No More Gain Setting — The Magic of 32 Bit Float Audio Recording and Why you May NOT Need It

Question. When using 32- bit float it’s said you don’t really have to worry about setting your gain because you can always adjust it in post. However, with 32-bit float you have over -700 db capacity and over +700 db capacity. How do you prevent the noise when increasing the really low audio levels in post since it picks up a lot of room noise?

Integer Bit depth doesn't change accuracy with level, only dynamic range
For floating it changes.

Your explanation further confirmed my understanding of 32bit floating. However my F8 don't have this feature, I usually do dual recording instead, do you think I should also turn the limiter on or it's unnecessary? Thanks

awesome video, thx!

I am trying to get my hand on some zoom f6 or tentacle audio files to see if it is feasible for my workflow.. Do you have any example files on hand i can download? The zoom f6 + tentacle are both on my radar!

Please let me know it’s compatibility with Shure SM7b

does protools work with 32

Gain still needs to be set imho. It affects noise quantity in a recording and that cannot be removed easily later. Again what is recorded and where has large effects on these things too. It's still true that the lowest gain to get the sound you need is best. Gain levels can be increased in post without any more noise being created. Not so during original recording where increased Gain can allow much more noise into the recording.
Yet 32float does mostly remove the issue of clipping during recording.

You said, when the gain is higher you capture more detail, the same I heard in another podcast. When the gain is too low, you only capture let's say 8 bit of dynamic. You can boost it in post without noise, but the missing detail is gone so the 32bit "don't care about gain" is a myth, isn't it? You should set your gain right with 32bit recorders, too?

Incredible video. Thank you for this information!

great information. well done. thanks

Onkel Spiegl, it took me a minute to digest all those technical information. its a loooooot . but very useful. i would love to see an actual test on the field just so we can have a better visual understanding.

German? Haha

Thanks for the video. It help me understand the advantages of 32-bit recording.

Great video as always. Do you know if when you pan hard left and hard right with the Zoom F6 whether it is a destructive modification or does it only affect the headphone mix output. The reason I ask is that I like to play around with the panning in post. Thanks again.

If you record in 32b float (as I always do), once you're in the ballpark of where your levels should be at, would it be "safe" to save those new leveled files to 24 bit for use in the rest of the post-processing chain?

I've always been "worried" that maybe, even if the EQ/Comp/Limiter would maybe only affect maximum 12-18 dB changes, maybe I still need those source 32 bit float in all the post-process until I bounce to a final product (44.1 at 16 or 24 if FLAC or whatever lossless)

As I understand it. We still need a healthy signal going into this box. That is to say, a signal where the sound in the air going into the microphone is loud enough to create a signal over the noise floor but not high enough to smash the physical metals and membranes at each other so they clip the signal at the analog stage. What 32 bit float is essentially doing is removing a necessity to set a floor and ceiling to in the Analog to digital conversion. Or rather. It records the full range in high enough bitrate that you can safely convert it to 24 or 16 bit with enough accuracy in post at whatever level you set there. And it does so with enough headroom and floor-accuracy to cover most microphones dynamic range that you can simulate a gain-knob at playback. What it does record with the sound is a separate volume control level signal that is applied during playback and that is what we can control in post when it comes to gain. In playback that gain is applied and the 32 bit samples are converted to whatever the bitrate is for the playback device.

What I am getting at is that the examples I see about what the 32bit float can do is impressive, but probably mis-representing what is actually being done back-stage in the box and in post. For those examples, the signal being recorded were all in a healthy rather constant level. Then the virtual knob turns it down or up for playback conversion. It is essentially like having a decently produced CD and showing that you can use a volume-knob on the cd-player to set the sound-level. What would be more interesting is if one would record something with a signal that is a bit too quiet and something a bit too loud with and try to save the sound during the mix in post. Because it sounds a bit too much like Zoom and others want me to think I can hook up a mic that needs phantom power, forget to give it phantom power and then I can recover that miniscule sound in post no problem.

I can’t believe I saw someone from the future.

When I lower the gain in either Audacity 3.0.2 (confirmed can edit 32bit float) or AudioDirector 11, it just flattens all the peaks to one equal level. Using a Rode VideoMic Pro+ with a Zoom F2 BT. I specifically got this combo since I record very loud concerts. Any suggestions? I see in your video you just adjust the gain and it works as expected. Thank you!

Very interesting

just a bullshit for people that cant set a gain.

32-bit floating is the RAW in photography. Have been using the F6 for 6 months now. Tips: DON'T plug the F6 to wall socket using Phone Charger, the NOISE is really LOUD. Just use Power Bank.

Very nicely done, Thank you. This may sound like I'm crazy but bear with me. Is there just one form of 32 bit float? - What led me to the question is you mentioned the Zoom F8 uses dual A/D converters per XLR input. Now the Zoom F2 only has a 3.5mm input but offers 32 bit recording and it sells in the UK (with Bluetooth) for only £179. So is the circuitry different? (apart from no XLR input) does it use dual A/D converters? Is the quality the same? - If you have time to answer, that would be great, Thanks, Peter.

Great.

Very useful and informative video!

Great explanation of bit depth and sample rates. Looking forward to more in this F6 series. Liked and sub'd.

I’ve just got the F6 after using the 24 bit H5 for a couple of years and I really like recording or more importantly processing 32 bit float. Nice summary of it, thank you 👍