Barry Hufker
Well-known member
Waiting eagerly (patiently, not so much 
)...
)...
Four KM-84 Clones - Blind Test
- Thread starter joulupukki
- Start date
Help Support GroupDIY Audio Forum:
joulupukki
Well-known member
Ok, the answers are posted at the bottom of the video description!
joulupukki
Well-known member
Edit: The goal of this “experiment” wasn’t to crown a winner or to say definitively which mic is the “best” but it was interesting to see how the KM-184 bubbled to the top.
Unofficially earlier today between here and YouTube comments it looked like this (for people who actually said they preferred one particular mic).
Last edited:
Definitely called them all wrong myself. Was quite interesting to see the variety of opinions on these.
rock soderstrom
Tour de France
That's pretty much exactly what I hear and think. I haven't read the answer yet. I think D is the much “bashed” KM184, which I know and appreciate very well.
Edit: pretty close to my expectations...Thank you joulupukki for this nice experiment! All microphones sound very similar, at least on youtube.
Last edited:
rock soderstrom
Tour de France
Ohh, I've come across that quite often online in the last few years.(check gearslutz) The KM184 would be too sharp in the treble and only the KM84 would be the real deal and so on. I never understood that, I think they are very good and universal workhorses in the recording studio and also live on stage. I've worked with them myself for long years and I think you get a lot of microphone for the money. Very flexible.
GeorgeToledo
Well-known member
- Joined
- Mar 3, 2015
- Messages
- 392
I honestly hope the KM184 naysaying keeps up because it makes them a good buy on the used market.
I have my drumset right now with two KM184 overhead and a 421-II on the bass drum, which as the internet knows, is absolutely awful.
I have my drumset right now with two KM184 overhead and a 421-II on the bass drum, which as the internet knows, is absolutely awful.
dreamix
Well-known member
Myself I dont like the km184 that much compared to the km84, but its really more of getting a more likeable eq-curve from the start with the km84, nothing you cant eq to get the same in mix... if anyone would want
imok brown
Well-known member
Well, the KM 84 is dead flat; doesn't have an "eq-curve".Myself I dont like the km184 that much compared to the km84, but its really more of getting a more likeable eq-curve from the start with the km84, nothing you cant eq to get the same in mix... if anyone would want
Anyone who expected same from KM 184 was disappointed.
dreamix
Well-known member
Im sure km184 is "dead flat" aswell, still dont sound the same, same with everything. The sonic differences is best described with eq curve in mind in terms of more mid, more highs, less low etc etc...
rock soderstrom
Tour de France
Yes, that's right.
Like the TLM103 (another “terrible” Neumann microphone ), the KM184 has a built-in EQ curve, so it's practically ready to mix. In my opinion, this works quite well, you immediately get high-quality results, which can be very advantageous, especially in time-critical production environments (TV and radio broadcasts, VO and media productions). On the other hand, this can perhaps also stand in the way of one or other application, but in the end I don't think the differences are so great that an EQ couldn't fix them. YMMV.
Like the TLM103 (another “terrible” Neumann microphone
), the KM184 has a built-in EQ curve, so it's practically ready to mix. In my opinion, this works quite well, you immediately get high-quality results, which can be very advantageous, especially in time-critical production environments (TV and radio broadcasts, VO and media productions). On the other hand, this can perhaps also stand in the way of one or other application, but in the end I don't think the differences are so great that an EQ couldn't fix them. YMMV.Now that @Joulupikki has revealed the microphones, I'd like to chime in and tell you something about the KM84+ and KM84++. But first of all, I want to thank Joulupukki for the many hours he spent testing my PCBAs, all the interesting discussions we had, and making all the awesome recordings and Youtube videos since March this year. It was a great learning experience for both of us! But it was also good to receive comments from others, in this case from all those who dared to speak out and left their opinions in this thread or on Youtube. In some cases your comments supported what we experienced ourselves, and in other cases they didn't. Which is fine too. Many thanks to all!
Back to the KM84+, KM84++, and the Microphone-Parts SDC-84 as used in this blind test. I will not elaborate on all the technical details. This will be explained on my website later. I'll briefly summarize the differences and point out some points that I think could have influenced the sound of the microphones and might explain what you heard.
KM84+: Takstar CM-60 body with a 3U cardioid capsule and 3U GZT-84 transformer. Circuit design follows the legacy KM84 design, but most parts are in SMT, mainly 0805. Instead of the pad, which the CM-60 does not have, I added a low-cut filter with a jumper to select -12 dB/oct or -6 dB/oct roll-off. Most important for Joulupukki was the RFI filter that effectively suppresses cell phone interference. I expected this design to sound the closest to the KM84.
KM84++: Takstar CM-63 body with a 3U cardioid capsule and 3U GZT-84 transformer. In the KM84++ circuit, I set the JFET to a much higher bias current to achieve lower noise and distortion and a somewhat extended low-end. And I added a CMOS polarization voltage generator, also for higher SNR. Finally, it has a -20 dB pad switch that the KM84+/CM-60 does not have.
SDC-84: I don't know how close the SDC-84 follows the legacy KM84 design. But at least it has a different capsule and transformer than the KM84+/++. And in the mic used in this blind test, Joulupukki removed the feedback capacitor C2. That makes it more sensitive of course, but also affects the Frequency Response and distortion. I think this caused the edginess of this mic, which could be appropriate on, e.g. guitar.
Joulupukki recorded the FR charts of both the KM84+ and KM84++ using his monitor speakers, which are shown in the snapshot below. I aligned the charts at 1 kHz and applied 1/6 octave smoothing. As you can see from the graphs, the KM84++ has a tad bit more output at lower frequencies. But the major differences can be seen in the upper octaves. As both impedance converter designs are ruler-flat in the upper octaves, I assume the differences should be attributed to the different back chamber and slot designs of the CM-60 and CM-63 bodies. As far as I remember, @kingkorg pointed out the importance of the back chamber and slot designs on the FR and this would only confirm that. Despite the bumpy top-end of the KM84++/CM-63, there were quite some people who liked this one best, which came a bit as a surprise to me. Anyway, for this reason, I want to experiment some more with other nose cones to study their effect on the FR. And I'm also going to test the circuits in the Alctron T-02A, which is much more similar to the CM-60 and the MXL603 in terms of back chamber and baffle design. But one thing is for sure: the 3u capsules will remain in any case! These are really good and do not have that hyped high-end. And off-axis, they are also quite flat.
Jan
Back to the KM84+, KM84++, and the Microphone-Parts SDC-84 as used in this blind test. I will not elaborate on all the technical details. This will be explained on my website later. I'll briefly summarize the differences and point out some points that I think could have influenced the sound of the microphones and might explain what you heard.
KM84+: Takstar CM-60 body with a 3U cardioid capsule and 3U GZT-84 transformer. Circuit design follows the legacy KM84 design, but most parts are in SMT, mainly 0805. Instead of the pad, which the CM-60 does not have, I added a low-cut filter with a jumper to select -12 dB/oct or -6 dB/oct roll-off. Most important for Joulupukki was the RFI filter that effectively suppresses cell phone interference. I expected this design to sound the closest to the KM84.
KM84++: Takstar CM-63 body with a 3U cardioid capsule and 3U GZT-84 transformer. In the KM84++ circuit, I set the JFET to a much higher bias current to achieve lower noise and distortion and a somewhat extended low-end. And I added a CMOS polarization voltage generator, also for higher SNR. Finally, it has a -20 dB pad switch that the KM84+/CM-60 does not have.
SDC-84: I don't know how close the SDC-84 follows the legacy KM84 design. But at least it has a different capsule and transformer than the KM84+/++. And in the mic used in this blind test, Joulupukki removed the feedback capacitor C2. That makes it more sensitive of course, but also affects the Frequency Response and distortion. I think this caused the edginess of this mic, which could be appropriate on, e.g. guitar.
Joulupukki recorded the FR charts of both the KM84+ and KM84++ using his monitor speakers, which are shown in the snapshot below. I aligned the charts at 1 kHz and applied 1/6 octave smoothing. As you can see from the graphs, the KM84++ has a tad bit more output at lower frequencies. But the major differences can be seen in the upper octaves. As both impedance converter designs are ruler-flat in the upper octaves, I assume the differences should be attributed to the different back chamber and slot designs of the CM-60 and CM-63 bodies. As far as I remember, @kingkorg pointed out the importance of the back chamber and slot designs on the FR and this would only confirm that. Despite the bumpy top-end of the KM84++/CM-63, there were quite some people who liked this one best, which came a bit as a surprise to me. Anyway, for this reason, I want to experiment some more with other nose cones to study their effect on the FR. And I'm also going to test the circuits in the Alctron T-02A, which is much more similar to the CM-60 and the MXL603 in terms of back chamber and baffle design. But one thing is for sure: the 3u capsules will remain in any case! These are really good and do not have that hyped high-end. And off-axis, they are also quite flat.
Jan
rock soderstrom
Tour de France
Very interesting Jan! Thanks
joulupukki
Well-known member
Thank you @jp8 for the chance to try out your circuits and great job with them!Now that @Joulupikki has revealed the microphones, I'd like to chime in and tell you something about the KM84+ and KM84++. But first of all, I want to thank Joulupukki for the many hours he spent testing my PCBAs, all the interesting discussions we had, and making all the awesome recordings and Youtube videos since March this year. It was a great learning experience for both of us! But it was also good to receive comments from others, in this case from all those who dared to speak out and left their opinions in this thread or on Youtube. In some cases your comments supported what we experienced ourselves, and in other cases they didn't. Which is fine too. Many thanks to all!
Just a quick note on this…
I also wondered if not having the feedback capacitor installed (C2) would cause a difference in frequency response so I sent Matt an email about it. He wrote back quickly (he’s always great about responding) and said, “The feedback capacitor does not change the frequency response.”
If that’s truly the case, it means that the MP SDC-84 in these tests, at least to me, always seemed to have a hyped high end in comparison to the other three microphones in this test. This was a surprise to me. I’ve got a stereo matched pair of the SDC-84 mics and they sound identical. So, for what it’s worth, that’s that.
Looking back, I wish I could have snuck in a few recordings of my Graeme Woller KM-84 build in Takstar CM-60 bodies but … my interface only had 4 XLR inputs that I could use at a time. Perhaps I’ll make an additional audio-only recordings with the KM84+, GW KM84, and MP SDC-84 just to see how those three stack up.
Oh, and the other thing I didn’t do in this comparison was to record a full bluegrass mix of guitar, banjo, mandolin, and bass using the same exact microphone to see how they’d stack up in a mix. It’d also be interesting to put a limiter and/or do some simple mastering of the mix. That always seems to draw out deficiencies in a mic in my experience – especially if a mic has too much self noise. I’ll have to get on that experiment!
I performed recently a somewhat similar test with a friend, very experienced singer who asked me to build a mic for him. So i thought we blindly shootout bunch of my premium built mics, c12, elam, u47... just for fun i added a few modded cheapies.
The friend knows nothing about microphones, and couldn't care less if it was 50$ modded mic, or 2000$ one. I added he shouldn't be surprised if he picked cheap one over say my premium u47 with original m7 (i reskinned myself, confirmed and measured to original specs).
Long story short, he picked u47 every time. There's something about Neumann. And i really hate to admit it. What made me happy tho was that close second was modded Perception p420, which is confirmed to have 1:1 response of u87.
The friend knows nothing about microphones, and couldn't care less if it was 50$ modded mic, or 2000$ one. I added he shouldn't be surprised if he picked cheap one over say my premium u47 with original m7 (i reskinned myself, confirmed and measured to original specs).
Long story short, he picked u47 every time. There's something about Neumann. And i really hate to admit it. What made me happy tho was that close second was modded Perception p420, which is confirmed to have 1:1 response of u87.
k brown
Well-known member
Matt would have been right if the JFET amplifier would have had an open-loop gain much higher than the closed loop. And if the output impedance driving the transformer would be sufficiently low. That is not the case. Without C2, output impedance rises, moving the RL high-pass frequency up. And if the transformer output is capacitively loaded, e.g. with 22nF RFI suppression caps or extremely long cables, it will reduce the upper bandwidth and increase harmonic and IM distortion. Some snapshots below show the effects described (pictures taken with my phone from PC display). I assume in @joulupukki's SDC-84, only the low end was affected by the removal of C2.
Jan
Attachments
joulupukki
Well-known member
Interesting. So with C2 at 2.2/3.9pF the high pass frequency shifts even higher? …and the mic would have an even higher sensitivity in the high end?
Similar threads
