Sound Device MixPre3 II tested with live birds? (1 Viewer)

Hi
Wonder if any out there had a chance to test the ability to record in 32 bit float (file )and especially interested what happens to far away birds being pulled up in the computer afterwards? Does noise etc. kill the birdsound or is it really as good and noiseless as they claim.
Myself using Zoom F4 and Telinga parabolic MK2 in field with a maximum of 24 bit files. Using mostly Audacity
Stein
See photos and listen to Northern Birdsong at (also a quick guide to filters in Audacity
http://fuglesang-troms.net

The usable signal will still be limited by your microphone. The real benefit of 32 bit recording is, you don't need to pay attention to recording gain. You can set a low gain in order to avoid distortion for loud signals and still get a really useable recording.

So, no worse than 32 bits of course, and more forgiving.

As stated, you limit with a Mixpre mkii will be your microphone. 32 bit float will take away worries of clipping as it allows for huge headroom and the ability to “pull back” in post.
32bit will also allow for more forgiving flexibility in processing To a degree.

If you are used to the standard type of recorder such as the Tascam low to midrange gear then the Mixpre can be a little daunting at first, it is a mixer/recorder so plan into your workflow some time to familiarise yourself with its operation.

I’m using a Mixpre 6ii. It is a truly awesome device that affords enough channels to produce some stunning results with the stereo image.

Another important consideration with 32bit recording is file size! While file storage these days is not an issue due to cheap storage media you will need to think about the extr time required to post process they really big files.

32-bit for birds

As an SD MixPre 3II user of eight months recording mostly birds, I can add my recommendation for 32-bit float. I was almost to the top of a steep sagebrush hill at daybreak one morning recording a couple of birds 50 feet away when a feathered singer opened up full throttle in a bush 8 feet away just over the top of the hill. That was a moment. When I opened the file later, the call of the spotted towhee was a good 6dB past the point of clipping. Instead of cutting the gain for the whole recording, I selected just the towhee's energetic bursts and pulled them back down unclipped just as clear as can be. The sounds of the other birds remained clear and further back in the stereo image. For best results, listen with headphones.


I like the observation that 32-bit recording does not replace top-notch mic preamps or low-noise, high-sensitivity mics. As the gateway to a recording, things won't get any better than the mic specs followed by the preamp specs. The good news is, both mics and preamps are getting better. Before I dive into thousands of dollars worth of high-end mics, I decided to try out the remarkable Clippy XLR EM172 lavalier mics from FEL Communications. Nearly as high sensitivity and low noise as some of the big mics. Better specs than my old DPA 4060s. I fashioned an Olson Wing mic array to mount the mics in and I was off to the races.

Right now the only place you'll get 32-bit float capability is from the Sound Devices MixPre line or the Zoom F6. I went for the MixPre. Macaulay Library of the Cornell Lab of Ornithology liked it the best out of the recorders they tested. Curtis Judd found little difference in the recordings made with the F6 and the MixPre II although the differences in the physical forms of the two recorders pulled him towards the MixPre II. If you need more than three XLR inputs but don't want to spend more than for the MixPre 3II, the Zoom F6 offers six XLR imputs for about the same price. Personally, I really like the MixPre II automatically backing up my recordings to a thumb drive that I can easily pull out at the end of the day and transfer to the computer. Happy recording.

I have decided to switch to a 32 bit float recorder, but primarily so that I can rescue clipped recordings (as described by Kwesting). I am interested in recording bird calls, which can be intermittent and given by moving birds, so it can be hard to get recording levels right!

Prior to purchasing a Mix-Pre, Sound-Devices sent me the following link, which shows the advantage of 32 bit float when the recording level is set to low https://www.sounddevices.com/low-signal-32-bit-float/

Sound Devices acknowledge that the example is a rather artificial/exaggerated. You would need to get your recording levels very wrong indeed and need to apply a huge amount of boost, to see this impact on the noise floor and the 'piano key' distortion in a 24 bit recording - in the example they apply >100db boost, which I think equates to boosting the signal over 8 billion times! You would also need to have very low internal noise in your system to really benefit - if you mic is noisy then 32 bit cannot magic the noise away. I think more importantly, you would need a really quite environment - if you boost the recording then wind, traffic, handling, or even other distracting bird noises are all going to be increased as well.

I am not an expert, but to record distant birds, I would opt for a parabola which will boost the signal of the target bird, without boosting other environmental sound, and set the recording levels as high as possible. Then (if necessary and with care) post process the recording to boost the target only - software such as Wavelab Pro (quite expensive) have spectral editing tools that allow you to select part of a sonogram and boost or reduce the levels of the selected area only (not sure if Audacity has this capability). I would imagine that 24 bit would be quite sufficient to boost part of the signal by 3-6db (2 to 4 times) without seeing a noticeable difference compared to 32 bit.