Mark,
Thanks. Trouble is, from what i've learned is that True Peaks are frequency
related (as well as loudness is).
I'll see what i can do!
Bouke
VideoToolShed
van Oldenbarneveltstraat 33
6512 AS NIJMEGEN
The Netherlands
+31 24 3553311
www.videotoolshed.com
----- Original Message -----
From: "Mark Spano" <cutandcover@gmail.com>
To: <Avid-L2@yahoogroups.com>
Sent: Monday, October 22, 2012 4:36 PM
Subject: Re: [Avid-L2] Re: EBU R 128 / Calm act, HIGHLY COMMERCIAL
> >
>> I have no idea how loudness an true peak relate, (and in fact i think
>> they
>> don't relate at all...),
>>
>
> Correct. True peak means just that - an actual value of PCM data that
> reaches -2 dBFs. Program material can have all sorts of peaks (that is
> what
> is great about this loudness requirement, as it allows for more dynamic
> presentation), just none of those peaks can go above -2 dBFs.
>
> What we do here (in Pro Tools) is run our loudness measurement (in real
> time, across the master bus), get the LKFS number, adjust overall gain to
> hit the spec number, and then peak limit at -2 dBFs. A plugin on my master
> in Pro Tools like Waves L2 can make both gain and peak limit adjustments
> (transparently without adding any compression, if you know what you're
> doing), so it's rather easy.
>
> Adding yourself a peak limiter might be putting your app into overtime
> cost
> (!). As long as you have the measurement for true peaks, you should be
> fine
> in just reporting it. You may consider adding a feature that can change
> the
> window of measurement for "momentary" loudness to shorter than 400 ms. as
> true peaks are instantaneous (i.e. one sample).
>
> -Mark
>
>
>
> On Mon, Oct 22, 2012 at 9:32 AM, bouke <bouke@editb.nl> wrote:
>
>> **
>>
>>
>> Terry,
>> Job is right on almost all points.
>> This thing can find places where there is either a momentary peak that is
>> too high (momentary means 400 Msecs), or Short peaks (short means 3
>> seconds)
>> This is BEFORE bringing the entire mix up or down to -24 LUFS (-23 for
>> Europe).
>> Since the peaks are relative to the current loudness, it does not make a
>> difference if you measure before or after.
>>
>> What is slightly more complicated as i initially thought (after analyzing
>> a
>> couple of long form pieces), and hence i disagree with Job a bit,
>> Short peaks happen more than i would think. Especially if a mix is
>> compressed, the Short peaks can occur even if you're 2 dB below our
>> trusted -9.
>> On an uncompressed mix its way less likely to let this happen, even if
>> you
>> go over -9.
>>
>> For the legalizer part, yes, it does 'legalize', but only the Integreated
>> Loudness (the overall measured loudness).
>> This is NOTHING MORE than lowering / upping the entire mix!
>> In theory it's possible to let this thing mix down the peaks, but i don't
>> trust a machine to do this.
>> (and it's very easy to go back, fix it yourself and try again. If you
>> have
>> the feedback on a problem and the TC, you'll probably hear what the specs
>> are bitching about..)
>>
>> But it's getting more complicated for the US.
>> A bit out of the FOX specs:
>> ---------
>> >>Fox requires that commercial material have a subjective loudness of -24
>> >>LKFS
>> <snip>
>> >>The FNC/FBN target subjective loudness level is -24 LKFS.
>> >>Instantaneous true audio peaks shall not exceed -2dBFS (BS.1771),
>> >>while average peak audio levels should not exceed -8dBFS (EBU Tech 3205
>> or
>> >>equivalent).
>> ---------
>>
>> First, LKFS is (about) the same as LUFS, where 1 LUF is called a LU.
>> Now we got that part covered.
>> Over here newspeak is to talk about LU's to describe loudness, relative
>> to
>> the overall loudness.
>> But these specs do not speak about the Integrated / Momentary loudness!
>> It's only Instantaneous / Average and are described in DBFS,
>> (to add to the fun, the EBU Tech 3205 is concidered obsolete...)
>>
>> >> Instantaneous true audio peaks shall not exceed -2dBFS
>> >>average peak audio levels should not exceed -8dBFS
>> I have no idea how loudness an true peak relate, (and in fact i think
>> they
>> don't relate at all...),
>> But i've got a gut feeling that with a Momentary loudness set to 8,
>> and Short to 2, you're way covered (probably this is too conservative...)
>>
>> But if you have the -2 / -8 dBFS covered already, you can just use this
>> to
>> get to the desired LUFS.
>>
>>
>> Bouke
>>
>> VideoToolShed
>> van Oldenbarneveltstraat 33
>> 6512 AS NIJMEGEN
>> The Netherlands
>> +31 24 3553311
>> www.videotoolshed.com
>>
>> ----- Original Message -----
>> From: "Job ter Burg (L2B)" <Job_L2@terburg.com>
>> To: <Avid-L2@yahoogroups.com>
>> Sent: Sunday, October 21, 2012 8:52 PM
>> Subject: Re: [Avid-L2] Re: EBU R 128 / Calm act, HIGHLY COMMERCIAL
>>
>> Terry,
>>
>> Yes, but the idea is not so much eliminating offending elements, as any
>> short peaks or loud or silent scenes are not the issue. Total overall
>> dynamics is. So it is better to mix by your ears, then use this tool (or
>> some of the pther more expensive ones) to measure it and bring it up to
>> spec.
>>
>> Apparently, Bouke's too will tell you where the parts of the mix are that
>> cause the needto normalize, so you could go back and tweak those.
>>
>> Bouke's tool otherwise seems to go by the same approach as Grimm Audio's
>> LevelNorm (AudioSuite plugin, Avid compatible, R128 or A/85). This was
>> specifically designed so that Avid editors would mix any way they want,
>> then
>> normalize the entire result to R128 or A/85. Even if you would mix to
>> older
>> norms (like European mixes used to be crushed flat at -9dBfs), or if you
>> want your mix to have a specific compressed sound and very limited
>> dynamic
>> range, LevelNorm or LoudnessChange will indeed legalize it to either R128
>> or
>> A/85.
>>
>> Most mixers tell me they need to rely less on meters and more on actual
>> creative mixing with these new standards. They need to set up their room
>> to
>> a certain dialogue level they like to go by, then mix as they feel is
>> right
>> creatively, rather than the other way around. Folks mixing for TV can
>> finally actually mix again, rather than fighting their way into a
>> brickwall
>> limiter.
>>
>> J
>>
>> On Oct 21, 2012, at 20:08, "Terence Curren" <tcurren@aol.com> wrote:
>>
>> > Bouke,
>> >
>> > You are just lowering the entire mix to pass specs right? The better
>> > approach would be to only lower the offending elements, but that can't
>> be
>> > done with a finished file. So your program is more like a legalizer,
>> > yes?
>> >
>> > --- In Avid-L2@yahoogroups.com, "bouke" <bouke@...> wrote:
>> >>
>> >> Guys (and Shirley),
>> >>
>> >> If you need to deliver to the new sound specs, this might be of help:
>> >>
>> >> http://www.videotoolshed.com/product/68/loudnesschange
>> >>
>> >> In short, it can analyze a mixdown (as a wave file), see potential
>> >> trouble
>> >> (short / momentary peaks) so you can fix them,
>> >> and render the file to a new file while upping / lowering the mix to
>> get
>> >> to
>> >> your desired LUFS.
>> >>
>> >> And it's blazing fast....
>> >>
>> >>
>> >> Bouke
>> >>
>> >> VideoToolShed
>> >> van Oldenbarneveltstraat 33
>> >> 6512 AS NIJMEGEN
>> >> The Netherlands
>> >> +31 24 3553311
>> >> www.videotoolshed.com
>> >>
>> >
>> >
>> >
>> >
>> > ------------------------------------
>> >
>> > Search the official Complete Avid-L archives at:
>> > http://archives.bengrosser.com/avid/
>> > Yahoo! Groups Links
>> >
>> >
>> >
>>
>>
>>
>
>
> [Non-text portions of this message have been removed]
>
>
>
> ------------------------------------
>
> Search the official Complete Avid-L archives at:
> http://archives.bengrosser.com/avid/
> Yahoo! Groups Links
>
>
>
| Reply via web post | Reply to sender | Reply to group | Start a New Topic | Messages in this topic (8) |
