I’m recording voice with the following equipment:
Zoom H5 digital recorder
FetHead in-line preamp
Sennheiser e835 dynamic mic
When recording through XLR, I’m puzzled by the fact that clipping occurs around -6db (as measured with the Zoom’s internal meter), and at 0db when recording through the built-in XY mic. The problem is that -6 isn’t a very loud source and it must be boosted in post, which is problematic for live podcasts, for example.
In fact. Yes that’s a problem. What happens if you take the FETHead out of the system? Dynamic (moving coil) microphones are almost impossible to break and the Zoom is a respected recorder. That leaves…
I’m perfectly clear the volume is going to go down, but does the clip point move? If you yell (Do Not Blow) into the microphone can you hit the expected clip points? I expect the clip point to return to normal.
I’m going to look up the FETHead. I’ve never met it. Most In-Line preamps need 48volt phantom power. Is there something there? Looking.
Koz
There it is.
Make sure your Zoom is set for 48v phantom power.
Koz
Actually I’ve solved this one. I had the “mono mix” function enabled, which seems to have a built-in pad. Disabling it did the trick. Also, when recording via USB, you must choose the “multitrack” option instead of “stereo” because stereo automatically activates the 6db pad. Of course there’s nothing about this in the manual.
As for the fethead, I’m not very impressed. It does boost the gain, but when you boost the audio in post without the fethead, the noise is only marginally worse. So the fethead doesn’t reduce noise, at least with the h5, it just adds gain. Either the h5’s preamp is quieter than I thought or the fethead is adding its own noise.
Noise floor is around -62 when speaking two inches from the mic, and -55 when speaking 4 inches away. (Measure with the ACX tool.) I’m still trying to understand the SNR concept. I’m recording in a booth with no ambient noise except my quiet laptop computer, which a cardioid dynamic shouldn’t hear. So why is the noise floor higher when speaking 4 inches from the mic than 2 inches?
I had the “mono mix” function enabled, which seems to have a built-in pad.
Mono Mix on the Zoom? I didn’t know the Zoom had that. That’s not an unusual feature of mono mix. It assumes you’re feeding both inputs and want to produce a mixdown to one sound channel without overload distortion.
So the fethead doesn’t reduce noise,
Noise is determined by the first thing the microphone signal hits. Noise will never get any better than that save post production noise removal, etc., which can have the problem of distorting the sound.
From the advertisement my impression is they’re solving the very common problem of “running out” when you try to make level with a dynamic or ribbon microphone. I have a Shure USB microphone adapter that I have never used at any setting but full up. I finally gave up and put it in the garage. As I wrote to the company, I would kill for ten dB more gain on this thing.
All of these devices have one common goal: prevent you from sending it back. If they overload, they will produce distorted audio which is immediately apparent. But if they have restrained volume, you will automatically assume you are doing something wrong and mess with it for an extended time, but not send it back.
When I use my sound mixer, there are three different places to set volume and two places to tell if the sound is distorted.
I know you’re itching to run right out and get one of these, right? I shot a panel discussion once where I set everything up and was ready to sit there and ride levels to make sure nothing crazy happened. Everybody else in the room assumed I was going to stand up and go do something else. It doesn’t work that way with non-automatic systems.
my quiet laptop computer, which a cardioid dynamic shouldn’t hear
But even if it does, unless you move something, it should be constant.
Automation can pop up when you’re least expecting it. Does the Zoom have Automatic Gain Control? If it does, your voice will seem to be relatively constant (independent of distance) but the noise level will pump. They do that. The goal is to keep your voice constant, not manage the noise.
All that and you can fake out ACX Check. ACX Check looks for the quietest half-second of performance and measures that for noise. If there is no quiet half-second it measures what it can get. So for accurate readings, you should freeze and hold your breath for a second to give the tool a fighting chance.
A recent audiobook reader submitted multiple test clips with room tone (background noise) segments and not once was quiet during the whole segment. Shuffling, tapping, breathing, gasping, scratching and mouth ticking the whole thing. Unsurprisingly, the “noise” wandered all over the place.
Koz
Unsurprisingly, the “noise” wandered all over the place.
But you have to be there. You can’t leave and cut that room tone into your performance. Your body is part of the sound dampening in the room. The room noise will actually be worse if you’re not there.
Isn’t this fun?
A recent post found find a noise which would pop up at random and ruin a voice reading. Turns out we were listening to their external hard drive spinning up. Does your laptop have an SSD?
Koz
I wanted to get to the bottom of this fishy Fethead gadget, since so many people are raving about it. I’ve been fooling around, comparing the noise floor of recordings made in different conditions, with and without the Fethead.
Here are some screenshots of four typical recordings made one after the other with Audacity via USB.
The first is with the Fethead and the Zoom’s dial set at 4 1/2. The second is without the Fethead and the dial set at 9. These are totally raw recordings: there has been absolutely no manipulation, not even adjusting the RMS level.
The third is with the Fethead and the Zoom’s dial set at 4 1/2. The fourth is without the Fethead and the dial set at 4 1/2. The RMS has been set to -20.4 to put all four recordings on equal footing.
There is sometimes considerable variation for no apparent reason between two recordings made in (what seems to me to be) exactly the same conditions. Compare for example the first and third recordings, where the first recording’s noise floor is 2db quieter, and the second and fourth recordings, where the fourth’s noise floor is 1db quieter. Considering these variations, the Fethead improves the noise floor anywhere from 3 to 6db.
The first is with the Fethead and the Zoom’s dial set at 4 1/2. The second is without the Fethead and the dial set at 9.
Right. And we should be crystal clear the goal of using the FetHead is to allow you to reduce the dial from 9 to 4-1/2. It’s a twenty-some-odd dB volume boost gain block. That is its reason for existence.
It’s slightly quieter than the MicPres inside the Zoom which will surprise no one. The built-in Zoom MicPre’s have never been anything to write home about. The built-in microphones always work better.
Six dB quieter should be about right. That’s probably the noise floor of the four field effect transistors inside the FetHead.
When you reduced the volume of the Zoom, the Zoom’s noise decreased by whatever 9 to 4-1/2 is, so the dominant noise in the system is now the FetHead. The first thing the microphone hits determines the best possible noise. The Zoom noise doesn’t vanish, it just becomes insignificant.
Koz
By my calculation…put down four carry the one…the Zoom noise went from -57 to -84. It’s still there, but is blown away by the new, louder noise value from the FetHead.
Koz
What surprised me is the variations between different recordings made in the same conditions. The Fethead is liable to not make any difference at all in some instances. For sure it will be useless for helping with difficult recording environments.
We’re doing this whole posting so far under the assumption nothing is broken. It might be.
When you’re using the Zoom as a MicPre and digitizer, is it using its internal batteries or the battery on the USB cable? Are you sure?
You never came back to assure us the Zoom is set for 48 volt phantom power. The FetHead likes 48, tolerates 24 and won’t work right at 12. The zoom can be set for all three.
Koz
Actually the original clipping problem was solved by turning off “mono mix” for recording to the Zoom (no computer). I did some other menu magic to get it to record properly on the computer. These adjustments also positively impacted the noise floor, but I couldn’t say by how much.
Since then I’ve just been blabbering about the Fethead. Do you see any other problems that need solving? Personally, I’m satisfied with a noise floor below -60.
When connected to the computer, it’s running off 5v usb. But I have also tested it with the computer off, running on battery power. The noise floor is lower without the computer, but nothing spectacular. With the computer in sleep mode, I can run it on usb power to avoid running down the batteries, and I haven’t seen any difference in noise, so the power coming from the computer is not causing any issues.
Phantom is on 48v.
Phantom is on 48v.
It’s good to know that.
I followed a highly respected consulting engineer around the company. He wasn’t doing anything I couldn’t do, except he took zero shortcuts and made no assumptions. He liked talking about it and I took notes. Most people hire him and he vanishes into the wallpaper until he’s done. I brought coffee.
I’m satisfied with a noise floor below -60.
So is ACX.
http://www.acx.com/help/acx-audio-submission-requirements/201456300
Koz