2021-2023 Isiris Update

[JonMarsh]

Thanks for the link-

I love spoilers, too!

I'll check that out with breakfast tomorrow- it's been a long busy day, and I'm a morning person- up at 4 today, and heck, I'm RETIRED!!

[JonMarsh]

Supply chain issues, but still some progress

well, things are moving along, but this is 2021, so you know there will be "issues"...

Solen showed all the big boy inductors as being in stock, but I got an email from them today, turns out they just may have been exaggerating that just a wee bit... as in all of them.

But since they are "Solen" inductors, they'll get the elves busy and build some, sometime soon, and they will get shipped... eventually.


OTOH, after 12 days, it looks like my Parts Connexion order will arrive today, with some, guess what, Solen AWG12 4.3mH inductors.

And as I already have some 2.2 mH inductors, plus some left over North Creek Inductors, that means it's time to get ready and layout and build a test board to evaluate all these puppies on.


Well, 'tis the season to be jolly AND to be productive, so, voila!




Those yellow rings mark the boundary of the space reserved for the bigger parts to check out; however, the one on the left needs a full inch more on the radius.


So, though I HAVE been drinking a little spiked egg nog this week, it's only after I've gotten the days tasks done, and certainly not before noon time!

Hmmm, might be reasonable to hook all these new boards up and see if anything interesting happens... all the Xmas stuff is done, thank heavens!

[JonMarsh]

UPS a knock-knocking on my door...

Soon as I hit save... my doorbell rings.




And there's bigger ones coming- the big boy 4.3mH is 6" in diameter.

[JonMarsh]

Slowly I turn, Step by step, inch by inch...

I think we're getting there, folks.

The only thing I did for this measurement was add one more resistor in parallel on the shunt side of the midrange L-Pad. That is, besides the new woofer and midrange crossover!







Now, until I finalize the woofer crossover inductors, and know the levels, and "finalize" the midrange level matching it, I'm not going to tweak the tweeter level. Actually, it's a pretty good voicing at this point, considering the overall power response of the system, and even how it sounds at the other end of the house (by the front door- totally opposite where the speaker is right now- this is a long deep floor plan.

What's always fascinating about this system is how much you can hear 90 degrees off axis, due to the dispersion characteristics resulting from the front midrange-tweeter baffle.

I'm planning to listen to a bunch more cuts later today, even though just from the test MacBook at this point, but it seems like things have really fallen into place with the revised crossover frequency and the overall network update.

:T

[tktran]

Jon,

That crossover looks like it costs a small fortune! Do you prototype with a DSP or software based filters and multi channel D/A convertors first, or do you simulate then build and tweak as you go along?

[JonMarsh]

VituixCAD is my go-to tool for most things these days, though with my heavy duty turntable packed up in storage, I'm not doing multi-polar simulations for this project, as I already have a good overview of the polar behavior of the drivers.

I do have a multi-channel DSP packed away and more amplifiers to experiment with that (for the line array system in development), but since I don't plan to use DSP's for the end build (DSP DAC's and digital clocks are not exactly state of the art...) I'm comfortable working with my measure/simulate/build/Measure rinse and repeat process. And I can implement stuff in "analog" that is hard or impossible to implement in any budget DSP I've seen, probably not even in the Accuton DSP I have.

The crossover topology is basically the same as for the Wavecor Ardents and the original Isiris; it's the quasi LR 3rd order alignment that expects specific wavelength related offsets for drivers to sum flat, as described in other threads.

This version is getting pricey, but remember, the whole project was originally pricey. I keep a running BOM for the crossover, and it's at about $1,1000 per cabinet now. That should be put in perspective, that the original Isiris build had a total BOM of about $10K. I really like what I'm hearing with this last update, and the driver expenditures have been slashed considerably.

My target was to bring in drivers + crossovers in the same range as the Wavecor Ardent, and I'm still very close to that. If I go with the big boy inductors I'm waiting on, then it will get pushed up a bit, but still well under $5K I think. That is meeting the component cost reduction goal I set, while improving many aspects of performance.

Of course, there are those other crossover component updates suggested... but then I am on a fixed income, these days, and ET's Kurosawa's need to be finished!

[dwk]

VituixCAD is my go-to tool for most things these days,

Hopefully not too OT, but knowing you're a Mac guy, how are you running VituixCAD? Are your running on an M1 machine? I picked up an M1 Air, but it's still not at all clear how well - if at all - many x86 Win apps run under a VM w/ translation.

[JonMarsh]

Well, I have one 2013 generation Mac Pro that is setup to dual boot, it's usually used in WIN10, and all my Windows CAD stuff is on that, like VituixCAD 1 & 2, Altium Designer, SIMPLIS, SIMetrix, DEX PCB, MicroCAP, TINA, etc.

I have a couple of M1 machines (Mac mini and 13" MacBook Pro) but because of the differences in system architecture from PC, they won't run x86 Windows- though some folks have been experimenting with getting Windows for ARM running. But as far as I know, VituixCAD would have to be recomplied into ARM format to run on that. I know Parallels is working on some things.

I have Parallels with WIN10 also on an iMac Pro, and VituixCAD runs fine on that.

I have an LG Gram 17" and a Dell XPS15; they were bought mainly as dedicated controllers for the Audio Precision APx555, though I use the LG for a lot of other stuff, speaker oriented, because the combination of specific screen resolution and size works better for that. It's also a backup system for Altium Designer.

Now, I can tell a funny story about CPU, GPU, and thermal power comparisons- I have Divinity Original Sin 2 on both the iMac Pro and the 13" M1 MacBook Pro. Both run it pretty well with the graphics and resolution max'd out. But the fans get cranking on the iMac Pro (10 core, mind you) pretty good, whereas on the M1 MBP I never hear the fan, and the chassis barely gets warm. Quite a contrast.

If I didn't have my financial resources tied up buying a house, I'd love to have one of the new M1 Max Macbooks- and eventually I will get another desktop if I need something more powerful than the 2020 Mac Pro I have. That I'm going to setup with WIN10, too, as the future WIN CAD system- with 4TB HD and 48GB of RAM, I don't think it will have any problems with that- I might even take the easier way out and go with Parallels again. But the 2013 generation one works so well, I have no real reason to update from that yet. Xenon processors and fast memory and fast SSD's will do that.

In fact, the last 5-6 years I used to do all my day job stuff on my own hardware, because the work supplied computer was such a joke- literally, an Admin special with a spinning hard drive. Same HP model as the admins got. Fine for email, not for anything else.

[Scottg]

..the big boy 4.3mH is 6" in diameter.

8awg chonker? 8)


I was figuring on these, but at .446 (vs. the 8awg's .26) they are more resistive:

https://www.erseaudio.com/Products/FoilQCoils12Gauge/EAF12-4300

..of course if you really wanted to "f-the-budget" :B you could do a search for these:

000-9430 Wax Coil 4,300mH +/-2% 0,19Ω +/-5% 8AWG

[JonMarsh]

Hey, thanks for the suggestions, Scott! I thought the AWG10 bad boys from Solen were pretty over the top- cost more than the woofers did for the 4.3mH and 2.2mH!

[Scottg]

If I didn't have my financial resources tied up buying a house, I'd love to have one of the new M1 Max Macbooks..

My computer building/purchasing was a bit further behind then yours. ops:

I had a built a gaming rig (around 2007) with some some SLI nvidia GPU's that lasted until about 2011, when the Witcher 2 showed me that it was no longer sufficient :cry: .. so I waited, and I waited and built a small home server in 2016 ("Live" Debian-based Linux system), but still no real gaming. Last year when the *Reverb G2 (VR headset) arrived I decided it was time - so I purchased it and then picked-up most of the pc-components (AMD Ryzen 5600 build) EXCEPT THE VIDEO CARD - which I still haven't gotten yet (because I won't pay scalper's pricing). So I'm still sort of stuck in limbo. :cry:


However the additional year + of added time is allowing me to do something a bit different with the system design:

I'm designing it for bare-metal booting virtual machines (..and "differenced" ones at that), so that each game effectively has its own separate operating system. Doing it this way allows for particular settings for each game (that don't have to be changed), a build with fewer services in the background (effectively only what each particular game requires and minimized services for the OS), no real OS degradation, very little chance of corruption (..the parent-differencing disks are read-only), along with the ability to easily save each game VM to the server (in case that "top" child VM that houses the game gets corrupted).

*I've also decided to ditch the monitor in favor of either a single large or multiple monitors - all in VR.

[Scottg]

Hey, thanks for the suggestions, Scott! I thought the AWG10 bad boys from Solen were pretty over the top- cost more than the woofers did for the 4.3mH and 2.2mH!

You're Welcome!

Yeah, good crossover parts can often be more expensive than the drivers. :cry:

It's ironic that you can purchase some really good moderately priced drivers (..like the 8 ohm NE180W), but to get the best from them - wind-up spending double or triple their price in passive crossover parts. ops:

Still, if the design can handle the added bit of resistance - those Solen foils are (comparatively) a good value. The Jantzen waxed 12 awg's are sometimes available at reduced pricing as well, and though still more expensive than the Solen's, are also a pretty decent value.

I guess you could always DIY the coils though and save more - at least for a "litz" design. (DATS is good for measuring this easily.)

[tktran]

Your Welcome!

Yeah, good crossover parts can often be more expensive than the drivers. :cry:

It's ironic that you can purchase some really good moderately priced drivers (..like the 8 ohm NE180W), but to get the best from them - wind-up spending double or triple their price in passive crossover parts. ops:

Still, if the design can handle the added bit of resistance - those Solen foils are (comparatively) a good value. The Jantzen waxed 12 awg's are sometimes available at reduced pricing as well, and though still more expensive than the Solen's, are also a pretty decent value.

I guess you could always DIY the coils though and save more - at least for a "litz" design. (DATS is good for measuring this easily.)

Btw, and trying not to get the thread too off topic, what's less good about the 4 ohm NE180W?

I like the NE123W-08, btw, fantastic little midrange...

[Scottg]

It could be down to sample or measurement differences, but the the 4 ohm is a little less linear overall, and its non-linear is a bit higher at lower freq.s and a bit higher overall with 5th order. I also think it's marginally less efficient (factoring-in a typical -3db from an 8ohm to a 4 ohm, where the 8 ohm here doesn't seem much lower than a decibel overall).

While it doesn't have the non-linear performance of the Purifi mid, it's still pretty impressive overall and it has substantially lower initial mechanical resistance from the spider and surround (though to get the best of it requires a low-loss cabinet interior).

[JonMarsh]

You're Welcome!

Yeah, good crossover parts can often be more expensive than the drivers. :cry:

It's ironic that you can purchase some really good moderately priced drivers (..like the 8 ohm NE180W), but to get the best from them - wind-up spending double or triple their price in passive crossover parts. ops:

Still, if the design can handle the added bit of resistance - those Solen foils are (comparatively) a good value. The Jantzen waxed 12 awg's are sometimes available at reduced pricing as well, and though still more expensive than the Solen's, are also a pretty decent value.

I guess you could always DIY the coils though and save more - at least for a "litz" design. (DATS is good for measuring this easily.)

Actually, I used to wind all my own coils, but left the coil winder and LCR meter at Twaale.

And since then the base price of copper has gone up a lot. So for now, I just have to pay the price like everyone.

OTOH, I still have an LCR meter, and do 100% incoming inspection on crossover parts, just like the real speaker companies do. Sleep better at night that way...

[JonMarsh]

Santa Solen & UPS delivered early!

It's nice and very surprising to get early delivery on something this time of year...

UPS was promising Thursday delivery on the inductors from Solen in Quebec... Uh, I checked my calendar App three times just to make sure this REALLY was just Wednesday!


So, plans for the afternoon changed, and the coils were swapped out- this version of the LF board had been laid out to accommodate up to the AWG10 inductors from Solen.





Of course, big coils may give you some kind of ersatz bragging rights, but I was more interested in measurements, and some listening, at least what I could do for today.

And yes, the resistance drop makes a difference...





Distortion is obviously dominated by the drivers, and a more complete investigation is justified, soon, including a comparison with the big Erse cored inductors- I have a 2.2 on hand, and another larger one I can unwind to create a 4.3mH. Possibly tomorrow!




Some other checks were done, at measurement axis parallel to the floor for the midrange and for the tweeter.

It still looks like trimming down the 800-1100 Hz level a couple of dB is advisable.

Things look pretty good- more data if all goes well tomorrow. Got other things on the task list.

Like I say, I never dreamed retirement would be so much work! And there's the Osiris top module to finish building, too!

[JonMarsh]

Basic inductor tests

One other thought to share at this point- a simple diagnostic regarding inductors that anyone with even a basic inexpensive LCR meter like a B&K 879B can perform-

• Test your incoming inductors- check to see if the accuracy spec is met- this of course assumes that your meter is reasonably accurate, and preferably has been calibrated against a known standard. But there are ways to work around that, and learn more
• Use one high quality high precision part from a reputable vendor as a reference standard- say, a capacitor in the 1 to 4.7uF range, and an inductor of 1mH.
• Measure the part with several test frequencies- for small inductors, and small caps, higher test frequencies may give more accurate results.
• But, compare inductors against their spec at two very "standard" frequencies- 1kHz and 10kHz.
• You will find that air core inductors will measure quite consistently, regardless of testing at 1kHz or 10kHz.
• Solid core inductors I test at 120Hz, 1kHz, and 10kHz. You may be surprised at the differences you get based on test frequency. This is an aspect of nonlinear core behavior and core losses.
• The closer a part behaves like an air core inductor, the better the core material behavior.

[JonMarsh]

Adjusting the midrange balance and reviewing far field testing

Per my last comment, the were two things I had in mind, after reviewing measured data at 1M and also after a listening session...

• Provide some inline equalization to reduce the midrange level from about 700Hz to 1200 Hz by a few dB. This was based on both the last measurement, which still showed a slight elevatgion in that area, and my subjective listening experience, especially at 10 ft off more (note that the room is very live in the midrange, so I had some concerns about putting too much emphasis on the tone at 20 ft ore more.)
• Conduct addiitonal measurements at what could be considered more like farfield- experience has shown that large cabinets with large baffles may reuqire 3x to 4x the baffle width to reach nominal far field response.

This is the updated simulation schematic with the LCR EQ network comprised of R24, C30, and L12. This was considered just a first cut, because after all, we're looking at a fine tuning sort of situation, and estimating the Q needed as well as the center frequency. Furthermore, these values were assembled from parts I had on hand! The degree of depth of the dip can be adjusted just by changing the value of R24.




This graph from VituixCAD shows the updated crossover filter transfer functions.





Measuring at 1 meter, it might appear that the dip EQ has overshot the mark. However, I suspected the more complete story would be told with a measurement at something like 2.5 meters.


1 Meter





2.5 Meter





Some additional tests were done at several listening heights- at 2.5 meters, the behavior was quite consistent between the nominal seating listening height (about 40") and at the midrange driver or Tweeter height. That was not expected, but appreciated.

I only had time for listening to a few cuts this evening before switching to other tasks, but I will follow up tomorrow. Initial results are satisfying.

[Scottg]

3-4 kHz looks a little "hot", but above that looks pretty good in-room. :T (..which is often a pain to get anything close to linear.) :B

[Zvu]

Actually, I used to wind all my own coils, but left the coil winder and LCR meter at Twaale.

And since then the base price of copper has gone up a lot. So for now, I just have to pay the price like everyone.

OTOH, I still have an LCR meter, and do 100% incoming inspection on crossover parts, just like the real speaker companies do. Sleep better at night that way...

I don't know about price of copper in US but in Europe it goes for around 10 euros/kg. I have a store at about 500m from my house so it's quite convenient.

I just have to find a good winder that can wind 2mm wire thickness.

[JonMarsh]

3-4 kHz looks a little "hot", but above that looks pretty good in-room. :T (..which is often a pain to get anything close to linear.) :B

This is something I noted, also, and is pretty easy to address.... it's sort of like closing in step by step on the target, and evaluating and documenting as I go along.

And realistically, it all ties in with the basic concept:





Now, what would have been helpful and interesting (to compare) would be if I'd had more time to document things as a went along in 2012 and 2013... but they were very eventful and busy years... on the work side, the immediate "family" side (GF especially), and trying to get this project off the ground. plus launch some Class D efforts with Hypex modules.

Interestingly, even with the first version, due both to the configuration and the Accuton C173-6-090 I see a midrange "bump" with the measurements I was using then.

For now, I'm happy with how the process of steady refinement is going. For a two way system, in a typical bookshelf enclosure, nailing things in one or two passes is pretty standard these days.

Something like this is a bit different, but we're getting there. Some day I'm going to get the hang of this...

I did go out to the storage units on Wednesday, thinking about pulling stuff out so I could get to the other cabinet at the back, but the work involved would be fairly ludicrous, so stereo will have to wait until after the move into "permanent" digs.

[JonMarsh]

Progress is our most important product...

OK, having a quiet Christmas Eve messaging a number of friends, but not ignoring the oppportunity here-

So, a very minimal change to the tweeter crossover, adding a 12.5 ohm resistor in parallel with the output shunt (why 12.5? It's a value I had, and simulated well in VituixCAD...)

So, once again, we'll look at 2.5 meter and 1 meter.


2.5 meter response 100msec window

1/24th octave smoothing



Room reflections are evident in the response plot with this relatively minimal smoothing and moderate window


1/6 octave smoothing




1/6 octave smoothing makes the trend line more apparent.



1 meter response 100 sec window


1/24th octave smoothing



At 1 meter, room reflections are at lower relative level

1/6 octave smoothing




Again, 1/6 octave smoothing clarifies trendiness for tonal balance.


In listening, this was pretty satisfying- nice overall balance, good clarity, very good pitch and transient definition in the bass- particularly the upper bass. These Daytons aren't bad at all!

. :T


It's at a point where more listening with something better than a MacBook and Motus interface is needed. But I think this is pretty close- probably need to have these in a more representative room, too, with carpet and my normal wall panels.

This is a pretty nice addition to my Christmas Eve experience!

I hope everyone is having a marvelous holiday!

[chrisn]

In room response looks great Jon

Merry Christmas

[Scottg]

2.5 meter response 100msec window

1/24th octave smoothing

[ATTACH=CONFIG]31638[/ATTACH]

Room reflections are evident in the response plot with this relatively minimal smoothing and moderate window

Yup, that's the Money Shot! :yesnod:

[draki]

Something I noticed and was trying to look up for an explanation in the forum but couldn't, hence: the values of the Y (height) coordinates in the crossover are zero for all drivers.

Why is this? Clearly the physical layout is different. I am genuinely intrigued and as I said I tried to find out if this was addressed/explained elsewhere. Sorry if it was and I missed it.

[tktran]

Yes the In-room response is very good.

I wonder what the dip between 100-200Hz is caused by? Is the the floor bounce, ceiling bounce, side wall bounce or front wall bounce (or combination)

Also, what’s the combined volume of the RSS315HF8A? Is that a sensitivity of 90dB/2.83V anechoic? They are flat to 50Hz, which is very good for a sealed unequalised design.

I find it a challenge to get the levels right in a 3 way.
What I like vs what measures flat. And left to my own devices I don’t seem to mind a boost [i]below[/] 100Hz, But I really detest a 100Hz boost, which sounds just too “boomy”

[JonMarsh]

It's a boundary mode from rear wall in what is nominally the dining area. Things were just setup for where there was space for now, in this rental. Normally I would be doing what most of the world calls a Cardas setup.

Today I will break out some good midrange electronics for listening- Tascam CD transport with AES/EBU output, Denafrips Terminator DAC, Halcro DM10 preamp, and Cambridge Edge W amplifier.

Of course, it still just be "Living Mono", his master's voice.

[Amplifier dude]

Beautiful build! I can just imagine how they would sound fully active!

[Zvu]

..................

I find it a challenge to get the levels right in a 3 way.
What I like vs what measures flat. And left to my own devices I don’t seem to mind a boost [i]below[/] 100Hz, But I really detest a 100Hz boost, which sounds just too “boomy”

Do you live in a brick-concrete house or in wooden house ?

[tktran]

I travel a bit, even times of covid, which I consider very fortunate to be able to do.

One residence is “double brick”. It’s a standard in that area of the world where there is lot of sand and limestone, so all houses are made with two layers of bricks for external walls, and one layer for internal walls.

Another is brick veneer, where the construction is mainly timber, with one layer of bricks on the external facing walls. All internal rooms are divided by drywall, I think is the northern hemisphere term, or plasterboard.

The third is a timber frameds and timber exterior too; with drywall interior partitions.

The walls seem make a difference in the bass (real or perceived) What’s been your experience?

[Zvu]

Quite real differences with different types of walls since thin ones absorb low energy very effectively. It is known way of treating room modes at low frequencies when adding more damping material isn't practical. As long as the inner walls can flex, they can and will absorb a part of energy. You probably enjoy the same amount of bass but you have to tweak it going from one place to another. Long live the EQ

[JonMarsh]

OK, I finally dragged out and setup some decent electronics- of course, some might say I'm cheating here a bit, too, because it's so ridiculously easy to setup a Cambridge Audio Edge W for bi-wire operation, using the input/output pass through capability and a short signal jumper cable!

While out at the storage units before Christmas I did come across my Monster Power controller/distribution unit (and some other fun stuff, like my basic small collection of test CDs!) so I brought that home to make hooking stuff up easier...





The setup is fairly simple here, no network music system, nothing but a CD player, DAC, preamp and power amp. No HiRes tunes, but for checking tonal balance and all the basics, CD's with a good DAC will get the job done. It's on a fairly tall stack of bins because I don't have another shelf in the house at this point, and it's a good working height while standing.





The simple setup minimizes the cable count and speeds plugging everything in- heck, it's been a while, what with all the moving!




So, ready to roll!





I figured I would listen to a cut or two off of each CD just to get a varied reference... some Hiromi (Another Mind, Spiral), some Best of Fourplay, some Stanley Clarke (Bass-ic Collection), some Sonny Landreth (Outward Bound), some Alice Gomez (Flute Dreams, Obsidian Butterfly), some Michael Hedges and Andreas Vollenweider (White Winds), and a little Keb Mo.

well, I screwed up... I put the Keb Mo "Slow Down" on first, and that's just what I did... slow down... and listen.... and listen some more.

Yeah...

:T

The rest can wait.

[ergo]

Happy to see such a good progress for the end of 2021! Many reasons to be envious, both for the stuff on pics as well as what's happening
But I must say your "equipment rack" does make me nervous to look at. I'm sure it is actually much more sturdy in real life.

[Steve Manning]

Happy to see such a good progress for the end of 2021! Many reasons to be envious, both for the stuff on pics as well as what's happening
But I must say your "equipment rack" does make me nervous to look at. I'm sure it is actually much more sturdy in real life.

I'm with you Ergo ..... I did a pucker seeing that Halcro sitting perched on that tower of storage boxes. 8O

[JonMarsh]

If you knew the way those boxes interlock together for stacking, you'd likely be a lot less concerned...

These are my go-to solution for organizing projects, storage, and moving. Never any guesses about what's inside, support very high weights (I could stack as high as line signal stuff and put the Edge W amp up their with no worries), and very durable- much more heavyweight than other storage boxes. These are the Mark Levinson and Magico of storage bins.

Come from England... Staples or Office Max are some of the best places to buy.

https://www.reallyusefulproducts.co.uk/usa/

Fourplay is on now, for my indoor walk.

[JonMarsh]

Best of Fourplay

Ok, I'm cheating again.

But one of my favorite sayings at work used to be, "If you're not "cheating" (thinking outside the box) you're not trying hard enough!"




How am I cheating?

This is a Hybrid SACD disk- it has a CD layer made from same master (*usually) as the SACD layer, just down converted.

Anyone that knows anything about the early SACD releases knows that Sony/CBS went back and remastered all the releases from the original tapes. So, those that are available in Hybrid SACD format know that these are the best CD releases available, in most cases.

This is not a Sony release, but a 3rd party effort, and like the other 3rd party efforts I know of, also went back to the original source material in order to put their best foot forward.


Now, for those SACD releases that are not Hybrid, well, there are also ways to "cheat". I have used special boards from AudioPraise in the Czech Republic to retrofit Pioneer multi-format players to output DSD converted to high resolution PCM on an SPIF data stream, and I have new boards to retrofit an Oppo 203 player, plus the player, in storage, to have multi-channel SACD output.

Now, these SPDIF data streams have the copy bit set, but with the right hardware and software (as I have for my Mac's) one can import the PCIM data stream into my DAW and record to disk. From there, I convert to the resolution and filter setting for down conversion I prefer- usually I have the PCM output at 176kHz, and down convert to 88kHz, which works well with the DCD-8 Brainstorm for re-clocking. I also down convert all the way to CD for using with portable music players and Apple Play. Think, a 44kHz version of Elton Jon's TumbleWeed Connection sourced from SACD, on my iPad or iPhone. Of course, I bought and own the original SACD disk, so there's nothing hinky here going on.

PM me if you're curious about the DAW tools I've used in the past- they may likely get updated. Yeah, all Mac stuff. One has changed hands to another company. My Audio capture interfaces are RME Fireface, Firewire interface to the Mac; these days, you have to use a firewire to USBC converter for that.

I haven't even assembled this most recent configuration yet, it's another little bit of deferred gratification for retirement, which if I hadn't bought the items then, I wouldn't be able to get them now. I had this sense about many things I've stashed for my retirement. Limited edition guitars, for example. Portable electronic Hammon clone, briefly made in the USA for an Italian company.

You snooze, you lose.

https://audiopraise.com

Anyway, regarding the Best of Fourplay on the Isiris...

• Very articulate and percussive reproduction of bass lines
• Percussion instruments very clean.
• Delightfully transparent midrange, on all instruments and voices. The PuriFi midrange "Excels" (bad pun) and blends beautifully with the BlieSMa tweeter- you can't tell where the crossover it at all- sounds like no crossover.
• Imaging? Who knows? This is Living Mono, his master's voice!

[tktran]

Yes sometimes I also take a break and do some listening, and then it's either too much listening, or quick, this doesn't sound right, back to the drawing board.

What are the pro/cons of having the dual woofers close to the midrange, vs closer to the ground. Or having one close to the ground and one close to the midrange?

When you design the crossover, do you take into account the floor bounce from the woofers? Or do you go anechoic and accept whatever bass boost you'll get from the room? (long live EQ, as Zvu said)

[JonMarsh]

Well, it all depends...

Yes sometimes I also take a break and do some listening, and then it's either too much listening, or quick, this doesn't sound right, back to the drawing board.

What are the pro/cons of having the dual woofers close to the midrange, vs closer to the ground. Or having one close to the ground and one close to the midrange?

When you design the crossover, do you take into account the floor bounce from the woofers? Or do you go anechoic and accept whatever bass boost you'll get from the room? (long live EQ, as Zvu said)

First, let me say that I often suggest to those that PM me on this topic, to get some Denovo modular MDF knock down cabinets and to experiment with configurations and do measurements themselves- assuming they have some basic measurement equipment that is capable of gated measurements with variable length windows and window filters.

Second, I can relate a bit about my own experience and evolution in thinking...

In the late 70's, I built systems with a sort of classic Allison configuration- woofer(s) at the floor, and columns of mid woofer, dome midrange, and tweeter above that, generally dome mid on the seated listening axis. There were pretty much no mid woofers that went high enough for what the day's best tweeters could handle on the low end. A heavy faceted baffle system like that with two 10" woofers on the floor and a narrower baffle width with bevels going up is a configuration I gave to Charlie Hansen in the mid 80's because they didn't meet my new wife's acceptance factor.

I've experimented with variations on things since then- in the late 90's, building a Wilson clone of sorts, and discovering for myself the desirability of elevating a dual woofer array...






And after the first gen Ardent build, before the Isiris or Wavecor Ardent, experimented with a vertical approach with one woofer near the floor, another higher just below the midrange... using a special variant of Scanspeak woofers and the Accuton C173-6-090 midrange, with a tweeter and waveguide arrangement derived from the Modula MT XE.







Guess what? This system had floor bounce issues... possibly due to the proximity to the floor! (This is not how I measured the system overall; measurement checks for drivers for crossover were done at 1meter and 2.5 meters)


But, you'll likely not guess what solved them- and reinforced my experience with the Wilson clone and the first Ardent.





Here's my report on the modified configuration lifted from the original thread:

OK, while I was talking to ThomasW earlier this AM, I stacked up my Modula Xtremes on a new pair of PE boxes, just to see how the room interface for LF would be with the revised woofer positioning, and the imaging and overall balance.

My brief listening impression was that the midbass was a bit smoother, and that we'd picked up a third to half octave in the bottom end. Now, as they were previously nearly boundary loaded on one woofer, I did not expect there would be an improvement in LF response picking the whole thing up off the floor. Seemed like a classic case of placebo effect- you hear what you wish you would hear.

The microphone doesn't lie, though, at least not if you treat it right, so that was next.

This is a 2 meter measurement, gated at about 160 msec, in room. Damn, if it doesn't look like the midbass is smoother and we picked up a third octave in the bottom end- with no other modifications.

Am I lucky or what? Like I said, Jed must enjoy me watching me walk around stumbling in the dark, barking my shins now and then and also coming on a little gold at other times.

Well, it seems to me we've at least got a valid argument for using a pair of the 1 cu ft boxes as speaker stands... but I think doing a full addon with the crossovers relocated will be in the cards now. Sheesh, if I'd come up with this response curve just by doing the full mod, I'd be happy, but now I suspect there may still be a little more performance to wring out of these with making more of the woofer modules available to the woofers.

Size is getting a little out of control; they're about 59" tall, only 1-1/4" short of an Isis, for comparison. Don't look anywhere near as nice as an Isis, but they're not $60K, either.

Well, and you can see what that led to in the end...

[JonMarsh]

And for today's agenda...

Anyone remember The Crusaders? Show of hands? Please?


OK, just one more thing that dates me... we used to use cuts from "Chain Reaction" as demo material in the B&O room at the Boulder Sound Gallery. Let's not discuss how long ago that was... :W

I might have that CD stashed away somewhere in storage... but!


Imagine what I came across last week!!





Well, you can't really say this is something "new under the sun"- but a five CD collection for the price of one CD isn't a bad way to end the year! And it may be a fun trip down memory lane.

And "Made in the Czech Republic"? Curious... I've gotten some cool stuff from the Czech Republic (AudioPraise), and Prague is my favorite city to visit in continental Europe, after Munich, but still, curious as a source of CD's!

[JonMarsh]

Sunday afternoon, New Years Day, just finishing a late lunch, and spinning "Images" from the Crusaders. Very nice music, and well recorded. if you're into this sort of jazz flavored R&B that they composed and play, this 5 albums for the price of one is a heck of a deal!

https://en.wikipedia.org/wiki/The_Crusaders_(jazz_fusion_group)

This kind of dates me in a different way, that their active career spans from the year after I was born until 2010. (that's 1952!). Images dates from 1978.

45 albums all in all...

[JonMarsh]

Another step in voicing

A few more days with sporadic opportunities to listen, often while having to be doing something else...

But thinking about this in the wee hours this morning, I figured I may not have gone quite far enough... been quite ruthless enough...

This "great room" is very reverberant, but seems to suck up bass and lower midrange. So... why not try adding another shunt resistor in parallel, for both the midrange and tweeter, and bring their level down just a bit more...

Exhibit A, updated schematic with an 8.2 ohms added to the shunt network on the midrange, and one 12 ohm resistor added to the tweeter shunt resistor network.




The updated estimated response profile.




No measurements taken yet, just doing some listening, starting off with Fourplay.

No one is going to call this speaker bass heavy, but for me, this brings the low midrange and upper midrange into better balance. And it sounds better anyway at the opposite end of the house!

I think it is finally to where I won't change anything until I have the 2nd speaker dug out from the back of storage (probably not until after the move, but who knows, I may get desperate!) and another crossover set built, I won't make any more updates.

But now I'll put some more Crusaders on, and some Michael Hedges!

[JonMarsh]

For any of you that might be "Crusader Curious", let me suggest after listening to this collection, I'd suggest trying "Images" - nice compositions, and dynamic punchy recording quality.


There's one that has some really nice guitar work, and I've got to go back and figure out which one sometime soon.. but so many other things on the plate, it's Russian Christmas today, and I'm sending out hand me downs today, a 12.9" fairly late model iPad Pro and two 4K Blu Ray players.

And there's some wood cutting to do this afternoon on the Osiris test cab.

[Mr.Ed]

Thanks for the recommendation Jon. My copy should be here today. Looking forward to it.
-Ed

[Finleyville]

This "great room" is very reverberant, but seems to suck up bass and lower midrange. So... why not try adding another shunt resistor in parallel, for both the midrange and tweeter, and bring their level down just a bit more...

Exhibit A, updated schematic with an 8.2 ohms added to the shunt network on the midrange, and one 12 ohm resistor added to the tweeter shunt resistor network.



No one is going to call this speaker bass-heavy, but for me, this brings the low midrange and upper midrange into better balance. And it sounds better anyway at the opposite end of the house!

I think it is finally to where I won't change anything until I have the 2nd speaker dug out from the back of storage (probably not until after the move, but who knows, I may get desperate!) and another crossover set built, I won't make any more updates.

You are making me very excited to start creating this design.

Do you currently plan on leaving the shunt resistor networks on the tweeter and mid as shown or will you combine those networks into a respective larger value for simplification? The idea of fine-tuning the levels of the tweet and mid for my specific listening room is too good to pass up. My intended listening space will be a 28'L x 18'W x 7.5'H basement. One long side wall and far short rear wall is drywall over concrete block. The other two are standard drywall and stud framing. It is all storage right now but once I set it up I am really interested in how the room acoustics will pan out.

Plus, I am very excited that you, Jon, could not hear the presence of the mid to tweet xover!

The only thing I am concerned about looking at your last measurement is the general, very slight downward tilt to the otherwise awesome flat response. While I am not a fan of speakers with a lot of high-frequency energy, i.e. Focal, I do need some very-high-freq presence that "tickles your eardrums" when those tiny cymbals crash or during electronic songs that have a type of crackling effects going on. Do you feel that this "tickling effect" is lacking with this current design?

[JonMarsh]

What I've been thinking about is keeping a switched resistor bank on the shunt for each, after I go over it one more time - it works very well over a certain range of adjustment, and I would choose resistor values to give consistent small steps. No reason not to do this for both midrange and tweeter. I wouldn't usually consider this, but this room has acoustics that are, shall we say, a bit different than what I've encountered personally in a home environment before.

Now, regarding the tweeter, I'd refer you to the voicing thread, as well as many textbooks on stereo speaker voicing, and the assumption for studio monitors of having a soft roll off in the top, as well as the psycho acoustics. These don't sound soft to me at all the top, and part of that is due to the power response that results from that baffle shape.

However, I'd also point out that I'm actually using EQ in the Tweeter crossover filter to shape the response and RAISE it with increasing frequency, as the T34 (either version) has a tendency to peak at about 2kHz and roll above that.

Note the filter response for the tweeter network:




So, between 4kHz and 15kHz we've actually got 2dB of lift.


I may look into seeing if I can come up with a way to make that adjustable too, but probably just two positions- one being the way it is, because it sounds clean and extended but not like a boom-tinkle box, if you know what I mean.