How to match tone with audio levels?
 

I'm submitting a film to a festival and they need ten seconds of tone matched to the audio level of my filim. I'm not sure exactly what this means. I use PP2. Is there anyway I can do this, if so how?

File-new-bars and tone

Roshdi, that was easy. Thanks. One other quick question. Do I set the tone at the peaks of the dialogue or do I look for a middle ground? Right now I have it set at -6 but some dialogue peaks at -1.

http://www-tc.pbs.org/producers/redb...on_8_20_07.pdf

http://www.dvinfo.net/conf/showthread.php?t=106701

So I have read all the linked threads, and I have read pages and pages of other sources over the past couple of years.

I have converted hours and hours of vhs and beta video to DVD for wide distributed use.

However, I still have no clue to the simple direct answer to the question in this thread. So to help, here is the chart from another thread:
http://www.dvinfo.net/conf/attachmen...3&d=1193850410

The first thing I don't get is why, if we are reading digital levels in the NLE and we are outputting digital, why do we have to reference anything to analog levels?

That said, the first purple scale on the right in the diagram seems to be what we see in the NLE and on the digital meters in audio programs. Using this scale, What level should the reference signal tone at the beginning of the digital output from the NLE be?

-10? is that right? counting down from the full scale 0

or maybe -20?

And, separate from that, what would be the correct average level be for the program audio. (It is understood that no peaks can go above 0 full scale.)

So to summarize, based on the full scale used in digital:
What level should the reference tone be?
What average level should the program be?

(Or to put it another way, what levels are the folks at the festivals expecting to see?)

Reference tone can be -20 or -12 depending where in the world you are. Fullscale tone (which I find much easier to understand and is the same everywhere in the world) is 0dBfs

For DVD releases you might do well to also read up on dialnorm
http://en.wikipedia.org/wiki/Dialnorm - even Dolby have problems with broadcasters setting correct levels. (Though it could be argued that's because dialnorm can be variable).

Reference tone should be at -20dBFS (NTSC land) or -18dBFS (EBU). This represents the 0VU level which is also the average signal level of real program material. A sinewave tone at 0VU will read the same level on a VU or a peak reading meter. On the other hand, pink noise, which statistically represents real material, that has an RMS average level of 0VU will read 0VU on a VU meter and about 8 to 10 dB higher on a peak power meter. So setting tone at -20dBFS on a digital peak reading meter and mixing voice & music program so it hovers around -12dBFS with occasional peaks to -10dBFS on the same meter really represents the same average signal level/perceived loudness. This is for conventional analog broadcast and DVD. You are correct that Dolby Digital AC-3 adds the whole notion of dialnorm to the mix.

The only use of -12 as a reference level tone I'm aware of is in the context of mastering music for CD.

The TV docus I have made here in EBU-land had to have the dialog at -20 to -12 dBFS range, all peaks hitting -6 dBFS or louder listed with time code on a piece of paper and absolutelly nothing above -3 dBFS. I quess those rules apply more or less everywhere.

Thank you all for the very helpful links and information. If I ever amount to anything in the film world I'm going to remember Chris and all the thoughtful responses here.
On that note, I have one other quick question. I shoot everything with my HV20 in HD, but I have an SD722T for sound. I read in the PBS link that if it's HD it must be -3 but if it's SD it can be -10. Since the project is in pp2 HD, I'm assuming I should stick with -3 but the sound is 24bit from the SD722 so I half wonder if I should keep it at -10. Any thoughts or comments?

Don't confuse 'never exceed' peaks with average signal levels. Tone still at -20 for NTSC even if HD. HD's Dolby broadcast chain has a much wider dynamic range than does the SD broadcast chain.

Thanks Steve, I'm working on the project now and I have yet another question. Right now I'm going by the master audio levels, which I'm dutifully making sure they don't go above -6 while keeping the dialogue in the -20 to -12 range. Here's my rub. Although the master levels are in range, sometimesthe third track which has dialogue in it goes to 0 at times. Which one do I go by? The master or the third track?

3rd track? Ultimately you should be mixed down to 2 tracks which are rendered to tape or DVD, etc. That's the one where you levels need to be set. If your 'third track' is being mixed into the final stereo mix you still need to make sure it's not clipped since it's going up to zero.

I'm in pp2 and working with 4 tracks that are labelled Audio 1,2,3,4. The master is to the right of this tracks with an L and R channel. So I will make sure the other tracks are not clipping. I figured that was the answer, just didn't want to face reworking things again.
It never ceases to amaze me how often I learn the right way by doing it the wrong way and having to start all over.

Notice also that when summing the tracks the output is higher, two equally loud tracks summed produce a track which is 6 dB higher in level, 4 tracks produce a master which is 12 dB higher. In real life the peaks do not usually coinside, but still you have to adjust the levels to avoid overloading the master.