Live next to madisound, compared Idunns and ZA5.3s, questions about power response

[bunshi]

I'm new here (see the bottom for unimportant information about me).

I'm debating about building Mark K's ER18DXT or buying the ZA5.3s since the cabinets are currently 77$ off. As a relative new guy to DIY speakers, the ER18DXTs seemed really attractive because of the waveguide tweeter, and Zaph as well as other respected designers have commented on the sound of waveguide speakers having a unique quality, like "wearing a good pair of headphones except with forward imaging".

Basically I wanted to compare a "waveguide" design to the ZA5.3s, because without listening to them I don't know how much I would care about power response. Since I can't listen to the ER18DXT, and I know Madisound has the Idunn speakers, I went there and listened to both speakers.

Based on everything I've heard about waveguide speakers I was expecting myself to be wowed by the Idunns, but I didn't really observe any real difference in the sound stage of the Idunns vs the ZA5.3s. In fact, I really preferred the sound of the ZA5.3s. I'm looking for an explanation as to what's going on?

1. I know the Idunns are crossed over relativity high, and the power response around the crossover (and the topology) could make the power response not representative of a well implemented waveguide design.
2. How does off axis response really affect on axis listening? Does an "omnidirectional" speaker always have a better soundstage than a "directional" speaker? I think the ZA5.3 has very controlled directivity in a certain sense, in that the primary lobe won't interact with the room very much as compared with a traditional design. Basically, I'm just really confused about speaker lobing and off axis response, because controling both at the same time seems to be contradictory. If you control lobing for less room interaction, you get less off axis response and vise versa. Please help me understand.
3. I understand that my preferences for each speaker could be based on other factors, but I can differentiate midrange clarity and bass output from sound stage. Actually, one of the main reasons I compared the Idunns and the SR-71s to the ZA5.3s is that I assumed I cared a lot about bass (I went there to prove to myself that I should buy the more expensive and labor intesive design). The bass extention of the ZA5.3s was very satisfying. You can hear that it doesn't reach a few more hz into the 40's, but it doesn't seem very obvious when listening to music. The midrange is clearer.

I wish I could have brought my measurement equipment with me so I could measure what was going on from my listening distance. I guess the main thing I want to ask the experts here is: "how important is power response for on axis listening?" I'm buying a new pair of speakers for primarily on axis listening. Thanks for everyone that participates in this awesome website. I've been using this forum for several years, but I haven't had to post because I've been able to learn by reading and searching. I've learned so much just by reading threads and looking at sites you make.

---
I'm a grad student in genetic engineering working on hard infectious diseases. I've built a few subs, and 5 Zaph B3N speakers from scratch, which is what I currently listen to. I do know how to measure non/linear distortion. I brought shrewd member of my lab (a friend) with me to madisound (who knows nothing about speakers) and he also observed that the ZA5.3s have better midrange. He's from Thailand, and I'm not good enough at describing things to explain what I personally define as "soundstage", but he "liked" the overall sound of the ZA5.3s more than the other designs too.

The ZA5.3 is cheaper design with less bass output and less directivity, so why does it sound better?

[JonMarsh]

1. I know the Idunns are crossed over relativity high, and the power response around the crossover (and the topology) could make the power response not representative of a well implemented waveguide design.
2. How does off axis response really affect on axis listening? Does an "omnidirectional" speaker always have a better soundstage than a "directional" speaker? I think the ZA5.3 has very controlled directivity in a certain sense, in that the primary lobe won't interact with the room very much as compared with a traditional design. Basically, I'm just really confused about speaker lobing and off axis response, because controling both at the same time seems to be contradictory. If you control lobing for less room interaction, you get less off axis response and vise versa. Please help me understand.
3. I understand that my preferences for each speaker could be based on other factors, but I can differentiate midrange clarity and bass output from sound stage. Actually, one of the main reasons I compared the Idunns and the SR-71s to the ZA5.3s is that I assumed I cared a lot about bass (I went there to prove to myself that I should buy the more expensive and labor intesive design). The bass extention of the ZA5.3s was very satisfying. You can hear that it doesn't reach a few more hz into the 40's, but it doesn't seem very obvious when listening to music. The midrange is clearer.

Not to be deliberately silly, but the Idunns are crossed over at 2 kHz, which isn't all that high as two ways go. Note the provided axis response curves on Madisound's site. Probably due to midwoofer diameter, they show some limitations in off axis response BELOW the crossover- in the sub 2 kHz area. This I'd find annoying.





With regards to on and off axis response, it's dangerous to generalize and you also must keep in mind room placement and acoustics. A properly implemented waveguide system may have more uniform power response off axis, especially above 10 kHz. But if room placement is a problem, that won't buy you much.

Zaph's 5" driver is quite nice, and due to its size will have better off axis dispersion in the upper midrange and presence area- this is part of what you may be hearing that you prefer. This doesn't tie so much into imaging, as it does good timbral balance.

Imaging I find depends on having good early arrival sound an minimizing late reflections and comb filtering in the response due to room acoustics. Room and speaker setup is important there. Of course, so is source material and components. To reproduce imaging, you're trying to setup a system that's transparent enough to reproduce the spatial cues and reflections of the original space the recording was made in.

In general, you're lucky here that you live close to Madisound and can listen to systems you're interested in. Go with your gut for now- but keep in mind room placement, how loud you may listen to music, etc.

[bunshi]

Jon,

Thanks a lot for the thoughtful reply. I didn't take into account the woofer size and it's affect on off axis response. Actually, I would still like to hear a really well implemented waveguide design like your Modula MT MkII. I've read and learned a lot from your threads so thanks.

I do feel really lucky to have madisound next to me. They have a very nice amp and preamp, and the music I listened to is all music I've been listening to for more than 10 years (just uncompressed CDs) and I am very familiar with it and know where to look for distortion in these songs.

In terms of the room at Madisound, it's a room designed for listening, lots of treatment and nice placement and nice stand height/seats. Closed doors. They have almost all Zaph's designs and I listened to the awesome ZRT for a while too (and I still want the ZA5.3s, they have such nice midrange for the price, and midrange is something I though I never would have cared about so much before I heard them). I've been meaning to listen to the SB12.3s. The only thing out of the ordinary that needs to be taken into account is that there are about 20~30 completed speaker designs in that room and all the woofers tend to suck the bass out (but it's all equal for the speakers).

Now it's starting to sound like an ad for madisound, but I like going there because the people let me look at the woofers before I buy, and also they let me use their inductance meter to unwind some coils once

I guess back to the original point I wanted to ask about, since I really haven't heard a constant directivity design yet, do you think off axis response makes a big difference on axis?

[JonMarsh]

Short response, yes, because you're almost always listening to the reverberent sound field in the room as well as the direct; speakers that have significant issues off axis just don't sound as "right", regardless of how ruler straight on axis is, unless you listen to them in an anechoic chamber or outdoors. It's always a design concern of mine.

How much that's important depends on setup and the room, and your listening style. But don't forget, unless you never listen to your system other than sitting in front in the sweet spot, that off axis stuff is important for "other" listening positions.

A quick easy way to judge this is to listen to a speaker in the next room- what you hear is just the averaged power response. Pretty revealing on that score, and indicative of what you may want to be concerned with if you listen to music while cooking in the kitchen!


Last, if I lived close to Madisound, I'd probably be in big trouble- like living next door to a deluxe high end candy store, or whatever... it's bad enough trying to stay away from their web site!

[Juhazi]

I can only warmly recommend the ERDXT to you. The 2nd order slopes make the power response better (more even) than the Idunn. X-o is quite simple and making the box tuning as recommended is just perfect for most rooms (not too heavy on bass). I've enjoyed them for 3 years now.

Jon's designs are great too, I'm waiting for the Arvo Pärt II to finish! I hope it will me my next speaker. The midrange is most important and THE difficult challenge for a 2-way speaker. I made Zaph's L18 first but it had the midrange-power response issue, I used the box for MarkK.

It would be heaven to live next to a shop like Madisound.

[5th element]

So far I've only seen the off axis response properly discussed here with regards to the typical wide-narrow-wide pattern associated with two ways, higher then optimal xover points and mid/bass beaming. One of the great benefits of a wave-guide design is that it will allow you to more easily optimise the power response around the xover point so that you don't get the typical dip around the xover point. But this, imo, isn't really the main reason for going with a wave-guide.

The main reason is for the region of constant directivity that a wave-guide provides. What this basically does, as Jon mentioned, is help maintain the off axis response higher up in frequency, but the important part here is that it only does this over a fixed range of off axis angles.

Imagine if you will that you've got a cone of sound being radiated out by the wave-guide and that this cone exists out to say 25-30 degrees off axis. If you observe the frequency response within this cone it will be flat, that is regardless of what angle you choose. This helps to maintain the tonal balance of the loudspeaker as you go off axis, which is nice too, but the real reason why you want to use a wave-guide is what happens as you go further off axis.

What goes on here is that the wave-guide essentially captures the far off axis sound and redirects into this main cone. What this does is reduce the amount of sound being thrown out into the room and reduces the amount of reflected sound in relation to the direct sound. As Jon mentions this is one thing that's important with regards to imaging and the ability of a loudspeaker to throw out a convincing spread of sound. This is where the description of headphones but without the image within your head comes from - headphones obviously have no reflected sound as they remove the room and this is exactly what a wave-guide attempts to do by controlling the directivity of the loudspeaker so that less sound is thrown out into the far off axis.

Now what's important here is that the extent to which a design can manage this is based solely on one thing - the size of the wave-guide. The bigger it is, the lower down in frequency it can control the directivity. The DXT is a nice tweeter and obviously offers some wave-guide benefits, but as it is pretty small its effect is rather limited and other design factors could easily outweigh the small benefits that it can provide.

As Jon mentions, even with a 2kHz xover point and 6.5" driver, you're still going to get some uneveness through and below the xover region when using the DXT. This is because the DXT is too small. As you will have probably read at some point, one of the benefits of a wave-guide is allowing you to match the directivity of the mid/bass to that of the wave-guide. This in turn allows you to use a higher xover point as you deliberately cross over when the mid/bass starts beaming as you allow the wave-guide to pick up where the mid/bass starts to droop. This only occurs however if the wave-guide and the mid/bass are matched well and this usually means using a wave-guide of a similar diameter to the mid/bass. As the DXT is far far smaller then the mid/bass in the Idunn, matching the directivity of the two isn't possible. In this regard the Idunn is exactly the same as any other 6.5" two way with 1" dome tweeter, in that you really need to cross over as low as possible to minimise the effects of mid/bass beaming.

The DXT is a nice effort, but to properly experience what a good wave-guide design can offer, you really need to listen to something with a wave-guide of at least 6" in diameter crossed over in an optimal way to a similar sized mid/bass. Of course designs like this are few and far between as mating a standard dome tweeter to a wave-guide is often a small challenge in and of itself and it adds a reasonable amount of complication to a design.

As a direct comparison loudspeaker, the only thing I can really think of at this moment is Jon's own design, which uses the same mid/bass as the Idunn in a very similar cabinet, but with a 6.5" wave-guide. I thought the design plans for this were in the mission accomplished section, but I can't see it there!

The bottom line though is that the Idunn is only really just introducing you to what a wave-guide design can do and as a result it's easy to see how you could prefer something else. After all what the ZA5.3 does well, it does extremely well, this is a very well put together design. There's also the chance that you might not like the way wave-guides sound, but if you really want to get a good feel for what a wave-guide can do, then you really need to listen to a bigger design.

[JonMarsh]

Personally, not to ding Seas or Madisound, but I think MarkK's design with the DXT is better done overall, and can be recommended for that driver.

The Modula MkII HE is available in a complete PDF, just email me or PM with your email to send it to anyone interested. I recommend using the Jantzen waveguide since the MCM is no longer available- better quality construction, but a little bit of a pain to get, as you pretty much have to email them and get it direct.

[dar47]

Great discussion on this comparison and it's scary how good others describe a good wave guide speaker. I can see how you would like mids on a MTM with 5" drivers crossed to a 1" dome and wonder how you like 7" MTM's. Of the 5" MTM's I have done I like the on axis sound and the extra db's they provide but I think if you compare a good 3way monitor with low distortion drivers to a wave guided design you will see they are close on axis and would prefer the wave guided off axis. Not sure if Madisound has Zaph's waveguided TMM design but that would give you a great comparison. Jon's Modula's are the best speaker I have done that sound great in a good room and bad.

[Juhazi]

I support 5thelement's text. The DXT is a tiny sort of a waveguide. ERDXT is perhaps the best 2-way implementation of it, at least in the 6.5" midbass category.

I think that too much directivity gives the headphones effect. I don't like that in my speakers and ERDXT doesn't do that. Optimal level of directivity is an opinion. More directivity helps in rooms with hard surfaces. Best horn tweeters give the most constant directivity but that's a different story.

One classic of commercial controlled directivity speakers is the Finnish Amphion Argon 2-way, the x-o point is 1,6kHz with a custom Danish tweeter. It was introduced in the 90's. Here the current "evo3" model http://www.amphion.fi/fi/tuotteet/argon3/ I have a pair of the smaller Helium acting as HT rear speakers.

And just for curiosity I digged for my in-home measurements of the Zaph L18, 0¤ 30¤ and 45¤ distance 1m. The curves for 2m would be different. Non-calibrated laptop used for this. Sorry, I haven't made a similar for the ER18DXT. And the laptop has been changed since too.

[bunshi]

Jon,

Thanks for helping me understand why a smooth off axis response is important. It makes more sense now. Waveguides control the off axis response by making them flat, and that has two effects: 1) increasing the size of the listening window, and 2) making the reflected sound come back flat, which is something we all hear. I could also see why a ragged off axis response is less desirable, even if on axis is nice. That ragged sound will eventually hit your ears with unpleasant sounding linear distortion.

What goes on here is that the wave-guide essentially captures the far off axis sound and redirects into this main cone. What this does is reduce the amount of sound being thrown out into the room and reduces the amount of reflected sound in relation to the direct sound. As Jon mentions this is one thing that's important with regards to imaging and the ability of a loudspeaker to throw out a convincing spread of sound. This is where the description of headphones but without the image within your head comes from - headphones obviously have no reflected sound as they remove the room and this is exactly what a wave-guide attempts to do by controlling the directivity of the loudspeaker so that less sound is thrown out into the far off axis.

5th, this is well written and easy to understand too. Thanks, because you've helped me untie the a knot of ideas in my mind about off axis response. The waveguide controls the direction and the flatness of the sound, and the MTM does this to a certain degree with vertical lobing. The thing that didn't make sense to me before that does now is that off axis response isn't bad when it's flat.

Juhazi, Jon, and 5th, thanks for the advice for designs to look into. I would like to hear the MCM or another larger waveguide sometime. Maybe I'll go to a DIY show. I'm still interested in the ER18DXT because, well it measures much better than the Idunn.

Actually, you guys nailed my questions, so thanks!

I guess a follow up question to this is "how does flat off axis response affect imaging characteristics?" Let's take two extremes, how would an omnidirectional speaker with a flat response in every direction compare to say Jon's example of listening to a perfectly flat on axis speaker in an anechoic chamber? Anyone have any reading recommendations for this topic?

[SpeakerGuy]

...I recommend using the Jantzen waveguide since the MCM is no longer available- better quality construction, but a little bit of a pain to get, as you pretty much have to email them and get it direct.

See here: http://www.diysoundgroup.com/waveguides/dw-62s.html

Only $5.50! I bought some, can't tell any difference from the earlier ones I got from MCM.

[craigk]

something else to keep in mind bunshi. don't let yourself get caught up in wave guide vs. flat baffle. just listen to the music, and let that guide you as to what you like. it is very easy to get an idea stuck in your head and then try to make it fit the situation. i have listened to speakers with and without wave guides that were great speakers. just because a speaker has a waveguide will not make it sound beter, but it might. as far as the low crossover goes ... there are several designs that use the mid to cover freqs up to 4000 hz. i know off axis response and beaming come into play, but a good creative design can go a long way. take a look at the unifield 1 loudspeaker. if you like pin point imaging an omni is not the way to go. if you like lots of atmosphere it might be your thing. if you look around there are not of omni designs out there. there are exceptions to everything, the mbl speakers are in my top 5 speakrs to own if i hit the lottery. one last thing, i reallly like mark k's speaker.