@wolf_teeth yup, and I use them both almost every day. I have alot of live concert LDs and tapes which I made in the 80s and 90s. Most are chrome and metal tapes, so they held up very well over the years, unlike me...
@ralphs99 These Parasounds idle at 300W each. They're room heaters for sure.
The active analog crossover for my large 3.5 ways is under development. I'm using Jensen I/O transformers which have very good linearity from under 5hz to well above 80k. I bought the rack enclosure from Penn fab. After alot of deliberation I've decided to use Signetics NE5532ANs for the buffers. They're quite good compared to many other op amps if used in their happy place of input biasing. The crossover for the CB3010-N8, M74A and T34B will be passive. Only the main HP and LP between subs and 3 ways is active. Depending on component choices and linearity of the 10s, I may also do a separate HP/LP between the 10s and mid/tweeters.
The active analog crossover for my large 3.5 ways is under development. I'm using Jensen I/O transformers which have very good linearity from under 5hz to well above 80k. I bought the rack enclosure from Penn fab. After alot of deliberation I've decided to use Signetics NE5532ANs for the buffers. They're quite good compared to many other op amps if used in their happy place of input biasing. The crossover for the CB3010-N8, M74A and T34B will be passive. Only the main HP and LP between subs and 3 ways is active. Depending on component choices and linearity of the 10s, I may also do a separate HP/LP between the 10s and mid/tweeters.
Last edited:
Regarding the D7608s, some of you are already finding out the complex dependency of rear chamber volume and lower mid performance. This is where the issues reside, as there is a strong correlation between chamber volume and lower mid HD/FR.
When comparing the wonky factory FR curve and corrected FR with any type of chamber between 350 - 750 ml, its apparent the D7608 was never intended to be used without some type of rear chamber. Adding to that is the felt on the motor structure, where things get even more weird. There is definitely some performance to be gained by optimizing this air flow restriction and the net chamber volume. The question is, how much snd what type of felt is needed for a given net chamber volume. The relationship I discovered is the felr should only be modified if the target lower mid HP is below 500 hz. Above this, the felt is pretty close to optimal for larger chamber volume. I'm honestly not sure what material they used for the rear felt, but it does appear to be some type of mix between synthetic and natural fibers.
After having experience with PMC IB1s, I've always felt there was more to be had from the mids than what they were getting. The main limitations are in the D7608 motor. It really needs more BL to control the diaphragm better and have a lower Qes for easier integration. Having tried to add magnets to the motor quickly revealed an already saturated gap, not further helping things. Perhaps a WG would be a better solution, but that requires much more experimentation.
For larger baffle widths, there is a bump in the D7608s FR tapering down to about 2.3 - 2.5k. Linearizing this is priority, but there's some free output there to offset some of the HD rise when crossing lower and asking for more output.
When comparing the wonky factory FR curve and corrected FR with any type of chamber between 350 - 750 ml, its apparent the D7608 was never intended to be used without some type of rear chamber. Adding to that is the felt on the motor structure, where things get even more weird. There is definitely some performance to be gained by optimizing this air flow restriction and the net chamber volume. The question is, how much snd what type of felt is needed for a given net chamber volume. The relationship I discovered is the felr should only be modified if the target lower mid HP is below 500 hz. Above this, the felt is pretty close to optimal for larger chamber volume. I'm honestly not sure what material they used for the rear felt, but it does appear to be some type of mix between synthetic and natural fibers.
After having experience with PMC IB1s, I've always felt there was more to be had from the mids than what they were getting. The main limitations are in the D7608 motor. It really needs more BL to control the diaphragm better and have a lower Qes for easier integration. Having tried to add magnets to the motor quickly revealed an already saturated gap, not further helping things. Perhaps a WG would be a better solution, but that requires much more experimentation.
For larger baffle widths, there is a bump in the D7608s FR tapering down to about 2.3 - 2.5k. Linearizing this is priority, but there's some free output there to offset some of the HD rise when crossing lower and asking for more output.
Really lowed all about John Curl´s Halo-series, like the JC-1 monos and a A51 , C1 etc.
Had a A21, A51 and the C1 preamp 2003. real powerful and good sounding gear 👍
2005 i shifted to Pass labs X350 mostly for Pass labs fantastic midrange.
@jawen I came from a Pass X250. It sounded wonderful in the mids on up, but left me needing more for my purposes. I was driving an older pair of Apogee Duets, which were just power hungry and had horrible impedance dips. I also have a JC5 offline which I'll be using for the top end. Its a great all around power house that never runs out of steam. The A21s are very good, but have a slight current deficit compared to the JC5. The shining star is however the JC2. Its a phenomenal preamp.
I tried a X250 first but it was not powerful enough at all for my Infinity IRS Sigma´s
The X350 was a completely different beast which felt twice as powerful with a tube-like wonderful midrange.
I later tested 7 strong different amp´s at home with a hifi friend, and all had small characteristics differences.
But all in all Krell 450-mxc monos won, no other amp could control the bass cone's micro movements just as well no matter how loud you play. ( 82 kilo McIntosh MC602 and Krell FPB 600 was the heaviest )
Pity that it hardly exist any Parasound stuff in Sweden, i really would have liked to own a Parasound setup today.
And i think they are beautyful also!
I find the D7608 reveals bad amplification and source more than alot of other mids. It allows fine details to be emphasized in the harmonic balance, specifically on piano and female vocals. The dome composition delivers a relatively high Qms. If you overdampen the chamber too much, it hurts the low level dynamics. In contrast, the M74A and M74B sound more clinical at lower volume levels. The M74S is better in this regard, but still not as open in the lower mids. I suspect their paper dome version M74P to be better this way.
Hi profiguy after a while, all best for you from here.. This dome thread have been golden!
Have i understood right that in low volume d7608 performs better, than bliesma's silk M74S?
Yes and its not getting any easyer with age hahaha 😉
Probably why most DIY guys end up building bookshelf speakers
Have you listening on the "old" JC-1 monos?
Seems like a real powerhorse!
400 watts RMS @8 O
800 watts RMS @ 4 O
1200 watts RMS @2 O
135 amperes peak
Bliesma is a real interesting mid dome, but very exensive in Sweden.
@jawen. Yes, based on my experience I feel that the D7608 has better low level resolution than most other 3" cloth domes. This includes the M74A/B and even the ATC SM75-150S. While the others can rip off your face, the D7608 has a level of detail in vocals and piano which reveal the small underlying information better. I'm pretty sure this is a result of the dome material and magnet system. Its a beautiful sounding mid all around. A pair of these in staggered filter arrangement would be impressive. This is what I wanted to build with a pair of NE315Ws, NE180W and T35C002 tweeters per side.
In your own opinion.
Can any of this smaller dome´s "beat" a 6,5 inch "ordenary" high quality cone midrange?
Say like a SS Relevator 18M ?
And the criterion is also real high SPL but still without "for the ears sensitive" distortion.
@tf1216 Thats an odd one. Can't make it out. I'm sure there are other Asian companies now that make custom drivers like this for OEMs. Would love to hear those.
@jawen Well, yes, that is my opinion and it stems from extensive listening to many drivers like this. How they integrate is likely more important than the overall performance ie. HD, FR linearity, etc. The D7608 has been around a long time with several small changes. The current guise is excellent based on dome construction and suspension. It can play loud if crossed correctly with the appropriate chamber / dampening.
The ATC and Bliesma mids are excellent, no doubt. Hence my purchase of M74As and T34B to keep the harmony between the two drivers. I've heard quite a few 3 way systems with large domes. The thing you'll always make out is the integration of LF and mid. The shift in directivity is obvious with many of them, moreso when the crossover is too low between them. You'd need a special 12" that can hang with the large mid dome. In the case of the ATC SCM50, it has too low of a LF- mid crossover point. The dome strains at higher levels more than what's comfortable to the ears. Their crossover is in the mid 300s Hz, which is IMO too low. You can hear the dome over the other drivers, including the separation in the LF - mid location. Yes, ir plays stupidly loud, but it has more glare in the mids that I'd like. Again, my own assessment. The SCM 25 in contrast is wonderful.
I wouldn't have purchase a bunch of Bliesma drivers if I felt they were overall inferior to other $100+ 3" mid domes. These Bliesmas can go very loud, but they need careful handling in the upper mids to avoid aggressive sounding upper midrange. Some interpret this type of sound as detailed or analytical. The problem here is how the mid Fs is suppressed without being that aggressive.
@jawen Well, yes, that is my opinion and it stems from extensive listening to many drivers like this. How they integrate is likely more important than the overall performance ie. HD, FR linearity, etc. The D7608 has been around a long time with several small changes. The current guise is excellent based on dome construction and suspension. It can play loud if crossed correctly with the appropriate chamber / dampening.
The ATC and Bliesma mids are excellent, no doubt. Hence my purchase of M74As and T34B to keep the harmony between the two drivers. I've heard quite a few 3 way systems with large domes. The thing you'll always make out is the integration of LF and mid. The shift in directivity is obvious with many of them, moreso when the crossover is too low between them. You'd need a special 12" that can hang with the large mid dome. In the case of the ATC SCM50, it has too low of a LF- mid crossover point. The dome strains at higher levels more than what's comfortable to the ears. Their crossover is in the mid 300s Hz, which is IMO too low. You can hear the dome over the other drivers, including the separation in the LF - mid location. Yes, ir plays stupidly loud, but it has more glare in the mids that I'd like. Again, my own assessment. The SCM 25 in contrast is wonderful.
I wouldn't have purchase a bunch of Bliesma drivers if I felt they were overall inferior to other $100+ 3" mid domes. These Bliesmas can go very loud, but they need careful handling in the upper mids to avoid aggressive sounding upper midrange. Some interpret this type of sound as detailed or analytical. The problem here is how the mid Fs is suppressed without being that aggressive.
Thanks for an detailed answer profiguy 🙂
In your opinion, is 600- 4000 hz a good "area" for theBliesma M74a ?
Even 500 hz, 450 ?
@jawen Sorry if I sounded too harsh on my last reply. I didn't mean to direct the answer at you like that.
The M74A can easily cross at 500 3rd order, even 2nd order if the levels don't get crazy. The VC underhang is more than 3mm p-p which allows for this. In fact I've seen dented domes on the M74A when they were played too loud and too low. There's easily 4-5 mm distance between dome and grille on those, so its capable of that much x- mech (at unknown distortion levels of course).
My target filters for the M74A are 500 - 3.5k LR2 using 2 domes and staggered filtering. Thats conservative IMO for this driver and you'd likely be able to go down to 375 hz LR4 with the right baffle step compensation and use of an LCR at Fs. I'm just not a fan of single suspension drivers (without a spider) being pushed hard. You can run into radial instability this way depending on how far the VC has to go for a given output level. I wouldn't go lower than 500 LR2 HP on a single mid. The LP is a little more flexible. 4k is no problem but you'll start running into directivity discrepancies between mid and HF if the tweeter diameter is smaller than 28mm. The baffle width also dictates this. I would recommend a 3.5k 2nd order LP on the M74A and also filter out the 10k jackhammer up top. That peak is nasty. It really needs an LCR even at reasonably low LP points, otherwise you may see down modulated odd order HD components in the upper mids. These are very hard to distinguish from typical elevated HD up top from breakup.
You asked about the SS 18M compared to the D7608. Thats a tough cone mid to beat and it happens to be one of my favorites. The major advantage is being able to cross it low. 150 hz LR2 isn't an issue with a 12" woofer. The top end is well controlled in terms of breakup and rolloff. It will sound more veiled than the D7608 or most other large domes, mainly due to earlier cone breakup onset, but also due to the use of cone dampening in the upper rolloff area. The multiple peaks up top are well suppressed but will ruin detail. I wouldn't LP that one higher than 2.5k. Its a very good sounding midbass driver and thats its strength. I'd say it compares well with much more expensive cone mids.
A very important thing with the M74A is suppressing the primary Fs peak with lower crossover points. This avoids most of the overexcursion issues when pushed harder. Its definitely recommended with HP < 500 hz LR2 or BW2. 500 hz 3rd or 4th order is ok without an LCR at Fs.
The M74A can easily cross at 500 3rd order, even 2nd order if the levels don't get crazy. The VC underhang is more than 3mm p-p which allows for this. In fact I've seen dented domes on the M74A when they were played too loud and too low. There's easily 4-5 mm distance between dome and grille on those, so its capable of that much x- mech (at unknown distortion levels of course).
My target filters for the M74A are 500 - 3.5k LR2 using 2 domes and staggered filtering. Thats conservative IMO for this driver and you'd likely be able to go down to 375 hz LR4 with the right baffle step compensation and use of an LCR at Fs. I'm just not a fan of single suspension drivers (without a spider) being pushed hard. You can run into radial instability this way depending on how far the VC has to go for a given output level. I wouldn't go lower than 500 LR2 HP on a single mid. The LP is a little more flexible. 4k is no problem but you'll start running into directivity discrepancies between mid and HF if the tweeter diameter is smaller than 28mm. The baffle width also dictates this. I would recommend a 3.5k 2nd order LP on the M74A and also filter out the 10k jackhammer up top. That peak is nasty. It really needs an LCR even at reasonably low LP points, otherwise you may see down modulated odd order HD components in the upper mids. These are very hard to distinguish from typical elevated HD up top from breakup.
You asked about the SS 18M compared to the D7608. Thats a tough cone mid to beat and it happens to be one of my favorites. The major advantage is being able to cross it low. 150 hz LR2 isn't an issue with a 12" woofer. The top end is well controlled in terms of breakup and rolloff. It will sound more veiled than the D7608 or most other large domes, mainly due to earlier cone breakup onset, but also due to the use of cone dampening in the upper rolloff area. The multiple peaks up top are well suppressed but will ruin detail. I wouldn't LP that one higher than 2.5k. Its a very good sounding midbass driver and thats its strength. I'd say it compares well with much more expensive cone mids.
A very important thing with the M74A is suppressing the primary Fs peak with lower crossover points. This avoids most of the overexcursion issues when pushed harder. Its definitely recommended with HP < 500 hz LR2 or BW2. 500 hz 3rd or 4th order is ok without an LCR at Fs.
Last edited:
You know im Swedish 😉 , so i noticed absolutely nothing of what you maby felt, and for me your always a gentleman.
Just WoW how enormous big real practical knowledge you have profiguy, i envy you.
Lies so much time in that knowlege, and you are one of the few i "understand" when you write.
Thanks!
best regards John
You know, I read that in some cultures its considered a compliment to talk down to a person you respect in a stern fashion. Maybe that's why its seen as a polite gesture to kick someone in the groin as a show of courtesy... Lol... Just kidding...
We really don't know nuch about Swedish culture in the US, just that Swedish meatballs are a very popular fiood over there. I'm sure thats not correct though, just like Germans mainly eat sour kraut... I should know all about that being from Bavaria.
We really don't know nuch about Swedish culture in the US, just that Swedish meatballs are a very popular fiood over there. I'm sure thats not correct though, just like Germans mainly eat sour kraut... I should know all about that being from Bavaria.
@jawen I'm flattered at your praising me. I just like to share my knowledge and experiences with people.
To the rest of you guys, I have a good piece of info for you - - - its a bit on the long side, so my apologies...
The fact of the matter is people don't always want a clinically perfect flat sounding speaker. While I have built flat FR monitors, I personally don't like that balance of sound. There are large compromises needing to be made when designing a perfectly flat speaker with perfect power response.
The crossover point to a typical 25 - 28mm tweeter is almost always too low in most 2 way and some 3 way speakers. Thats a design choice which usually makes the tweeter sound rough and harsh 99% of the time, even though the tweeter can handle it in theory according to HD measurements. That doesn't sound so enjoyable to the ears when its played loud and elevates listening fatigue.
Unless a larger diameter tweeter and/or WG is used, the corrective action to make the tweeter sound more balanced and less fatiguing would be to raise the HF crossover at a minimum to 3k LR2/BW2 filter, 2.7k with BW3 and 2.5k with LR4/BW4. I'm not a fan of higher than 2nd order filters, so that means at least 3k with a LR2 or BW2 electrical crossover.
My preferred HF HP is actually 3.2 - 3.5k and having the horizontal off axis response dip slightly at that frequency. This allows fine tuning the speaker by angling it in or out towards the sweet spot. It reduces the 3.5k range off axis (preferably by a few dB), which sounds far better to the ears, especially at higher volume levels (aka Fletcher Munson or BBC dip). The falling Harman curve everyone chases helps a little here, but it doesn't account for the natural Fletcher Munson dip which is more pleasant sounding and actually closer to the way we hear at a given SPL. Thats honestly the secret to building a great sounding home listening speaker that will sound natural and less fatiguing than a flat FR monitor type speaker. Yes, the engineer intends to have the recording sound a specific way, but unless the recording is played back at the same SPL, it won't sound the same in terms of overall FR balance. This is especially true with most 80s pop recordings that are mixed so they stand out on FM radio, mainly at a lower volume.
While it may be important to some if they mainly listen off axis and/or need wide even coverage angles, this also creates the same type of compromise as mentioned above. I respect the fact some people need flat FR over a wide horizontal listening window. Unless you're building a studio monitor or PA speaker, this isn't really that important and again, allows for a more relaxed, less fatiguing natural sound at higher volume levels.
Along those lines, waveguides have come a long way for dome tweeters and can compensate for some of the lower midrange roughness the tweeter would otherwise exhibit using the same low crossover point without a WG. They can also fix the directivity jump from the mid or woofer if they're large enough to cover the lower mid frequencies down past the chosen HP slope.
So to finally make the point I intended to, choosing a crossover point around 3 - 3.5k and allowing for a slight 2 - 3 dB dip in that area actually sounds alot more pleasant to the ears and facilitates a more relaxed, natural FR balance, specifically at higher SPLs. While that may be viewed as a highly subjective opinion, I've built most of my home audio, casual listening speakers using this trick with great results. This is based on the input from most people using compared listening tests. Almost everyone gravitated towards the speakers with the slight mid dip. Let's also not forget the fact that having most of the mids come from a single driver greatly improves midrange coherence and imaging accuracy with live, unamplified and unprocessed acoustic recordings. Smoother relative midrange phase response is improved as well. So based on those two major advantages, I'd say its more beneficial to employ a slight mid dip around 3 - 3.5k and select a.crossver point which supports this. It only takes 2 - 3 dB dip to do the job. It will however exclude the use of larger than 5 - 6 inch cone mids, but its right in the butter zone for most 2 - 3" mid domes.
I you're skeptical about all this, try listening to some louder music with and without EQing a 2 - 3dB dip around 3.5k. I think you'll be pleasantly surprised.
Sorrt.for rhe long post, but I feel this applies strongly to the preference of using mid domes.
To the rest of you guys, I have a good piece of info for you - - - its a bit on the long side, so my apologies...
The fact of the matter is people don't always want a clinically perfect flat sounding speaker. While I have built flat FR monitors, I personally don't like that balance of sound. There are large compromises needing to be made when designing a perfectly flat speaker with perfect power response.
The crossover point to a typical 25 - 28mm tweeter is almost always too low in most 2 way and some 3 way speakers. Thats a design choice which usually makes the tweeter sound rough and harsh 99% of the time, even though the tweeter can handle it in theory according to HD measurements. That doesn't sound so enjoyable to the ears when its played loud and elevates listening fatigue.
Unless a larger diameter tweeter and/or WG is used, the corrective action to make the tweeter sound more balanced and less fatiguing would be to raise the HF crossover at a minimum to 3k LR2/BW2 filter, 2.7k with BW3 and 2.5k with LR4/BW4. I'm not a fan of higher than 2nd order filters, so that means at least 3k with a LR2 or BW2 electrical crossover.
My preferred HF HP is actually 3.2 - 3.5k and having the horizontal off axis response dip slightly at that frequency. This allows fine tuning the speaker by angling it in or out towards the sweet spot. It reduces the 3.5k range off axis (preferably by a few dB), which sounds far better to the ears, especially at higher volume levels (aka Fletcher Munson or BBC dip). The falling Harman curve everyone chases helps a little here, but it doesn't account for the natural Fletcher Munson dip which is more pleasant sounding and actually closer to the way we hear at a given SPL. Thats honestly the secret to building a great sounding home listening speaker that will sound natural and less fatiguing than a flat FR monitor type speaker. Yes, the engineer intends to have the recording sound a specific way, but unless the recording is played back at the same SPL, it won't sound the same in terms of overall FR balance. This is especially true with most 80s pop recordings that are mixed so they stand out on FM radio, mainly at a lower volume.
While it may be important to some if they mainly listen off axis and/or need wide even coverage angles, this also creates the same type of compromise as mentioned above. I respect the fact some people need flat FR over a wide horizontal listening window. Unless you're building a studio monitor or PA speaker, this isn't really that important and again, allows for a more relaxed, less fatiguing natural sound at higher volume levels.
Along those lines, waveguides have come a long way for dome tweeters and can compensate for some of the lower midrange roughness the tweeter would otherwise exhibit using the same low crossover point without a WG. They can also fix the directivity jump from the mid or woofer if they're large enough to cover the lower mid frequencies down past the chosen HP slope.
So to finally make the point I intended to, choosing a crossover point around 3 - 3.5k and allowing for a slight 2 - 3 dB dip in that area actually sounds alot more pleasant to the ears and facilitates a more relaxed, natural FR balance, specifically at higher SPLs. While that may be viewed as a highly subjective opinion, I've built most of my home audio, casual listening speakers using this trick with great results. This is based on the input from most people using compared listening tests. Almost everyone gravitated towards the speakers with the slight mid dip. Let's also not forget the fact that having most of the mids come from a single driver greatly improves midrange coherence and imaging accuracy with live, unamplified and unprocessed acoustic recordings. Smoother relative midrange phase response is improved as well. So based on those two major advantages, I'd say its more beneficial to employ a slight mid dip around 3 - 3.5k and select a.crossver point which supports this. It only takes 2 - 3 dB dip to do the job. It will however exclude the use of larger than 5 - 6 inch cone mids, but its right in the butter zone for most 2 - 3" mid domes.
I you're skeptical about all this, try listening to some louder music with and without EQing a 2 - 3dB dip around 3.5k. I think you'll be pleasantly surprised.
Sorrt.for rhe long post, but I feel this applies strongly to the preference of using mid domes.
@profiguy and any others that have input!
I'm not close to your level of experience or knowledge but work as an acoustic consultant so have a reasonable grasp of some items. I've noticed that many 3" dome commercial (domestic rather than studio) speakers tend to crossover higher than theory suggests as ideal and tend to be around the 5kHz area. The only speaker l've found that cross over a 3" lower is the ATC but their drive unit is within a waveguide and most are not and I believe the ATC has poor directivity at it's upper freq end. Do you have any thoughts on why? Is this felt to provide a more cohesive mid-band (Phase/freq response) with all frequencies within the ear's most sensitive region being handled by 1 drive unit?
My current speakers are Yamaha NS1000M - X-over 5 kHz. (whilts reported to be a 3.5" I believe it is actually a 68mm dome plus surround)
I used to own Ruark Crusader 2's (Dynaudio D76AF) which were 4-4.5kHz from memory.
I built some analogue active 3-way speakers using Volt B250.8/D76AF/D260 Esotec in the 90s and later went early digital x-over. I preferred cross-over set around 4.5-5 kHz. I still have the drive units.
I'm looking at trying to start a DIY "modern take on the NS1000M" based on M74B-6 and probably T34B-4 but considering a baffle design that resembles the Snell A (another favourite speaker of mine). I've today noticed Hifijim's similar baffle design. A higher cross-over freq may favour the T25B although I'm tryting to make the speaker very sensitive to still partner with low powered valves so the loss of senstivity is not welcome - I do also like to very listen loud at times so power handling/Max SPL of the T34 is appreciated. I'll always be chasing the bass SPL/sensitivity to match the upper freq but that's another challenge for another day/month/year - for now I can adopt 4x10" Volt B250.8 which I already have - 91 dB/1W so 97 dB/1W with a pair per channel running at 4 ohm without BSC..
I'm not close to your level of experience or knowledge but work as an acoustic consultant so have a reasonable grasp of some items. I've noticed that many 3" dome commercial (domestic rather than studio) speakers tend to crossover higher than theory suggests as ideal and tend to be around the 5kHz area. The only speaker l've found that cross over a 3" lower is the ATC but their drive unit is within a waveguide and most are not and I believe the ATC has poor directivity at it's upper freq end. Do you have any thoughts on why? Is this felt to provide a more cohesive mid-band (Phase/freq response) with all frequencies within the ear's most sensitive region being handled by 1 drive unit?
My current speakers are Yamaha NS1000M - X-over 5 kHz. (whilts reported to be a 3.5" I believe it is actually a 68mm dome plus surround)
I used to own Ruark Crusader 2's (Dynaudio D76AF) which were 4-4.5kHz from memory.
I built some analogue active 3-way speakers using Volt B250.8/D76AF/D260 Esotec in the 90s and later went early digital x-over. I preferred cross-over set around 4.5-5 kHz. I still have the drive units.
I'm looking at trying to start a DIY "modern take on the NS1000M" based on M74B-6 and probably T34B-4 but considering a baffle design that resembles the Snell A (another favourite speaker of mine). I've today noticed Hifijim's similar baffle design. A higher cross-over freq may favour the T25B although I'm tryting to make the speaker very sensitive to still partner with low powered valves so the loss of senstivity is not welcome - I do also like to very listen loud at times so power handling/Max SPL of the T34 is appreciated. I'll always be chasing the bass SPL/sensitivity to match the upper freq but that's another challenge for another day/month/year - for now I can adopt 4x10" Volt B250.8 which I already have - 91 dB/1W so 97 dB/1W with a pair per channel running at 4 ohm without BSC..