As a DOP, it is perhaps the ultimate irony of our craft, that audiences will tolerate any degree of bad picture, but never bad sound. From the first frame of a movie or broadcast program, bad sound communicates the feel of an amateur production that will quickly drive viewers to the cinema exit, the refrigerator for a snack, or their iPhone 16e to check their email.

Needless to say, it is a most unwelcome burden for DOPs to have to worry about capturing proper audio. Still, as we’re seeing more and more in the broadcast and non-theatrical arenas, we are having to do just that. From placing a simple lapel mic on talent and patching into the camera, to setting a level, limiter, or pad, these days, on many shoots, we have met the audio recordist, and, sadly, they are us.

Not surprisingly, given this schizophrenic condition, the quality of audio can suffer. For DOPs, a bad cable, loose connection, or a mis-set menu option, can easily escape our notice. How many DOPs when also managing sound have come to discover only later that, despite our best efforts, we’ve inadvertently captured feeble, unusable tracks.

Considering all this, the advent of in-camera 32-bit float recording is a godsend that offers significant operational advantages to today’s solo camera operators.

For years, camcorders and DSLRs/DSLMs have offered multi-channel balanced inputs via a supplemental (usually optional) XLR module. In a sign of a new trend, the Panasonic Lumix GH7 and S1RII camera models also offer a top-mounted XLR module but with a key difference: The adapter module enables 32-bit float recording directly into the camera - a feature that will no doubt become standard in future cameras and camcorders across the board.

Truth be told, 32-bit float is hardly a new creation, dating back to at least 1985 (IEEE754). Largely dismissed by professionals and relegated to the consumer ranks, 32F has languished on the sidelines for decades. Still, given the emerging IP environment and the dramatic changes in how many DOPs work, the operational benefits of 32-bit float audio can no longer be ignored.

For the DOP/ersatz sound recordist, 32-bit float offers a slew of advantages not the least of which is its massive 1528db dynamic range. Compared to 24-bits’ static 144db of dynamic range, the 32-bit signal can float to accommodate extreme high and low passages. The 32-bit float system does not encode data directly from a microphone (or other analog source) but rather describes the magnitude of the inputted signal as a mathematical expression.

In traditional 24-bit recorders, a single analog-to-digital converter (ADC) must accommodate the inputted signal’s entire dynamic range, so audio levels must be precisely set and regularly monitored while shooting to stay within the 144db range.

In 32-bit float, typically two ADCs work in tandem to create a single stream. One ADC is optimized at low-gain to accommodate the loud passages while the second ADC is high-gain biased to better accommodate the soft passages. The chicanery enables the capture of signals with a far greater dynamic range than would be possible encoding the 24- or 32-bit signal directly.

In practical terms, for DOPs operating solo on documentaries and news magazine-type shows, 32-bit float recording means we no longer need to closely monitor audio levels – or wear a cumbersome headset while shooting! To the DOP on these types of run-and-gun programs, the operational advantages of 32-bit float audio capture cannot be overstated.

To actually benefit from 32-bit float, however, DOPs must be sure to maintain an all-copper connection from the microphone into the recorder. If the recorder is in camera this precludes the use of a traditional wireless setup in most cases as the transmitter has its own gain stage that interrupts the flow, and re-introduces the risk of clipping (or under-modulation) into the receiver.  Several vendors have overcome this with stand-alone wireless systems that place the recorder inside the transmitter, or in an additional body worn device that sits between the mic and the transmitter.

Which brings up another critical consideration at the source. Despite 32-bit float’s enormous dynamic range, the clipping of audio is still a risk. A cheap lavalier mic, for example, with a insufficiently strong acoustic element and feeble 96db dynamic range may clip the inputted signal to the 32F field recorder or XLR module – and undermine the goal of the 32-bit float strategy.

While a single system that records directly to the camera is ideal, a dedicated 32F recorder within wireless setups may provide some advantage. Systems with the recorder integrated into the transmitter typically adopt a dual signal flow with the live audio feed being transmitted to the receiver in the traditional manner, whilst the 32F file (recorded to a micro SD) serves as a kind of safety net; the file, if need be, can be utilized in post if the analog levels at the wireless receiver are incorrect and result in clipping or under-modulation. Some also feature integrated time-code generation to help align and synchronize the 32F replacement tracks in post.

For years, professional sound recordists have pooh-poohed 32-bit float as an overhyped gimmick and not the cure-all it portends it to be. They point out that 24-bit recordings provide more than enough dynamic range (144db) to capture everything from the faintest whisper to a roaring locomotive. Regardless of the system, bit-depth, or strategy employed, they say, the quality of the recording can only be as good as the microphone or preamp used, and the skill of a professional recordist.

Still, for the DOP/solo operator with so many other things on our mind, 32-bit float offers a high-degree of solace and convenience. The reassurance it offers and immunity to clipping can be an important safety net, and help us focus on our primary task of creating a compelling visual story.