Metal vs Softy Shootout

[augerpro]

Almost $600 for XO parts for 2 pairs of speakers 8O Using ClarityCaps for the resonant traps on the tweeters really pushed up the price, and I only used them for half the value! I hope these sound kicka** for that kind of money. Whichever pair doesn't really interest me is going in a nice enclosure and hopefully I can seek it recoup some of my costs.

[augerpro]

deleted

[Nathan P]

Nice

[augerpro]

I received the parts for the prototype XO. Those 2.5mH perfect layer 14 guage inductors are HUGE! At least 4" across I bet. And the ClarityCaps are monstrous too. These crossovers will be outboard but if someone built a normal speaker of these you'd definately have to compensate in box volume. I'm really looking forward to testing these out :P

[---k---]

Yeah, the Khanspires have a 2.75mH in it.

[augerpro]

I need some advice guys. Here is a pic of the XO for the HDS tweeter:




The two large inductors sharing the terminal strip are for mating the tweeter with either the HDS woofer or RS150. I just swap between the two inductors. My question is even though only one inductor will be used at time, will the proximity of the other do anything funny? It's a lot of metal close by so I was kind of concerned...

These are nearly finished, just have to finish two baffles. I'll only have couple days to listen to them though since the refinery I work at is shutting down for maintenance at the end of the month and I'll be working for thirty days straight :M maybe 45 days even...in twelve hour shifts... the money is great but that kind of continous work just sucks the life out of you. Anyhoo, Thomas stock up on gas now because prices are going up in a month :E

[joecarrow]

My question is even though only one inductor will be used at time, will the proximity of the other do anything funny? It's a lot of metal close by so I was kind of concerned...

Those are both air-core inductors, right? In that case, there's nothing to worry about. If you shorted the the unused inductor to form a closed loop, on the other hand, it might have some quantifiable effect- but the way it is now I don't see any way for the presence of this non-connected inductor to influence the circuit.

[JonP]

Yep... should be no problem. It's a lot of metal nearby, but it's not ferrous (magnetic) metal, so it won't affect the other inductor's value. Like Joe said, since it's not in the circuit (both ends connected) or not shorted (a "shorted turn" would make it soak up energy from the other inductor) it should have no influence at all.

[augerpro]

Awesome guys, thanks!

[augerpro]

Just thought I'd give a little update on this project. I've finally finished my music room and started listening to these. I have them level matched and running through an A/B switch box I built. Being able to instantly flip back and forth is very helpful comparing speakers, our auditory memory is extremely short. Right now I'm just trying to find how they differ, then try to discover why they differ, then start tweaking with EQ to see what the effect is.

First observation: speakers with similar design and similar response sound similar! I know, a very profound observation :P

There are differences though. Both the Metalhead and Softy do some things better than the other. Eventually I'd like to determine how to improve the areas where each falters, and determine which drivers I am able to practically able to get the highest SQ and performance out of. Anyway here is a short review of each, please excuse the "audiophile" terminology, I hate the vagueness but it's all we've got in the spoken word:

Metalhead- I really like the SS 9800. Cymbals, chimes and female vocals have a really nice weight and texture. Nothing harsh here at all, these are very refined sounding, I guess Stereophile would say they have good blackness between notes or some such. Combined with the steep rolloff of the RS150 this speaker is a little laidback. There is fine line between refined and recessed though. This mellow quality made the tonality somewhat unnatural though, even though both the 9800 and RS150 seemed very detailed. The soundstage was not wide but did seem deep, especially with a gospel type chorus these sounded great. Deep soundstage and all the individual voices were audible and with good texture. The RS150 is great with electric guitars. And decent bass too.

Softy-Currently I'm favoring this one. Now I just need to find why I like it and how to replicate that in future designs. These are more forward than Metalhead, and could be fatiguing with certain material after a time. However the soundstage was huge compared to the Metalhead. My small room felt much larger when these were playing. They had an almost reverberant quality. Also everything sounded more natural on these. The treble detail was very good and had a nice sparkle. Not as weighty as teh 9800 but more spacious and natural. Voices, strings, flutes were all more natural and "alive". Still undecided on whether the 832873 or RS150 is better with lower midrange stuff. Whereas my impression of the Metalhead is laidback, refined, and nice texture, the Softy is BIG, natural, sparkly.

I think the woofers actually contributed more to the spaciousness and naturalness than you would think. I listened with just the woofer circuit on and teh rolloff of the RS150 was clearly contributing to the laidback and slightly unnatural quality of that speaker. Funny thing is the summed response doesn't suggest that this should be the case, afterall the 9800 is contributing at similar level as 810921-and lower in frequency, which would make it sound brighter in my mind. But that wasn't the case.

Hopefully the spaciousness of the Softy is not related to it's tendacy to be a bit forward. I want to see what changes to the response will make it sound more refined, while still keeping the spaciousness and naturalness. So far I really like the 832873, it's response is smooth enough that you can play with the voicing to get it where you want. The 810921 is very nice too, but you really need to be careful shaping the response I think so it doesn't become too "in your face". I'd like to see if tweaking the response will yield the same kind of weight and texture as the 9800, or if that's just a quality inherent to the 9800.

Again the 9800 is very smooth. I would like to see what response tweaking could open it more like the 810921, but since it's $150 I don't think I really want to know. Then I'd be forced to use this expensive tweeter in my future projects...

The RS150 is has very good midrange. Textured, detailed. Electric guitars and big stringed instruments sound powerful and real. Not as good with voices and the higher notes on stringed instruments compared to the 832873. My basic assumption is that most of the differences between these drivers are response related and that any of these drivers could be made to have the strengths of the others with proper response shaping, XO point, and rolloff. Unfortunately for the RS150, the breakup limits how high you can cross this driver, and so limits the flexibility while voicing the system. In this regard teh 832873 is the better driver to work with, and the cost is relatively the same when you consider the RS150 needs a notch filter. Unless you were using multiple RS150s and cost was an issue anyway. I still have a lot more testing I need to do though, that RS150 certainly does the lower midrange well, perhaps better than the 832873.

Anyway I'm very happy so far. Either of these sound leagues better than anything at BB or even alot of the dealers around here. Soon I'll throw my Rocket RS250s in to the mix and see if I've created a truly reference quality bookshelf.

[---k---]

Great thoughts.

Though, the way you wrote it focuses on the drivers as being the causes, but I wonder if it is more the implementation. Thinking about my Khans with the RS150 (yes it is a 3-way....) and CJD's RS150 MTM. Mine I definitely would not call laid back. I think they are a bit forward. And neither project would I describe as having a small soundstage. That was something that struck CJD when we compared his MTM to the Ascend 340s back when - his had a much bigger soundstage. Both were MTMs... hummm...

It always seems that I'm reading Stereophile-like people say that lower-order, particularly first-order, crossovers produce the most open, biggest, realistic sound. Maybe there is something to this. Unfortunately, metal cones don't like first order.

I look forward to your further experimentation and seeing if you can find that secret that opens up the soundstage and brings them alive.

[mazurek]

Not sure if this is relevent. The speakers I had with Peerless 830875 distorted at higher volumes, its possible that the rising third order distortion of the mid could contribute to some harshness at louder listening levels.

I am currently enjoying my speakers with the HDS tweeter, guests have given this one some praise. It is crossed LR4 at 1700 Hz, with a notch for diffraction around 1800. There's a couple songs that caused me to pause and ask what happened with these tweeters, but upon further investigation (my room mate turned me on to Audacity) it was always something where for example the cymbals in a song digitally clipped, or it was a lower bit rate MP3.

Have you thoroughly investigated off axis performance of each speaker, I saw your figures but the scales are quite compressed. Also, I have found the best balance on my speakers by setting the tweeter levels relative to the woofer according to either a +/-15 or 30 polar average or just slightly off its axis.

[cjd]

Play with box stuffing and wall treatment! You may have the 150's overstuffed (or under-boxed - they open up as you push to the larger box sizes, something David Ellis noted in Chicago with my MTM vs the TM with TBFC+RS150). Of course, if you have NO treatment in the box, you can't have it overstuffed. Though I still recommend playing with stuffing and wall treatment if that is the case. It can change things in interesting ways - I recently received a note from someone that had me working up a version of my MTM with some BBC dip, turns out box stuffing and wall treatment proved to solve his issue.

C

[augerpro]

Hey guys thanks for the replies! A lot of good observations :T

I should emphasize that any differences are as much relative as they are absolute. I don't know if many people would call the Metalhead laidback, maybe smooth. You could listen to these at higher volumes for a while and not get fatigued I think. Also I don't want to give the impression that the SS9800 is dull. I think if most people heard they would say "ooooh". It has sparkle, just not like the 810921.

-k->my underlying assumption is that most of the differences I hear are simply response related, and could be EQed whether through the actual EQ or tweaks to the XO. This certainly true for the woofers I think, the tweeters a bit less so since ways to shape their response are more limited. I don't know that the Metalhead's soundstage is really that small, but compared to the Softy it is. I wonder if the response that is causing your khans to be a bit forward is directly related to the size of the soundstage? The Softy shares both of those qualities too, I hope I can tone them down a bit but still keep the open soundstage. Maybe this one of those areas where we must comprimise according to taste. What do you think contributes tot his quality? Lots to learn there I think... Perhaps a higher woofer rolloff and different tweeter account for the differences between yours and mine? I hope to get to a diy meet at some point and listen to yours and cjd's designs so I have more references to compare to. For all I know people could listen to these and say "wow those suck!" I hope not, but...

mazurek>The difference in soundstage is apparent at all levels. The effects of distortion *so far* have been subtle, both of these speakers will get pretty loud, and sound good doing it. I did take offaxis into account when designing these. Just offaxis to the inside should be quite flat. But I still need to run final measurements and THD sweeps, I just got the system up and running this week so haven't had the chance to verify the response. Maybe some enlightment awaits within these measurements.

cld>I definitely do have to investigate stuffing and wall lining.I used the 1" PE foam for teh back wall and some wool/synthetic batting from MCM for the rest of the sides. I like that batting for smaller speaker I built, but it may not be the best for larger speakers. The Qb is slightly different, with the Metalhead being slightly higher, around .7 I think. I tried to get teh Q as close as possible but these woofers are just so different. I do need to mess with the stuffing to see how that effects things, since the Metalhead bass is much more pronounced and pleasing. That makes it hard to compare the midrange since the difference in levels down low color how I hear the upper frequencies I think. I think you are correct in that part of the unnaturalness I hear may be a box resonance or coloration. More investigating...

These are just my first impressions so far. I have a lot more work to do to discover what *is*. Then I'll move on to *why* it is. Then I'll see how can change it.

[Hank]

lower-order, particularly first-order, crossovers produce the most open, biggest, realistic sound. Maybe there is something to this. Unfortunately, metal cones don't like first order.

I get here sporadically, and as a complete novice (don't design crossovers, just cabinets), from the volumes I've read on the forums, especially here with the very generous knowledge-sharing of Jon and Thomas, I think there may be a LOT to the quoted statement. I summize the if one likes the sound/performance of metal cones then one will end up with complex, high component count crossovers. I'm not saying that's bad, I'm just trying to head toward general conclusions. (Oh no, I just had a flash that tells me I shouldn't have posted this and should don my flame suit - nah, not on this forum :-) ).

[Dennis H]

One thing that's measurably different with lower/higher order crossovers is the vertical polar response. One could make a case for that's why they sound different.

[augerpro]

Not sure if this is relevent. The speakers I had with Peerless 830875 distorted at higher volumes, its possible that the rising third order distortion of the mid could contribute to some harshness at louder listening levels.

How did this sound to you when it distorted?

[mazurek]

How did this sound to you when it distorted?

I took a little while to diagnose this problem because it sounded somewhat like the tweeter was bright or a little harsh, which as I understand is often due to frequencies at maybe 2-6k, I would put false detail frequencies on the higher end of the range and above (IMO). I ended up listening to just the tweeters and the harshness went away. Then I listened to just the woofers (at loudish levels maybe 90-95dB), and the bright or harsh feel remained. Not that anything was terrible, but some music didn't sound as clean as I thought it would (complex music and obnoxious singers like The Shins guy). In my case it made sense that the woofer was pretty good at lower levels, but its distortion pattern matched what I heard at slightly higher levels. In my newer speakers I have a similar crossover topology to the old (LR4), but drivers that scale distortion with volume better, and they do not exhibit the same problems.

Who knows if this is relevant to your situation? Good thing you like to measure to find out.

I'm glad you built two new reference speaker systems, it sounds like you are avoiding the common audiophile pitfall of blaming everything you hear on inferior electronics and music, all while being secretly dissatisfied and waiting to upgrade.

As a side note, it sounds like you have some very good speakers now. Try using the thuneau arbitrator program to make your system transient perfect, it is a pretty cool effect.

[JonP]

Just thought I'd give a little update on this project. I've finally finished my music room and started listening to these. I have them level matched and running through an A/B switch box I built. Being able to instantly flip back and forth is very helpful comparing speakers, our auditory memory is extremely short.

A great study... something I'd like to do someday...

A thought occurred to me... So, I'm thinking you've got two pairs of speakers in the room, with the A/B switch setup. The two speakers would have to be in somewhat different locations. It may be pretty close, but fundamentally you have them in different positions. There might be changes in the FR due to placement effect, how the dispersion works in the room might be noticable with the walls closer or farther, etc...

Probably a minor effect, but you might want to swap their positions to see how much influence it has.

[augerpro]

Hi guys! I'm still experimenting a lot with these two speakers and learning a lot...playing with an external EQ has helped me to define my tastes, which I've found is very much for smooth, flat response on and off axis, with perhaps a slight dip from 1k-3k depending on teh room and drivers. That frequency range more than any other really hurts my ears if it's too prominent. I have a new crossover for the Peerless combo that I'm going to try and I'm going to port the box. Which brings me to my question for the sub gurus here. I modeled a box where group delay peaks at 15ms @ 35Hz and port velocity is 25 m/sec @ 35Hz at xmax. How do those numbers sound?

[augerpro]

No one has any input? I know Thomas must have the info locked in his head :

[ThomasW]

Sorry I missed the question. I've only been looking at the thread occasionally, so I'm not sure exactly what it is you're building.

The tuning specs are for a RS150? And this is a plot of the box?

[cjd]

Thomas: he was asking about sub tuning...

my question for the sub gurus here. I modeled a box where group delay peaks at 15ms @ 35Hz and port velocity is 25 m/sec @ 35Hz at xmax. How do those numbers sound?

[ThomasW]

This is a question about this sub?

https://www.htguide.com/forum/showthread.php?t=26988

[---k---]

Did you happen to see the recent thread at PE by JeffB? He did something very similar, but comparing cheep drivers to expensive drivers.

You should check it out, but the thing from his thread that stuck out in my mind was his statement (and this is paraphrased from memory) that the initial impression was that cheep drivers sounded better with more sparkle and flash and the more expensive drivers sounded dull. But more careful listening showed the cheep drivers were pumping out distortion, the cheep drivers sounded like an image with a sharpen filter applied - That was probably the best analogy trying to convey how something sounds that I've ever read. While the expensive drivers had greater resolution with more detail. And, a lot varied based on listening volumes.

Definitely worth checking out as I think it really is a parallel adventure.

[augerpro]

Hi guys. I was actually porting the Softy enclosure from this thread, a Peerless Nomex 5" driver. I just figured the sub heads would have a good idea of how much group delay and port velocity was too much. From my post above,"I modeled a box where group delay peaks at 15ms @ 35Hz and port velocity is 25 m/sec @ 35Hz at xmax"

Ryan> interesting comments and I considered that too. What was making the 810921 sound more detailed than the SS9800? I don't think it was distortion, they are both very close in that department. It's not just level because I EQ'ed them to be about the same level. And it's real, not an illusion. With my A/B switch box I can flip back forth instantly (invaluable tool) and hear the difference. It's a characteristic I guess...I know SS voices their tweeters for a particular sound and it is VERY pleasing. Never harsh, always smooth and detailed. If all I listened to is Rock I'd pick the SS9800 hands down. Led Zeppelin would be too fatiguing with the 810921. And so would any rock band from the 70s and 80s. The SS9800 is very detailed, better than my XT25s in my Rockets I think. And it has weight, like the stick hitting a cymbal actually has force you know? It's just doesn't seem as detailed and real as the 810921, but then the 810921 is a strange tweeter, you really must hear one. And A/B it against something else. That's the other thing, instant A/Bing makes small differences much more noticeable. I don't think I could be as sure of the differences if I couldn't A/B them instantly. If I heard them hours or days apart I would probably just hear two very high quality tweeters, one being a bit smoother and teh other being more spacious. So I don't want to overblow the differences. And I'm not done, I'm still playing with the EQ, and have a new XO design for the Peerless combo that I did to try to get the smoothness of teh SS9800/RS150 combo, but retain the other qualities. I like the sound of the 810921, but I also know I don't like anything that can fatigue my ears. Months from now I may decide the SS9800 is better to own, simply because I can listen to it for hours, even if it doesn't have that last bit of spaciousness or detail. Who knows...

[ThomasW]

This driver?

http://www.madisound.com/catalog/product_info.php?cPath=45_228_254&products_id=1612

Wouldn't port it at 35Hz under any circumstances.

In addition to the group delay and Mach numbers, it's important to see the power level and frequency where the driver runs out of Xmax, then stay well clear of those with a small driver like this.

The alignments posted on that web page seem reasonable, sort of....

[augerpro]

That's the driver. I tuned the box (9.8L) to 45 Hz. At xmax (about 35 Hz or so and 30-40 watts) those were the numbers I came up with for GD and port velocity.

[ThomasW]

Fb is way too low IMO.

With a real woofer/subwoofer it's ok have have a bit of a dip/saddle between the driver roll-off and the output of the port since room gain will fill it in. With this midrange there's not going to have enough output to pressurize the room. This means little if any room gain, so there would be a hole in the FR between the driver output and port.

The 65Hz Fb mentioned on the Madisound web page looks about right for this driver.

[---k---]

I wasn't trying to say JeffB's results explained your results. I was just pointing out an interesting thread of a similar experiments. Thought you might find it stimulating.

It will be interesting when CJD gets his speakers done and my new center. While his will be the inwall version, and mine somewhat onwall. I would imagine the final FR should be the very similar. The only major difference will be his uses the Seas tweeter versus the RS28 in mine. Usually we run out of time too early when we get together, but we should be able to do a similar listening tests.

[augerpro]

I wasn't trying to say JeffB's results explained your results. I was just pointing out an interesting thread of a similar experiments. Thought you might find it stimulating.

I hope I didn't come off like I was refuting what your were saying Ryan! I'm glad you brought up that thread, I'll have to go search for it at PE.

It will be interesting when CJD gets his speakers done and my new center. While his will be the inwall version, and mine somewhat onwall. I would imagine the final FR should be the very similar. The only major difference will be his uses the Seas tweeter versus the RS28 in mine. Usually we run out of time too early when we get together, but we should be able to do a similar listening tests.

Should be a very informative comparison, please post your impressions since those are two tweeters I've never heard. If you can try to use an A/B switch. Being able to switch instantly is really eye (ear?) opening. I made a simple one using McGill SPDT and DPDT switches from amazon, and one of those project boxes and 100watt LPads from PE. I think I used 10 and 12 guage wire, and the LPad was bypassable.

[mazurek]

I'm curious about your tweeter comments mainly because I would really like to make the HDS tweeter work very well in my application. I am interested if tweeter geometry causes the difference in observed response due to differences in dispersion. The Seas TBFC aluminum tweeter I have is much flatter than my HDS tweeter, and the HDS tweeter has a sharper faceplate. I wonder if some of your observations regarding tweeter differences are due to top end dispersion and to some extent the design axis at which the top end is made flat. I'm wondering if you tried varying your listening axes to compare on and off axis for the speakers, or top end equalization.

I'm trying to get cymbals as good as possible on my speakers. I have not been used to the sound I'm getting with these, and don't know whether it is right or wrong. I tried playing around in the lower tweeter frequencies, but I think the frequency range of the cymbals that have been catching my interest are in the higher tweeter frequencies. I'm curious if what I am hearing is due to the fact that the peerless HDS seems to be designed to have a flat high frequency response slightly off axis. I feel that is a valid choice because I never listen exactly on axis, but maybe I am not used to hearing speakers that are flat to that frequency off axis.

[Jim Holtz]

I'm not sure if you guys saw this post by Dennis Murphy or not, but he had some opinions about the HDS tweeter. Click here

Jim

[mazurek]

I'm not sure if you guys saw this post by Dennis Murphy or not, but he had some opinions about the HDS tweeter. Click here

Jim

I saw that discussion, and some interesting points were raised; I feel augerpro has an excellent opportunity to evaluate some of the issues with actual measurements on two systems (along with off axis response). I have some measurements, but only one system and one set of ears so I am reluctant to say too much. Another thing that could factor in is rising 2nd order distortion above 5kHz that you see at higher SPL levels.

[augerpro]

I'm not sure if you guys saw this post by Dennis Murphy or not, but he had some opinions about the HDS tweeter. Click here

Jim

Just goes to show how much taste is involved. ThomasW thinks the SS9800 is the best tweeter he's ever heard, and few would dispute that. And while I really like the SS9800 too, there is no doubt in my mind that the Peerless sounds cleaner and more open, usually more detail too although on some stuff it's pretty much the same. Although the 27TBFC is tempting to try from all the positive response.

[augerpro]

I saw that discussion, and some interesting points were raised; I feel augerpro has an excellent opportunity to evaluate some of the issues with actual measurements on two systems (along with off axis response). I have some measurements, but only one system and one set of ears so I am reluctant to say too much. Another thing that could factor in is rising 2nd order distortion above 5kHz that you see at higher SPL levels.

That's the thing, the differences seem to defy measurement. Both are low enough in distortion that most would probably discount the difference as audible. The only real difference there is slightly higher 2nd order on the SS9800 below 3000Hz and slightly higher tall order harmonics for the 810921 throughout the spectrum. So maybe you have something there.

Response doesn't really seem to explain it either, the baffle dominated the response and I made further tweaks with EQ and I can tell you that even if you EQ'ed both of these flat, the difference would still be apparent. It is certainly a distinct tweeter adn I think everyone owes themselves the chance to hear one. Hopefully we can get diy get together here in Denver and I can bring something that uses this tweeter.

[augerpro]

Fb is way too low IMO.

With a real woofer/subwoofer it's ok have have a bit of a dip/saddle between the driver roll-off and the output of the port since room gain will fill it in. With this midrange there's not going to have enough output to pressurize the room. This means little if any room gain, so there would be a hole in the FR between the driver output and port.

The 65Hz Fb mentioned on the Madisound web page looks about right for this driver.

Thanks Thomas. You don't think I'll be any danger of unloading the driver under the tuning frequency with music? That kind of scares me and is why went lower with the tuning frequency. If not then I'll try 65Hz. And what kind of group delay numbers and port velocity numbers should I be looking to stay under?

[mazurek]

One more idea, I'd love to see you try this if you have time. If you equalize both to have a flat on axis response, what does the ungated >1meter in room response look like?

[ThomasW]

Thanks Thomas. You don't think I'll be any danger of unloading the driver under the tuning frequency with music? That kind of scares me and is why went lower with the tuning frequency.

Unloading is certainly a problem but so is over excursion especially with a driver having only 3.5mm of Xmax. Typically a driver like this is used as a mid not a midwoofer, or it's used with a sub and an 80Hz XO point.

And what kind of group delay numbers and port velocity numbers should I be looking to stay under?

The numbers you posted are fine but it's an unrealistic Fb for that driver

[Jed]

Augerpro. You seem to really like the 9800 and the numbers look good as well as your listening impressions. Would you recommend it over an RS28?

I think the HDS tweeter's added detail you are hearing is directly correlated to it's FR. For example, it has that rising response above 10k and you also seemed to have a different/hotter FR summing for the "softy" design. Also, don't discount the mids contribution to detail in the treble qualities- it's a sum of both parts kinda thing. A more controlled test of a tweeter would use the same mids. Really interesting non-the-less. Don't take that as criticism or anything.

Jed

[augerpro]

Augerpro. You seem to really like the 9800 and the numbers look good as well as your listening impressions. Would you recommend it over an RS28?

I think the HDS tweeter's added detail you are hearing is directly correlated to it's FR. For example, it has that rising response above 10k and you also seemed to have a different/hotter FR summing for the "softy" design. Also, don't discount the mids contribution to detail in the treble qualities- it's a sum of both parts kinda thing. A more controlled test of a tweeter would use the same mids. Really interesting non-the-less. Don't take that as criticism or anything.

Jed

I've never heard the RS28 so I can't give an opinion Jed. I've heard it supposed to be a copy of the SS 9800, but who knows. I'll tell you this: it's super easy to get great sound from. Just keep the response flat and use an Lpad to adjust the level to taste and voila! magic! Price wise it is not a tweeter for beginners, but it's such an easy tweeter to work with.

You know I've EQ'ed this tweeter 20 ways from Sunday and it's very difficult to ascribe what I hear to anything other than an intrinsic characteristic. I'm also swapping tweeters so that is an interesting perspective too. The 810921 still sounds relatively the same with the RS150, although it's clear the Nomex woofer is made to work well with the 810921 since they have similar sonic qualities. I think with proper voicing this combo would be excellent. The Nomex/9800 combo was nice to, but this particular XO wasn't as good as the others.

I have a lot more listening and tweaking and swapping to do before I'll have any final judgment on any of these drivers. So much more testing to come...

This project was to determine the drivers I was going to use for a complete 5.1 system to replace my Rockets. The original idea was to use the Vifa D26 but now I'm not sure. I'm going to build some small speakers with that tweeter and see how they compare with 810921. If I really like them than it'll be MTMW mains using the Nomex mid and RS270 woofer and WTMW center using the small Nomex mid and either RS150 or RS180 woofers, and MT rear surrounds. If the 810921 still stand out I'll use that for the mains and the D26 for all of the other speakers.

Hopefully I can build a decent 3 way by this summer and I can get to diy meet and get some actual feedback on them. And I need to hear some other good speakers, your Lineup series is high on my list.

[augerpro]

Unloading is certainly a problem but so is over excursion especially with a driver having only 3.5mm of Xmax. Typically a driver like this is used as a mid not a midwoofer, or it's used with a sub and an 80Hz XO point.
The numbers you posted are fine but it's an unrealistic Fb for that driver

I'll try the higher Fb Thomas, thanks!

[Jed]

Augerpro,

I think you will really like the D26 and it sure makes a good value statement. I recently sold my Visaton Ke25sc tweeters to get replaced by something like a Scanspeak 9800 or Vifa D26.

The D26 tested better than the Ke25sc from 2k on up. It has such low high order distortion. I'm still fighting the "snob" in me that says if it's expensive it has to be better. I know that isn't true but I have to unlearn years of ignorance and brainwashing by hifi loudspeaker salesmen.

Through the years, when I started getting interested in DIY loudspeakers back in 1996- some notable tweeters that I really liked were the Focal Series of inverted domes. They had a certain immediacy to them where cymbals, chimes, etc just were so dynamic they exploded into the acoustic space. I wouldn't say the D26 has that quality- they just sound "clean." From a scientific perspective (note I'm an art teacher!) clean is the desirable outcome, however it isn't always the most "exciting" to listen to. As you stated though, this quality (in your experience the 9800) is less fatiguing over the long run, so I can listen to D26s for hours on end.

The question is, can one have both aspects of sound that I am looking for- clean and ultra detailed sound or is the detail just "distortion" that at first is like sweet candy at first but causes a stomach ache if consuming too much?!

Jed

[mazurek]

Here are some measurements of the HDS tweeter that I think may be relevant (with a calibrated microphone). These are all in cabinet, 1/6 octave smoothing, I didn't calibrate the levels so those are arbitrary.

The top curve is on axis. The middle curve is 15 degrees horizontal off axis. The bottom curve is my chair.

I see about a 2.5dB peak at 15k on axis, 15kHz is easily audible to me. However ungated in chair measurements show a smooth response.

Also attached is what the system frequency response (crossover 1.7kHz) and harmonic distortion is at my chair (2.5meters away, between my speakers). Drive level was 90dB. This also gives a mechanism for higher detail, there is some increased second order distortion in audible frequency ranges.

So does anyone think these add to the story augerpro tells, or measurements don't tell everything.

[Operandi]

It didn't seem apparently clear to me from reading this topic backwards... Is this just an "experiment" or is the goal of this to bring both the Metal and Softy to a "finished project" status? Because I certainly think they should be.