Standards for normalization levels

[Bob Pierce]

I should have mentioned that the deliverable was via .mp4 files which encode the audio AAC. The videos sound fine on my Macs, but the clients all have PCs, if that’s a clue.

http://www.lightstreamassociates.com

Mac Pro 2.66 – 8GB memory –
Mac Book Pro 2.33 Duo –
FC Studio 2 – Kona Lhe
Adobe Production Suite CS4 –
Sony Multiformat 14″ – Panasonic 42″ Plasma –
Ikegami HLDV7 – PVW EX-1

[Jerry Hofmann]

I’ll bet -1db isn’t enough distortion can happen in digital audio as low as -4db or even -6db… I set my peak average to about -12db and have never had a complaint. There’s not much loss in dynamic range, and it sounds pretty full bodied to me when played back on any computer or tv. I am referring to levels in FCP on the digital audio meter.

Jerry

Apple Certified Trainer, Producer, Writer, Director Editor, Gun for Hire and other things. I ski. My Blog: https://blogs.creativecow.net/Jerry-Hofmann

8-Core 3.0 Intel Mac Pro, Dual 2 gig G5, AJA Kona SD, AJA Kona 2, Huge Systems Array UL3D, AJA Io HD, 17″ MBP, Matrox MXO2 with MAX Cinema Displays

[Andrew Kimery]

-1 is really loud, IMO. By comparison most broadcast work I’m involved with doesn’t allow peak levels above -12 or -10.

-Andrew

3.2GHz 8-core, FCP 6.0.4, 10.5.5
Blackmagic Multibridge Eclipse (6.8.1)

[Michael Gissing]

Yes to make it sound similar to typical content, aim for a lower level and perhaps use a bit more peak compression and limiting to achieve that.

That said, when your clients play a CD on the same media players, they will have to deal with levels that go all the way up to zero so in theory, their system should be able to handle your levels. Many people just turn the level down on their speakers instead of lower the level in the player.

[Bob Pierce]

Thanks for your input, Guys. I understand why you might have -10db for broadcast, but these are video files that will likely end up as streaming video, so I would have thought that mastering hot was the way to go. I’m curious to know what others think about normalization levels for web video.
Thanks!
Bob

http://www.lightstreamassociates.com

Mac Pro 2.66 – 8GB memory –
Mac Book Pro 2.33 Duo –
FC Studio 2 – Kona Lhe
Adobe Production Suite CS4 –
Sony Multiformat 14″ – Panasonic 42″ Plasma –
Ikegami HLDV7 – PVW EX-1

[Adam Duplay]

This is an interesting thread. I have delivered content intended for the web that was mixed at -12dB, as we typically do for broadcast, and received complaints that the web material was too quiet.

Slightly off topic, the same holds true for DVDs. Compare the audio levels of your DVD exported from a -12dB timeline with Hollywood DVD’s and you will see that your audio is not nearly loud enough.

The standard -12dB for broadcast mantra, in my view, does not apply to digital delivery.

Any one else?

Adam

Adam Duplay

[Michael Gissing]

Without proper dynamics control (ie compressor/ limiting) your -12 or -10 FCP project will sound quiet. If you learn to use STP as a minimum, to finalise a mix you will be able to achieve broadcast levels and a superior sound that seems the right volume. If you just bang your levels up to -1, it is likely you will cause clipping and distortion in many inferior computer based playback environments.

The term normalize is misused. It just adds gain to a clip until it reaches a designated maximum. This is very different to dynamics control like compressor limiters that also control maximum levels, but more importantly, assist in controlling apparent loudness. I can tell you that I would be able to make a program that peaks at -10bdfs sound just as loud as your unprocessed ‘normalized’ mix at -1dbfs. Importantly, my mix will not only sound better, it won’t cause clipping distortion on playback

[Bob Pierce]

Michael,
You are right – I remastered the audio limiting it down to -12db while increasing the gain and sure enough, the subjective volume of the sound was just about the same as the less compressed audio mastered up at -1db. One does sacrifice some of the nuance of dynamics, but I think this may solve my immediate problem.

Since I have your attention, one last question about this: I typically do my mix in Final Cut, striving for a good balance between voices, nat sound and music. I then export out a stereo mix aiff file and then do the final normalizing and compressing. The only problem with this approach is that the balance of the audio mix is altered – usually the subtle music is suddenly overpowering. I know I can do multitrack mixing inside STP but I always found it an awkward workflow.

Thoughts?
Bob Pierce

http://www.lightstreamassociates.com

Mac Pro 2.66 – 8GB memory –
Mac Book Pro 2.33 Duo –
FC Studio 2 – Kona Lhe
Adobe Production Suite CS4 –
Sony Multiformat 14″ – Panasonic 42″ Plasma –
Ikegami HLDV7 – PVW EX-1

[Michael Gissing]

I export audio via OMF and use Fairlight with a host of onboard and outboard dynamics & limiting. FCP is a lousy audio edit system as it is frame based editing with little choice in cross fading edits. Also it has no bussing and track based processing.

STP is there and free. It isn’t Fairlight but it is a big improvement on FCP. You must mix whilst listening to all your EQ & dynamics processing. Doing a mix and then applying dynamics later will not sound the same. Apart from that different elements need totally different settings on a compressor. Music needs a slower attack and longer release. Voice needs fast attack and fast release.

And the final balance needs a final compressor limiter with thresholds set higher to finalise and envelope the sound. A staged approach to dynamics is required. FCP cannot do this properly and edit and dissolve accuracy based on frames is too coarse.