VERY generous house curve 😎.
I plan on running 2 of the mids linked above, and 2 more 7-inch drivers to manage the transition from thunderous bass to the high frequencies. I will need a lot of juice between 80hz and 200 hz to do it properly. The box recommendations for the 7s are similar in having a high Qtc.
On paper the higher Qtc is a good thing for this setup. I get a boost in this mid-frequency range BUT, what looks good on paper doesn't always sound good to the ear. I do not want to hear the speakers, I want to hear music. In the low range, I typically associate "invisible speakers" with transient response, which is achieved with a LOW Qtc (~.5). I am not sure if this applies to the midrange the same way it would to a subwoofer.
I don't really want to build multiple boxes to test. Do you have any experience with this? Thanks gang!
TLDR: Why is Madisound (ultra-high-end speakers) recommending an enclosure with a Qtc of .8 for my mids?
https://www.madisoundspeakerstore.com/approx-5-woofers/scanspeak-15w/8530k-01-5-revelator-woofer/ (.3 cuft + 20% fill)
LONGER VERSION:
I am building a 4-way system using active crossovers (MiniDSP Flex8). I was big into hi-fi car audio when I was young. I am fanatical about sound quality but I also want the music to have... impact. Read: I will be using a VERY generous house curve 😎.
I plan on running 2 of the mids linked above, and 2 more 7-inch drivers to manage the transition from thunderous bass to the high frequencies. I will need a lot of juice between 80hz and 200 hz to do it properly. The box recommendations for the 7s are similar in having a high Qtc.
On paper the higher Qtc is a good thing for this setup. I get a boost in this mid-frequency range BUT, what looks good on paper doesn't always sound good to the ear. I do not want to hear the speakers, I want to hear music. In the low range, I typically associate "invisible speakers" with transient response, which is achieved with a LOW Qtc (~.5). I am not sure if this applies to the midrange the same way it would to a subwoofer.
I don't really want to build multiple boxes to test. Do you have any experience with this? Thanks gang!
If you cross this to another driver then you'll be making it as flat as you like. Just because the box is a higher Q (0.8 isn't that high) doesn't mean you can't undo the fast transition afterward.
This is a midwoofer designed for sealed boxes.
Perhaps you misinterpreted the context of Madisound’s recommend.
If you want a midwoofer for a vented cabinet choose the K00 model.
If you want a midwoofer for a sealed box choose the K01 version.
If you buy this K01 driver and want to get a QTC 0.5 you can just use a bigger box.
If you want a pure midrange use the 15M (or 18M)
Perhaps you misinterpreted the context of Madisound’s recommend.
If you want a midwoofer for a vented cabinet choose the K00 model.
If you want a midwoofer for a sealed box choose the K01 version.
If you buy this K01 driver and want to get a QTC 0.5 you can just use a bigger box.
If you want a pure midrange use the 15M (or 18M)
Last edited:
I dont think you'll notice the difference in Qtc between your target and the actual response when it interacts with all the room modes. However, if you are crossing it in the rolloff region, keep Qtc under 0.7 for best linearity.
Pretty vague description of what you intend building. Can you be more specific? Brand, unit type, x/o points, etc...
Thanks for the response, everyone!
@profiguy considering the house curve I'm after - a high Qtc creating a bump in response around 100HZ will actually help me.
@tktran303 - I will be putting these in a sealed box. I do not know if I want at Qtc of .5 - that is the question of this post. I want SQ AND I need a bump at 100HZ.
Restatement of the question If a Qtc > .7 degrades sound quality WHY would Madisound recommend a box with .8?
@Lojzek -
EQUIPMENT IN PLACE NOW:
Mini DSP Flex 8 - this is a 4-way active crossover and EQ. I also own a MiniDSP microphone for measurement. I can move the crossover points whenever I want. Tons of flexibility.
Morel MDT33 tweets - crossed over at 1900 hz - 48db/octave
ScanSpeak 15W/8530K-01 (2 of these with plans to get 2 more) - Crossed over at 1900HZ and 100hz at the moment.. Would like to consider bringing these down to 90 or 80. 24 db/octave
ScanSpeak M18WO-27-08 (6 of these (3 in each speaker) - I will tweak the cross-over as needed but planning on crossing these over at somewhere between 80 and 100 hz on the low end, probably a 12db slope. On the high end will take them to 300HZ with a 6 db slope. Considering how much midbass I need to keep up with the subwoofers I intend on sending these about 250 watts per channel.
2 12" Image Dynamics IDMax subwoofers - 600 watts on each. Crossed over at 100HZ, 24db slope. Sealed box. Qtc = .7
1 18" Dayton Ultamax subwoofer - 1200 watts - Crossed over at 50hz with 6db slop with a harder taper at 100HZ. Ported box tuned to 18hz.
@profiguy considering the house curve I'm after - a high Qtc creating a bump in response around 100HZ will actually help me.
@tktran303 - I will be putting these in a sealed box. I do not know if I want at Qtc of .5 - that is the question of this post. I want SQ AND I need a bump at 100HZ.
Restatement of the question If a Qtc > .7 degrades sound quality WHY would Madisound recommend a box with .8?
@Lojzek -
EQUIPMENT IN PLACE NOW:
Mini DSP Flex 8 - this is a 4-way active crossover and EQ. I also own a MiniDSP microphone for measurement. I can move the crossover points whenever I want. Tons of flexibility.
Morel MDT33 tweets - crossed over at 1900 hz - 48db/octave
ScanSpeak 15W/8530K-01 (2 of these with plans to get 2 more) - Crossed over at 1900HZ and 100hz at the moment.. Would like to consider bringing these down to 90 or 80. 24 db/octave
ScanSpeak M18WO-27-08 (6 of these (3 in each speaker) - I will tweak the cross-over as needed but planning on crossing these over at somewhere between 80 and 100 hz on the low end, probably a 12db slope. On the high end will take them to 300HZ with a 6 db slope. Considering how much midbass I need to keep up with the subwoofers I intend on sending these about 250 watts per channel.
2 12" Image Dynamics IDMax subwoofers - 600 watts on each. Crossed over at 100HZ, 24db slope. Sealed box. Qtc = .7
1 18" Dayton Ultamax subwoofer - 1200 watts - Crossed over at 50hz with 6db slop with a harder taper at 100HZ. Ported box tuned to 18hz.
Greets!
Simply to keep the box size reasonable since a 0.5 Qtc requires an infinitely huge box.
Math wise it's ~0.866 Qtc before any damping, so either damp 'to taste' in situ or use DSP and/or by XO design and/or build the biggest box you can tolerate.
SQ wise down low is very room dependent, so even 0.7 Qtc isn't good enough for some folks in their rooms. The best way to test is with an impulse response with DIYers historically relying on the click test with 'stretched' as in 'TIGHT AS A DRUM' stretched.
Attachments
Why would you cross those 15W’s that low? Rather aim for 200-300Hz and reduce cone movement and thus distortion. And the enclosure size becomes irrelevant too.
you're kind of in a fiddling while Rome burns situation. With your very ambitious plans I hardy think a small difference in the qts of the mid will matter, you will be putting out bigger fires.
@GM - thank you. Useful. So is the conclusion here that the primary reason a .3 cuft box would be recommended here would be to keep the box small? A box with better SQ would be much larger and the effect would be negligible in the midbass arena - which is what these are for. Do I have it right?
@mordikai - lol. You might be right but if I'm going to spend the time building a box I want to do it right. Right now I'm running 2 of the 15ws in two oversized enclosures (don't ask why). I am running these with the subs and Morels and it sounds pretty darn good but the 15ws are not keeping up. I want to feel a guitar pluck and I don't want it to sound like it's coming from speakers and I especially do not want to hear the subwoofers.
@markbakk - interesting question. Worth testing. See comments above - they do serve the purpose pretty well they just can't keep up on their own. Once I have the 7" drivers incorporated I can try what you're saying however, if that creates the best result I want to know I can bring the crossover back down and they'll be in the right box.
Again - thanks all!
@mordikai - lol. You might be right but if I'm going to spend the time building a box I want to do it right. Right now I'm running 2 of the 15ws in two oversized enclosures (don't ask why). I am running these with the subs and Morels and it sounds pretty darn good but the 15ws are not keeping up. I want to feel a guitar pluck and I don't want it to sound like it's coming from speakers and I especially do not want to hear the subwoofers.
@markbakk - interesting question. Worth testing. See comments above - they do serve the purpose pretty well they just can't keep up on their own. Once I have the 7" drivers incorporated I can try what you're saying however, if that creates the best result I want to know I can bring the crossover back down and they'll be in the right box.
Again - thanks all!
Hi, thinking it a minute, Q of a mid driver in a multiway speaker has to do with crossover. Since you have DSP you can tailor it the way you want and there shouldn't be much of a difference as long as the box is in the ballpark, right? DSP allows you to design the box with some other criteria than the Q. If you had passive xo the Q would affect the crossover making your life easier or harder.
Make the construct so that DI is nice, then utilize the DSP for any house curve, either as global EQ or just adjusting gain of the ways.
Additional thoughts that affects:
If kick bass is your goal then make sure there is enough headroom in the system = cone area, you have more than most but calculate to make sure, you don't want to fall short. Also, construct the system so that first early reflections in your room and positioning don't make huge null at the listening spot, they usually do. Also keep group delay at bay, but the flex8 has FIR? so shouldn't be a problem.
Have fun 🙂
ps. thinking a minute what the peak in response from high Q alignment is? its resonance, the cone moves further with less power. On a closed box the cone is moving the same for any given output SPL regardless of Q and thus difference doing a peak in frequency response with box alignment or EQ is how much amplifier power is involved. More amplifier power through voice coil means more heat and eventually compression when approaching limits of the system. In this sense high Q system would be more punchy at the limit of operation, more output before compression. But, if goal is high sound quality and punchy bass, no compression, you want to stay well below limit of operation, which means you build so big of a system the Q is irrelevant.
Make the construct so that DI is nice, then utilize the DSP for any house curve, either as global EQ or just adjusting gain of the ways.
Additional thoughts that affects:
If kick bass is your goal then make sure there is enough headroom in the system = cone area, you have more than most but calculate to make sure, you don't want to fall short. Also, construct the system so that first early reflections in your room and positioning don't make huge null at the listening spot, they usually do. Also keep group delay at bay, but the flex8 has FIR? so shouldn't be a problem.
Have fun 🙂
ps. thinking a minute what the peak in response from high Q alignment is? its resonance, the cone moves further with less power. On a closed box the cone is moving the same for any given output SPL regardless of Q and thus difference doing a peak in frequency response with box alignment or EQ is how much amplifier power is involved. More amplifier power through voice coil means more heat and eventually compression when approaching limits of the system. In this sense high Q system would be more punchy at the limit of operation, more output before compression. But, if goal is high sound quality and punchy bass, no compression, you want to stay well below limit of operation, which means you build so big of a system the Q is irrelevant.
Last edited:
Your application is a waste of those K01 Revelators. Those drivers are specifically designed to dig deep in a small sealed box. To that end they sacrifice sensitivity. You'd be better off using drivers that are more sensitive, if you only want to go down to 100 or 300Hz. A driver such as the 6ND430 from Eighteen Sound or one of many similar drivers from B&C, BMS, Faital Pro, etc would be the trick.
Last edited:
Yeah more sensitivity means less power required, less all kinds of distortion and compression. Get even more sensitivity with bigger ones like doubling diameter of each of the ways drivers in the plan. Here is one speaker that is made to kick https://www.diyaudio.com/community/threads/shadow-of-the-colossus-build-thread.357825/ and I bet it kicks even with low listening volume.
ps. anyone have pointers how much voice coil heating affects? in order to figure out how much compression a system would have? Goal being reach some particular SPL capability without audible effects of the heating. From this we could calculate how sensitive system is needed and what size drivers / ways is required.
edit. here is one article about it https://www.stereophile.com/content/hot-stuff-loudspeaker-voice-coil-temperatures-page-2
ps. anyone have pointers how much voice coil heating affects? in order to figure out how much compression a system would have? Goal being reach some particular SPL capability without audible effects of the heating. From this we could calculate how sensitive system is needed and what size drivers / ways is required.
edit. here is one article about it https://www.stereophile.com/content/hot-stuff-loudspeaker-voice-coil-temperatures-page-2
Last edited:
I'm too lazy today to do any math but if efficiency of typical loudspeaker driver is about 1-2% per the article it means 98% of power fed to it turns into heat in the voice coil, almost all of it. Simplifying time out of equation and approximating that ten times power means ten times heat increase. From the graph in the article ten fold increase in voice coil temperature increased Re so much that the system becomes 4.7db less sensitive?
So, remembering that most people hear 1db difference in level / EQ we could take that as some sort of simple audibility threshold. one could calculate how much power some voice coil can take before it heats up so much so that 1db change occurs. Without touching calculator to find exact value I'll assume it is roughly doubling the heat from room temperature? Certainly, if system output is reached with say 10W of amplifier power instead of 100W there is 10 times less heat introduced into the system, possibly no audible compression from heat.
Uh, must use calculator, found one http://www.hometheaterengineering.com/splcalculator.html
Assuming we want to listen 85db level and account for 15db peaks, we'd like to reach 100db at listening spot with 10W of amplifier power. System sensitivity would need to be roughly 96db/W/m. Some 15" drivers can reach to this around 100Hz but not much below so one needs multiple / bigger drivers / big boxes below 100Hz to reach such system sensitivity, perhaps 6x 7" per side. Horns increase sensitivity and so on 🙂
Here is reason why big systems sound more impressive, less compression, less distortion, mo dynamic sound.
I hope some one comments on my posts to prevent misinformation, just thinking out loud 🙂
So, remembering that most people hear 1db difference in level / EQ we could take that as some sort of simple audibility threshold. one could calculate how much power some voice coil can take before it heats up so much so that 1db change occurs. Without touching calculator to find exact value I'll assume it is roughly doubling the heat from room temperature? Certainly, if system output is reached with say 10W of amplifier power instead of 100W there is 10 times less heat introduced into the system, possibly no audible compression from heat.
Uh, must use calculator, found one http://www.hometheaterengineering.com/splcalculator.html
Assuming we want to listen 85db level and account for 15db peaks, we'd like to reach 100db at listening spot with 10W of amplifier power. System sensitivity would need to be roughly 96db/W/m. Some 15" drivers can reach to this around 100Hz but not much below so one needs multiple / bigger drivers / big boxes below 100Hz to reach such system sensitivity, perhaps 6x 7" per side. Horns increase sensitivity and so on 🙂
Here is reason why big systems sound more impressive, less compression, less distortion, mo dynamic sound.
I hope some one comments on my posts to prevent misinformation, just thinking out loud 🙂
Attachments
Last edited:
@cowanaudio - you are right. I purchased them like 15+ years ago and was going to use them in a car. The subs are always in the back of the vehicle and so you need some amazing drivers up front that go low in order to draw the sound of the bass up front. That is really one of the underlying reasons for this post. I am trying to find a way to utilize them well - they are spendy little guys.
Should I cross them over lower? Not at all? My perception is that the lower I can go with the midranges the more likely the subwoofers disappear sound-wise.
@tmuikku - thank you for your analysis! You've gone into some things that are new to me.
I did some rough calculations awhile to this effect and it seemed like by adding the 6 7" drivers with around 600 watts the midrange would keep up to the bass. My calculations were far cruder than what you have gone into. I simply started with the sensitivity of the drivers and added 3db for each doubling of driver quantity and power. I concluded that with 8 midrange drivers and that many watts - if it's not loud enough, I've got problems... 🙂 BUT, it seems what you are saying is that driving the speakers with a lot of wattage - even if they are capable is not as good as fewer watts across more speakers.
It does appear that the FlexEight has FIR: "IIR (10bands/outputs) + FIR (Stereo bank on input @ 2048taps for impulse response correction)", additionally I am often driving the audio from a computer where I have REW and Equalizer APO so I'd guess I could manage it from there. That said, I have some homework to do on it. FIR is a new idea to me.
@GM - 👍 . thanks for the confirmation,.
Should I cross them over lower? Not at all? My perception is that the lower I can go with the midranges the more likely the subwoofers disappear sound-wise.
@tmuikku - thank you for your analysis! You've gone into some things that are new to me.
I did some rough calculations awhile to this effect and it seemed like by adding the 6 7" drivers with around 600 watts the midrange would keep up to the bass. My calculations were far cruder than what you have gone into. I simply started with the sensitivity of the drivers and added 3db for each doubling of driver quantity and power. I concluded that with 8 midrange drivers and that many watts - if it's not loud enough, I've got problems... 🙂 BUT, it seems what you are saying is that driving the speakers with a lot of wattage - even if they are capable is not as good as fewer watts across more speakers.
It does appear that the FlexEight has FIR: "IIR (10bands/outputs) + FIR (Stereo bank on input @ 2048taps for impulse response correction)", additionally I am often driving the audio from a computer where I have REW and Equalizer APO so I'd guess I could manage it from there. That said, I have some homework to do on it. FIR is a new idea to me.
@GM - 👍 . thanks for the confirmation,.
My calculations were based on assumptions and approximation, so better do it yourself your plication in mind 🙂 sunday hobby thoughts
I'm assuming you want good big speakers with nice punchy bass and designing from there. Put bandwidth and SPL requirements based on your preference on paper and calculate what the system needs to be. Good sound needs to be for ear and hearing system so room taken into account. DSP allows any kind of house curve as long as the system DI is nice and there is enough capability in the system.
Enough capability is found with the calculations. Lots of amplifier power is one way but probably means more compression and distortion than bigger system that needs much less power for same SPL which in theory translates for better sound quality, here aso the Q seems to become irrelevant. Quite simple high level reasoning. Then figure out if its practical to build and take compromises as you refine it further so that you reach your goal for the system. Prepare quick prototypes and verify performance is there before building expensive boxes 🙂
Last edited:
Power compression is certainly one concern for midranges. But cone excursion is another. The motor of even Revelators isn’t completely linear. Likely someone has Klippeled them, but in general, you want minimal cone excursion for midrange drivers (I won’t start about nonlinearities induced by the surround).
Gut feeling tells me that you should cross those 15W’s above 300Hz if you want to reach high levels (well above 100dB SPL at the listening position). Simulate these things in WinISD, Basta! or other apps to get insight In what power levels and which cone excursions are needed for the desired SPL.
Gut feeling tells me that you should cross those 15W’s above 300Hz if you want to reach high levels (well above 100dB SPL at the listening position). Simulate these things in WinISD, Basta! or other apps to get insight In what power levels and which cone excursions are needed for the desired SPL.
This all changes when you combine your box with your crossover, so you can begin with a smaller box in this regard..