Small portable recorders have really established themselves on the market now, and my old Zoom H2 (firmware 1.90) is becoming a bit worse for wear. I've had it for a long time and found it to be an incredible useful device to have in my recording kit, even though I only think of it as 'consumer' as opposed to my 'professional' sound kit. That said, when used in the right way, I've got good recordings with it. I always combine it with my 302 or 442 mixer and have it running as a backup recorder when I'm running sound to camera - such that I have a 2 channel version of the mix (if not the iso tracks) incase there is a problem with the link to camera, or the camera itself.
The Nightmare before Christmas
Recently I was using the H2 (and 302 combo) to record some very dynamic sounds and I started noticing some audible clipping on the peaks on the headphone output of the H2 and something very odd about the meters on the H2 - they appeared to be limiting at about -8dB below full scale. So I did the obvious checks of the limiters on the H2 (they were off) and checked my limiters on my mixer and the limiting on the H2 was happening even before my limiters had kicked in on the mixer outputs. Something was not right.
I did all the usual checks, and the recording level was set to its usual value of '77', which aligns the input of H2 to match -18dB tone with the tape output level of the sound devices mixers. So all seemed fine technically, yet the meters seemed to be hitting some form of brick wall limiter and the audio was clipping.
Expect the unexpected
I returned home and started experimenting with the H2 and researching, and found something fairly gutting: when recording to the H2 Line input, the input level control does not alter the gain structure, but just shifts the recording range and "offsets 0dB" - meaning that it clips at a value less than 0dB. It's such a terrifyingly stupid concept and so fundamentally wrong, that I had to fully test it out to confirm what I was hearing. These tests were carried out with me speaking into a MKH60 mic and steadily increasing the gain until the output from the mixer +20dBu.
|H2 Line Input Level 100|
The first test is with the line level set to 100 (it's default setting): -18 dB does not correctly align, on the H2 meters, it correlates to about -12dB But you can see as the signal increases from the mixer, the recorder records all the way up to 0dB(FS); and essentially clips in the usual manner - but means that I'm (in essence) losing 6dB of recording range as this is now 'over' the range of the recorder and with any other recorder I'd just reduce the input gain, such that the range matched my analogue mixer.
|H2 Line Input set to 77|
Now the input recording level has been reduced to 77, which nominally lines up -18dB tone on the scale of the H2. Yet as the signal increases from the mixer, the recorder starts clipping the signal at about -7dB(FS). So, even though I've supposedly reduced the input gain of the recorder, it is now internally clipping the signal at about -7dB(FS) - and this clipping sounds much worse than that at 0dB with the internal limiters enguaged.
And as the recording level is reduced to 60, the effect is even more pronounced.
|H2 Line Input set to 60|
Pay your money / Take your choice
So now I have the dilemma of what to do, I've used the recorder for years for back ups and rarely had to dip into them and also those recordings have been much less dynamic than those I've made recently. There are two ways to work now to try and keep the H2 as a reliable backup: Firstly keep the recording level at 77, and to set the limiters on the Sound Devices Mixers to be cutting in before -7dB(FS) and so prevent the clipping with the H2. But this is then adjusting how I record just to suit the quirk of my backup recorder; which feels like the tail wagging the dog to me. Or Secondly, leave the levels at 100 on the H2 and let the internal limiters (which are not great) on the H2 stop the peaks that are beyond -6dB(FS) from the mixer - which to be fair would usually be few and far between. Neither is really ideal.
I think the simple answer is that I've got what I paid for, the £120 recorder is a consumer piece of equipment with flaws that I have to live with - but at least I know about them now. When (or if) I decide to use it again, I'll have to be very aware of the limitations.
13 January 2013