Four KM-84 Clones - Blind Test

[k brown]

I'll be the first to admit that the sound of a microphone is mainly determined by its capsule and capsule surroundings (head basket, back chamber, vents etc). I think I've mentioned that more than once. And that is exactly why I like the 3U cardioid capsule, in combination with the right SDC body having the right back chamber and the right vents design (also mentioned by 3U Audio, btw). To my taste, it sounds balanced and in shootouts on Youtube, I could not hear a difference between the 3U and an original KM84.

You mentioned how you appreciate the flat frequency response of the KM84. Well, I think that answers your question from post #97: the CM-63 is not flat and people preferred the flat(ter) response of the 3U capsule in the blind tests.

Whether the KM84, or 3U capsule or any other capsule for that matter would sound better with other electronics? Maybe... But we should first define what is "better". Less THD? Some like a bit of "color" and use it as an artistic effect. More headroom? Some use the transformer LF compression at high SPLs to their benefit on snares. There is a place for almost any type of circuit. That's also why I designed several other circuits having more headroom, lower distortion and lower noise than the KM84. One of them being the KM84++. The other three, one with transformer and two w/o, outperform the classic Schoeps in many ways. I'll publish their specs in due time, but will not share the schematics. Whether they will also sound any better or different? I'll leave that up to my son to decide, for whom I build them.

Although some microphone properties are a matter of taste, there are, however, several properties that you typically want any microphone to have: low noise, good RFI rejection, high CMRR, good moisture protection, in most cases high headroom and high reliability. Maybe these are more important properties of head amp circuits than sound? Could be... But of course: YMMV and it is not forbidden to be a heretic in that respect.

It was the KM 84's off-axis response I referred to.

I also take the published responses of most affordable mics with more than a grain of salt. My own ears tell me that the CM-63 hasn't as much of a dip as shown in their docs.

It would also be very interesting to compare response of the CM-63 capsule when fitted to a CM-60 body (with it's different slots) to that on it's own body.

 

[jp8]

For those of you who'd like to build the KM84++: I have made the PCB for the CM-63 available on PCBWay now. Link: https://www.pcbway.com/project/shareproject/KM84_for_Takstar_CM_63_Microphone_524d4f22.html

 

[inkster]

Interesting post.
I liked B and D, B I liked on voice and D on instruments.
Thanks for providing the PCB files!

 

[joulupukki]

It's been a few weeks and I'm a little out of practice with REW, but I got a CM4 and did a quick, highly nonscientific frequency sweep to compare against @jp8 's KM84+. Not bad! The SPL is aligned at 1K to see the comparison more easily.

 

[micolas]

This is all just for fun – nothing too scientific here, but I definitely learned which mic(s) I'd want to use next time I record an album. In this video I didn't include my Graeme Woller KM84 build because I feel like it sounds nearly identical to the MP SDC-84.

The microphones I compare in this video are (in no particular order):

1. Neumann KM-184
2. Microphone Parts SDC-84
3. KM84+ from @jp8 (in a Takstar CM-60 donor mic)
4. KM84++ from @jp8 (in a Takstar CM-63 donor mic)

Can you pick out the KM-184? Do you have a favorite? I know the KM-184 is slightly different than a KM-84. I'm working on finding one to A/B. But in the meantime...

I'm sorry, I'm late to the party.
I was absent from the group for a while (actually I left temporarily when @Wordsushi announced that he would leave groupdiy, a fact that made me very sad, for me he is a very good, generous, altruistic, passionate man, always ready to help everyone, a perfectionist, a professional worthy of all admiration)
But back to the topic:
I averaged all sources.
I haven't read any other posts in this thread.
So,
I liked them more in the following order:
D C A B
D sounds the best on most sources, for me it's the most musical, conveys the most emotion, I'd say it's almost radio-ready, as if I wouldn't do any major editing
C : I like it
B sounds little more artificial
This is what I heard with my tired ears, damaged in decades
And now, I can't wait to find out the identification!

*Edit:
I saw the revelation of the comparative test.
Hats off to Neumann!
I will definitely read all the posts in the thread now.

 

[joulupukki]

I'm sorry, I'm late to the party.
I was absent from the group for a while (actually I left temporarily when @Wordsushi announced that he would leave groupdiy, a fact that made me very sad, for me he is a very good, generous, altruistic, passionate man, always ready to help everyone, a perfectionist, a professional worthy of all admiration)
But back to the topic:
I averaged all sources.
I haven't read any other posts in this thread.
So,
I liked them more in the following order:
D C A B
D sounds the best on most sources, for me it's the most musical, conveys the most emotion, I'd say it's almost radio-ready, as if I wouldn't do any major editing
C : I like it
B sounds little more artificial
This is what I heard with my tired ears, damaged in decades
And now, I can't wait to find out the identification!

*Edit:
I saw the revelation of the comparative test.
Hats off to Neumann!
I will definitely read all the posts in the thread now.

Thanks for chiming in with your thoughts! Yeah, Neumann did well in the comparison. When I had it to try out I was definitely impressed with its smoothness.

I was also saddened to learn about Wordsushi’s departure. I still watch his YouTube videos. He brings so much to the table in terms of research, experimentation, and thoughts. But, I totally understand his position and reasoning. It’s easy for forums to overtake one’s time. I keep coming back here for the fascinating information I learn from all of you! Kudos to you all!

 

[joulupukki]

Wow, I don't know what was up with my REW measurements the other day, but here's a new set of measurements comparing these microphones against a Line Audio CM4. Read the graph key for which mic is which. If we're going solely off this non-scientific FR capture from REW, @jp8 's KM84++ comes closest to the CM4. The Microphone Parts SDC-84 has significant bumps in the higher frequencies in comparison. Worse high bumps are the CM-60 and CM-63.

 

[k brown]

Wow, I don't know what was up with my REW measurements the other day, but here's a new set of measurements comparing these microphones against a Line Audio CM4. Read the graph key for which mic is which. If we're going solely off this non-scientific FR capture from REW, @jp8 's KM84++ comes closest to the CM4. The Microphone Parts SDC-84 has significant bumps in the higher frequencies in comparison. Worse high bumps are the CM-60 and CM-63.
View attachment 142318
View attachment 142319View attachment 142320
View attachment 142321
View attachment 142322
View attachment 142323

The CM-63 graph certainly reflects what my ears have been telling me - that the upper-mid / lower treble dip is rather less than the published graph shows.

I encourage folks that have both, to try the CM-63 capsules on the CM-60 bodies; I don't know if it's because of the different body slots, the fact that the CM-60 body rings less, or both - but I'm liking this combo best. I know the '63 has better RF protection, but so far mine haven't picked up any noise. Try it and see what you think.

Also, the CM-63 capsule is nearly as flat as the Line Audio, when the front grille is removed. A pair of these, plus a pair of stock capsules covers a lot of bases. The top lift of the stock capsule is well suited for distant Classical pair, and the modded ones for close / spot mic'ing. The un-modded CM-60 capsule ain't good fer nuthin'.

Also a tip for the CM-60 bodies: Not sure if it's oxidation, paint overspray,, or both but I've experience some 'uncertainty' with XLR shell not making good contact with the inside of the body. Completely addressed by fine-sanding the XLR shell and where the inside of the body contacts it, then making sure the three set screws are firmly set.

 

[joulupukki]

I put the CM-63 capsule on the CM-60 mic and I'm not noticing that much difference. Equally bad. I haven't tried modifying the CM-63 capsule though, like you mention.

 

[k brown]

I put the CM-63 capsule on the CM-60 mic and I'm not noticing that much difference. Equally bad. I haven't tried modifying the CM-63 capsule though, like you mention.
View attachment 142368

Very interesting - thanks for doing that; I would have expected the very different rear slots of the two bodies to make more difference than that. Polar pattern probably a bit different.

'Bad' only for close mic'ing; that kind of lift is very useful for mic'ing in the diffuse field. Not too different from Rode NT5 (whose published graphs are of course highly smoothed); lift just starts about an octave higher (which is nice, actually - no boost until 6kHz, which is about where the Neumann KM 83 omni 6dB lift begins).

That strong dip at 12kHz must be an anomaly of your test rig; too much of a coincidence that it's in all the capsules you tested.

Still, would be interesting to try removing just half of the front grille material (enlarge the slots or holes).

 

[k brown]

Interesting test; just tried enlarging the outer-ring holes of the CM-63 grille to 3/32", and (taking my lead form the DPA grille style) deepening every other slot to the bottom of it's bevel, then re-beveling it, and indeed the top boost now sounds almost exactly between the no-grille mod and the stock one. This, of course, requires disassembling the capsule.

Top boost now seems closer that of the Neumann KM 143.

 

[jp8]

Just like @joulupukki , I have acquired a Line Audio CM4 to use as a reference mic when measuring cardioid microphones. But unlike joulupukki, I use a foam booth to perform microphone measurements. This is because I do not have an acoustically treated room, and doing measurements outside or in the living room is out of the question for me. My setup and method are described here. I have measured the CM-60 and CM-63 with the stock capsules and also swapped capsules with their Takstar siblings and with the 3U Audio Cardiod capsule. The results can be found on my website here. My take-outs from these measurements: CM-60 and CM-63 are peaky in the 7kHz to 10kHz region, with 90 degrees off-axis response deviating from the ideal 6 dB of a cardiod mic in the HF range. CM-60 body with CM-63 capsule or vice-versa do not yield improved frequency responses. If you want a flat response combined with a good separation between the on-axis and off-axis response at higher frequencies, the 3U Cardiod capsule on the CM-60 or CM-63 is a much better choice. But I'll leave it to the dear reader to judge whether the 3U capsule on these microphones is anywhere close to a KM84 and a good alternative or not.

Now it would be interesting to see to what extent joulupukki's and my measurements yield the same results. Joulupukki kindly shared his REW .mdat files, so I could combine them with my measurements into one chart. In REW, I divided joulupukki's DUT charts by the CM4 reference, which yields a difference plot. If we assume the CM4 is flat, then this difference plot equals the frequency response of the DUT. I then combined this with my frequency response measurements. I applied 1/6 smoothing to all charts and they were aligned at 1 kHz. The results are shown below.

I expected to see the largest deviations in the LF area, but I saw that Joulupukki's measurements gave structurally higher values in the HF area. I have no direct explanation for this, but this will undoubtedly be due to the different setups for measuring. I will still see if I can also come to similar results when measured differently. When we take a look at all the pictures, we see quite similar frequency responses. But even though the curves measured on the east and west sides of the Atlantic appear to be the same in shape, due to the HF drop-off in my curves, you can come to different conclusions about which microphone has the flattest curve. From joulupukki's graphs, one would conclude the CM-63 + 3U would be the flattest, and from mu curves the CM-60 with the 3U capsule.

From top to bottom, the graphs below show the stock CM-60, stock CM-63, CM-60 with 3U capsule and CM-63 with 3U capsule. In each graph, the frequency response with the highest HF content is from joulupukki. The difference is at most a few dB, so perhaps this is also within a to be expected bandwidth of these types of measurements. But if anyone has a reasonable explanation for the structural difference, I'd be glad to know.