Friday, July 25th, 2014 | by matthew mcglynn
The third session of our “$1k Ribbon Mic Shootout” features a sensational vocalist named Molly Reed. She sang a song that I believe is by Celia Cruz, called Te Busco. The downloadable session file contains audio samples at 4-inch and 8-inch working distances, and a “control” track from a Neumann U87. Read on for MP3 audio samples that you can audition blind.
Studio & Session Setup
For all the reasons mentioned in the sax ribbon mic shootout, we kept exactly the same setup in the room, with a “dead wall” 30 inches behind the microphones. (Click that sax link for more details.)
We knew that some of these microphones like a bit of distance from the source, but others like to be much closer. We recorded Molly at two distances, 8 inches and 4 inches. The 8-inch distance worked better for most of the mics, so that’s what you’ll hear below, mostly.
As in the sax test, we ran the passive ribbons through a Cloudlifter CL-2 to boost gain. Some have commented that this signal path is somehow unfair, due to the input impedance of the Cloudlifter. I think this is a misunderstanding.
All preamps impose a load on the ribbon of a passive ribbon mic. There is no standard for input impedance; it can range from 200 Ohms* to over 20k Ohms*. My two preamps are rated at 1600Ω and 2000Ω; the Cloudlifter’s 3000Ω rating is no more or less correct than the others.
And it is irrelevant that the actives mics don’t see the 3kΩ impedance of the CL, because their active stages set whatever load on the ribbon is deemed ideal by their designers. At the risk of oversimplifying: the active mics don’t care what the input impedance of the preamp is.
Some have claimed that the über-clean gain of the Cloudlifter gives these passive mics an unfair advantage. We have always tried to use high quality, very clean preamps for our mic shootouts, like the Martech MSS-10 ($1700) in our character voice ribbon mic shootout with Corey Burton, the Manley VOXBOX ($4100) in a voiceover tube mic shootout, or the Hardy M-1 ($2700) in our vintage mic smackdown, the U87i vs. U87Ai test. Had I used one of those high-cost preamps instead, but turned the gain way way down for the active mics, would that have been more fair? I guess you could argue the point.
Whether you have a high-dollar preamp, or a low-cost Cloudlifter, the audio you’ll hear below is representative of what you’ll get from these microphones. Enough said.
Audio File Management
We set preamp gain for every ribbon mic individually, then adjusted RMS levels later in Pro Tools.
Because this vocal passage has a large dynamic range, the RMS approach can yield differences if you spot-check levels at arbitrary points across tracks. That is, even if two tracks have identical RMS levels (which these do), the level of any single section of them (such as the opening words) might differ in level. Beware of this as you audition the tracks; it is a common psychoacoustic phenomenon that the louder track sounds better.
We present MP3 tracks here for convenience of playback, but recommend that for critical analysis you download the session files and gain-match the sections you’re comparing first.
WAV files were output from Pro Tools at their native resolution (24/48), then converted to high-bitrate MP3 via Sample Manager.
“Te Busco” Excerpt Lyrics
Te busco perdida entre sueños
el ruido de la gente
me envuelven en un velo.
Te busco volando en el cielo
el tiempo te ha llevado
como un pañuelo viejo.
Y no hago mas que rebuscar
en lugares tan extraños
que no puedo dar contigo.
Use the audio players below to hear the six ribbon mics at the 8-inch distance, plus the backsides of the R92 and R-101, plus the N22 at 4 inches.
If you’re making a purchase decision, we recommend that you download the session files so you can hear the “control track” (the U87) and the 4-inch track for every mic.Microphone 1
These notes are based on extensive, or arguably obsessive listening to the raw (24/48) audio tracks, through reference headphones.
Let’s start with the control mic, a 2013 Neumann U 87 Ai. I found it to sound very good on these tracks. I can hear the mic’s presence boost in the upper mids, but this boost does not harm Molly’s relatively low-pitched voice, to my ear. Rather, it gives the vocal track a nice sheen. In short: The U87 sounds like an expensive microphone. (As well it should.)
The differences between the condenser and most of these ribbons are less pronounced than what I heard during the sax test. However, the U87 is rolled off in the lows, while most of the ribbons have a more pronounced low-frequency response, even at an 8-inch working distance. Said another way, in these a capella tracks, the ribbon mics nearly all sound warmer and fuller in the lows than does the U87.
A few of these ribbons beat the U87 in clarity, as well. “Clarity” often means high-frequency detail, but that’s not how I mean it here. What I mean is that the best of these ribbons have a 3D quality that makes the U87 sound a bit flat.
Now let’s get into the specific models…
The Cathedral Pipes Seville seems to be the standard-bearer for vintage tone. It sounds thick and rich here, if maybe a little veiled. It sounds like it might be happier a few inches farther away from the source, or with a combination high pass (low cut) and high shelf boost EQ.
Note that the first half of this track is about 2dB louder than typical of the other samples, due to the RMS matching approach.
I also hear a few plosives in this track, despite our use of a Steadman metal-mesh pop filter; I suspect this is a side-effect of the mic’s somewhat exaggerated response below 100Hz. (I did a quick experiment and found that a high-pass filter makes a noticeable improvement to the Seville vocal track.)
The rear side of the R-101 sounds very nice on Molly’s vocal. It is slightly less present around 10kHz than the U87, although that could be easily corrected via EQ if desired. And it is warmer in the bass, due to the U87’s rolled off low end. The overall response is smooth and balanced.
At 8 inches, the AEA N22 tracks the U87’s response remarkably closely over the vocal range. Outside of that range, the N22 has fuller bass response — which is saying something, considering that the N22 has less output below 200Hz than any other ribbon mic in this test — and the U87 has fuller high frequency response.
I prefer the N22 to the U87 on Molly’s vocal. The N22 makes the U87 sound a bit pushed and phasey in the highs. It is a subtle thing, but I think the N22 has more detail, or higher resolution than the U87.
Curiously, I did not find the N22 to sound better at 4 inches than 8 inches. The reduced distance increases the proximity effect, but I didn’t find the 8-inch track to be lacking in bass. The 4-inch track loses clarity. A high-pass filter helps, but not enough — the 8-inch track is superior.
The two sides of the AEA R92 are subtly different from one another here, moreso than on the John Douglas’ saxophone. The rear side has audibly more output in the lows (below 100Hz). But once that is rolled off, the two sides are all but interchangeable, by which I mean you could punch in one for the other and I wouldn’t notice.
Without processing, I think the front side sounds better; the back side is a bit heavy in the bass, making it sound too dark. With the aforementioned rolloff, both sides sound good. But I prefer the N22 to both of them — the N22 has more presence in the mids, and that seems to flatter the source.
The Mesanovic Model 2 is brighter than the U87 on Molly’s voice. That’s not something I’d normally expect to write about a ribbon microphone! And I’m not sure this bright response necessarily helps the mic on this vocal track, as it opens the door to sibilance in a way that most ribbon microphones do not. Listen to the last words of the first line, “entre sueños,” especially the ess sounds.
I am sensitive to exaggerated ess sounds, because they seem to poke out and distract me from the rest of the voice. So, just as I’d reach for a high-pass on some of these other mics, I would want to roll off the highs on this track. As a quick experiment, I set up a 6dB rolloff above 10kHz; it worked very well to bring the Model 2’s highs back to earth.
But to be fair, the Model 2 shows no signs of distortion or overload, and sounds reasonably balanced here (if a bit bright for my taste). And its boosted low bass response gives it a sense of intimacy, as if the singer were standing closer than she actually was.
The Samar Audio AL39 tracks the U87 very closely in frequency response over the vocal range. Like all of these ribbons, the AL39 has more low bass than the U87. Unlike most of them, it has a flat-to-rising frequency response above 15kHz.
But it is not the frequency response that sets the AL39 apart — in fact, in terms of frequency content it is hard to distinguish from the U87. Rather, it is the mic’s response to dynamics that distinguishes it. Loop the very first words of the sample, “Te busco.” Compare the U87 to the AL39, and I think you’ll find that the AL39 track has more life, more vibrancy; the ribbon mic track is more dynamic. It is a subtle thing, but audible.
I hear the same effect from the Model 2 and N22; all three of these make the U87 sound a bit flat or two-dimensional.
What about EQ?
People who love ribbon mics are not afraid to EQ them. Yet we’re presenting these tracks in their dry, unprocessed form. This can be misleading, because in a real session you would use EQ on all of these tracks.
In the notes above I’ve made some comments about EQ changes that I found to be beneficial to specific microphones, just to illustrate the point. In a future installment we hope to present tracks both “dry” and “EQ’d.” For this vocal session, you’ll need to download the session files and apply your own processing.
The audio clips above are MP3s. The format provides a useful balance between sound quality and convenience of playback. If you’d like to listen in higher resolution, download the PT9 session here (544 MB, *.zip).
If you’re not a Pro Tools user, you can still use the above-linked session file; just open the “Audio Files” subfolder of the session folder. Drag them into your DAW of choice.
The session contains 4-inch and 8-inch tracks for all these ribbons, plus the U87 paired with each ribbon, at 24/48 quality.
The following examples illustrate the naming convention:
- “F8-U87-Seville.wav” is the U87 during the Seville test, at a working distance of 8 inches.
- “F4-R92R.wav” is the Rear side of the AEA R92, at a working distance of 4 inches.
As I’ve stated, the U87 sounded very good on Molly’s voice. But I think several of these ribbons sounded better: warmer, more detailed, higher resolution, more dynamic, more emotive. The Samar AL39 and the AEA N22 both delivered such results, without needing drastic EQ to sound great.
Table of Contents
- Introduction to the new ribbon shootout
- Frequency Response and Proximity Test
- Six ~$1k Ribbon Mics on Alto Sax
- Female Vocals
- Two voices, six ribbons (male voiceover)
- Acoustic Guitar (coming soon)
- Drum Overheads (coming soon)