Gearslutz

Gearslutz (https://www.gearslutz.com/board/)
-   Post Production forum (https://www.gearslutz.com/board/post-production-forum/)
-   -   Radio commercial, tips? (https://www.gearslutz.com/board/post-production-forum/105598-radio-commercial-tips.html)

Kenadon 24th January 2007 01:04 PM

Radio commercial, tips?
 
Im doing a 10 second radio commercial that includes just a speech (no music or else). I was just wondering if you had some tips on the mixing and else. It was recorded with Rode Nt2000 to PT 24bit 44khz, so the quality shouldn't be an issue. So far I've gotten a quite good sound with c4 - eq - de-esser - rvox - some presence reverb, but i'm having some issues with "s" 's, as I'm not sure how they sound on the radio, and how much I should squash them. I also have another commercial that has speech, and under it a movie trailer soundtrack that also includes speech (in a different language). I was thinking about doing a volume automation to drop the trailer volume when the recorded speech comes in and also cut some frequencys that the recorded speech uses, so they wont mash. Any ideas on this I'd also appreciate.
Thanks.

paully 24th January 2007 04:10 PM

Quote:

Originally Posted by Kenadon (Post 1089610)
... I was thinking about doing a volume automation to drop the trailer volume when the recorded speech comes in and also cut some frequencys that the recorded speech uses, so they wont mash.

>>>>>>>>>>>>>
If the speech and trailer volume are on seperate tracks (I assume they are at this point), this is usually done with a compressor that has a side-chain input. Use some of the speech track's signal, mabee an unused channel aux. send, as the compressor's trigger. Run the trailer track through the compressor, then make adjustments so that when the speech is active, the trailer is 'ducked' or reduced. Much easier (and more precise) than trying to encode manual volume changes. Unless it's really annoying, I wouldn't mess with the speech EQ. That's where the money is!

Best, Paul

Kenadon 24th January 2007 07:10 PM

Quote:

Originally Posted by paully (Post 1089835)
>>>>>>>>>>>>>
If the speech and trailer volume are on seperate tracks (I assume they are at this point), this is usually done with a compressor that has a side-chain input. Use some of the speech track's signal, mabee an unused channel aux. send, as the compressor's trigger. Run the trailer track through the compressor, then make adjustments so that when the speech is active, the trailer is 'ducked' or reduced. Much easier (and more precise) than trying to encode manual volume changes. Unless it's really annoying, I wouldn't mess with the speech EQ. That's where the money is!

Best, Paul

ah yes, i knew there some better way of doing that, i'll try that for sure. thanks!

paully 24th January 2007 08:07 PM

I accidently stumbled onto this old thread that may be of help.

https://www.gearslutz.com/board/low-end-theory/28692-tv-commercial-loudness.html

Best, Paul

Gibby 30th January 2007 06:57 PM

I do alot of radio and TV commercial mixes. I find that slightly compressing the vo in stages, with NO compression or limiting on the mix-buss actually makes everyting sound louder, bigger, and better when it goes on air. If you squish the hell out of everything, then the station limiting will make it sound small. The only EQ i use on the voice is low-cut at 70 hz and +1 or 2 db at 10k. use a de-esser only if u really have to. if yer having problems with "s's" alot then maybe u shud use a different mic. I always insert the EQ after the compressor. i went thru alot of trial and error and finally my mixes are sounding great on air. elemental audio (now RND) and sony oxford work great for radio and tv stuff. U can draw the volume changes in your DAW, works better for me than the side-chain/ducking technique. Cheers.

Inner Light 2nd February 2007 06:11 PM

I do tons of Radio commercials, and promos. on the voice track i remove everysingle breath, Compress the voice, eq after comression, and have a limiter plug at the end to kinda glue the mix together.

We're a music / talk radio station and have gotten tons of great recognition on our sound quality, witch actually surprises me. But unlike most other stations our limiting is almost non existant very minimal , to retain the dynamics in our classical, jazz, world, and pop music. Some times the installation engineer would complain or scream that we were off the air cause he heard nothning, but it was just a very low passage of classical music. I just laughed and told him this isn'T you hit's fm type station. a fact that I'm reall proud of

Garret 2nd February 2007 07:08 PM

I wouldn't recommend side chain/ducking because that in effect doesn't really give you what you need. When the voice is loudest the music trailer will be at its lowest and at the voices lowest the trailer will be at its loudest. You want the opposite for clariity. I recommend riding the fader of the trailer underneath for best results

Inner Light 2nd February 2007 07:23 PM

I've always equated ducking as a live thing, if the announcer or dj, wants to do a cut in or something, never as a production thing. There is way more control with automation.

AdAudioInc 2nd February 2007 07:49 PM

In a ten second spot - if I have bad "S"es, then I'll go through and manually lower the volume of each "S". Simply make a cut before and after each "S" so that the "S" itself is an event and lower the volume to what sounds good.

paully 2nd February 2007 09:18 PM

Quote:

Originally Posted by Garret (Post 1107709)
I wouldn't recommend side chain/ducking because that in effect doesn't really give you what you need. When the voice is loudest the music trailer will be at its lowest and at the voices lowest the trailer will be at its loudest. You want the opposite for clariity. I recommend riding the fader of the trailer underneath for best results

>>>>>>>>>>>>>
A couple of problems with this. Announcements should not vary dramatically in level.. certainly not enough to cause the music track to waver. That's the responsibility of the vo announcer and the engineer (also why the voice is usually compressed during tracking). With that said, ducking becomes very effective.

Riding gain (via DAW fader automation and in lieu of auto-ducking) is of course the way to go, but it takes experience, is time consumming, and is not as spot-on as auto-ducking. Truth be known, manual level automation is of course the ultimate compressor, without the artifacts heh .

Inner Light: Cudos to ya'. Stations like yours that are actually preserving the dynamics of what they send out are to be commended. Very rare in this day and age. Unfortunately most commercial producers want it either loud or louder. Then there's the station's compression prior to broadcast! Oh well..

Best, Paul

RobMacki 2nd February 2007 09:33 PM

My background has been radio production for over 12 yrs.
Reading through the posts I have a similar approach as Inner Light on almost every point. (I don't eliminate every breath but most of them and as always, it all depends)

With ssssssss I usually control them with a combination of Dynamic EQ followed by gentle or surgical EQ followed by gentle DeEss followed by Phoenix. The whole time doing this I am focused on the full range of the voice. One of my pet peeves is hearing a DeEsser.

In terms of ducking, It's quicker and easier for me to just volume automate. I've also found I can have more control over each instance of GR instead of dealing with a thresholds, attacks and releases.

Inner Light 2nd February 2007 10:06 PM

Quote:

Originally Posted by paully (Post 1107977)
>>>>>>>>>>>>>


Inner Light: Cudos to ya'. Stations like yours that are actually preserving the dynamics of what they send out are to be commended. Very rare in this day and age. Unfortunately most commercial producers want it either loud or louder. Then there's the station's compression prior to broadcast! Oh well..

Best, Paul

thanks man, we are kinda purists in a sense, our whole strategy is a little backwards from any other stations but it's working out well.

I remove all breathing because I don't always get to work with pro's and i have to eliminate space and save some time, and you'd be surprised how much time you can gain by doing this, all the while keeping it natural sounding. I've never even used a de-esser, wish i had sometimes, but i get by

RobMacki 2nd February 2007 11:46 PM

Quote:

Originally Posted by Inner Light (Post 1108076)
I remove all breathing because I don't always get to work with pro's and i have to eliminate space and save some time, and you'd be surprised how much time you can gain by doing this, all the while keeping it natural sounding.

I know what you mean. When you have 45 sec of copy to cram into a :30 gruudge you do what ever it takes to please the client and come out the hero.mezed

I have a few clients who actually write good copy and know the value of letting the spot breathe with the music.kfhkh

paully 2nd February 2007 11:56 PM

I'll bet a few of you guys know the old #2 pencil routine for disbursing sibilance during tracking. Still a very effective trick, and can eliminate the need to de-ess at the end. I work in Logic, and a very narrow EQ band can be quickly inserted to bat down sibilance, then bypassed when not needed. No level reduction needed.

Paul

Inner Light 3rd February 2007 05:36 AM

Quote:

Originally Posted by RobMacki (Post 1108272)
I know what you mean. When you have 45 sec of copy to cram into a :30 gruudge you do what ever it takes to please the client and come out the hero.mezed

I have a few clients who actually write good copy and know the value of letting the spot breathe with the music.kfhkh


most of the times it's the actual text that's way too long.

Some of the voice over talent i work with know that i'll cut out all the space so they even take longer, and get the words pronounced clearly and take to 3 seconds pause. Then I have a short cut that selects, cuts, and remove silences. then i check to make sure it doesn't cut any important syllables. I even go as far as gaining the ends of thier phrases, where they naturally drop in volume.

AdHawk 3rd February 2007 10:59 PM

Quote:

Originally Posted by Garret (Post 1107709)
I wouldn't recommend side chain/ducking because that in effect doesn't really give you what you need. When the voice is loudest the music trailer will be at its lowest and at the voices lowest the trailer will be at its loudest. You want the opposite for clariity. I recommend riding the fader of the trailer underneath for best results

You´re absolutely right about this, but there are a couple of ways to make it work.

I used to do it this way:

When the VO is edited, make a copy of it to another track.
Assign the copied track to a bus called "ducker" or some other brilliant name.
Put a brickwall limiter such as L1 on the copied track and set it up to completely slam the track flat.
Use this copied track to sidechain your comp.

Nowadays I do it this way:

Same setup with a separate track fore sidechain signal.
On that track I have put a sinewave audio file, no limiter.
A few simple edits/fades on the sinewave gives me quick control of the background.

JK

Mattrix 4th February 2007 03:34 AM

sinewave (tone)
 
Quote:

Originally Posted by Jan Karamell (Post 1109876)
You´re absolutely right about this, but there are a couple of ways to make it work.



Nowadays I do it this way:

Same setup with a separate track fore sidechain signal.
On that track I have put a sinewave audio file, no limiter.
A few simple edits/fades on the sinewave gives me quick control of the background.

JK

I thought i was the only one that does that? anyone else use this technique?

duplicate voice track - use grid 500ms handles then fades for smooth up and down ducking - convert to tone depends on how much Db for ducking - send track to a sub to side chain compressor, set comp attack and release for smooth or quick response. The purpose of this technique or limiter squashing is if the voiceover file was left with peaks and valleys the background volume level will go up and down and sound lumpy, hope this helps.

Matt

matucha 4th February 2007 11:29 AM

As there is no good solution in cubase for sidechaining, I'm riding volume of the background with automation, once it gets under your skin how the things work, it is very easy and fast almost without listening to draw the automation ;-).

I also hate that 45s into 30s where there is no single moment without "Blablablablablablabblablb BLAH blablablablab labblablbalbalablablablablabla blabbla blbalbal ablablablabla blablabbla blbalbalablablablabla BLAHBLABLA BLAH!!".

With just enough time for making intro, intermezzo and outro, it gets much much better in the impact.

And everything depends also on the quality and structure of the background music, because some pieces let you do really killer stuff in terms of sound and impact and some are plain sh*t that only muddies "your" voiceover and takes your work down.

paully 4th February 2007 04:02 PM

Quote:

Originally Posted by matucha (Post 1110664)
...I also hate that 45s into 30s where there is no single moment without "Blablablablablablabblablb BLAH blablablablab labblablbalbalablablablablabla blabbla blbalbal ablablablabla blablabbla blbalbalablablablabla BLAHBLABLA BLAH!!"...

>>>>>>>>>>>>>>>>>
kfhkh heh heh heh heh heh

Paul

gsilbers 6th February 2007 07:34 AM

mix with a good multiband compressor and an L2 in the output mix buss. that was the setup where i used to work doing TV and radio Ads for sears, MCds Citibank etc.
set up the limit to 1db or so below the limit. if its TV and its -8db, go for -9db.

make it sound loud and THEN check the mix on crappy speakers. generic computer speakers might to do the work.

not much too it really just make it sound as it does in radio and TV. sounds silly i know, but if u come from a musical engineering background then youll be all like "maybe an avalon pre will make it sound more transparent , and blablabla" which of course matters in its element but in Ad post and TV is just to make it work . so if it sounds clear and loud, great.. ship it! if u think it needs a Desseser, then it needs a deseer... ship it, the client needs it yesterday! always of course, dialog is king. so if u have a music background , lower it and even more if it has sung parts. but at the end u have to think like if u where the consumer watching tv or listning radio. its your 1st thought that counts, which is the same 1st thought that TV viewers will get/have.
you always have to hear dialog nice and clear, no matter whats behind. u hear on TV ads that have rage against the machine really low and the VO really loud. and u you know the agency spent like 80 grand for that license. what a waste.





we recorded with a neuman tl103 and focusrite red pre stright to PT.

RobMacki 6th February 2007 05:40 PM

Quote:

Originally Posted by gsilbers (Post 1114387)
dialog is king. so if u have a music background , lower it and even more if it has sung parts. but at the end u have to think like if u where the consumer watching tv or listning radio. its your 1st thought that counts, which is the same 1st thought that TV viewers will get/have.
you always have to hear dialog nice and clear, no matter whats behind. u hear on TV ads that have rage against the machine really low and the VO really loud.

Clients freak out if their message is not coming through loud and clear.hittt
One thing that works for me and especially if there is any voice in the music or in the case of the OP, "speech", is to boost the presence range (male near 3k female near 4k) then gently carve out that freq in the music or whatever is under. You can then get away with a louder music bed that does not compete with the VO. But placement of the VO musically is what is one of the most important factors IMVHO.

peterc 9th February 2007 01:31 PM

Quote:

But placement of the VO musically is what is one of the most important factors IMVHO.
Yes! Sliding the voice slightly to allow a beat through or cover up a dud one is such a keen trick. The rhythm of the voice & music have to work together.

Mix on some speakers that are typical radio style speakers, the correct playback is important. Hopefully the stations dont screw it up completely.

Peter

Inner Light 9th February 2007 05:44 PM

I never have a voice start on a strong downbeat, i always slide it after because it takes away the listeners attention from the voice. Not Good.

if you do your job correctly it should just flow nicely.

kladdekagan 18th February 2007 10:55 PM

There are a couple of issues to think about when producing for radio. One thing is that the radio station might process your sound in a way that makes your great sounding mix sound dull, low or lack punch. From my experience processing your audio to hard might work against you. Ie, use of to much limiting or multiband compression may sound good and loud in your monitors, but on air your commercial sound lower than the rest. Most stations where I live (Sweden) use an Orban Optimode on the stations outfeed and the sound that you deliver have to be produced to suit that. The problem is that different stations use different settings because of format or taste. Remember, it's the ON AIR sound that is important, not what you hear in monitors. And here you have to learn by doing. Produce a commercial, listen to it on air, the next time you produce change what went wrong in the first place.

If you produce a lot for the same station you eventually will learn how to process your sound. The problem arise when you produce a commercial that is to be broadcasted on several stations, or if it is a national campaign. Then you can't tailormake the sound, you have to go for a sound that will work OK on most stations.

I usually compress the VO track, 3:1 and 2-3 gain reduction just to tighten the sound. If needed (mostly it is) I eq around 3-4 khz for precence and around 8-10 khz for air, just 1-3 db. I use only the plugins that come with Logic, except for the master fader, I think they work just fine for radio. I ride the faders on background music and ambiences/effects. On the master fader I use a good limiter (before the L2, nowadays Izotope Ozone) set to moderate settings. With Ozone I get a few more db out of the sound compared to some other limiters I have tried. Although it might be tempting to use the other great stuff on Ozone like the multiband, I do not do that for radio commercials. I have learnt that the simpler the better.

Regarding problems with SSS i don't experience that anymore. Decent microphones, acoustics, cables and preamps most of the time eliminates that. You don't need really expensive stuff for radio, just decent. I use a Rode NT2A microphone, Grace 101 pre amp and RME Fireface 400 running into Logic. And all Monster cabling. Can't remember when I had to de-ess the last time, but that must be many years ago. And I record A LOT of different voices into the same system. I had more problems when I started with radio many years ago. Then I used a Neumann U87 and unfortunately cheap cables and pre amps. A petty for this expensive mic :-( Had to de-ess all the time.

I have read the comments above regarding text. It's a very common problem that the text is too long. Either the client is too eager to tell the WHOLE story, or the writer is unexperienced writing for radio. It is also a common problem that the client first buys the media time (like 30 or 45 sek, sometimes just 20 sek) and that the time is too short for the creative solution. In an ideal world the client should first produce a script and a commercial that can market his produkt/company in a good way, and then buy whatever lenght of media time needed.

This is just my point of view, from a Swedish perspective.

Cheers,
Mikael

matucha 19th February 2007 11:46 AM

about levels

I've searched the net for voiceovers and I found everything from nonlimited stuff with voc way up above "background" to squashed to dead (-6dB RMS in wavelab).

My exp. is that I gain something by getting to ~ -10dB RMS (in wavelab) and going over is only going to sound smaller on radio. 2Buss compression and limiting helps me have background and VO glued together, but it is more thing of what is going out of the monitors.

My clients wouldn't accept anythin under ~ -12dB RMS wworried

arkmabat 13th April 2010 10:13 PM

This thread has been very informative. Thanks guys! thumbsup