MTM RS180 Seas H1212 tester finished\Too revealing? (threads merged)

[4thSeason]

Well I have endevored to make a MTM similar to the Modula and Nat but using simpler crossovers for my first project. I will attach some initial pictures of the cabinets the completed and the networks. This baffle was a guinia pig for the Jasper and it seems I cut my tweet cutout a 16th too big and too deep but thats why we practice. I will add all my pics to date along with the crossover I used and and the speakerworks plots. Hopefully the Breakup mode with these crossovers will be well enough taken care of.

My goal in this project was actually to get my feet wet and challage my friend who said I couldn't build a speaker that sounded better than his Boston VR-3's I am hoping personally to have them sound in line with my personal favorites in that price range Vienna Acoustic Baby Grands. I guess we will see.

Next I know that these look rough like I said testers. The box will remain the baffle is gone that is why it looks like I steel wooled it. Oh wait I did when I was testing finishing

[ThomasW]

Looks okay, perhaps a little too much foam....

What crossover did you use? It's certainly minimalist.

[4thSeason]

Can't really take credit . Jay from the board here helped me with a network
Heres what I have. Just fired the tester up...Ugh my Yamaha Rv-795A does not like this at all. protection mode at about 25% volume. Guess I need to run back out to the Lab and get my M70. Will post info tonight. As for the foam Thomas, I do have it in front and behind the baffle but it is not glued in the front just wedged open with a piece in the top so it may come out. I also treated the inside of the box with rubber coating.

[4thSeason]

Just hooked up my M70 and C50 after I returned to the Lab last night to get some equipment capable of running this. I must say that intially I am impressed. After both speakers are finished and I have then properly set up I will post my and my partner in Crime Bills opinion in case mine seems like I have ownership admiration as to how this crossover is working. I know the drivers are a proven design and I am hoping that somehow I can listen next to a Modula to A-B those amazing CE filters. I know it is way to early to make any realistic determinations but for 125$ apiece for these I think that so far so good. Notably I used the 2.4 ohm resistor intead of the 2.0 ohm for a bit more laid back sound. As I write this I am listening to Brothers in Arms right now and keep thinking wow these things get some low end- wow! I may try the 2.0 Ohm resistor and am wishing that I had an easier way to attenuate up and down. Was considering using a Lpad in there to test. I will post more as I can

[4thSeason]

Made a mistake in posting my Speaker work curve there seems to be a slight dip in the response near the crossover frequency. This is due to an incorrect driver offset when I created the parallel driver. I will post the correct curve when I go in on Tuesday.

[augerpro]

Made a mistake in posting my Speaker work curve there seems to be a slight dip in the response near the crossover frequency. This is due to an incorrect driver offset when I created the parallel driver. I will post the correct curve when I go in on Tuesday.

How about post the actual response too?

[4thSeason]

As soon as I get my M-audio Pre from Newegg I will Not sre about the soundcard I have and I am sending back my Behrinnger 802. Will post as I go ,going to order veneer today also not sure what to do whether vinyl for my first or Melamine or real veneer.

[4thSeason]

Just got both speakers up and operational yesterday and got them connected this morning. I really haven't gotten a chance to listen much or play with placement but in the brief time I have I really enjoy what I have heard. I will do some measurments early next week I surmise.

Initial impressions are thee things play low and loud. Very very clean sounding. Not really able to comment on imaging and stage.

[4thSeason]

first project Too revealing?

Recently finished my first project MTM H1212/RS180 and although I am overall very pleased with how the project came out there have and are some challanges.

Some material sounds very very good with these. I.E. Brothers in Arms, Tracy Chapman. Steely Dan Cant Buy a thrill. But as I continued to demo these there started to be some glaring problems. First CD that sounded BADDDD was Tool Undertow which I have always enjoyed to demo in my Mobile audio systems. It sounded awful.

This past weekend I drug them up to a friends house at the request of my best friend and we A-B'd them with another friends Boston VR3's.

Listened to Brothers in Arms- my best friend Bill said they sounded very good.
Very open very natural and commented that sounded like they were at least 10 feet furter away than the Bostons althought they were sided by side. We then listened to the rest of Bills Demo material. Some sounded very good, some enhh some very bad. ZZ Top sounded okay but the bostons sounded much better other then they completly come undone at high volume where my MTM's don't
Crash Test Dummies God Shuffeled his Feet- Sounded much better on mine other than the missing stength in the kick drum area.
Suzan Vega and Michel Jackson Thriller sounded very bad like they were mixed horribly

Okay now my question. Why? More technically what makes them more revealing/less forgiving. How can I adapt? I don't want to throw away all of my CD's to buy just high quality stuff I don't like. And I am not ready to build a whole new set of speakers yet

Any direction as to which part/parts in the design define this will help me. It's been awhile since I have been on because of personal demands, but I plan on doing a whole thread on the build so everyone can see the design development and tests when I complete them.

[Dennis H]

Have you tried padding down the tweeter a bit more?

[4thSeason]

Have you tried padding down the tweeter a bit more?

No I am however preping to takethe baffles off and use a 2 ohm resistor unpadding the tweeters a bit. I originally ordered 2.0 2.4 2.7ohm variants and started in the middle. Both Bill and my other friend felt that the tweets were a little restrained perhaps though not really lacking.
I guess to dive a bit deeper as my friend Bill and I were discussing why are speakers (for the masses) like VR3's so program insensitive. I don't want to lose what I have in the way of image, stage, and overall clarity but I also would like to be able to listen to Disturbed without being disturbed by the output.

Oh also if amplification were a matter in our equation I use a Yamaha M70 and Greg demoed using a Sunfire series II. For pres we both had a Yamaha RX-V795 but I have used my Yammy C-50 pre to test also

[Jed]

Why? More technically what makes them more revealing/less forgiving. How can I adapt? yet

Take some off axis measurements and see if there is a peak in the response at 2k. If so, you'll have to flatten the power response to make them smoother from 1-5k. You can also try a slightly steeper slope on the midrange to get rid of any harshness.

[4thSeason]

Take some off axis measurements and see if there is a peak in the response at 2k. If so, you'll have to flatten the power response to make them smoother from 1-5k. You can also try a slightly steeper slope on the midrange to get rid of any harshness.

I will take some measurements and post in a few days.

Harshness? I never said anything about harshness. If theres any I haven't heard it and unless we were talking of my friends VR'3's I said recordings sounded like they were mixed badly.

[---k---]

maybe it is the metal cone break-up that your simple cross over isn't taming very well. The VR3s have poly woofers that probably arn't as detailed, but don't suffer from breakup as bad.

[Jed]

I will take some measurements and post in a few days.

Harshness? I never said anything about harshness.

I interpretted "less forgiving" as harshness, but perhaps you find it neutral? Not sure what the problem is then. It'll be interesting to see your measurements from a variety of mic locations. Then we can shed some light on what you are experiencing. Otherwise, it's all semantics.

[4thSeason]

maybe it is the metal cone break-up that your simple cross over isn't taming very well. The VR3s have poly woofers that probably arn't as detailed, but don't suffer from breakup as bad.

Ryan

I had seriously thought about that but I read and article(where I can't remember but I'll try to ref) that said that after 30 db down is inaudible. So I can't say that I don't believe it's that, but I have read other people say if I remember correctly that the Modulas were discerning, maybe i'm wrong. Another point is I don't feel I am hearing distortion- I think it sounds like I said earlier all over the place, like bad mixing, Like on Tools' 10,000 days it sounded to me like I had three or four speakers in different locations on each side playing certain areas of the bandwidth I know this sounds strange but perhaps I am not articulating what I hear properly

And also I would think that if there were a breakup problem, the music I played would be immaterial those frequencies where the breakup is present would still be produced. I know this sees weird but a song for instance that I know isn't audiophile type material is Yellos Jungle Bill It sounds awesome and has a very broad spectrum of sound lots of horns drums and vocals and bass guitar. That should sound bad too but it doesn't. I feel stupid trying to explain this Tracy Chapman acapella awesome. Suzan Vega Acapella nauseating.

I know this sounds a bit weird but could I have and area that is too flat. I know from talking to my Best friend Bill, we've both been Car audio enthusiasts for 20 years and he has even placed in many Iasca events that flat sounds flat and unappealing , sorry for the rant

Bri

[Brian Bunge]

Tracy Chapman acapella awesome. Suzan Vega Acapella nauseating.

I know this sounds a bit weird but could I have and area that is too flat. I know from talking to my Best friend Bill, we've both been Car audio enthusiasts for 20 years and he has even placed in many Iasca events that flat sounds flat and unappealing , sorry for the rant

Bri

That sounds like it could potentially be cone break-up to me. Think about it. Tracy Chapman and Suzan Vega's vocal ranges are completely different. Chapman has a bit of a deep voice, more like a man. Vega's voice or more like a typical woman's voice and would be in the range where cone break-up is potentially occuring. Saying that Vega's voice is nauseating sure sounds like cone break-up to me.

Does it have a sort of a shouty sound, kind of like someone's voice blaring at you through a bull horn?

[PoorboyMike]

That sounds like it could potentially be cone break-up to me.

That would be my guess too. Most rock music has lots of distortion in it anyway, so if you haven't tamed that cone break-up good enough, I'm pretty sure music like Tool will amplify it.

I have a pair of Nat P's and I've noticed on music like Alice and Chains, I don't like to play it as loud as I used to on my Inifinitys. It still sounds good at normal to moderate volumes, I just don't like the sound of it cranked up. Part of it is probably my crappy room acoustics, but I think the revealing nature of those aluminum drivers has something to do with it also.

[jimangie1973]

Take some off axis measurements and see if there is a peak in the response at 2k. If so, you'll have to flatten the power response to make them smoother from 1-5k. You can also try a slightly steeper slope on the midrange to get rid of any harshness.

I agree that this may be the issue. The tweeter maintains it's output level off-axis at 2 kHz, but loses output as the frequency rises. If the overall sound is a little forward with a perceived lack of energy in the very top octives, you may need to alter the tweeter response to account for this.

In addition to measuring, a great way to figure these things out is to get a 30 band eq and play with it until you get the desired sound. Then aim for that response alteration in your crossover simulation. I like to measure, and if I see a potential problem, eq it out and listen to the result.

Based on the simulation, I wouldn't think the woofer breakup would be audible. It looks to be attenuated very well.

Good luck!!

[Brian Bunge]

I agree that this may be the issue. The tweeter maintains it's output level off-axis at 2 kHz, but loses output as the frequency rises. If the overall sound is a little forward with a perceived lack of energy in the very top octives, you may need to alter the tweeter response to account for this.

In addition to measuring, a great way to figure these things out is to get a 30 band eq and play with it until you get the desired sound. Then aim for that response alteration in your crossover simulation. I like to measure, and if I see a potential problem, eq it out and listen to the result.

Based on the simulation, I wouldn't think the woofer breakup would be audible. It looks to be attenuated very well.

Good luck!!

Hey there! Another Brevard County resident and just up the road from me! It sounds like you are experienced with taking measurements. Maybe we can get together sometime. I always love talking audio with local enthusiasts!

Oh, and welcome to the forum!

[4thSeason]

I am getting alot of good stuff here. Jed, since i am new to the whole home DIY stuff testing has been somthing I held off on or slow to get to. Yesderday I ordered a USB Pre. I have real RTA and a Behringer 8000. Do you guys think I should do an Anechoic room at out lab to test or just test here in the "real world"?

[---k---]

I don't have that the cds you mentioned, so I can't say. Do they sound like crap with a really good pair of headphones? lots of music is just mixed like crap, and you can't tell on more forgiving/less quality speakers. I'm not sure car audio is a good point of reference.

[4thSeason]

I don't have that the cds you mentioned, so I can't say. Do they sound like crap with a really good pair of headphones? lots of music is just mixed like crap, and you can't tell on more forgiving/less quality speakers. I'm not sure car audio is a good point of reference.

Well I can't say on headphones but My Friend Bill and I went on a road trip and demoed Vandersteens I believe 5B and 2C's same type of thing happened but they even made Dire Sraits sound bad. best CD we had was Tracy Chapman. All others sounded horrible. And as far as car audio. I would concur that most consumer level stereos sound just ok but don't tell Richard Clark that Car Audio isn't a good ref point

Bri

[cjd]

You've got negative summing at the low end of the tweeter, which is going to make power response pretty poor. Integration in general is a little loose.

I'd also like to see impedance response - your amp might be a little unhappy causing a little more harshness to poke its head through.

[4thSeason]

Thanks all again for the good insight. For the first time now since I have built these I have connected my sub. I haven't wanted it to be part of any equation in listening to music. Just threw in 10,000 days fired up my velodyne DPS-12 set the crossover @ 40HZ and gave it a little volume and I gotta say it did help the music come together a bit maybe the low end perceptually skews my hearing away from the trouble area or maybe it helped mask it. HMMM now I know why I didnt want to hook up my sub until I figured out where I was having my challanges.

[---k---]

That those CDs sound like crap on other high-end speakers should be a big hint.

I would listen to CJD.

[Dennis H]

Jed, since i am new to the whole home DIY stuff testing has been somthing I held off on or slow to get to.

Ooooh. I assumed the curves you posted were based on measurements. If the whole thing is simulated, your actual response may be very different.

Edit, okay the thread merge clears things up a bit. FWIW, people who have actually measured the drivers in boxes get results differerent from Jay's sims and don't get as much breakup suppression with a simple XO. Baffle effects are a big question as well. You really need to get some measurements and may need to revise the XO. The good news is the Modula MTM and Natalie P are proven designs using the same drivers.

[4thSeason]

Ryan
I agree but what I guess my original question was opted toward was why are speakers so revealing mine included not whats wrong with my speakers.
We could speculate ad nauseam about this subject I'm wondering how I can reduce the effects.

[4thSeason]

headed out for the weekend for the UFC fight. I am sure when this weekend winds down My friend Bill will jump in and let you know what he heard and compare them to the VR-3's and perhaps his Klipsch KV-4's

Many thanks again you guys bring so much knowledge here its uneal

[cjd]

You've got a big emphasis 1-2kHz with anamolous response due to negative summing just above that (this is always a bad thing by the way) and breakup that's already tough to deal with sitting a little high.

I wouldn't call your descriptions "revealing" so much as "too forward" - I'm not surprised.

Did you run a Hilbert transform? What process, in other words. If you haven't even done that much, you could be *way* off.

C

[jimangie1973]

You've got negative summing at the low end of the tweeter, which is going to make power response pretty poor. Integration in general is a little loose.

I like to try to get the phasing as close to in-phase as possible at the crossover frequency to minimize this negative summing. With the crossover design shown, there's only about 3 dB acoustical summing at the crossover frequency, which means the woofer and tweeter are almost 90 deg out of phase. This could be much worse (or possibly better) if the woofer-tweeter acoustical offset assumption was not exactly correct.

You could reverse the tweeter polarity and listen. Who knows, you may get a better overall response.

[cjd]

Eh - 90 degrees out of phase still should not cause negative summing. That means something closer to 180 degrees out of phase.

[probillygun]

Lets try to approach this from a different angle then, lets say for arguments sake Brian's speakers were measured, tested, and belived to be "Ok".

One of the things we really want to know from a philosophical standpoint is;

What specificaly about a loudspeaker design in general could make his speakers sound more revealing and less pleasing to the ear with lesser quality recodings and another speaker for the masses, like a Boston VR-3 or B&W 604 sound "Better" with a much wider selection of mediocore grade recordings?

IE; Could it be differences in the material the drivers are made of? Grade of crossover componets or slopes? Lower distrotion drivers? ect....Please be as general or as specific as you like with your opinions, I will enjoy hearing your opinions either way on it.

Thanks.

[Dennis H]

lets say for arguments sake Brian's speakers were measured, tested, and belived to be "Ok".

But they weren't so it's a moot point. No measurements were ever made by the builder or the designer. My guess, he has some peaking in the upper mids or lower treble -- what Chris called 'too forward, not too revealing.' Talking about cone materials, etc. doesn't make much sense if we don't even know if the speaker has a flat frequency response. We also don't know how much the mid's breakup is suppressed. So let's stick to the big stuff and look at the small stuff once we have all that covered.

The 'Speaker Camp' thread is a perfect example. JonMarsh is one of the most talented designers around. He designed a crossover based on his own measurements, not something he downloaded off the internet. Yet he and Thomas weren't entirely pleased when they listened and Jon ended up cutting the mid and tweet by 3dB to get it to sound right. To think you can download some measurements off the net and design a finished speaker without ever building it or listening to it is either ignorance or arrogance, take your pick.

[---k---]

I think if you read though Loudspeaker Cookbook or Loudspeaker Design 201, you would have a better chance of understanding everything that has been said. I get the impression that Dennis was a little irritated in that last post, and looking at the answers that have been given, I think it is justified.

I find it interesting, and you're not the only one who does this, but you're looking for a pair of speakers that makes "mediocre grade recordings" sound "better". To do this sacrifices the sound quality of good recordings. It is tradeoffs.

If you want your bad recordings to sound good, get some cheep dashboard speakers.

[Mudjock]

I think I can offer some insight here.... A couple years ago, I was working on an RS150-H1212 MT using a very similar approach to yours. The high pass was second order electrical and the low pass was an inductor with a "bottomless" notch filter much like yours. I feel like I went through a similar struggle to what you are currently encountering. If I padded the H1212 enough to not cause pain on poorly recorded music, it sucked the life out of the extreme top end. After a dozen or so iterations, I finally wound up padding the tweeter down an extra dB from what modelled flat - but more importantly changed the shape of the H1212's response with a parallel inductor. This mitigated any peakiness in the 2-5 kHz region while retaining the "air" at higher frequencies. I demo'ed these at a DIY meet in Chicago, where they finished in a virtual dead heat with cjd's MTM (a much larger and significantly more expensive speaker) for sound quality.

I have thought about what made this particular design so tricky and came up with a couple of thoughts.

1. Offsetting the H1212 close to the baffle edge (my design has that, too) raises an on-axis diffraction peak around 5 kHz which unfortunately corresponds with a slight peaking in the H1212's natural response.

2. The notch filter may not attenuate the cone breakup enough.

Since then, I have tried another RS150 MT (with the B&G Neo3PDR). In that design, I went with a third order electrical filter on the RS150 and found the design to be considerably less finicky. I am not sure how much of that has to do with the tweeter vs. the low pass filter.

So, based on my experiences, I would try:

1. Modify the padding on the H1212 to include either an inductor in parallel with the tweeter or a cap in series to flatten out the response in the 2-5 kHz region.

2. Consider a 3rd order filter on your RS180's. I'm not sure how easy that will be to implement - but maybe worth some simulation work.

[probillygun]

When I listened to Brian's speakers, I thought they sounded good with Dire Straights, "Brothers in Arms" and Crash Test Dummies, "When God shuffled his feet". These are 2 of my reference recordings that I use to judge if a speaker will play well the type of music I normaly listen to, and these recordings are higher Quality than most other CDs I listen to for that type of music. It didnt sound like anything was missing other than a more solid kick drum that I prefer, and some lower resonance I heard which I believe to be coming from cabinet construction.

However, when I played some of my other reference recordings I use, like Yello, ZZtop, Susan Vega...the sound was definately not as pleasing to my ears anymore. It sounded flat and dull, almost like the speakers were bringing out something in the recording that wasnt there when I listen to these same CDs on Boston VR3s.

The tweeter I think sounds good on all my recordings, and I belive the sound I dont care for is coming from the midrange area.

I was hoping to get some insight as to what might be going on.

Thoughts?

[Jed]

When I listened to Brian's speakers, I thought they sounded good with Dire Straights, "Brothers in Arms" and Crash Test Dummies, "When God shuffled his feet". These are 2 of my reference recordings that I use to judge if a speaker will play well the type of music I normaly listen to, and these recordings are higher Quality than most other CDs I listen to for that type of music. It didnt sound like anything was missing other than a more solid kick drum that I prefer, and some lower resonance I heard which I believe to be coming from cabinet construction.

However, when I played some of my other reference recordings I use, like Yello, ZZtop, Susan Vega...the sound was definately not as pleasing to my ears anymore. It sounded flat and dull, almost like the speakers were bringing out something in the recording that wasnt there when I listen to these same CDs on Boston VR3s.

The tweeter I think sounds good on all my recordings, and I belive the sound I dont care for is coming from the midrange area.

I was hoping to get some insight as to what might be going on.

Thoughts?

Simple, peaking in the female vocal range and/or 1-5k region. Take some measurments on and off axis and you'll see a rise in those areas, possibly.

[Mudjock]

That 1kHz to 5kHz area is what I'm talking about, too. The simulation from the previous page shows peaks around 3 kHz and between 5 and 6 kHz. My design had similar issues at around 2 kHz and 6 kHz (slightly different baffle width and layout). There was no way to fix that response with a single series padding resistor. If you pad that enough to bring the peaks in line with your target response, you will suck the life out of it - and if you let the peaks ride high, you will be running for the volume knob to turn them down. Granted, these are low distortion drivers and on well recorded material that will show through - but if you cycle through a wide enough variety of material, the response issues will manifest themselves.

On a more general note, as one moves to higher resolution, lower distortion drivers - the difference between good and bad recordings becomes much more evident. I tend to put in a good recording and be blown away by the details I can hear and then put in an "old favorite" and wonder what happened. The truth is that there are no details to be resolved or the those details are not exactly pleasant to listen to...

[probillygun]

That 1kHz to 5kHz area is what I'm talking about, too. The simulation from the previous page shows peaks around 3 kHz and between 5 and 6 kHz. My design had similar issues at around 2 kHz and 6 kHz (slightly different baffle width and layout). There was no way to fix that response with a single series padding resistor. If you pad that enough to bring the peaks in line with your target response, you will suck the life out of it - and if you let the peaks ride high, you will be running for the volume knob to turn them down. Granted, these are low distortion drivers and on well recorded material that will show through - but if you cycle through a wide enough variety of material, the response issues will manifest themselves.

On a more general note, as one moves to higher resolution, lower distortion drivers - the difference between good and bad recordings becomes much more evident. I tend to put in a good recording and be blown away by the details I can hear and then put in an "old favorite" and wonder what happened. The truth is that there are no details to be resolved or the those details are not exactly pleasant to listen to...

Ok, Mudjock, thanks for the input! Now I can see this being a real possibilty. :T

BTW thanks for contributing something constructive to this discussion and not critisizing and being arrogent about knowing more than I do.

[Dennis H]

Sorry if I came off as cranky last night. Clearly the problem is with the crossover. Other successful designs have used the same drivers. We can't really say exactly what the problem is without some measurements of the tweeter + filter, the woofers + filter, and the whole system. The nice thing about an MTM is the design axis is the tweeter and all the measurements can be done from one point.

[cjd]

well, I've described a couple times where I suspect the biggest problem lies: the negative summing evident at the low end of the tweeter which will result in some nasty peaking in that region as you move off-axis.

[Dennis H]

Chris, are you basing that on your own measurements (which I believe -- top half of your WWMTM) or the (questionable) ones posted for this speaker? I don't have a lot of faith that the posted curves get the offsets and the baffle effects right as they are simulating all that rather than measuring it.

[cjd]

Good point. The fact that the posted network causes negative summing on even these sketchy measurements suggests there is a very high liklihood of problems, but... who knows what they are. If I remember I'll run it through some data I have handy.

C

[fjhuerta]

Basically, I agree with most of that everyone has been saying in this thread.

I found the same issue with my MTM's (RS180 - RS28). The issue was - I had a very nasty peak at 2 KHz, and the simulation simply wasn't modeling it. Only real life measurements showed it - and it was quite ugly, a 3 dB peak with a very narrow bandwith. Ouch.

Only after taking real life measurements in baffle, modeling, and then re-measuring it, could I get rid of this baffle effect.

And even then, I found that I found the aluminum drivers to be extremely revealing on some badly recorded music (ie - too "forward"). Classical music sounds absolutely incredible, some pop - rock is too strident.

Anyway, you may even want to use a bit of a BBC dip. I tried it with my speakers, and it helped - a lot.

Now that I've built a couple of poly / paper cones, I find myself preferring their "warm" sound (OK, I know it's not "warm". Let's just say it has a different distortion profile, or a BBC dip is way easier to implement, or you don't have to worry about breakup nodes) to aluminum. BUT - and I'm pretty sure of this by now - I find that a well designed metal cone driver loudspeaker sounds a lot more accurate than a "pleasing" plastic-poly one. Pick your poison...