Javier's RS180/RS28A MTM & mic cal thread....

[Dave Bullet]

Hi Javier,

Thanks for taking the trouble to post the results and comparison. It gives me some comfort my 27TBFC/G based calibration is "good enough". In my case I decided to splice my measured calibration above 1KHz with a generic file under 1KHz since I couldn't guarantee room effects being the culprit at low frequencies. Looks like I chose correctly.

Thanks,
David.

[fjhuerta]

I'm tired of trying to get good sound out of this design.. help.

Warning: long, desperate rant ahead.

I've been literally fighting with this design for years. At this point, I know I should just throw the cabinet away and save the drivers for something else. I just can't do it because I feel "beaten" by this thing.

In case you are wondering, it's the good ol' Natalie-P. I've had lots of help from a lot of people at this point, but no matter what I do, I've never been able to be truly satisfied by them.

I now have the calibrated microphone. I've taken in-box measurements outside. I have made harmonic distortion sweeps. I've tried the original Nat-P crossover, a modified Nat-P crossover 4th order, 6th order, everything from 1.5 to 1.8 KHz crossover point... it's no use. I've tried flat on axis frequency response. I've tried a little extra attenuation on the tweeter. I've gone for flat off axis. It's always the same.

What I hear: an exaggerated midrange - high frequency response. If I raise volume too much, the midrange and highs become piercing. At lower levels, the speaker is quite fine.

Why I think it's wrong: because I've never had, or built, any other speaker that is as bothersome in the midrange as this one. I'm currently using a pair of Monitor Audio Silver RS 8 all-aluminum driver speakers, and I'm saddened to find out they never hurt my ears at all. I know the RS drivers should be so much better than the ones Montior Audio uses - I just can't understand why they sound like they do in my design.

At this point, I'm willing to just measure the Monitor Audios, and aim for such a frequency response in the Dayton design, and be done with it.

Why I don't like my idea: because somehow, I thought flat frequency response on axis, or the smoothest power response, were the way to design a speaker. If it isn't, then how on earth am I supposed to make a good speaker? The thing is - maybe, just maybe, the problem is that modern music is just badly mastered, and it sounds really bad on a "flat" speaker. Or maybe using a pair of 7" metal cone drivers with a tweeter on the middle is not the best idea. Or perhaps my equipment is trashed, and it has serious problems in its frequency response. Or maybe I simply dislike the distortion profile of the finished speaker. I don't know.

Maybe speakers really need to be "voiced" according to their harmonic distortion, etc. If that's so... then I'm willing to believe speaker building is more of an art, and not a science.

Perhaps I should just store away these speakers, and give them a try in a couple more years.

Sorry for this. Rant over.

[augerpro]

It's entirely possible the voicing or sound of the individual drivers just doesn't agree with your taste.

I'd be curious to see the measurements of the individual MA drivers if you ever get the chance.

[Hdale85]

Its hard to believe your so unhappy when so many others have been more then happy about their Nat P's. Maybe you just need a different design?

[Jed]

Try the modula MTM crossover instead as it's a better design overall.

[Hdale85]

But it needs some amazing electronics to really hear the difference I've heard.

[---k---]

Has anyone else that listened to your NatP and shared your opinion? Its too bad your so far away and it isn't possible to get a second opinion on them. I just keep thinking there has to be something wrong in that crossover.

[JonW]

... I thought flat frequency response on axis, or the smoothest power response, were the way to design a speaker. If it isn't, then how on earth am I supposed to make a good speaker?

Javier,

Sorry for the troubles you're having. I've had plenty of frustration with the hobby, at times. (Measurements, anyone?) Just one thought that comes to mind from a recent experience: Have you considered phase when started changing the speaker? For my Spassvogels, I ignored the phase issue when making my first crossover in test cabinets. I ignored it purely out of ignorance. [I've got lots of that to spare, in case you need any extra. ]The speaker did sound quite good to me. But later, after I got the final cabinets built, I remade the crossover with phase a prime consideration. Not just frequency response. Now the speakers have a real magic that was lacking before. The phase of the original NatP crossover should be plenty fine. But a lightbulb went off above my head when I learned about the phase issue. So I thought I'd pass it along.

It sounds like you need another set of experienced ears to hear your speakers. Another thing that will possibly help is for you to have a listen to many other speakers that are of good design. Other DIY speakers might be OK. But maybe it's best to go to some stereo shops and listen to lots of speakers. That will help tell you where you stand with your NatP's.

Good luck. :T

-Jon

[Dennis H]

Javier, from the pics I've seen of your room, it looks pretty 'bright' with wood floors, etc. Have you done any RT60 measurements, or the ghetto version, the old hand clap test? Room treatments might be your friend.

[dlr]

Any chance you'd be able to post measurements of each section with the crossover in place and the system response?

Dave

[---k---]

Javier, from the pics I've seen of your room, it looks pretty 'bright' with wood floors, etc. Have you done any RT60 measurements, or the ghetto version, the old hand clap test? Room treatments might be your friend.

You can download a trial version of ETF5 that has RT60 capabilities. If you can do measurements, using ETF to do the RT60 will be easy.

[JonW]

Javier, from the pics I've seen of your room, it looks pretty 'bright' with wood floors, etc. Have you done any RT60 measurements, or the ghetto version, the old hand clap test? Room treatments might be your friend.

Yeah. Or maybe a quick test by bringing them into another room that has more rugs, cushy furniture, pillows, etc. See if they sound any different.

[dawaro]

Just for shitz and grins if you would like are the actual measurement files that Jon used for the design. That way you can load the design up and see what he was seeing when he built it. You can also compare his driver measurements to yours to see if there is a major difference. The measurements are in the zip file attached. Per Jon the levels are low due to clio so you made need to scale them to 90db or so.

[fjhuerta]

Thank you all for your opinons (and bearing my long winded rant).

This weekend, I had a couple of hours to test a couple of things. First of all, I decided to see what kind of frequency response the Monitor Audios had vs. the Nat-P. I assumed the MA's would have a far warmer presentation (better bass response), recessed midrange and a touch too much highs.

I couldn't have been more wrong, had I tried.

d

The black trace is the MA; red trace is the RS180 / 28a MTM (it has a 4th order crossover at 1.5KHz). The measurements were taken at 110 cms, with a 6 ms gate. If I'm understanding the graph correctly, the Dayton speaker has *less* midrange energy, and more bass response.

After reading Dennis comments, I began wondering whether the room was the culprit. My initial measurements with SynRTA did indicate a bit of excess energy at 2 KHz, so I lowered the tweeter response slightly. I then found the spot that measured the flattest at my listening position, and took some 1/6th octave measurements. I did the same with the MAs. I really was expecting that, this time, there would be far more bass from the MAs, and a subdued midrange.

Again... wrong!

Monitor Audios:



Dayton MTMs:



Side by side comparison, MA's in black:



These measurements were taken at my listening position. As you can see, measurements clearly favor the Dayton MTMs. They are flatter, they have better bass response, they are smoother. At this point, I'm only left with a couple of thoughts.

1) My hearing is completely wrong (my wife listened to the speakers earlier today, and she thought they were perfect. Same thing with a couple of friends with only a casual interest in audio).

2) The upper woofer may be causing some additional reflections that I didn't capture with the RTA analysis.

3)The extra energy at 600-800 Hz from the MA's are fooling me.

I also did some very informal distortion sweeps using ARTA, anc couldn't find anything wrong with the speaker.

At this point, I think I'll just accept the fact that I don't like the sound of the overall design and move on. It's probably a combination of the room / driver layout, but unfortunately, there's not a single room where I have carpet, or lots of sound absorption, so I guess I'll never know.

At least I'm feeling better now.

[ThomasW]

The black trace is the MA; red trace is the RS180 / 28a MTM (it has a 4th order crossover designed by Jay Kim, and slightly modified by me)

So all this time you haven't been listening to Jon's NatP crossover, or did Woojae Kim decide to come to your 'rescue' and offer a 'fix' for Jon's design?

[augerpro]

Javier when you say you expected the MA's to be more prominent in the bass, subdued in the mid, and a little bit bright in the treble, I think your graphs show almost exactly that. The bass isn't obviously different, but a number of things can make bass subjectively loud. Otherwise the MA does indeed have a lower level in the midrange (150-600hz) and tipped up treble. So I'd say your impression matches the data pretty good. And that is not always typical since one on axis measurement does not give a complete picture of the total power response.

[Jed]

This is a confusing thread, Javier. First you had issue with the NatP sounding bright but you show FR measurements from a different designer?



Please clarify because we'd like to offer suggestions, just we need to see what's going on with the correct crossover in place.

[Dave Bullet]

One thing that Jay does as other designers do, is offer voicing options with the circuits they provide. Whether the Nat P crossover or some modified version is used, the measurements show a reasonably flat response for the Dayton driver combo. Of course, we aren't seeing distortion nor phase so there could be other problems.

To me, it shows (unless the measurements are wrong) that voicing is a process all listeners should put *any* design through.

As has been noted, room differences, placement and our ears and preferences mean a single design cannot satisfy all, hence the use of voicing.

Cheers,
David.

[Dennis H]

but unfortunately, there's not a single room where I have carpet, or lots of sound absorption, so I guess I'll never know.

The room is possibly your most important audio component. Speakers are about equal to the room and amps, etc., are way less important. Javier, you seem really committed to this hobby so it's too bad you don't have any interest in fixing what's possibly your biggest sound problem, the room.

Edit: you don't need to make the whole room absorptive. You just need to kill all the first reflections from both speakers off all the surfaces -- floor, ceiling, the 4 walls.

[Dave Bullet]

Dennis, I agree with what you are saying (it would be wise for me to anyway!), but what is interesting is that (I assume) the monitor audio and dayton speakers are being played in the same room. One has the perceived problem and the other doesn't. I therefore presume the problem can be fixed in the speaker rather than treating the room.

Javier - it would be good if you can answer if and why you can / cannot fix or change your room treatments to answer Dennis' suggestion.

Thanks,
David.

[Jed]

I suspect the problem Javier is having is the result of using a crossover design that doesn't fully attenuate the breakup peaks of the RS180s or take into consideration the power response in the critical range from 1-5k. I'm saying this because it doesn't appear he's using the Modula or NatP crossovers. You might have a flat FR, but that says little about how other things are handled like I mentioned above.

[fjhuerta]

So all this time you haven't been listening to Jon's NatP crossover, or did Woojae Kim decide to come to your 'rescue' and offer a 'fix' for Jon's design?

I used Jon's design for about 4 months, and since then I've built about 10 different crossovers. The current one sounds better than any other I've tried -less midrange glare and highs.

The current design does not fully attenuate the break-ups. But I was suggested to try and lower the crossover point, since 1.8 KHz supposedly is a bit too high for the RS180s. My experience was a positive one - midrange glare was lowered quite a bit with this change.

I'll post the crossover diagram and FR plots later today.

I'd like to emphasize the following - I'm sure my problems have nothing to do with Jon's design (which never worked for my specific build - I'll also post the simulation results with my in-box data), but with my room - electronics - particular drivers - baffle layout, etc. 99.9% of satisfied Nat-P builders can't be wrong!

[ThomasW]

This thread has been closed per request of the OP