Your biggest problem is not under your control - and that is the settings each particular station uses on their stereo processor.

One lap through the FM band will tell you all you need to know about this. Just like in commercial mass-market CDs, LOUDNESS RULES. So, after you bust your buns maxing out the levels on your CD, the stereo processor will try to blast it even harder (louder).

The typical stereo processors are multi-band affairs (some as many as 5 bands) that can be peaked, tweaked, and geeked until the sound that comes out has little to do with what went in, other than the notes.

Can you tell I work at a classical public radio station yet? ;-)

Grab a copy of Bob Katz' book 'Mastering Audio' and read the appendix on radio processors - written by the two guys that are responsible for most of this (Bob Orban and Frank Foti) to learn what happens when your CD airs.

I would aim for a K-12 to K-14 range. Expect the bass to get PUMPED, too.

Thanks dpd classical music is ok but not for me. I've actually done quite a bit of reading on this subject ,but was wanting to know what me's in this forum are sending to radio stations.Sounds like its a little too personal Rolling Eyes .Most mixes maxed out for cd are fine (if mastered diligently)as long as your not clipping the 30 cent da converters in most consumer cd players.I really don't want to have a mix maxed out which sounds great in the studio and on cd players but on the radio sounds distorted and has holes punched all through it by broadcast limiters and the like.I'm planning to get the main singles mastered by a top guy in Sydney AUS anyway so it won't be long before I get a better understanding of broadcast levels.I and my clients are very happy with my mastering for cd,all I'm trying to do is get a better understanding of what dfferent engineers are doing to mixes for broadcast.I expect that for these broadcast stuations a master should probably not exceed - 1 dbfs peak and not exceed much more than + 2-4 vu also keeping peaks of freqs above 5-6 khz to a minimum reasonable level,I'm probably wrong but at least I feel I have a starting point to work around,feel free to tell me I'm wrong.It's taken a few years for me to be confident with final eq and levels for cd but I am mastering my mastering slowly but surely.Broadcast levels are somthing which are critcal as far as I'm concerned and is the next chapter for me to learn.Bob Katz may be a brilliant engineer but that is the usual answer and I'm sure most of the older me's here did'nt just go read bob katz book.It may be usefull as a reference but if everyone followed the same rules every master would sound the same or very similar. Oh well if no one wants to talk audio,has anyone been fishing lately and caught anything?Hows the weather?Has'nt rained here for 8 weeks. Also hope another group master project is started on this forum would be a good learning experience with constructive criticism.Cheers :D

Contrary to popular belief, louder mixes do NOT sound better on radio, they tend to sound worse. I won't get into all the crazy reasons behind this because the technical reasoning would be an entire chapter of a book. But suffice to say, most CDs sent to radio are too loud. CDs that are a little quieter than average with more dynamics will sound better when broadcast. The station has plenty of multi-band compressors and limiters and other goofiness to make damn sure all the songs are the same perceived volume so even if you send in a CD that is 10dB lower than everything else, it will still sound just as loud. The key is to keep as much of your dynamics as possible so that all the compression going on at the station isn't just recompressing everything. I always try and put a lot of dynamics in my mixes and compress as little as possible and I'm usually successful in getting the labels to go with mastering engineers who don't compress the crap out of the mixes and shoot for a nice average level instead of trying to make the CD the loudest thing on the planet. And when i hear one of my tunes on the radio they tend to slam a little harder than other tunes. They are the same perceived volume 'cause the station makes sure of that, but they slam a little harder. If you really want to know why this is the case, there's a great chapter on it in the Bob Katz book.

heathen22 wrote: Bob Katz may be a brilliant engineer but that is the usual answer and I'm sure most of the older me's here did'nt just go read bob katz book.It may be usefull as a reference but if everyone followed the same rules every master would sound the same or very similar.

Two comments:

1) The chapter I referenced in Bob Katz's book is directly related to your question.

2) I would recommend that you use paragraphs to improve the readability of your posts

Okay – it is late at night…so I hope this will all make sense…

I have mastered quite a bit for radio, whether it be classical, jazz, folk, rock, etc. I am fortunate to work in the best of both worlds in regard to the question that has been posed...being an audio recording and mastering engineer, as well as working in close association with several radio stations, or various genres. Furthermore, I have an intimate understanding of broadcast audio processing and have setup broadcast air chains for several radio stations across the United States, including the most recent - WGUC-FM 90.9 FM in Cincinnati. In addition, one of my closest friends is a broadcast engineer and processor designer who developed numerous products for the broadcast industry. He has provided me an even greater understanding and education.

The previous posting from Chris Carter is dead on. Louder mixes do NOT sound better on radio. CDs that have greater dynamic range and are a little quieter will indeed sound better on the radio. For radio broadcast masters, there are a couple of things to keep in mind. One is, do not use a lot of limiting and/or clipping. Due to the implementation of phase rotators in combination with the limiters/clippers in broadcast audio processors and the nasty 50 or 75 microsecond pre-emphasis curves (75 in the US), the more clipped audio an audio broadcast processor sees, the more IMD the broadcast audio processor will produce. This is especially so if the source material has a great deal of content in the mid-high frequencies (4 kHz and above) that is limited/clipped and/or accentuated. Bob Orban, basically the father and inventory of some of the finest broadcast processors, has emphasized in his application notes that any clipping of the audio before the broadcast processor will be exaggerated severely.

A brief education on broadcast audio processors…
Broadcast audio processors original intent was to provide a consistent average listening level and allow a station to color (automatically EQ) the sound to make their station(s) sound unique and provide some ‘punch’ on the dial. Also, the processors were designed to minimize the distortion byproducts cause by the pre-emphasis curve and to keep a station from going over 100% modulation. However, when these processors were invented and the pre-emphasis curve implemented, there were NOT a lot of recordings with a great deal of higher frequency content nor where they heavily limited/clipped as they are today. So, the amount of IMD that these processors created wasn’t nearly as great as we hear today. As time evolved, broadcast audio processors have become far more advanced. However, with the unfortunate proliferation of “louder is better,” these audio processors are often abused. Also, many broadcast engineers, nor mastering engineers, truly understand how a broadcast audio processor actually works as they did 15 years ago…the science is simply not shared/taught as much, making the problem worse. In addition, some broadcast manufacturers say that their boxes can go louder than their competitors, which further magnifies this problem.

So – I’ll attempt to provide a little bit of education on broadcast audio processors now. Following is an edited post I made to a group of mastering engineers a few years ago, including some of the worlds best. It describes many of the problems of audio processors as well as one of the great things that keep them from sounding even worse – distortion cancel clipping.

Here we go…

The basic gain structure of most broadcast audio processor is as follows:
1) An AGC (Automatic/automated Gain Control). These generally are set to gain ride the audio on the average of 8-12db. Consider this a mild compressor with variable attack/release set with an extremely low threshold, with plenty of make-up gain.
2) An equalizer.
3) 75us or 50us Pre-emphasis (75us in the United States)
4) Compressor/limiter section (either multi-band or wideband)
5) A steep Low Pass Filter (LPF) at/about 15kHz, with minimally –30db at 19khz. (This is used to protect the FM pilot – which is another conversation)

Through the above series of audio events, the audio signal gets compressed then gets a lot of its higher frequency content boosted due to the 75us Pre-emphasis which boosts HF at about +3db at 2120Hz, and +17db at 15kHz. Ouch! Now, when you have program content such as pop music with a lot of high frequencies, or just the human voice producing sibilance around 6-7kHz, those high frequencies get clipped before hitting the airwaves. Now, this ‘clipping’ of that 6-7kHz material produces THD. In this case, that THD is around 18kHz or so. Big deal, the LPF gets rid of it anyway. If the ‘clipper’ affects frequencies around 3kHz, the THD is around 9khz. However, the signal has to get de-emphasized (opposite pre-emphasis = ~–3db at 2120Hz, and -17db at 15kHz) by the listener’s receiver. That’s reducing that 9kHz crap by 12db or so below the regular audio content. Therefore, for all intensive purposes, THD is not the problem.

The problem comes when you have ‘clipping’ of say sibilance occurring at 6 & 7khz. The clipping produces sum, and difference distortion. The sum of 6 & 7khz is 13khz. Like the THD, the sum isn’t really a problem due to the de-emphasis curve on the listener’s receiver. It’s the difference that’s the killer. The difference between 6 & 7khz is 1khz. That will be audible since the de-emphasis curve hardly touches it. Plus, the ear is more sensitive to midrange. If the difference were lower than 1khz, it wouldn’t be touched by the de-emphasis at all, making HF distortion audible and unpleasing (sibilance splatter). It can be easy to hear on some radio and TV stations. It sounds like sizzling in a frying pan when someone is very sibilant or with high frequency content audio.

Nice thing is that Bob Orban (if I am not mistaken) invented distortion cancel clipping. This is where a side chain of audio processing basically exists. The audio is pre-emphasized and applied to a clipper and a 15 kHz low pass filter. The distortion is the difference between the clipper’s input and output, and then is fed to a 2 kHz low pass filter. Now you have a sample of the distortion below 2 kHz. This is subtracted from the main audio that came from the clipper and 15 kHz low pass filter. So distortion below 2 kHz, (most of your difference IM distortion that produces the “splatter”), is reduced. This is one of the most powerful techniques in broadcast processors today that allow them to get louder with less distortion. However, this invention can only go so far and is now becoming defeated by the great amount of overly squashed, limited, clipped – essentially slammed CDs that are commonly fond in the marketplace.

So, long story is, the propensity of highly squashed, limited/clipped CDs is indeed adding to the poor sound or radio today. So, make your CDs sound great but DON’T be as loud as the next guy and don’t go limiter/clipper crazy.

Alex and Chris thanks,was a good read,very informative.I like a good dynamic mix,the sound of an L2 thrashing the guts out of a nice mix is not that pleasing.I have had good results with them but just as a brickwall limiter without raising the gain too much.I'll leave it off for any radio masters I may attempt and see what happens.My al smart c2 would be a better choice for that I guess,using a bit of comp and boost,the mix would still be quite dynamic.
dpd thanks and sorry I did'nt mean to confuse you,yeah I was'nt so good at english lessons at school,but I did get upset if I got lower than 94% in a science exam,cheers.
Thanks guys.

^^ OK, Jason Fox! :D

Alex - that may rate as 'Post of the Year'. Thanks for taking the time to explain that. I'm not sure I would have had the patience to put something like that together.

We just installed a new Orban processor the other night in preparation for our conversion to HD Radio (IBOC). Today the jazz guy tells me he hears and echo in the headphones. I tell him to switch the monitor from Air to Program in the studio - latency in the new processor is excessive enough to notice.

dpd wrote: ^^ OK, Jason Fox! :D

Alex - that may rate as 'Post of the Year'. Thanks for taking the time to explain that. I'm not sure I would have had the patience to put something like that together.

We just installed a new Orban processor the other night in preparation for our conversion to HD Radio (IBOC). Today the jazz guy tells me he hears and echo in the headphones. I tell him to switch the monitor from Air to Program in the studio - latency in the new processor is excessive enough to notice.

Thanks for the comments...all I want to do is educate and make audio better for others.

Which Orban processor - 2300 or 8500? I have extensive understanding of both the Orban product and IBOC, if you would like assistance on either, I can explain a lot (email me off list) and get rid a lot of the so called misguided propaganda. Bob Orban actually updated the 8500 and wrote software for us here in Cincinnati. We were beta testing the software which is now released and has some amazing functions which are not commonly know (such as on the fly switching/optimization of music and voice audio processing). I have written 2 custom presets for the 8500 as well by the way, one for WGUC and one for another station.

Some other tidbits...WGUC here in Cincinnati was the first public radio in Ohio to utilize IBOC, one of the first to utilize multi-casting and we are likely the only radio station in the US testing all three major proponents of surround sound broadcasting on HD as well. I am on the NRSC Digital Audio Broadcast subcommittee SSATG, so I could assist you a lot if you'd like. I'd actually like too...because, not to burst your bubble and no offense to Indianapolis either, but I have heard WHYI in Indianapolis many times - they need major audio help! One of the poorest audio managed major public radio stations I have heard in a long time. Haven't heard very good audio processing throughout much of Indiana (commercial or public radio) either. Is there anything I can do to make a difference there...I promise results.

Damn AlexAudio!! :shock:

Are you always this rude, or is this something new?? I've been lurking on this board for a long time, but didn't even get an account til today when I saw this post.

The dude compliments you and you trash him in return?! :shock:

Does this guy sound like he needs your help? Did he ask for your help?

Oh, and by the way - I used to live in IN (Ft. BH) - I've never heard of a WHYI! I've heard of WHTI - classic rock - we called this station "music for old white dudes by old white dudes"

Peace 8)
SB

My apologies if I came off rude, that was not my intent. Dpd – I welcome your comments.

Alex

alexaudio wrote: [quote=dpd]^^ OK, Jason Fox! :D

Alex - that may rate as 'Post of the Year'. Thanks for taking the time to explain that. I'm not sure I would have had the patience to put something like that together.

We just installed a new Orban processor the other night in preparation for our conversion to HD Radio (IBOC). Today the jazz guy tells me he hears and echo in the headphones. I tell him to switch the monitor from Air to Program in the studio - latency in the new processor is excessive enough to notice.

Thanks for the comments...all I want to do is educate and make audio better for others.

Which Orban processor - 2300 or 8500? I have extensive understanding of both the Orban product and IBOC, if you would like assistance on either, I can explain a lot (email me off list) and get rid a lot of the so called misguided propaganda. Bob Orban actually updated the 8500 and wrote software for us here in Cincinnati. We were beta testing the software which is now released and has some amazing functions which are not commonly know (such as on the fly switching/optimization of music and voice audio processing). I have written 2 custom presets for the 8500 as well by the way, one for WGUC and one for another station.

Some other tidbits...WGUC here in Cincinnati was the first public radio in Ohio to utilize IBOC, one of the first to utilize multi-casting and we are likely the only radio station in the US testing all three major proponents of surround sound broadcasting on HD as well. I am on the NRSC Digital Audio Broadcast subcommittee SSATG, so I could assist you a lot if you'd like. I'd actually like too...because, not to burst your bubble and no offense to Indianapolis either, but I have heard WHYI in Indianapolis many times - they need major audio help! One of the poorest audio managed major public radio stations I have heard in a long time. Haven't heard very good audio processing throughout much of Indiana (commercial or public radio) either. Is there anything I can do to make a difference there...I promise results.

Alex - we probably ought to talk offline. I'm up in Fort Wayne. We are getting ready to turn on IBOC and Tomorrow Radio (e.g. multicast). IIRC, we are using the 8500 - it just arrived about two weeks ago. I'm so busy with updating the station's ProTools and AVLT systems I almost didn't notice the Chief installing all the new stuff in the Transmitter Room.

dpd wrote:

Alex - we probably ought to talk offline. I'm up in Fort Wayne. We are getting ready to turn on IBOC and Tomorrow Radio (e.g. multicast). IIRC, we are using the 8500 - it just arrived about two weeks ago. I'm so busy with updating the station's ProTools and AVLT systems I almost didn't notice the Chief installing all the new stuff in the Transmitter Room.

Feel free to give me a private email - I'll shoot you my # and we can chat. Interesting is that this weekend, I am conducting a pilot cross compatibility test of composite surround sound systems for radio broadcast, in cooperation with the local AES and NPR. It is being done to ITU-R specs and the results should be interesting. That is another conversation all together.

Alex

dpd wrote:

Alex - we probably ought to talk offline. I'm up in Fort Wayne. We are getting ready to turn on IBOC and Tomorrow Radio (e.g. multicast). IIRC, we are using the 8500 - it just arrived about two weeks ago. I'm so busy with updating the station's ProTools and AVLT systems I almost didn't notice the Chief installing all the new stuff in the Transmitter Room.

dpd -

Not sure if it was you or someone else that tried sending me a private email, but I have not subscribed to that service as of yet (don't see the $20 justified as of yet). Feel free to just go to my webpage and send me an email through there...
http:// - go to the contact us section... Thanks.

Alex