Post audio levels

[Rick Reineke]

http://www.aice.org/...C_a_85-2009.pdf is good reference,... 'post' levels..

Totally different from acquisition levels... as hot as possible with out exceeding 0dB (EASY) but allowing enough head-room and/or riding gain. (NOT SO EASY)

[Henchman]

The CALM act is amed at both TV programming and Commercials.

The standard is -24db +- 2db. So a maximum of -22 minimum of -26.

I have found that with a calibrated level of 80, I can hit the spec every single time, without needng to measure it during the mix.

I start off by getting the dialog to hit around an average of -20 or 0 VU. And once my ears have adjusted, I don't need to really see the meters at all the rest of the mix.

[Philip Perkins]

One last pitfall to watch out for is the pink noise you use to calibrate.Nearflied monitors or monitors suitable for small rooms are normally not full range and have quite a severe roll-off below about 40Hz. If you are using standard 20Hz-20kHz pink noise to calibrate, part of the energy (the octave below 40Hz) being output by your D to A converter is not being output by your speakers, so you will have to increase the output of your speakers to achieve the desired SPL meter reading. In other words, your speakers will sound louder than the SPL meter reading indicates. I can't tell you how many small studios fall into this trap! The solution to this problem to use compensated band limited pink noise. Tomlinson Holman used to distribute a 500Hz - 2kHz pink noise file specifically for this purpose.

This is a bit of mis-information. C weighting is a flat curve... for the most part. A 85db(a) pink reference is going to be basically the same at 400hz as it is 10000hz. If anything c weighting has a slight downward curve in the lows and would create the opposite problem you have described above. The biggest reason for using pink noise is so the energy difference between the octaves is the same so missing frequency ranges won't be an issue.

Unless you are measuring with a white noise reference, a speaker eq shouldn't make any difference. Every speaker manufacturer and every model of speaker have different eqs.

Also, just because I'm curious and your profile doesn't have any info, what are your qualifications?

A little background about me. I have a degree in telecommunications (concentration in audio) a degree in acoustics, and most of a degree in computer science.

Thanks for all that good info, including that about your qualifications. Could we know your name too? Sorry if I missed it in an earlier post.

thanks

phil p

[Philip Perkins]

Chris calandro. Also I want to reiterate I'm not trying to be snooty. I just want to know what his background is and where he got his info.

I got no issue w/ snootitude as long as the snoots are signed. Thanks again for your info on this and other threads. As you may have figured out, many of us just have a disorganized accumulation of experience to guide us instead of a formal technical education, so members with both, such as yourself, are doubly valued here.

phil p

[ccsnd]

I have always been one to go the extra step to figure out not just how it works, but why.

I think the greatest thing I learned in college was that - anybody can figure out how to use something. But knowing why and how it works allows you to improve and use a tool in ways others wouldn't think of.

[AudioPostGuy]

Chris calandro. Also I want to reiterate I'm not trying to be snooty. I just want to know what his background is and where he got his info.

My university education (and yours) are irrelevant in the face of current industry practice and standards organisations, which dictate theatrical and broadcast calibration levels. I'm happy to tell you that I've 20 years professional experience in the audio post industry, if that somehow makes the information I've posted more valid. I do understand you want to be able to research or verify the info I've given though, I would as well. So, I have two points raised by your post, to which I would like to respond:

1. "The biggest reason for using pink noise is so the energy difference between the octaves is the same" - Agreed. "..so missing frequency ranges won't be an issue." - I don't agree. I think you misuderstand pink noise. Take any octave, anywhere in the frequency spectrum and the energy it contains will be the same as any other octave in the frequency spectrum. However, the energy contained in say 2 octaves of pink noise will not be the same as the energy contained in say 6 octaves of pink.

2. "Where did he get his information" - I would like to quote from Dolby (a print-mastering manual), who largely dictate worldwide cinema and theatrical mix stage calibration: "There are two settings for test noise: “PINK” or “FILTERED.” If no bass redirection is chosen (“NONE” - default), calibration of the monitoring system should be made using the “PINK” test noise option. This option uses a full-band pink noise signal. If a Bass Redirection option other than “NONE” is selected, calibration of your monitoring system should be made using the “FILTERED” test noise option. This will compensate for the absence of low frequencies in your main monitoring speakers."

The ATSC supplied pink for calibration is 500Hz-2kHz, which they reccommend over full range pink, as does Tomlinson Holman (THX) for smaller rooms. Maybe the folks at ATSC, Dolby and THX don't have qualifications to compete with yours but to be honest, that's irrelevant. If you want to get into the fine detail of room calibration, it gets complicated, even to the point of different brands of pink noise suitable for testing. Avid's pink noise generator (in Protools) is about 3dB out for example. And, we haven't even mentioned the various x-curves, which obviously also affect calibration levels. At the high end, room calibration is as much an art as a science and in the vast majority of cases the best that can be done with a small room is to get it "in the ball park". To be balanced in my info, it should be noted that some very well respected audio post practitioners disagree with using compensated band-limited pink for calibration.

You can accept blindly all the info I've posted, research it's accuracy and decide for yourself or you can dismiss it out of hand as nonesence. I've just posted pertinent info which I've gathered over the years and I thought might be useful to others, what you decide to do with it is entirely your decision.

G

[ccsnd]

I'm in the pool of people who think this is total mumbo jumbo. Pink noise is used specifically so issues like eq curves make no difference in meter readings. Db© was created to have a flat curve and thus measures at the highest db at any frequency.

That being said,

Percieved loudness all comes down to your speakers curve. Your speakers curve will also dictate your meter readings. Many speakers manufacturers (I won't say which) have a tendency to scoop their mids. Sometimes where a lot of dialog lies. When running pink noise through these speakers your going to hit your reference (because of the high or low frequencies) earlier giving making your mix sound low. THIS is where studios drop the ball. They mix to their settings and speaker eq, but when it gets to the final stages (or the theater) the mix ends up being way off.

Having good well tuned speakers is the most crucial part of not only mixing, but setting up the room.

I never got my my certifications but I have been involved with setting up a few prominent Dolby/THX rooms. I have heard the theory about how the low end roll off effects the meter reading before. Myself and many certified techs agree that theory is totally illogical, as that is not how db© works.

I would equate this to monster cables. Some people insist that they make a difference.

I do agree that avid pink noise is crap.

[ccsnd]

Also I should mention the ATSC 500-2k was made to help combat the problem of poorly eq.-ed rooms. You don't think it is a coincidence that those are the frequencies where it is generally recognized that dialog lives?

It was an alternative to get lesser quality equipment and rooms closer to the spec. Not a replacement for full frequency pink in proper mix stages.

[AudioPostGuy]

...When running pink noise through these speakers your going to hit your reference (because of the high or low frequencies) earlier giving making your mix sound low.

Not if your neafields can't reproduce the low frequencies you're not, this is the point I was making!

I would equate this to monster cables. Some people insist that they make a difference.

We'll have to disagree on this one too. They obviously do make a difference, as anyone with a bit of measuring equipment can verify. However, the difference, IMHO, is way below the threshold of audibility.

Pink noise is used specifically so issues like eq curves make no difference in meter readings.... Your speakers curve will also dictate your meter readings.

Huh?

I'm in the pool of people who think this is total mumbo jumbo. ... I have heard the theory about how the low end roll off effects the meter reading before. Myself and many certified techs agree that theory is totally illogical, as that is not how db© works.

Let me see if I understand the situation correctly: You don't like that I've come on to this forum and trodden on your little toes, so you try and call me out by demanding my qualifications and source of info. I quote you directly from Dolby themselves but nevertheless I'm talking mumbo jumbo and Dolby's theory "is totally illogical" and what eminent source do you use to refute and insult me (and Dolby)? ... You and a few of your tech mates! Sorry, I'm not even mildly impressed, I prefer to stick to the recommendations of Dolby, ATSC and Tom Holmann thank you very much!

My bad though, I thought this was an open forum where professionals could discuss issues and their understanding of issues, rather than your own private little fiefdom where whatever you say is the gospel. At least I've discovered early rather than wasting a lot of time/effort before finding out. Enjoy your illusion of importance, I'm outta here!

[TheBlimp]

peace

[ccsnd]

I never demanded anything. I polity asked, and then gave my background.

I never said anything was "the gospel". The reason I asked your qualifications was because I wanted to know where you got your info from. Not because I think I'm smarter or better than you.

[ccsnd]

Also, this is an open forum. A lot of things that get posted on here are opinions, but some are facts. (what db© is, what pink noise is).

There is no reason you couldn't use a slice of pink noise as a reference. But my opinion is that if you need to use only a section of pink to get a proper reference, then you have other issues in your room that should be addressed first.

There are reasons to use different approaches, but the reasons you mention are not ones I'm familiar with. Again, that is why I asked about your background.

There is a big difference between someone with krk's in their bedroom talking about how to set up a room, and someone who built skywalker ranch talking about how to build a room.

[Marc Wielage]

The CALM act is amed at both TV programming and Commercials.

Correct, but... CALM is a broadcast standard, not an industry mix standard. The standards I see mixes now being referenced to are the EBU R128/ITU 1770/ATSC standards. What's a concern is, there are cable channels just playing back shows through a box, and if the meter comes up at a specific LUFS number, they'll bounce the mix as being bad. I know of cases where bad-sounding mixes make it through, and good-sounding mixes get rejected.

The good news is that these metering standards are widely available as plug-ins (some very low-cost), and it's not that much trouble to do one playback pass to see how it goes through the meters.

I also again refer people to Marti Humphrey's lengthy discussion about mix room levels on the Gearslutz Forum.

[Henchman]

Yes, CALM is a broadcast standard. And us re-recording mixers are required to deliver mixes that adhere to that standard. If they don't, QC sends them back.

And as I stated earlier, if you've mixed enough shows, you don't really need the meters at all.

[ccsnd]

What's a concern is, there are cable channels just playing back shows through a box, and if the meter comes up at a specific LUFS number, they'll bounce the mix as being bad. I know of cases where bad-sounding mixes make it through, and good-sounding mixes get rejected.

Coughhhhhh coughhhhhh coughiscovery coughhhhh coughhhhhhh

[Henchman]

Coughhhhhh coughhhhhh coughiscovery coughhhhh coughhhhhhh

None of the broadcasters make calls on the quality of the mix.

QC is strictly for techical issues.

Once the producers of a show sign off on the mix after playback, creative input is finished.

[Marc Wielage]


Yes, CALM is a broadcast standard. And us re-recording mixers are required to deliver mixes that adhere to that standard. If they don't, QC sends them back.

No, shows are intended to meet the requirements of EBU R128/ITU 1770/ATSC. They are not exactly the same thing, but they have the same intentions. The LUFS numbers will at least tell you where and when the mix will probably fail QC:

I believe the Discovery Channel was the first to implement the ITU standard, but pretty much everybody has. The CALM act was mainly intended by Congressmen to stop loud commercials in America, which is a separate problem from a film or TV show with exceptionally-wide dynamic range.

[Henchman]

Anytime every broadcaster adopts a certain requirement for a mix to pass QC, it's a broadcast standard in my book.

If my mix does not meet the spec as handed down by the network, it fails QC and gets sent back.

The standard was arrived upon by the networks together.

I was at the AES presentation about 3 years ago, when it was explained in depth.

Discovery actually has an even lower level. -28 if I'm correct.

But all the networks are requiring mixes to be delivered that meet the spec.

Commercials and shows HAVE to meet The Same spec.

The networks adopted this to get the programming in line with the CALM act, so that The commercials and tv programming levels are the same. As you kow,wen a show or commercial is received, its ingested, measured, and afkusten accordingly.

[Marc Wielage]

Anytime every broadcaster adopts a certain requirement for a mix to pass QC, it's a broadcast standard in my book.

Correct. But CALM is the Commercial Advertisement Loudness Mitigation Act:

http://en.wikipedia.org/wiki/Commercial_Advertisement_Loudness_Mitigation_Act

EBU R128 is the actual standard that says, average levels have to be at this number, peaks have to be no more than this number. That's the spec that will get your mix kicked back, not CALM. (Before that, it was the Dolby LM-100 specs.)

What I don't yet know is how far the ATSC specs differ from EBU R128, or if they're the same.

[cinetj]

Should the spl meter be set to A or C weighting for this exersize?

C weighting. And also I think another great forum could this time help this amazing discussion group. I know that some people here are familiar with gearslutz. For all of those that are not, here's an incredible sticky about calibration and levels for cinema, tv and dvd: http://www.gearslutz.com/board/post-production-forum/229741-standard-mixing-levels-movie-theater-dvd-broadcast-tv-commercials-etc.html

Tons of great information.