recordinghacks



Neumann Microphone Shootout, U87i vs. U87Ai

Saturday, March 19th, 2011 | by


Vintage mics are always better!

Aren’t they?

Neumann U 87 AiNeumann U 87

In this episode of The Microphone Show, we put a vintage Neumann U87i against the current production version of the mic, the Neumann U87Ai. Our U87-vs-U87 shootout included sung vocals, voiceover, and acoustic guitar.

What’s the difference between the mics?

U 87 AiIn short, the original U87 had no DC-DC converter, and so the capsule was polarized at about 46V, which is all the 48V phantom power (or onboard battery pack) would provide. The ‘A’ revision replaced the battery pack with a DC-DC converter, allowing the mic to generate the 60V that the K67 capsule was designed for.

The voltage change raised the microphone’s sensitivity significantly — from 8mV/Pa to 28mV/Pa (in Cardioid mode), which is worth about 10dB of output level. It also increased the signal-to-noise ratio by 3dB.

Signal Chain

Because the U87Ai’s output level is ~10dB hotter than the U87i’s, any comparison of these two mics requires a sonically neutral preamp. Otherwise you’d be hearing 10dB of color from the preamp on the U87i track, which would likely mask the subtle difference in the sound of the two microphones.

For this reason we sought a high-fidelity, neutral preamp. Typically we’d reach for a Millennia Media pre, as the sonics of the HV series gear are well-regarded. For this test, however, we wanted a preamp with a transformer on the input, due to the widespread conviction that “Neumanns love transformers.”

(Perhaps we’ll test that notion in a future episode.)

John Hardy M-1 PreampWe went with the Hardy M-1, a 4-channel pre that meets all the required specifications:

  • Transformer input (using Jensen JT-16-B transformers)
  • Neutral sonics
  • High fidelity
  • 4 channels of awesome

The pads and filters on both mics were disabled. Both were set to Cardioid for all tests.

The M-1 was run directly into the studio’s HD converters. We adjusted preamp gain to produce equal signal levels from both mics in Pro Tools.

Audio Files

We output 24-bit, 48 kHz AIFF files from Pro Tools. Downsampling (to 44.1 kHz), bit-reduction (to 16-bit), and format conversion (to MP3) were done in Bias Peak.

Voiceover – Mark Keller

Neumann U 87 U 87 i AIFF Male voiceover: U87
Neumann U 87 Ai U 87 Ai AIFF Male voiceover: U87 Ai

Sung vocals – George Merrill

George sang a chorus from “Waiting for a Star to Fall,” a 1988 Billboard hit from George’s duo, Boy Meets Girl. (See the original video here.)

The Vocal samples are RMS matched to within 1/10 dB — and yet phrase-for-phrase there are audible level differences. If you want to analyse these closely, open the AIFF files in your DAW, select a short phrase from each, gain-match them, and then compare. (See below for one such excerpt.)

Neumann U 87 U 87 i AIFF Male vocals: U87
Neumann U 87 Ai U 87 Ai AIFF Male vocals: U87 Ai

Acoustic Guitar – Mark Keller

Neumann U 87 U 87 i AIFF Acoustic guitar: U87
Neumann U 87 Ai U 87 Ai AIFF Acoustic guitar: U87 Ai

The Listening Test

I found it interesting that the members of the listening panel had such different perceptions of these two mics. And yet, in the end, everyone agreed that the differences were so subtle that a mismatched pair of U87i/U87Ai could be used for stereo recording without problems.

Listening closely with headphones, I can hear a difference in the vocal tracks between these two microphones. The Ai has ever so slightly more air. It sounds more “open” at the very top end. The older mic, in comparison, sounds a tiny, tiny bit more compressed. The effect is most noticeable on the voiceover tracks, but I can hear it on the phrase “be what you” in the sung vocal line too.

I’ve isolated some excerpts to make this easier to hear. In both the tracks below, you’ll hear the U87Ai, then the U87i, and then the whole thing repeats:

Taking a Stand

In the video, I stated a preference for the newer mic, without going into detail. What I hear from the Ai that I prefer is its additional “air” on the vocal samples. On acoustic guitar, I don’t hear enough difference to identify either mic in a blind test.

Couple the slightly more appealing sense of air with the higher output level and lower noise of the U87Ai, and the choice is clear. For my money, the U87Ai beats the vintage model. I am no longer holding out for a U87i. The new model delivers the characteristic Neumann U87 magic with superior performance specifications.

Special Thanks

We owe a great deal to the people who contributed valuable gear for this test:

Kenny Evans of Mesa Recording for the loan U87Ai and the Hardy M-1.

Mike Pappas, Chief Engineer of Denver’s KUVO Jazz Public Radio for the loan of his personal U87i, which was serviced by the Neumann factory immediately prior to our shooting date.

Christopher Currier of Sennheiser USA for arranging the loan of Mike’s U87i.

Thanks also to the artists who contributed to the session:

Mark Keller of Loudvile Studios for his guitar and voiceover work

Composer George Merrill for amazing first-take a cappella vocals. See George’s full discography via AllMusic.com.

And finally, thanks to the cast and crew!

Jody Banks, Technical Director
Josh Petersen, Technical Engineer
Nissa Brehmer, Photography & Camera Operator
Alphonso Suerte, the masked announcer guy
Paul Simmans, Audio Engineer
Stephen Hart, Audio Engineer
Ana Sophia Dunham, Audio Engineer

Mathew Trogner is the Producer and my partner in crime. Mathew handles the engineering, the direction, all the video editing, and another hundred details I don’t even know about. If you like what you see, it’s largely due to his work. Thank you, Mat!

A word from our sponsor

Front End AudioThe production costs of The Microphone Show are nontrivial, so I am pleased to say that Episode #2 is sponsored by Front End Audio. I’ve worked with the Front End Audio guys for a while, and I can say they have some of the best prices in the industry, and their service is second to none. If you’re in the market for a U87Ai (as I am!), your first stop should be Front End Audio’s Neumann product listing.

Your Turn

What’s your preference? Which U87 did you prefer?

By the way, within a week after the filming of this video, Paul Simmans bought a brand-new U87Ai. And I’ll be next, unless Mathew beats me to it. 🙂

Posted in Microphones, Shootouts, Video, voiceover | 27 Comments »




27 Responses to “Neumann Microphone Shootout, U87i vs. U87Ai”

  1. Ian!

    March 21st, 2011 at 2:29 pm

    On the whole, my vote is for the Ai.

    I feel like it has more depth to its sound. The Ai definitely has a top end sparkle that the older model doesn’t. That really translates to difference in perception in the bottom end/low mids for me, but on a second listen, that isn’t really too big an ACTUAL difference, just a perceived one.

    On acoustic guitar, I can see either one working wonderfully, depending on the mix.

    Vocals – same thing. For his voice, I would probably pick the Ai though.

    Voice-over – I’d tend to pick the older one for more a classic radio sound. That slightly understated top end is more reminiscent of a SM7 or RE20 than the Ai is. But again, it probably depends on the voice.

    If I was shopping for one? Ai. To my ears, notch out a few things in the top end and you can really approximate the sound of the original.

    It also doesn’t hurt that I’m REALLY used to the sound of the Ai. Use them all the time on Orchestras at the University.

  2. Steve Faul

    March 23rd, 2011 at 12:09 pm

    Interesting how people seem to prefer the vintage for VO’s. I agree. The difference is very subtle, but the Ai had just a touch more upper-midrange presence to my ear. It’s so subtle, I wonder if just putting a windscreen on the Ai would give you the vintage sound. Maybe. Either one would be a great addition to my studio.

  3. Jordan Reynolds

    March 24th, 2011 at 7:04 am

    I agree with you Matthew. I actually prefer the Ai too. It just really has this warmth and presence that’s not as apparent in the u87i. I guess this is good for us considering the u87ai is much more accessible and probably cheaper!

    By the way: EXCELLENT work on the production of this video. I was extremely impressed with everything from the camera work, sound, editing, content, etc. High quality content! Well Done!

    -Jordan

  4. Stephen

    March 29th, 2011 at 6:08 am

    Anyone else wonder what the Neumann factory “serviced” on the older 87 before this test? If you know Matthew, please tell.

    I raise this question because many people claim the current Neumann stuff is not up to par with the old stuff (obviously your comparison suggests otherwise). So I’m curious if any “servicing” could have actually brought the older 87 toward the new 87s sound (as opposed to an unserviced or serviced by a 3rd party).. I know I’m really shooting in the dark here, and I do hate to go on hearsay. Neumann lineage is not something I know a ton about although their recent release of a mic that’s supposed to be reminiscent of a “67” but contains no tubes does kinda make me scratch my head. (i have a 67, fluke “inherited”, and the compression, particularly on drums is VERY noticeable).

    Excellent review! again, it’s probably a trivial detail above, but i’m just curious.

  5. matthew mcglynn

    March 30th, 2011 at 7:31 pm

    @Stephen, if your concern is that the U87 is an Ai in disguise, one look at the PCB will confirm that it is not. There’s no DC converter, but there is a battery pack.

    The reality is there are over a dozen U87 circuit revisions, both before and after the official ‘A’ designation. I don’t know whether any of those intermediate revisions made an audible difference in the sound of the mic, but certainly the ‘A’ revision does. Whether you like the ‘A’ sound more or less than the U87 sound is subjective.

  6. Chas Payson

    May 12th, 2011 at 9:00 pm

    For what it’s worth, I have a circa mid 70s U87i and a late 80’s U87ai. The 87i was re-capsuled a few years ago by Neumann. It sounds wonderful and forgiving it it’s reduced sensitivity is our go to mic. slightly more full bodied and well rounded than the newer model.

    I suspect the wear and tear on the capsule may have more to do with the sound difference when gain matched and will be curious to see how our 87ai responds when I send it in for capsule service.

    Furthermore, I can definitely say that on my little Macbook speakers the 87ai blew the 87i out of the water on the guitar when watching the video. So much so that I think I may have permanently strained my eyes trying to see if it was mic’d in a different spot.

    Really enjoyed the shootout.

  7. Justin

    May 17th, 2011 at 11:27 am

    First of all, great clips. Thanks for hosting this.

    I feel two of our panelists was using the terms “warmth” and “body” in an unconventional way when they suggested the ai had more of either. They also and used the word “brighter” in a kind of backwards way as well.

    In my opinion, Matthew Tragner and Paul Simmans’ ears were spot on. However I’m going to have to respectfully disagree with 98% of the comments that came out of Stephan Hart’s mouth. I feel Tragner and Simmans were making the opposite (and correct) statements when compared to Hart, but managed to sidestep calling attention to their fundamental disagreement about the mic’s tonal colors.

    Thankfully, we all have our own ears. You be the judge! Again, great shootout. It’s fascinating to note how differently each of us use language sometimes…

  8. C Bret

    May 18th, 2011 at 10:28 pm

    Using the laptop speakers, I preferred the older mic. As far as the low-mid bump, it seemed around 200hz on the AI which was not at all unpleasant. The AI also seemed to have more of a 4k boost. These two factors gave it a “windy” sound to me – something I would use for pop female ballads. Drums, ac guitar, and male vocal, I would lean to the 87-i. The “fanciest” mic I own to compare is a Rode K-2, which sounds fuller ( 87-i) as it ages, and has been serviced by Rode, also to it’s benefit. All of the recordings are beautiful, and you are welcome to send either mic and pre to……..

  9. Rick Spyder

    November 7th, 2011 at 4:57 pm

    What is recording all about?
    To me it’s capturing the emotion detail and character of the source.
    I was secretly hoping that the vintage u87i would satiate the expectancy of “older is better”.
    As a u87ai owner I was pleasantly surprised that i loved the sound of it in every situation
    and feel confident that it fulfills my criteria of a mike that can make great recordings.
    Cheers everyone.

  10. Torben Lysholm

    November 29th, 2011 at 3:18 am

    Thanks for testing and putting all this stuff out. Excellent! It’s very helpful and interesting.

    Are you absolutely sure it’s the same capsule in both mics?

    Listening to them, I’d agree. However, the old schematics for U87 and U87i specify a KK87 capsule and the U87Ai has a K67 capsule in it.

    If your U87i has had the capsuled replaced by Neumann at some point they have maybe replaced the KK87 with a K67 and it’s not really in it’s original form anymore?

    You often hear people liking the old version “a lot” better. To me, your test here suggests that “a lot” of them would probably disagree with themselves in a blind test. (Love it when that happens!) BUT knowing that the capsule means everything to the sound of a mic, maybe this test doesn’t shed the right light on whether a vintage U87 sounds better than the new Ai model.

    Note how Peluso distinguishes between the capsules K87i and K67 (Both replacements for U87:
    http://www.pelusomicrophonelab.com/parts/Capsules.html

    Any thoughts? Should there have been a different capsule in the U87i to make it the test you really wanted?

    Again, cheers to all of you for testing for all of us. 🙂

  11. Torben Lysholm

    November 29th, 2011 at 3:39 am

    Just investigated further and found more information.

    You mention in the video that the U87i has to be gained more than the U87Ai. How is the impedance in the U87i? Its manual states that “by changing the soldered connections on the output transformer” you can reduce the impedance from 200 Ohms to 50 Ohms. Not only will that mean a 6 dB drop in level, it may also very well alter the sound slightly. I experience that often when building gear myself and trying out different transformers in different configurations. The slight difference you all notice but also agree is very subtle could certainly be attributed to the difference in impedance/configuration of the output transformer.

  12. Mike Nott

    February 11th, 2012 at 11:32 am

    Did you find out what year were both microphones from?

    For the ai – was it a brand new one?

    Btw – great show!

  13. Mike

    March 3rd, 2012 at 9:53 pm

    I would prefer the old 87i every time. The Ai was ok at vocals, but sounds quite harsh.
    The old u87 has that “timeless” factor lacking in the new one.. as a musician, that’s a more desirable, even if the new one is more “detailed”. Not saying that the new Ai is a bad mic, but maybe I’d prefer something like Blue Kiwi to that, which is a little cheaper.

  14. John

    October 8th, 2012 at 10:17 pm

    Yeah, the old one is way less hyped, and sounds more natural. Does a great job of filtering out the subtleties I DON’T want to hear.

    The AI just sounds on edge, tense, and angry to me. This is after a LOT of listens on B&W’s. The old one sounds more to me like you hear it in a room, the new like your ear is against a speaker. The I nice and relaxed and comfortable, the AI full on “YOU WANT AUDIO?!!!! HERE’S YOUR AUDIO!!!!”

    Once again, specs do not tell the whole story. The ai should be better, just like digital should have been better than analog, but to me, it just isn’t. Ya know…sometimes you can hear too much. Just like you can see too much.

    Difference between soft warm ambient lighting, and a 3000 watt spot blaring on the subject.

    Also, what was dude on couch (producer guy) hearing? I hear the exact opposite. Dudes behind the console? Those guys have ears.

    I feel like the producer on said couch got fooled by the oldest newbie listening perception trick known to man, and that is that more is better. It has to be right? More polarizing voltage, more output, more high mid and air / treble. Mo is betta yo!!!
    Um…not for me in this instance.

    “More” often sucks the “class” outta stuff real quick. People back when the I was made, listened like their life depended on it. The AI seems like it was designed for improved specification, just cuz they could.

    More is better.

    Great test, and thank you very much for that. Producer guy included.
    Love the site!
    john

  15. Daniel Gildenlöw

    February 12th, 2013 at 5:10 am

    Preferred the 87i by FAR. It had a nice and compact sound without coming across as “boosted” in top and bottom. The new one felt more “aha” in a way that I would not like to have in a mix – it tries too hard to impress, if you catch my drift.

  16. Heinrich Hanf

    March 10th, 2013 at 4:28 am

    As a musician I also prefer the U87i especially when it comes to recording female vocals. I’m working for months now together with an outstanding opera singer (dramatic & lyric soprano) and she simply fell in love with the remarkable natural sound of my old Neumann microphone, which makes it very easy, to place her voice in the mix with orchestral music. Top classical singers don’t like their voices to be “idealized” by boosting highs or lows.

  17. delacroix

    March 31st, 2013 at 9:28 am

    Well, the sound of the new U87AI is too much present ? so juste sing half a meter from it ! The sound of the new U87AI has +1db too much of high frequencies ? Put a mask between the mic and the mouth or cut with an Eq.
    I wasn’ t a big fan of the original U87i despite I use a tubetech and a big Neve mixinx desk. The new U87ai is an improvement for me

  18. delacroix

    March 31st, 2013 at 9:39 am

    Also, it’s funny to read people making a choice for hi end microphone just listening with their laptop and with different takes ! This test is a no sense, you should have recorded EXACTLY the same performance with the 2 microphones. Actually they are not the same, the frequencies change , so how can you hear the very small details between the 2 microphones ? That is just Impossible

  19. delacroix

    March 31st, 2013 at 9:45 am

    One question , was the singer exactly at the same place between the takes ? As you should know , the sound of a microphone change a lot , sometimes juste with 2 inch … You will have more room or more voice

  20. steven

    August 15th, 2013 at 2:57 pm

    Secondly, the test is completely subjective. Unless your telling or assuming that we can absolutely be guaranteed that in all cases the guitar, the singing, the vocals, exactly the same pitch and durations where achieved, to the microsecond by the performer’s voice and or instrument, and to the millimeter the position of the performer to the the mike, those factors WILL always mean the outcome is always subjective.

    Of course one could use digital technology, (yes even for voice), as well as maintaining exact distance from the U87’s, to guarantee, that the playback was exactly the same in both tests.

    Then comparing the signals recorded on both U87’s would be a simpler task and a much more rewarding one.

  21. matthew mcglynn

    August 16th, 2013 at 10:06 pm

    @steven, all gear tests are subjective. This is as true in your studio as it is in ours. If you never test two different mics on a singer because such tests are useless for comparing tonality, musicality, etc., then more power to you — you can get by with a single mic for every source, I guess.

  22. James

    September 14th, 2014 at 11:33 am

    Thanks for the ear candy. To my ears the ai sounds very close to the original on vocals, although it has slightly less low end and slightly more high end ‘air’. Both have the trademark 87 forward mids though. On the acoustic guitar however I MUCH prefered the original. The ai sounded much more brittle and thin and faster attacking, it lacked lacking warmth and veers towards ‘plasticy’ imho.

  23. John Peluso

    October 14th, 2014 at 11:54 am

    The difference in capsules from the U87i to the U-87A is that the older I uses a 4 wire capsule with both diaphragms back plates isolated electorally from each other. The newer U-87A uses the same capsule as the U-67 with both back plates connected together thus only a 3 wire capsule.

  24. Melani

    August 10th, 2015 at 6:32 pm

    Ai is clearer and more intelligible in every way. It’s not even close. If that were a mic pre – not a microphone – we’d all be wondering what happened to the 12k+ range on the pre and be sending it back…

    Horses for course of course…

  25. Mike F

    June 14th, 2016 at 6:00 pm

    I prefer the older mic in all three tests. Much more natural. The only difference I hear is the peak at 10k on the ai, which I never liked (I have one, and usually just eq it out). I guess one person’s “air” is another person’s “fizz”.

  26. Bill Brekke

    June 23rd, 2017 at 7:43 am

    I think the hissing sibilants are harsher on the newer mic.

  27. Rob Ellis

    November 12th, 2017 at 2:59 pm

    For voiceover, I use an 87ai that is about 17 years old based on serial # (in the low 80000s)….for this application it has never let me down. Clear, forward and elegant without being harsh. The only time it is ever harsh is when I am.

    In MY world for VO it has no LDC peers and believe me I have tried them all.

Leave a comment