• Get involved.
    We want your input!
    Apply for Membership and join the conversations about everything related to broadcasting.

    After we receive your registration, a moderator will review it. After your registration is approved, you will be permitted to post.
    If you use a disposable or false email address, your registration will be rejected.

    After your membership is approved, please take a minute to tell us a little bit about yourself.
    https://www.radiodiscussions.com/forums/introduce-yourself.1088/

    Thanks in advance and have fun!
    RadioDiscussions Administrators

Normalizing Audio For Digital Automation

R

RadioActive05

Guest
I'm about to start the process of ripping a huge music library from CD to .wav for use in a digital automation system (we're still debating which one to use).

I wanted to try to get a head start and get as much audio ripped as possible, storage is not an issue so I prefer to stick with the .wav/PCM format vs compressing to mpeg..

My question is; should I normalize the .wav files?? I've heard people say to normalize to 27% or -10.0db, I've heard leave the .wav files as is because normalizing messes with the dynamic range and thus defeats converting to .wav..

Our music library contains CD's that were produced in the 80's (lots of dynamic range) through the latest stuff (hard limited).

Anyone have any suggestions?? I was looking at audiograbber as a means of ripping from CD to .wav. Audiograbber has a normalization option.

To normalize or not to normalize, that is the question....

Thanks..
 
> My question is; should I normalize the .wav files?? I've
> heard people say to normalize to 27% or -10.0db, I've heard
> leave the .wav files as is because normalizing messes with
> the dynamic range and thus defeats converting to .wav..
>
I suspect different people are using varied definitions for "normalize". In CoolEdit 2000, NORMALIZE ups the gain of the entire selection being normalized by an equal amount. The DYNAMIC RANGE does not change.

If you LIMIT or COMPRESS, then the dynaimc range changes. When I read your question, I think you are confusing the two concepts.

You may need to wait until you select your automation system before deciding at what level to normalize. If the system has built-in recording protocols that record tracks input through the automation system at a level other that 100% normalizing, then I would want tracks that I produced to be at that same level so that all the sources enter the audio chain at about the same level.

I suspect that a number of users/operators choose to record certain categories of audio (commercials, music, voice track, etc.) at different levels so as to make sure some categories always over ride others during corss fades.
 
> Anyone have any suggestions??

This word is all you need: WaveGain

<a target="_blank" href=http://www.rarewares.org/others.html>Download it</a>, use it with the "radio" option, and enjoy!
<P ID="signature">______________
noiboc.jpg

"This is the New York Emergency Broadcast System satellite channel. They took the crosstown bus."</P>
 
Well let’s look at it this way. What will you be doing with the automation system? Will it be used for a terrestrial broadcast? A background music system? A party DJ service? Before knowing the answer to that question, it’s difficult to determine the best course of action. However there are some general considerations.

Terrestrial Broadcasting: I would forget normalizing altogether in this application, because most of the processing in the air chain will squash everything flat. In that case I would only consider using an Auto Gain Controller before hitting the processing with the audio. The reason being the AGC will help ensure the audio levels of all programming will remain in that “sweet spot” the rest of the processing works best at. I work in terrestrial radio and I have been known to normalize an occasional selection via CoolEdit2000. Generally I only do it to avoid reports that audio levels are too hot or too low (some on board simply have not grasped the knowledge of the AGC box and how it is there specifically for that reason).

Internet Broadcasting: This is somewhat similar yet somewhat different. Most of your audience will be in offices with cheap PC speakers. I again recommend using AGC here vs. normalizing, for this reason. Also, if you can find some decent multiband processing to put after the AGC, that does not have clipping, I recommend using it. You don’t need a ten thousand dollar box from Optimod here. There is affordable multiband processing, both software and hardware based, that will work for this application.

R

> I'm about to start the process of ripping a huge music
> library from CD to .wav for use in a digital automation
> system (we're still debating which one to use).
>
> I wanted to try to get a head start and get as much audio
> ripped as possible, storage is not an issue so I prefer to
> stick with the .wav/PCM format vs compressing to mpeg..
>
> My question is; should I normalize the .wav files?? I've
> heard people say to normalize to 27% or -10.0db, I've heard
> leave the .wav files as is because normalizing messes with
> the dynamic range and thus defeats converting to .wav..
>
> Our music library contains CD's that were produced in the
> 80's (lots of dynamic range) through the latest stuff (hard
> limited).
>
> Anyone have any suggestions?? I was looking at audiograbber
> as a means of ripping from CD to .wav. Audiograbber has a
> normalization option.
>
> To normalize or not to normalize, that is the question....
>
> Thanks..
>
 
Don't forget to remove the dead spots before and after the audio on the tracks!



> > My question is; should I normalize the .wav files?? I've
> > heard people say to normalize to 27% or -10.0db, I've
> heard
> > leave the .wav files as is because normalizing messes with
>
> > the dynamic range and thus defeats converting to .wav..
> >
> I suspect different people are using varied definitions for
> "normalize". In CoolEdit 2000, NORMALIZE ups the gain of
> the entire selection being normalized by an equal amount.
> The DYNAMIC RANGE does not change.
>
> If you LIMIT or COMPRESS, then the dynaimc range changes.
> When I read your question, I think you are confusing the
> two concepts.
>
> You may need to wait until you select your automation system
> before deciding at what level to normalize. If the system
> has built-in recording protocols that record tracks input
> through the automation system at a level other that 100%
> normalizing, then I would want tracks that I produced to be
> at that same level so that all the sources enter the audio
> chain at about the same level.
>
> I suspect that a number of users/operators choose to record
> certain categories of audio (commercials, music, voice
> track, etc.) at different levels so as to make sure some
> categories always over ride others during corss fades.
>
 
HI -

We are a terrestrial college radio station preparing for a studio move/major upgrade. Our upgrade is scheduled for this spring, I wanted to start to get some of the minor things out of the way (like ripping our library) before we get in to the thick of things this spring/summer. We have a solid air chain, Omnia and Compellor combination which sounds fantastic with our current CD format. We used use NexGen101 for automated overnights but it is not as stable as we'd like and the interface is not volunteer/student friendly.

I'm going to rip our music into PCM for now and worry about normalization when we finally decide on our automation software. Our IT staff has provided us with more than enough HDD space so I have no problem using PCM for our audio.

The reason I asked was because I was talking with a rep from Digilink last week and was told that it would be a good idea to normalize the audio. I just did not want to double the effort by ripping the CD's and then having to normalize in the future. By the way, we will not be going with Digilink..

Thanks for all of your suggestions!!
 
I don't think you need to normalize every track. However, you do want to have the levels to be pretty even across all tracks, so normalizing the lower tracks will help. While any earlier poster is correct that the processing of the station will squash almost everything, you don't want to rely completely on the processing or an AGC to completely even things out.

Generally, even back in the record and tape cartidge days, it was good production practice to have the levels of all sources pretty much even across the board. This helps not only the board ops and jocks, but also makes the programming sound more consistent. The really fanatic production folks also re-eq and adjust the density of spots and on-air production elements to have them also sound the same so you don't have a thin spot back to back with a really dense spot.
 
Keep in mind in those days AGC was not common. With an AGC box properly set, you can get away with dramatic changes in audio levels from source to source. Carts were usually mastered with single band compression which helped keep the levels tight and improve the S/N ratio of the tape. But again, very little AGC existed and MultiBand processing was two or three bands (as opposed to five to six bands in current processors). Since the original poster indicated they have a Compellor, they don’t really need to normalize the audio files.

Our own setup is Compellor, STL and then Orban 8200. Since the 8200 is at the transmitter site, the internal AGC is defeated.

R

> I don't think you need to normalize every track. However,
> you do want to have the levels to be pretty even across all
> tracks, so normalizing the lower tracks will help. While any
> earlier poster is correct that the processing of the station
> will squash almost everything, you don't want to rely
> completely on the processing or an AGC to completely even
> things out.
>
> Generally, even back in the record and tape cartidge days,
> it was good production practice to have the levels of all
> sources pretty much even across the board. This helps not
> only the board ops and jocks, but also makes the programming
> sound more consistent. The really fanatic production folks
> also re-eq and adjust the density of spots and on-air
> production elements to have them also sound the same so you
> don't have a thin spot back to back with a really dense
> spot.
>
<P ID="edit"><FONT class="small">Edited by RobertBass on 11/09/05 02:27 AM.</FONT></P>
 
> Keep in mind in those days AGC was not common. With an AGC
> box properly set, you can get away with dramatic changes in
> audio levels from source to source.

AGC is a very useful tool, but even with today's technology it's not perfect. I heard a station which was running their liners and jingles much "hotter" through the board than the music. As a result, every time a jingle was followed by a song with a soft introduction, this caused a very audible "fade-in" effect on the air as their processor's AGC gradually adjusted itself to the drastic volume change.

For broadcast use, there is absolutely *no* downside to RMS-level normalization of your audio sources, other than the time required to do it. In my other message here, I posted a link to WaveGain, which is all you need to perform RMS normalization of WAV files. A similar tool called MP3GAIN is also available from that web site, which performs *lossless* RMS normalization of MP3 files, without needing to re-encode the MP3 data, thus there is absolutely no quality loss and the process is fully reversible.
<P ID="signature">______________
noiboc.jpg

"This is the New York Emergency Broadcast System satellite channel. They took the crosstown bus."</P>
 
My advice is to wait until you decide on an automation system and rip and encode with that particular system.

My other unwanted advice is to figure out what your hardware issue happens to be that is causing the instability of NexGen and get it fixed. Call prophet...get them on the stick.

If your staff at the station can't handle the interface of NexGen, they'll find relatively few systems any easier to grasp.

My two cents. Carry on.



> Don't forget to remove the dead spots before and after the
> audio on the tracks!
>
>
>
> > > My question is; should I normalize the .wav files?? I've
>
> > > heard people say to normalize to 27% or -10.0db, I've
> > heard
> > > leave the .wav files as is because normalizing messes
> with
> >
> > > the dynamic range and thus defeats converting to .wav..
>
> > >
> > I suspect different people are using varied definitions
> for
> > "normalize". In CoolEdit 2000, NORMALIZE ups the gain of
> > the entire selection being normalized by an equal amount.
>
> > The DYNAMIC RANGE does not change.
> >
> > If you LIMIT or COMPRESS, then the dynaimc range changes.
>
> > When I read your question, I think you are confusing the
> > two concepts.
> >
> > You may need to wait until you select your automation
> system
> > before deciding at what level to normalize. If the system
>
> > has built-in recording protocols that record tracks input
> > through the automation system at a level other that 100%
> > normalizing, then I would want tracks that I produced to
> be
> > at that same level so that all the sources enter the audio
>
> > chain at about the same level.
> >
> > I suspect that a number of users/operators choose to
> record
> > certain categories of audio (commercials, music, voice
> > track, etc.) at different levels so as to make sure some
> > categories always over ride others during corss fades.
> >
>
 
> My other unwanted advice is to figure out what your hardware
> issue happens to be that is causing the instability of
> NexGen and get it fixed. Call prophet...get them on the
> stick.

I agree, Nexgen is one of the more stable automation software out there.



> If your staff at the station can't handle the interface of
> NexGen, they'll find relatively few systems any easier to
> grasp.
>

Very true, and besides, if their a college student going onto the "real broadcast world" they will face "complex" automation systems, and WinAmp or something simple just will not be used.<P ID="signature">______________
Lenks
Program Director/Music Director
X Music Online
The X
Today's Best Music
http://www.xmusiconline.com/</P>
 
If the release time on the AGC is too slow, that will happen.

R

> AGC is a very useful tool, but even with today's technology
> it's not perfect. I heard a station which was running their
> liners and jingles much "hotter" through the board than the
> music. As a result, every time a jingle was followed by a
> song with a soft introduction, this caused a very audible
> "fade-in" effect on the air as their processor's AGC
> gradually adjusted itself to the drastic volume change.
 
> If the release time on the AGC is too slow, that will
> happen.

It wouldn't help much if it was faster and it would probably hurt (pumping). The transition from hot audio level to soft audio level will always be heard no matter how fast or slow release times are and no matter which processor you use! As will transition from soft to hot, though this is much less obvious...

There's really no substitute for constant audio level coming from console. Only that will provide smooth and consistent on-air sound - quality and professionalism people will consciously or subconsciously recognize. I prefer board operators and for other reasons as well. If we're talking computer automation, normalization is really the least you can do. RMS better than peak.


Regards,
Goran Tomas
 
> Keep in mind in those days AGC was not common. With an AGC
> box properly set, you can get away with dramatic changes in
> audio levels from source to source. Carts were usually
> mastered with single band compression which helped keep the
> levels tight and improve the S/N ratio of the tape. But
> again, very little AGC existed and MultiBand processing was
> two or three bands (as opposed to five to six bands in
> current processors). Since the original poster indicated
> they have a Compellor, they don’t really need to normalize
> the audio files.
>
> Our own setup is Compellor, STL and then Orban 8200. Since
> the 8200 is at the transmitter site, the internal AGC is
> defeated.
>
> R

All of that is true, although in the 1980's we had an AGC/compellor prior to the 8100 in the airchain. Proper consistent levels were still important. If you are running completely automated, no AGC is going to fix a song completely obliterating a liner or vice-versa -- its still GIGO. IMHO, I have yet to see an AGC completely resolve the issue of really low sources have a "thinner" presence on the air.
 
Status
This thread has been closed due to inactivity. You can create a new thread to discuss this topic.


Back
Top Bottom