Tascam DR-05 settings

My guess would be that FCPX, upon splitting the stereo pair, pans both mono files center. Depending on where your original audio levels are (and where are they, and how much are you increasing them?), they are somewhat "doubling" the sound level since it's two copies of the same source. That equals a jump of approximately +3dB.

Your dialog levels should not be that loud. They should be recorded, and mixed, to average between -20dB and -12dB.
 
Alex H. said:
Your dialog levels should not be that loud. They should be recorded, and mixed, to average between -20dB and -12dB.
Click to expand...

I read in instructions manuals and other forums that the levels intermittently in the yellow zone are ok. If I put them between -20dB and -12dB, they are well below that zone.
 
What the you're seeing referred to as "generally okay" means it is still below clipping. It has nothing to do with standardized mix levels, though there are only standardized levels for broadcast and, a little more loosely, film. If your material is destined for the Internet, there is no standard. But even for something as loose as the Internet, adhering to a standard process is still very helpful. It leads to consistency of your produced material.
 
Yes, setting the audio to Dual Mono centers the pan in FCPX. Yes, I find that audio peaks that intermittently show yellow in FCPX still sound fine to me. I bow to Alex's far greater knowledge and experience and do follow that -12 to -20 guideline, but for web video (which is 90% of what I produce) I sometimes end up with dialog that is mostly in the upper half of that range. And, I find the sound is still great even with a few (not frequent) very loudest peaks maybe bumping up to -9 or -6 but never higher.

In an earlier post you asked about the limiter on the DR-05. I just meant that if you're using manual audio levels I would always turn on a limiter if one is available just as insurance against unexpected clipping.
 
ddixon said:
I... do follow that -12 to -20 guideline, but for web video (which is 90% of what I produce) I sometimes end up with dialog that is mostly in the upper half of that range. And, I find the sound is still great even with a few (not frequent) very loudest peaks maybe bumping up to -9 or -6 but never higher.
Click to expand...

-6dB peak is safe for web video.

Again, the web is kind of the wild west of audio/video: pretty much anything goes. There are web videos with sound so low that it's barely audible even with the volume cranked, and videos where everything is crammed right up to 0dB. When I produce for Internet distribution, I create my mix basically to broadcast spec with a brick-wall limit at -12dB. When my mix is done, I raise the overall level to peak at -6dB (again, brick wall limiting). By starting with a broadcast spec, I know that I can get the mix polished for the end user. By raising the mix to a web-safe -6dB peak after it's done, I know the end user will be able to listen comfortably without either straining to hear or blowing out the speakers.
 
Alex, it's good to see you say it that way - make the recording on the safe side, then boost in post. That's what I usually end up doing too because of paranoia about distortion. That small a boost doesn't cause noise floor problems for me - passing traffic is much more a factor than that :-)
 
Rick R said:
I think there is some confusion here over acquisition levels and post levels.
Click to expand...

That should be explained, eh?

There are two general schools of thought that I see on acquisition levels (or, the recording levels during production). One is to aim for record levels that are close to post-production mix levels, so dialog will generally average between -20dB and -12dB with sudden peaks as high as -6dB or even -3dB. The other is to record as hot as possible without clipping, leaving audio post to reduce the level if needed.

Personally, I go with the former. I use -10dB as my target as often as possible, but many digital recorders with sparsely numbered scales have a mark at -12dB instead. It's what I was taught early on with digital recording, and every production company I've worked with expects this. Properly gain-staging the entire recording chain from mic to mixer to recorder (or camera) means ideal signal:noise ratio. So if levels need to be raised just a little in post, there should be no considerable noise that's also amplified. The benefit is that this practice leaves a sufficient amount of headroom which minmizes the risk of clipping. Yes, recording too low can have adverse effects when levels need to be raised in post, but recording in 24-bit as opposed to 16-bit helps to ease that. Recording a the proper level, though, minimizes the need to raise levels in post.

Recording as hot as possible without clipping leaves little headroom. That's a safety net without which I do not like to fly.

As for mix levels, -12dB is also more commonly marked on dB scales in NLEs like Final Cut even though -10dB peak is typical network spec. My mistake for using -12dB exclusively in my previous posts. (Posting from an iPhone near the end of a 9-hour road trip leads to poor though process...) Since I do all my audio post in ProTools, I have meters with -10dB marked. For broadcast and web, I'll use a brick-wall limiter on the mix, and I'll often use compression and the Waves UltraMaximizer for web mixes just to squeeze the DR a little.

Aquisition levels and mix levels may share a common peak, but mix levels often differ in dynamic range. I have done post for networks that want the DR squashed in the final mix. I've mixed shows that had between 6 and 8dB of DR, which is all about perceived loudness and wanting the shows to have an "in your face" sound.

Keep in mind that this is all regarding audio for video and film. The world of music recording has a different theory, and recording and mix levels are different as well.
 
Last edited:
You must log in or register to reply here.
Back
Top