Jump to content

Floating Point v Fixed Point WAV Files


jon_tatooles

Recommended Posts

2 minutes ago, John Blankenship said:

If someone thinks that the use of 32-bit floating point precludes them from the need to understand and apply proper gain-staging, they're akin to the kind of person who would lay an electric hair dryer on the edge of the tub while bathing.   

 

Translation:. Disaster could accompany your ignorance at any moment. 

 

That is pretty clear :) It will make your life easier, but good audio will still require a pair of ears and some quality gray stuffing between them. 

 

Remember the common language version of the 1st Law of Thermodynamics: "There's no free lunch".

Link to comment
Share on other sites

2 hours ago, Boomboom said:

SO, another possibly dumb question: how would I ride my tracks using 32bf ?

I mean the gain on ISO tracks vs, say, L & R tracks according to what I see on the meters ? 

I don't get how it will all ''behave'', especially usual visual references I use for my gain structure..? 

I think I got the theory not too bad, but it's to translate that into my day-to-day workflow in relation to the machine I'm having difficulty to figure out.

I believe the work flow is something like this...

Switch to 32 bit, record as you normally do... Set up your gain staging as normal... if something too hot blows out the track, it can be brought back in post..

I believe that this is best used directly into an input with a mic, and not having radios with limiters.. or track limiters engaged, otherwise it will hit the limiters before the saving graces of the 32 bit can help out...

Probably best used boom into over the shoulder rig, effects and wild dialog quiet to screaming... no limiters, but, your best attempt at trim and level setting.

Until I know more, this is how I see the system working in the real world.. We are about to know more as this progresses I am sure..

 

Link to comment
Share on other sites

So I'm going through this article:

 

32-bit Float Files Explained

https://www.sounddevices.com/32-bit-float-files-explained/

 

And about 3/4 of the way down, I read:

"Files recorded with 32-bit float record sound where 0 dBFS of the 32-bit file lines up with 0 dBFS of the 24- or 16-bit file."

 

So basically, with 32-bit float, or at least with SD's implementation (?), we'll meter the same, right?

 

-20dBFS (or whatever) for tone

0dBFS for full scale

 

And will the meters appear the same, with segments in green, but moving to yellow and then red as we approach/exceed zero?

 

And to be clear, limiters are not an option when recording 32-bit float? I get the "you don't need them" bit, I think. But are they disabled?

Link to comment
Share on other sites

I can mostly just guess here, but:

 

- gain staging before audio enters the recorder is as important as ever, maybe more so. 

 

- once the audio has entered the recorder, there’s not much you can do. Any gain changes are applied digitally. 

 

- the analog front-end likely will have a limiter. 

 

- the digital side will not have a limiter as it cannot be clipped. 

 

- the max output of the preamp relates to 0dBFS (presumably on the meter, to), but you can apply as much digital gain after that as you like, because there is an almost endless amount of headroom. 

 

- yes, I think in everyday work you would just ignore the whole 32-bit float side of things, and can just be grateful it’s there when you need it. In fact, I understand you can switch to record at 24-bit altogether 

Link to comment
Share on other sites

A few things on this:

  • If you are in a "bag drop" scenario, this is an ideal application for 32-bit float recording with the MixPre II. Whether you (grossly) under record or over record, you can scale the signals after the fact as required.
  • When in 32-bit float mode, limiters are not available on MixPre II's. Nor are they required since no analog voltage from a microphone signal can overload the AD converter.
  • Yes, when you are using a wireless system that has its own gain stage, you must set up a valid gain stage for that device.
  • In 32-bit mode, metering is identical to 24 bit recording, so if you see overload, a signal that is above 0 dBFS will be clipped if converted without changing gain post-record to 24-bit. So if you are using an application that does not recognize 32-bit float, signals over 0 are an issue. You would need to adjust gain and convert to 24-bit in a DAW that correctly interprets  32-bit float.
  • 24-bit recording is available, with limiters, and with extremely wide dynamic range, just as we have in the past.

Here are preliminary data on some common editing environments and their 32 bit compatibility (we recommend independent verification for your specific setup, etc.):

 

 

DAW/NLE

32-bit float support

Limitations

Adobe Audition 2015.2.1

Y

None

Adobe Premier Pro 2015.4

Y

Waveform displays can still show as clipped when signal is reduced below 0dBFS.

Apple FCPX 10.4.6

Y

Waveform displays can still show as clipped when signal is reduced below 0dBFS.

Apple Logic Pro X 10.4.6

N

-

Audacity 2.02

Y

None

Avid Media Composer 8.6.5

N

-

DaVinci Resolve

N

-

Izotope RX7

Y

None

Protools 12

Y

Waveform displays can still show as clipped when signal is reduced below 0dBFS.

Reaper64 5.979

Y

None

Steinberg Cubase LE 9.5

Y

Output can still distort if signal is not reduced below 0dBFS (MAC OS Mojave)

 

I believe we are at very early days with this, and I am certain we will continue to evolve how these are handled at all stages of production.

 

 

 

Link to comment
Share on other sites

Thanks Jon, for the clarification. 

 

Two quick questions: 

- when you say „waveforms can still show as clipped“ that really only refers to how they are drawn and not how they sound, right?

 

- Media Composer not supporting 32-bit float is a real bummer, but maybe less so for the target buyer of the MixPre. Is it possible, or can it become possible, to record only the mix tracks in 24-bit, but everything else in 32?

Link to comment
Share on other sites

7 minutes ago, Constantin said:

Two quick questions: 

- when you say „waveforms can still show as clipped“ that really only refers to how they are drawn and not how they sound, right?

 

- Media Composer not supporting 32-bit float is a real bummer, but maybe less so for the target buyer of the MixPre. Is it possible, or can it become possible, to record only the mix tracks in 24-bit, but everything else in 32?

 

Correct. Files are not overloaded, though their waveform display shows how it would be interpreted as integer (24-bit).

 

MixPre II record mode is either/or, so that isn't feasible, though you could perform a re-record/re-mix on the device to re-record the mix at levels below 0 dBFS.

 

Link to comment
Share on other sites

In Avid/ProTools workflows even if the AAF from editorial is 24 bit as long as the source files in the sound roll are 32 bit, and the TC from the exported AAF is intact the dialogue editor should be able to work in the 32 bit session and replace those 24 bit tracks.  Providing the TC is intact when it hits the dialogue editor Pro Tools Ultimate/HD allows you to easily replace that by reading the TC and auto replacing it and importing from the sound roll.  Replacing the AAF files from the sound roll is pretty common workflow practice on larger scale projects anyway for a variety of reasons.  I will say that having the option to record both formats simultaneously would be an excellent future feature to ad to our flagship recorders X3, Deva 24, Scorpio. Etc

 

 

Link to comment
Share on other sites

Join the conversation

You can post now and register later. If you have an account, sign in now to post with your account.
Note: Your post will require moderator approval before it will be visible.

Guest
Reply to this topic...

×   Pasted as rich text.   Paste as plain text instead

  Only 75 emoji are allowed.

×   Your link has been automatically embedded.   Display as a link instead

×   Your previous content has been restored.   Clear editor

×   You cannot paste images directly. Upload or insert images from URL.

×
×
  • Create New...