[hskiray]

Hello,

I made a test record between analog and digital recorders to
understand the audio reference levels but I have a few questions.


1. I sent 1kHz test tone signal at 0VU via analog audio mixer (the
output peak-to-peaklevel is 2.20 volts of 1kHz signal at 0VU) to the
analog BetaCam, DV and XD-Cam recorders.

2. I recorded 1kHz test tone signal and then some speech to these 3
recorders.

3. The signal levels we 0VU on the analog BetaCam and -18dB on the
other digital recorders.

4. I playedback the records one by one and the loudness of the records
were same.

5. I calibrated the reference level of Avid at -18dB and I transferred
all the audio respectively to Avid.

6. The loudness of the records were still same.


Now, here is the important questions:

1. When I capture a music CD to Avid, I saw that the audio level was
around at 0dB. But the levels and waweforms of the sound I receorded
were too low when it compared with music CD. So, what should be done?
Must I normalize the sound as higher as music CD? Or, must I down the
level of music CD?

2. When I finish the editing (render or mixdown) and when I want to
send back my work to the analog or dijital recorders, what should be
reference levels?

3. Should be different the levels of a finished work if this work for
broadcasting or mastering? (I mean if I send it to an analog BetaCam,
its level is 0VU but if I send this film work to a digital recorder,
its level is -18dB or 0dB?).

Would you please help me?

Best Regards,

Selcuk.

[David Gravereaux]

hskiray wrote:
Hello,

I made a test record between analog and digital recorders to
understand the audio reference levels but I have a few questions.


1. I sent 1kHz test tone signal at 0VU via analog audio mixer (the
output peak-to-peaklevel is 2.20 volts of 1kHz signal at 0VU) to the
analog BetaCam, DV and XD-Cam recorders.


In Vrms that equates to log^-1(4dBu/20)*0.775V = 1.228 Vrms. 2.2 Vpp
doesn't sound wrong. I wish I hadn't left my calculator at work


2. I recorded 1kHz test tone signal and then some speech to these 3
recorders.

3. The signal levels we 0VU on the analog BetaCam and -18dB on the
other digital recorders.

Yes, those peak meters measure dBFS, called dB full scale or the level
under full bit. And you see that full bit is placed 18 dB over 0VU.
That is proper.

4. I playedback the records one by one and the loudness of the records
were same.

5. I calibrated the reference level of Avid at -18dB and I transferred
all the audio respectively to Avid.

6. The loudness of the records were still same.


Now, here is the important questions:

1. When I capture a music CD to Avid, I saw that the audio level was
around at 0dB. But the levels and waweforms of the sound I receorded
were too low when it compared with music CD. So, what should be done?
Must I normalize the sound as higher as music CD? Or, must I down the
level of music CD?

Please do yourself a favor and don't eat your headroom. The "less often
used" area above 0VU is called headroom. Don't be a lemming like all
the other mastering engineers with marketing guns to their heads who
must have their peak to average ratios at 4dB.

Personally... the clipping level of your mixer should match the
clipping level of your digital recorder. Then let dBFS fall where it
lands, which probably will be about -18dBFS unless you got a sweet early
seventies Neve which clips @ +3 above 32Vrms (missing calculator
rather than the average 16Vrms. Thus, a Neve 8032 might have -31dBFS
setting! BWaa Haa Haa 8^)

IMO, of course.

2. When I finish the editing (render or mixdown) and when I want to
send back my work to the analog or dijital recorders, what should be
reference levels?

0VU = -18dBFS

Sounds good to me.

3. Should be different the levels of a finished work if this work for
broadcasting or mastering? (I mean if I send it to an analog BetaCam,
its level is 0VU but if I send this film work to a digital recorder,
its level is -18dB or 0dB?).

I say send it clean, uncompressed and with proper headroom maintained
and let them do with it what they do for the policies they have to
implement.

Would you please help me?

Best Regards,

Selcuk.

[Richard Crowley]

"hskiray" wrote ...
1. When I capture a music CD to Avid, I saw that the audio level was
around at 0dB. But the levels and waweforms of the sound I receorded
were too low when it compared with music CD.

Commercial music CDs are notoriously compressed to get
maximum perceived "loudness" within 0dBFS. It is to be
expected that naturally recorded audio will not sound as
"loud" as a commerical music CD.

So, what should be done? Must I normalize the sound
as higher as music CD? Or, must I down the level of
music CD?

That is an artistic decision that depends on what you
are doing with the music from the CD. Now if you were
to come here and say that you are trying to use this
music as "background music" under a dramatic video
of dialog or something, that would be more like a
technical question. But since we have no idea how the
music is used in your mix, it would be only speculation
for us to recommend anything.

2. When I finish the editing (render or mixdown) and when I want to
send back my work to the analog or dijital recorders, what should be
reference levels?

Transfering the edited material back to a digital recorder
via a digital path (Firewire, etc.) is a no-brainer. It will take
care of itself.

For transcribing back to an audio recorder, you would use
the same method of selecting the operating level on the
digital scale for "0db" in the analog world. Typical points
are -20dBFS for professional camcorders, and -12dBFS
for consumer equipment (because of the poorer overall
SNR). Reference: Jay Rose's books on audio for digital
video production, and his recomendations on the production
sound newsgroup: news:rec.arts.movies.production.sound

3. Should be different the levels of a finished work if this work for
broadcasting or mastering? (I mean if I send it to an analog BetaCam,
its level is 0VU but if I send this film work to a digital recorder,
its level is -18dB or 0dB?).

You could start by asking the customer (TV network, broadcaster,
etc?) what their preferred standard is. Many of them have
published specifications.

Otherwise, you could choose a commonly used reference point
like -20dBFS and use that for the tone & bars and specify the
reference point on the label on the tape.

[David Gravereaux]

Richard Crowley wrote:

Otherwise, you could choose a commonly used reference point like -20dBFS
and use that for the tone & bars and specify the reference point on the
label on the tape.

I fully agree.

FWIW, in the spec describing the AES digital transmission protocol[&],
it references two nominal alignment levels for byte 2 - bits 6,7 of the
channel status data that could be used by transmitters.

SMPTE RP155 for -20dBFS = 0VU and EBU R68 for 18.06dBFS = 0VU

[&] page 15, "AES standard for digital audio engineering -
Serial transmission format for two-channel
linearly represented digital audio data", AES3-2003.pdf, Audio
Engineering Society.