RMS for video production?

[Vitaly]

Hi,

As far as I know when it comes to audio for video clips a typical practice is using the same master which was done for CD. Obviously it is cheaper but wrong, I believe. Once, I was watching an engineer putting together a compilation DVD and when I looked at audio tracks I saw that all squashed ones where considerably turned down. Unfortunately, we didn't have time to talk.

My question is: is there some 'standard' RMS/Crest factor for video's audio? Or which RMS you follow when you master for video/TV/DVD?

Thanks,
Vitaly.

[bob katz]

Originally Posted by Vitaly

Hi,

As far as I know when it comes to audio for video clips a typical practice is using the same master which was done for CD. Obviously it is cheaper but wrong, I believe. Once, I was watching an engineer putting together a compilation DVD and when I looked at audio tracks I saw that all squashed ones where considerably turned down. Unfortunately, we didn't have time to talk.

My question is: is there some 'standard' RMS/Crest factor for video's audio? Or which RMS you follow when you master for video/TV/DVD?

Thanks,
Vitaly.

For BROADCAST:

HBO and the broadcast networks have very explicit rules. The current rule for HBO (if I recall correctly) is -20 dBFS average, and maximum PEAK -10 dBFS. You read it right: NO PEAKS ABOVE -10 DBFS or the master will be rejected.

This is equivalent to taking your entire metering system and turning it and your monitor system up 10 dB. Nominally, a 10 dB crest factor recording turned down so it never peaks above -10 dBFS.

--------------------------------------------------------------


NOW FOR DVD, for Dolby Digital-encoded material, there is supposed to be dialnorm to deal with this issue, but don't get me started :-). My best recommendations are: Good crest factor, nice dynamics, open sound, clear, unsquashed transients, which means anything between a K-20 and a K-14. But there is already a loudness war going on DVD, use your conscience.

BK

[TinderArts]

The -20dBFS level is the reference for the DigiBeta deck used by many of the broadcast networks in the 90s. 48k is the sample rate for those decks as well. With HD, many new storage options are used but the DigiBeta reference level was followed for compatibility.

[jayfrigo]

Originally Posted by TinderArts

The -20dBFS level is the reference for the DigiBeta deck used by many of the broadcast networks in the 90s. 48k is the sample rate for those decks as well. With HD, many new storage options are used but the DigiBeta reference level was followed for compatibility.

The -20 level comes from long before digibeta and has to do with analog parts of the chain still in use. Remember, broadcast has to be compatible with old, local TV stations, analog dubs, outdated gear with limited headroom, satellites, and all manner of variables and multiple steps between the TOC (operations center) and your home. Even new digital standards are robust, saftey-conscious, and versatile. One major feature of Dolby E is that it can withstand multiple encode/decode passes alsong it's trip to your home. Multiple codec trips through standard AC-3 would be a disaster.

[Vitaly]

Just a 10 db crest factor for broadcast? It means that all my striving for dynamic range is doomed. It is sad.

Thank you all for your answers!

[minister]

just picking up on Jay's good post...

for BROADCAST, there are standards but they differ from network to network in the U.S. in Europe they are fixed at a -9dBFS ceiling, but NZ i do not know.

for THEATRICAL there are different standards and rules for trailers.

for VIDEO release, there are no standards. a DVD movie or DVD-A music release can go close to full code.

Originally Posted by Vitaly

Just a 10 db crest factor for broadcast? It means that all my striving for dynamic range is doomed. It is sad.

are you telling me your music mixes have at least 20dB of dynamic range???

[Vitaly]

Since last year, I'm trying to master my own music (Ambient/New Age) circa -14db RMS and I haven't heard a single complaint from DJs or radio editors. Surprisingly, nearly everyone admires how good it sounds. Inspired by it I started to convince our Hip-hop/R&B artists (officially, I'm 'all-in-one' engineer for a hip-hop studio) to do the same. Now we're at -12db RMS (from -9). Guess what? Again, nobody complains. When I saw that guy putting DVD together I thought 'I can improve my sound easily'. And I hoped that things are close for broadcast. But I was wrong. Anyway, for the next project I'm gonna make a special 'master for radio' which will be circa -14db RMS and see what will happen.

So, to answer your question - no, my mixes don't have 20db of dynamic range, but they tend to

P.S. Some funny story which happened to me. Once, I came to my friend (she's a photographer) to take some pictures of mine for my CD cover. To help her to understand what my music is like I brought my CD. She put it in her CD changer and we started working. In a middle of our session my CD finished and the next CD, which was in a chain, started playing. It was R&B stuff and, as you can guess, it was much more louder then my music. Guess what my friend did? She turned the volume down! Knowing how hard a mastering engineer was working to make that CD 'hot' and seeing how easily all his attempts were crossed over I couldn't resist and started laughing. The irony was that mastering engineer was me.