Help understanding how loud things should be

I

illusiumd

Member
What's the best way to ask this question in AV-geek-speek? Need to know jargon so I can google ;)

When I go and listen to a great podcast (TAL - Fresh Air) I'm obviously listening to great audio. Then when I go to listen to my audio it blows out the speakers or is too soft.

I understand there's a whole host of other things going with limiting and compression and NR, but right now I'd just like to start with consistent volume. Need some general guidelines for: What DB should voice be at? What DB should the underlying music be at. Thanks!
 
First, I have noticed that a lot of NPR podcasts are mastered really, really low. We stock up on "Wait Wait... Don't Tell Me!" for long road trips, and always have to turn them way up (Wait wait... can't hear you...). Now, aside from that...

1) What dB level should dialog and music hit? Well, what are you aiming for? Is this dialog for film, video, theatrical release, broadcast, Internet, or podcast? The end goal matters.

2) For proper mixing, you need to have a treated room and a calibrated speaker system. When it comes to the difference between dialog and background, well... they shouldn't fight, and the background shouldn't be so low that it really isn't heard at all in comparison to the dialog.
 
Okay, the general "rule" (though standards can still differ from region to region) is that your mix should peak around -10dB, and no sudden peaks (loud noises, explosions, etc.) should go past -6dB. Your dialog/VO will probably average around -22dB with peaks at -10dB, though for Internet distribution some folks like to reduce the dynamic range a little (if you do this, be careful and conservative).

The relative level of background music (underscore) to dialog/VO is kind of subjective, so long as the music and voice aren't fighting with each other, and it also depends on how much the music is driving the piece vs. how much it's just keeping it from being too dry with voice only. For instance, if the piece is high-energy (like a sports promo), the music may be much higher in the mix though still not overpowering the voice. If it's a documentary, the music may sit further back in the mix as a mood-setter.

Assuming you're using proper audio monitors, and that your room is treated and set up for critical listening, monitors should be set up to play back 85dB SPL (SPL = Sound Pressure Level = acoustic amplitude... you can get an SPL meter at Radio Shack). That's for large mixing stages, though. You're probably in a smaller room, yes? So, 82dB SPL? You'll need to get a pink noise file to play back and use to calibrate your monitor level.
 
Shows uploaded to NPR are in MP3 format and have a maximum peak of -6 dBfs. This is the spec for independent productions anyway. It all goes on a big central server and local stations download it from there. Now the regulars may have a different requirement but I doubt it since it's all going through the same grinder.

I don't know of any average spec. and I don't believe there is one.

Alex is giving good advice. The 85 SPL is a film mix stage spec and will probably be VERY loud at -6dBfs in a small room. Edit stations and such generally aim closer to 79 SPL. Really if you set it up so -6dBfs is REALLY loud and you don't change the master level you should be about right. most radio folks I know don't work in calibrated spaces, most film folks do.
 
For Internet/web based stuff, I tend to mix to about -3dBFS or -2.5dBFS in case any transcoding requires more headroom. Just set your brick wall limiter at that point. For dialog, I agree that it *can* average around -22dBFS with peaks around -10dB.
 
Steve House said:
I've always thought Jay was one of the clearest writers around. What do you find confusing?
Click to expand...

Some of the more technical jargon went over my head as a new soundies. But I guess, like video, everything takes time. I probably will have to read this book again to understand it better.
 
You must log in or register to reply here.
Back
Top