Review of the Zoom H4N Handy Recorder (for linguists)
The Samson Zoom H4N Handy Recorder is one of the most popular field recorders available today. There are several reviews available on the Internet, mostly geared towards musicians, field reporters, and those interested in recording live shows and ambient sounds of nature. There is also a lot of interest in the Zoom H4N among sociolinguists, field linguists, and oral historians. The recorder seems to combine many of the desirable qualities, such as a small form factor, decent price, good battery life, an XLR interface for professional microphones, phantom power, and the ability to use the recorder as a USB recording interface. However, many questions still remain about the recorder's ability to provide quality-critical recordings of speech.
In this review, I am addressing the needs of speech, language, and hearing professionals, which is why I am going to focus on the Zoom H4N's ability to interface with close-talking microphones, the acoustic characteristics of the pre-amplifiers, and its noise performance. I am, therefore, going to only briefly mention other functionality, and leave out the features that are of little or no importance to us.
Before we begin
If you'd rather skip the review and want to know whether the Zoom H4N can make great recordings of speech, skip to this part. You will find out that, yes, it can, but there are some conditions that need to be met...
Figure 1. The Samson Zoom H4N recorder - front; Figure 2. The Samson Zoom H4N recorder - back
The recorder makes an excellent first impression. Having handled quite a few field recorders, I must say that the Zoom H4N really impressed me in its overall external look and feel. It fits perfectly in the palm of my hand, the buttons are laid out logically, all the buttons, rockers, and dials are in the right places. Nothing is too big or too small. The LCD is large enough, clear, and easy to read. The external surface seems a tiny bit rubberized, which makes an impression of a quality fit and finish. It's still mostly plastic, but it somehow seems better than that. Access to the batteries and memory card (SD/SDHC card) is very easy; all the controls easily accessible, too.
Figure 3. The Samson Zoom H4N recorder - left and right side
An impressive spec sheet
The Zoom H4N has an impressive spec sheet. The recorder has a pair of high-quality built-in microphones, which are designed primarily for ambient sound recording. The H4N can record 24-bit WAV PCM audio and MP3 files at several different bit rates. It can be interfaced with just about any microphone on the planet (at least in theory), including dynamic microphones, condenser microphones (phantom power is available on board), and plug-in power microphones. To top it off, the recorder can be interfaced with a PC and act as a USB hub, though only at 16-bit, 44,1 kHz settings. Musicians will be happy to find a number of useful tools, such as a metronome, a tuner, the ability to record four tracks (two internal and two external mics), and a few digital signal processing options (a low-cut filter, a limiter, a pre-record buffer, and others). For me personally, most of these features are bells and whistles; I am mostly interested in pre-amplifier performance.
Figure 4. The Samson Zoom H4N recorder - LCD readout with real-time levels
It's all in the preamp…
Speaking of pre-amplifier performance, Samson make a very bold statement, which I took the liberty to use as this section's heading: "It's all in the preamp…" Clearly, Samson understand the importance of pre-amplifier performance, and use it as part of their marketing campaign. If you can really pack premium pre-amplifiers into such as small package, at such a low price, the Zoom H4N could be as close to the Holly Grail of field recording as any recorder I'd ever seen. This is precisely why I decided to focus on electrical and acoustic performance in this review, as I am sure most field recording professionals would.
Figure 6. The Samson Zoom H4N recorder - Microphone/line interfaces; power adapter input in the middle
Where the Zoom H4N really shines
The Zoom H4N is an excellent ambient sound recording device. It has a pair of really nice microphones with switchable polar patterns. It can be mounted on a tripod to minimize handling noise. It can withstand high sound pressure levels without clipping. Files can be easily transferred to a PC for further editing. If this were the only use case for the H4N, the review could stop here and declare the recorder a winner.
Figure 7. The Samson Zoom H4N recorder - the built stereo microphone with varying polar patterns
The Zoom H4N has a very good user interface. You start the device by sliding and holding the power switch (you can also use it in the "Hold" position to lock the device down). Several of the main functions are accessible through dedicated buttons or rockers. All other functions are available through a menu-based user interface. The dial on the side of the recorder (Figure 3) functions as a scroll wheel and doubles up as a push button to enter selected commands. The menu tree is very logical and easy to figure out through self-discovery, without the need to read the manual. By the way, the manual is unnecessarily long and complicated. Let's look as some of the Menu options in more detail.
The most important menu options are included in:
Input - phantom power, low-cut filter, limiter, etc.
System - date, time, battery type, etc.
Mode - stereo, 4-channel (two internal and two external mics), MTR (for overdubbing and recording extra tracks)
I really like devices with dedicated buttons for the most important functions. The Zoom H4N does not disappoint. There are dedicated buttons for Play, Pause, Record, Channel, File Type, etc. (see Figures 1 and 3).
The Zoom H4N is one of the few field recorders available today that offers such a broad array of interface types. It has a pair of microphone/line inputs (Figure 7). It's a smart, space-saving design. The microphone inputs are standard balanced XLR inputs (Figure 6); the line inputs are standard unbalanced 1/4-inch inputs. The XLR inputs can be supplied with phantom power for condenser microphones. In addition, the Zoom H4N has a stereo 1/8-inch microphone interface, which can be supplied with plug-in power for use with stereo microphones such as the Audio-Technica AT8022.
The Zoom H4N has a stereo headphone/line output with a dedicated loudness control so you can monitor your recording with zero-latency. This drains battery power somewhat but you can turn it off when not needed. You can also connect a remote controller, a mini-USB cable, and a A/C power adapter.
There does not seem to be a true mono mode available. The Zoom H4N outputs stereo files only, but if you're recording with only one microphone, you can set the microphone to "mono output" which will put the signal in both channels, resulting in so-called "dual mono." This is not a deal breaker for me, but it is odd. If you need to work with mono files because your software demands it (e.g., Kay Pentax Multi-Speech), you will have to convert the files to mono before analysis. Also, if you record in mono, you are forced to "waste" 50% of your SD card space because of the obligatory 2-channel audio file. However, memory is relatively inexpensive today, so it probably is not a serious problem.
The Zoom H4N does not have dedicated recording level adjustments for each channel separately, except in the MTR mode - very odd, and quite restrictive.
The recorder has a 1/8-inch (3.5 mm) stereo headphone input, which is not compatible with many professional headphones, thus requiring the use of a somewhat bulky adapter.
The stereo microphone input is on the bottom of the device, which does not really bother me, but it is an unusual spot for a microphone interface (Figure 2).
The Zoom H4N uses a recording level scale of 1-100. It is unknown to me whether this represents a linear or non-linear scale. It is odd at first glace, especially to those of us who are used to working with a dB scale. However, it turns out to be useful for calibration. Say, you've calibrated your microphone / pre-amplifier combination to find its sweet spot, you can simply remember the level value from 1 to 100 to always easily use the same settings. Well, I am still not sure how I feel about it, but in the absence of a click dial or any other equal-interval gain control, the 1-100 scale will do.
The included effects, such as high-pass filter and compressor/limiter operate in the digital domain only. It is not clear from the marketing literature and can be misleading. Analog filters and limiters are typically more robust than their digital counterparts. While it is nice to have such effects available, I would advise that you turn them all off.
File naming is limited only to either the factory default (e.g., MONO-0001.wav) or date, which is somewhat disappointing. Say, you use the Zoom on location, with different speakers, under different circumstances. I find it necessary to customize file naming to suit my needs. The Fostex FR-2LE has an advantage here.
What we need
I have often heard people ask: "Why pack so many useless features into the device and leave so many crucial ones out?" The answer is very simple: the Zoom H4N is not designed for us. It is designed for mass-market consumption, and its feature set reflects that. Besides, it is simpler and cheaper to offer software-based effects than to design and implement state-of-the art analog electronics. Until recently, it used to be a common misconception to think that digital is always better than analog. While it is true that digital technology is superior to analog tape in terms of frequency response and dynamic range, the analog circuitry is still the most critical part of the design. It also happens to be the most expensive. Phantom power is a good example. It is relatively easy to design and build a phantom power supply unit operating on 110 V AC, but it is significantly more difficult to build one operating on two AA batteries and achieve the same level of quality. Unfortunately for us, the manufacturers have to cut corners, and the marketing department scores yet another win over the engineers.
What we need is a small, rugged, battery-operated digital recorder with two XLR inputs, ample gain, low noise, and a stereo headphone output. We need so little, don't we? Yet, very few recorders have made us happy. Can the Zoom H4N meet our needs? Let's take a closer look at audio quality.
Real-world noise performance
With medium-sensitivity microphones
Linguists, speech scientists, field recordists, oral historians, and others in the business of recording speech, will be most interested in using the Zoom H4N with external microphones. Typical conversational speech levels are relatively low (50-60 dB SPL), so it is of paramount importance to provide the microphone with adequately high gain, without appreciable noise or distortion. For example, the Audix HT5 (reviewed here), a medium sensitivity condenser microphone, requires the recording level of 80 (on the H4N scale) to achieve the peak voice level of around -12 dBFS. The value of 80 is specific to the Zoom recorder and indicates a high gain value, close to the maximum available gain of 100. Figure 8 shows an FFT of the real-world self-noise generated by the Zoom H4N and the Audix HT5 microphone under the testing conditions described here. Let me point out that, for the sake of comparison, I typically normalize the audio file with speech content to the RMS of 70 dB SPL, unless otherwise stated.
Figure 8. FFT of self-noise of the Audix HT5 microphone with the Zoom H4N recorder; recorded at the level of 80; calibrated (normalized) to 70 dB SPL
Microphone sensitivity is directly related to self-noise (see my article on how to measure and deal with self-noise). Some high-sensitivity microphones, such as the Sony ECM88B (11.2 mV), are likely to produce less noisy recordings with the Zoom H4N, but many of us would like to be able to use lower-sensitivity condenser microphones, such as the Shure Beta 53, because of their superb acoustic performance. And what about dynamic microphones?
With low-sensitivity microphones
Dynamic microphones are likely to put even more strain on the pre-amplifier and to have significant difficulty reaching adequate recording levels. Unless the H4N's microphone pre-amplifiers are of exceptional quality, one should expect increased noise levels and possibly even some distortion finding their way into the recordings. The literature provided by Samson does not seem to include the Equivalent Input Noise data (Zoom H4N owner's manual in PDF format), but I do not think that you should worry about it too much. It is surprisingly easy to overestimate the value and usefulness of technical specifications; noise performance being influenced by a number of factors beyond what the EIN value alone might indicate.
The Sennheiser HMD25-1 is my dynamic microphone of choice for recording speech in noisy environments. Being a dynamic microphone, it has the typically low sensitivity of 1 mV/Pa and it naturally requires higher pre-amplifier gain than most condenser microphones. In order to obtain peak conversational speech levels of -12 dBFS, I had to turn the Zoom H4N's gain all the way up to maximum level of 100. Though I was not able to reach -12 dBFS, I was able to record speech at an adequate level for most purposes (RMS of 63.68 dB SPL). However, pre-amplifier noise significantly increased, as compared to the Audix HT5. Figure 9 shows a spectrum of the noise print. The spectrum is normalized to the RMS of 70 dB SPL to allow comparison with other tests on this website.
Compared to the spectrum in Figure 8, the Zoom H4N produces a significantly higher level of noise with the dynamic microphone. Unfortunately, this is expected in this type of device. The Zoom H4N, despite the marketing claims ( "It's all in the preamp…"), is not designed with dynamic speech recording microphones in mind.
Figure 9. FFT of self-noise of the Sennheiser HMD25-1 microphone with the Zoom H4N recorder; recorded at the level of 100; calibrated (normalized) to 70 dB SPL
Compared to Fostex FR-2LE
The comparison between the performance of the Fostex FR-2LE and the Zoom H4N with the Sennheiser HMD25-1 microphone is particularly revealing. The Fostex FR-2LE is much quieter than the Zoom H4N. It is, in fact, really quiet. It would be hard to beat the Fostex FR-2LE's performance in this price range (below $600). The question you need to ask yourselve is whether you ever plan to use dynamic microphones. If not, you probably should not worry about this particular instance of noise performance. For me, however, this is a hard pill to swallow. I am not sure if I would like to be forced to work with medium to high-sensitivity condenser microphones only.
Figure 10. FFT of self-noise of the Sennheiser HMD25-1 microphone with the Fostex FR-2LE recorder; recorded at the level of 100; calibrated (normalized) to 70 dB SPL
We have so far agreed (or I have tried to persuade you) that the Zoom H4N works best with condenser microphones of medium to high sensitivity and low inherent noise. Condenser microphones require so-called "phantom power" to operate properly. Phantom power is generally agreed upon to have the voltage of 48 V. Most condenser microphones are designed to work with phantom power, but some may come with their own battery-powered units. Many modern condenser microphones operate on small voltage values of, say 2 V, but they are supplied with special phantom power adapters (see the Shure RPM626 unit, for example) to make them compatible with 48 V. As I mentioned earlier in this article, high-quality phantom power operated from a low-voltage source, such as 2 AA batteries, is not easy to engineer. It is important to take the marketing claims of 48 V phantom power onboard with a grain of salt, especially when it comes in a small field recorder. This is why I often recommend that you choose a microphone that has the option of being powered by its own dedicated battery-powered power supply, such as the Audix APS911 unit. Not only is it going to save your recorder's battery power, but also it is going to supply the microphone with just the right amount of clean voltage. Sadly, few microphone manufacturers offer such options any more.
I tested the Zoom H4N with and without phantom power to see whether there would be any difference in self-noise performance. I used the Audix HT5 microphone and the APS911 unit. The FFT graph in the left panel of Figure 11 shows noise performance with battery power, while the graph in the right panel shows an FFT of the noise print obtained with onboard phantom power (scroll down to the bottom of the page to download these files). I deliberately did not normalize the files to give you the raw data (both recorded at the H4N level of 80). I took 1 second worth of the noise print in each scenario and obtained the RMS value. Both the FFT and the RMS values demonstrate that phantom power adds a significant amount of noise to the recording. The increase in noise is statistically significant, which does not mean that it should be significant to your specific needs. My personal opinion is that I would much rather use a microphone with its own dedicated power supply, but if I didn't have one available, I would not hesitate to use the onboard phantom power, despite the obvious increase in noise. I'd rather have a slightly noisy recording than no recording at all.
Figure 11. Comparison of noise spectra of the the Zoom H4N recorder and Audix HT5 microphone with battery power (left panel) and onboard phantom power (right panel)
I suspect you would like to know whether you should buy this recorder for your own research use. Let's suppose that you really like the features related to your personal preference: size, the user interface, and the price. Now you want to know whether the recorder can give you quality-critical recordings. The answer is "It probably can." However, you do not have much flexibility when it comes to your microphone choices and recording technique. Here's a recipe for moderate success:
Pick a close-talking microphone of low self-noise and medium to high-sensitivity (e.g., Audix HT5 or Beyerdynamic Opus 55 Mk II).
Do not use on-board phantom power – power the microphone with its own battery power supply or external phantom power supply unit.
Do not exceed the recording level of 80.
Figure 12 shows a spectrogram of the phrase "buy a large barrel of good beer" recorded with according to the recipe above. I used the Audix HT5 omnidirectional headset microphone, a stand-alone Audio-Technica CP8506 48VDC 4-Channel Phantom Power Supply, the Zoom H4N recorder at level 80, and a 24-bit, 48,000 dual mono file type. The recording has excellent spectral detail with no appreciable noice or distortion. It also sounds good in every way.
Listen to MP3 at 128 kbps:
Figure 12. A spectrogram of the phrase "buy a large barrel of good beer" recorded with according to the recipe above. Yes, the Zoom H4N is capable of recording excellent speech signals.
If you want better quality (and perhaps you are like me, and you sometimes do), I recommend that you use an external pre-amplifier, such as the Sound Devices MixPre, and the line-level input of the Zoom H4N. You will end up with very clean recording with perfectly adequate signal-to-noise ratio.