Soliloqueen's k87(k67) and k47 capsules

GroupDIY Audio Forum

Help Support GroupDIY Audio Forum:

This site may earn a commission from merchant affiliate links, including eBay, Amazon, and others.
So say, for non b stock within 2-3 weeks would be a reasonable guess?
i think so personally. i've got tons of both retail and OEM orders to get through now that the k87 is good to go. god help me. thank god i didn't spend anyone's money so i actually still have the money to ship all of these. i held on to all the money from the unshipped orders and funded all the production and development out of pocket using profits from the k47s. things probably would have gone faster if i actually spent people's money, but that felt gross. joke's on past me now though since it worked out better that way.
 
Last edited:
i think so personally. i've got tons of both retail and OEM orders to get through now that the k87 is good to go. god help me. thank god i didn't spend anyone's money so i actually still have the money to ship all of these. i held on to all the money from the unshipped orders and funded all the production and development out of pocket using profits from the k47s. things probably would have gone faster if i actually spent people's money, but that felt gross. joke's on past me now though since it worked out better that way.
Great, thanks Ari!

Can i place order on the website as normal even though it will say 'Backorder'?
 
I've temporarily disabled the cardioid flat k47s because funnily enough, they're just aren't enough QC failures to sell this time. Our manufacturing on these has gotten too tight to have three bins. We have to go down to two. All of the current orders will get bumped up one bin for free and starting with the next batch of flat k47s I'm going to be making single-sided versions with only one side metalized.
 
@soliloqueen If you're willing to share (totally understand if not), I'm curious how many capsules you've sold of each model. I always wonder how big our little DIY mic scene is, and number of capsules sold seems like a decent proxy.
I'm going to hit 2000 capsules pretty soon! Not a huge amount given the monumental startup expenses and the issues I've run into but enough at least to keep the lights on. Most of those have been in just the last year. It's not a huge amount and it's normally not an amount that I would have trouble shipping but I've got so many plates spinning at once. I think I need to start delegating again, but with a part-time employee rather than outsourcing anything.
 
Last edited:
2000 is quite the feat! I used to own a small audio electronics business and I know first hand how much work stocking, shipping, and inventory management can be on top of designing, customer service, assembly, etc, etc, etc.

I work in the logistics industry now, and we have entire teams dedicated to managing shipping for our large customers. It’s one of the hardest things to get right.
 
skinned the backplates from my friend's machine shop. forgot that i compensated the drawings for the big factory's loose bits because they refused to take normal feedback. I didn't remember the changes I'd made before we started so we ended up accidentally making plates where the 5k boost is flipped upside down, which is hilarious. but look at the consistency! this is 10 measurements! That's insane! The only differences here are spacer (plastic) and tension related. the plates are identical to within like 2 micron. i have NEVER seen k47 plates this consistent. i am so excited to move to "in-house" (well, closely supervised) parts. it would be fun to make a video tour of the place and show off the machining

1732431061395.png
(K47V in red)

We did a few different styles of rings too, machined, as benchmarks. Here's 1 capsule with peek rings with recessed screws and another with FR2 rings like the second gen K47 from the 60s:
IMG_0021.jpg

to my knowledge, the only replica k47 with fr2 rings ever. i wanted to see if it would take a CNC milling easily. it's a bit smoother than real fr2 k47s though, since it's actually machined and not just a raw sheet with holes punched in it.

These are way, way, way, way way comically more expensive to make than the current capsules. Like twice as expensive at least, but honestly it pays for itself in the lower hassle, way higher consistency and having more control over the supply chain. right now we're messing with different ways to insert the plastic center termination into the plate. We tried traditional press fitting (unstable), refrigerated press fitting (plug falls out if acrylic at freezing temps, but more stable with a low expansion plastic, but it's difficult to dry the condensation out of the plates after) and now just for fun we're on tiny threads with like the ck12. we'll probably end up doing traditional press fitting with epoxy or something. i'm having a lot of fun messing around with exactly how to do this.

my friend has admittedly never done mass production before. he's used to under 10 unit orders of high precision tungsten tooling. he's not quite sure how he'd do making 100 brass plates, so we're doing a lot of brainstorming to try and bring efficiency up. the good news is that unlike the mass production factory i was using that makes all the plates for alctron etc, he can definitely do a ck12, He's crazy, so he'd probably find it fun. Now that my friend is on board, a CK12 is way more viable, probably assembled in the US.

Right now, we're brainstorming ways to achieve the flatness in bulk. It wasn't too hard for him with his precision grinder with a few units, but making 100 would be tedious. he, certifiable lunatic that he is, asked me what's stopping us from making a flat honing machine that runs as an attachment on a milling or lathe engine. I told him that would be extremely hard, and that we'd have to be- and he said "An ultra high precision tungsten carbide tool shop" and that shut me up and now he's in a cave somewhere trying to build a lapping machine from scratch. this guy is nuts. he is sure that he can make gears and have it do both sides at once and everything. perfect energy for this project.
 
Last edited:
dude was like "we can't use a lathe for the m7, it'll be concave" and I'm like "so what then" and he says "WE END MILL IT WITH DIAMOND" (coincidentally, someone else also suggested this to me at the same time)

We did, in fact, end mill it with diamond. Not only did he manage to achieve the critical flatness this way, he was able to double cut the gap so that the center tap is back up at the height of the edge.
PXL_20241006_061837233.NIGHT.RAW-01.COVER.jpg
 
Last edited:
So cool! Excited for what's coming up. It would be awesome to get a peek at the facility, if you ever end up taking those videos.
dude was like "we can't use a lathe for the m7, it'll be concave" and I'm like "so what then" and he says "WE END MILL IT WITH DIAMOND" (coincidentally, someone else also suggested this to me at the same time)

We did, in fact, end mill it with diamond. Not only did he manage to achieve the critical flatness this way, he was able to double cut the gap so that the center tap is back up at the height of the edge.
View attachment 140485
skinned the backplates from my friend's machine shop. forgot that i compensated the drawings for the big factory's loose bits because they refused to take normal feedback. I didn't remember the changes I'd made before we started so we ended up accidentally making plates where the 5k boost is flipped upside down, which is hilarious. but look at the consistency! this is 10 measurements! That's insane! The only differences here are spacer (plastic) and tension related. the plates are identical to within like 2 micron. i have NEVER seen k47 plates this consistent. i am so excited to move to "in-house" (well, closely supervised) parts. it would be fun to make a video tour of the place and show off the machining

View attachment 140464
(K47V in red)

We did a few different styles of rings too, machined, as benchmarks. Here's 1 capsule with peek rings with recessed screws and another with FR2 rings like the second gen K47 from the 60s:
View attachment 140463

to my knowledge, the only replica k47 with fr2 rings ever. i wanted to see if it would take a CNC milling easily. it's a bit smoother than real fr2 k47s though, since it's actually machined and not just a raw sheet with holes punched in it.

These are way, way, way, way way comically more expensive to make than the current capsules. Like twice as expensive at least, but honestly it pays for itself in the lower hassle, way higher consistency and having more control over the supply chain. right now we're messing with different ways to insert the plastic center termination into the plate. We tried traditional press fitting (unstable), refrigerated press fitting (plug falls out if acrylic at freezing temps, but more stable with a low expansion plastic, but it's difficult to dry the condensation out of the plates after) and now just for fun we're on tiny threads with like the ck12. we'll probably end up doing traditional press fitting with epoxy or something. i'm having a lot of fun messing around with exactly how to do this.

my friend has admittedly never done mass production before. he's used to under 10 unit orders of high precision tungsten tooling. he's not quite sure how he'd do making 100 brass plates, so we're doing a lot of brainstorming to try and bring efficiency up. the good news is that unlike the mass production factory i was using that makes all the plates for alctron etc, he can definitely do a ck12, He's crazy, so he'd probably find it fun. Now that my friend is on board, a CK12 is way more viable, probably assembled in the US.

Right now, we're brainstorming ways to achieve the flatness in bulk. It wasn't too hard for him with his precision grinder with a few units, but making 100 would be tedious. he, certifiable lunatic that he is, asked me what's stopping us from making a flat honing machine that runs as an attachment on a milling or lathe engine. I told him that would be extremely hard, and that we'd have to be- and he said "An ultra high precision tungsten carbide tool shop" and that shut me up and now he's in a cave somewhere trying to build a lapping machine from scratch. this guy is nuts. he is sure that he can make gears and have it do both sides at once and everything. perfect energy for this project.
 
I put my Flat k47 mic next to an m149 yesterday. With all the limitations that come with A/B testing already, this was only meant to be a quick microphone comparison. Adding the variables introduced by the different circuit and body designs, this is not an objective way to compare the capsules. Still, I thought the results were interesting enough to share.

(I apologize for the mediocre string performance. If anyone here is extra sensitive to pitch, I recommend extreme caution 😅)

These files come straight out of the PT session and need better gain matching

A
View attachment A.1.wav
B
View attachment B.1.wav

Both mics are going through AMS Neve 88rs channel strips. If I remember correctly, the m149 required a little less gain. They were standing next to each other, maybe about 5m/16ft in front of a 4 piece string section.
My mic is a cardioid only OA m49 with a 6s6b-v and a UTM 0549 transformer, the body is taken from a Slate ML-1. If you want to hear it on a more professional performance, with no pitch issues, here it is on drums. Model names for A and B are also at the bottom of that same post.

I would love to hear your thoughts.
 
B sounds really band-limited to me in both of your examples... It has basically no high end at all. Something up with the mic? it's really aggressively present at like 2.5k. Positioning maybe?
 
skinned the backplates from my friend's machine shop. forgot that i compensated the drawings for the big factory's loose bits because they refused to take normal feedback. I didn't remember the changes I'd made before we started so we ended up accidentally making plates where the 5k boost is flipped upside down, which is hilarious. but look at the consistency! this is 10 measurements! That's insane! The only differences here are spacer (plastic) and tension related. the plates are identical to within like 2 micron. i have NEVER seen k47 plates this consistent. i am so excited to move to "in-house" (well, closely supervised) parts. it would be fun to make a video tour of the place and show off the machining

View attachment 140464
(K47V in red)

We did a few different styles of rings too, machined, as benchmarks. Here's 1 capsule with peek rings with recessed screws and another with FR2 rings like the second gen K47 from the 60s:
View attachment 140463

to my knowledge, the only replica k47 with fr2 rings ever. i wanted to see if it would take a CNC milling easily. it's a bit smoother than real fr2 k47s though, since it's actually machined and not just a raw sheet with holes punched in it.

These are way, way, way, way way comically more expensive to make than the current capsules. Like twice as expensive at least, but honestly it pays for itself in the lower hassle, way higher consistency and having more control over the supply chain. right now we're messing with different ways to insert the plastic center termination into the plate. We tried traditional press fitting (unstable), refrigerated press fitting (plug falls out if acrylic at freezing temps, but more stable with a low expansion plastic, but it's difficult to dry the condensation out of the plates after) and now just for fun we're on tiny threads with like the ck12. we'll probably end up doing traditional press fitting with epoxy or something. i'm having a lot of fun messing around with exactly how to do this.

my friend has admittedly never done mass production before. he's used to under 10 unit orders of high precision tungsten tooling. he's not quite sure how he'd do making 100 brass plates, so we're doing a lot of brainstorming to try and bring efficiency up. the good news is that unlike the mass production factory i was using that makes all the plates for alctron etc, he can definitely do a ck12, He's crazy, so he'd probably find it fun. Now that my friend is on board, a CK12 is way more viable, probably assembled in the US.

Right now, we're brainstorming ways to achieve the flatness in bulk. It wasn't too hard for him with his precision grinder with a few units, but making 100 would be tedious. he, certifiable lunatic that he is, asked me what's stopping us from making a flat honing machine that runs as an attachment on a milling or lathe engine. I told him that would be extremely hard, and that we'd have to be- and he said "An ultra high precision tungsten carbide tool shop" and that shut me up and now he's in a cave somewhere trying to build a lapping machine from scratch. this guy is nuts. he is sure that he can make gears and have it do both sides at once and everything. perfect energy for this project.
Oh man, this thread is getting good again

Grateful that you’re pulling back the curtain and showing us the entire process of developing and refining these capsules in such detail

The commitment and obvious enthusiasm makes it obvious to me: you will be producing the best microphone capsules on earth before too long
 
basically no high end
The lesser extension into the high frequencies is quite noticeable.
it's really aggressively present at like 2.5k
as well as the aggressiveness around that range.

Something up with the mic?
Considering how early it is in my mic journey, it's very likely that I did something wrong. Assuming I assembled the board correctly, some other issues could be excessive foam at the base of the head basket, a mechanical problem I had with the capsule raiser of the Slate body which angled the capsule about 3/4 degrees down, or the acoustics of the head basket, which I know needs to be on the bigger size with k47s.
Unfortunately, with my primitive soldering skills, I can't assume I assembled the board correctly... Maybe I didn't handle the hi-z area properly, but the possibilities are endless, so I might go back to do some further testing and bother the original thread instead of this one.
Positioning maybe?
I don't know about that. The 2 mics were almost touching each other, and were far away from the source enough to make their position much less of an influence on these factors.

I did think my build was clearly sounding inferior, but, excited for having my first mic work at all, especially on the first try, I didn't consider it to be defective until reading your comment. I just considered it inferior.
I was already skeptical about sharing the A/B test around it to a capsule specific thread, but if it's the case the mic isn't even working, I might end up deleting the previous post until I'm back with the an appropriate device.
 
The lesser extension into the high frequencies is quite noticeable.

as well as the aggressiveness around that range.


Considering how early it is in my mic journey, it's very likely that I did something wrong. Assuming I assembled the board correctly, some other issues could be excessive foam at the base of the head basket, a mechanical problem I had with the capsule raiser of the Slate body which angled the capsule about 3/4 degrees down, or the acoustics of the head basket, which I know needs to be on the bigger size with k47s.
Unfortunately, with my primitive soldering skills, I can't assume I assembled the board correctly... Maybe I didn't handle the hi-z area properly, but the possibilities are endless, so I might go back to do some further testing and bother the original thread instead of this one.

I don't know about that. The 2 mics were almost touching each other, and were far away from the source enough to make their position much less of an influence on these factors.

I did think my build was clearly sounding inferior, but, excited for having my first mic work at all, especially on the first try, I didn't consider it to be defective until reading your comment. I just considered it inferior.
I was already skeptical about sharing the A/B test around it to a capsule specific thread, but if it's the case the mic isn't even working, I might end up deleting the previous post until I'm back with the an appropriate device.
Oh no no I didn't mean to be mean especially if it's your first build 😭 that's such an accomplishment!! I would suggest removing the foam first.
 
also, give the high-z area a good cleaning with 99.9% isopropanol

I’ve definitely had compromised performance due to an errant fingerprint in that area!
 
Ari, does the “flat” frequency response of the flat-k47 change the polar response of the capsule compared to a Neumann k47? Does it make high frequency response wider than the Neumann?
 
Ari, does the “flat” frequency response of the flat-k47 change the polar response of the capsule compared to a Neumann k47? Does it make high frequency response wider than the Neumann?
Yes, the flat k47 is almost a true cardioid. kingkorg tested this. The polar patterns are considerably different than a K47 (at least in pure 180 degree testing, the rear lobe is 6dB quieter and rear response is flatter than the k47). More stable and "correct" i suppose. That's why it hates being in cylindrical baskets so much compared to a regular k47 and why i always warn against it. a regular k47 is too narrow in the high end to pick up the reflections back from a cylindrical basket. the flat k47, being a more true cardioid, can pick them up (detrimentally). It's why I suggest the flat k47 primarily for M49 builds, where this isn't an issue. The flat k47 sounds great in an m49 body and pretty good in a u47 body, but anything smaller it starts to get grody. When I put one in a studio projects B1 it was legitimately unusable. Keep in mind that I am really mean and hyper-sensitive when it comes to sound. It might not be bad to someone else. But overall, it's very much a "scales depending on the attention you pay to the rest of the mic" capsule.
 
Last edited:
That’s about what I was expecting you to say! Ok, so my conceptualization tracks.

It can also sometimes be a good thing to have top end fall off a bit more when there is bleed, but it is subjective and context dependent.
 

Latest posts

Back
Top