Setting levels at Post production.

Discussion in 'Recording' started by Haamilton, Nov 16, 2012.

  1. Haamilton

    Haamilton Active Member

    Another question i've been having recently;
    how do i know that my sound is loud enough or too loud?
    if it's too loud, during mastering it muds
    if it's too soft, during mastering it's too sharp
  2. kmetal

    kmetal Kyle P. Gushue Well-Known Member

    Boston, Massachusetts
    Home Page:
    average monitoring level is about 85 db. as far as levels go, i record to about -15/20, and my mixes end up around minus 10/12ish. if it's going to mastering i leave it there, if not strap on some bus stuff and be done distortion free, and w/ some dynamics. talking about digital full scale output bus meters. and just a typical radio shaft db meter or a phone app.
  3. Haamilton

    Haamilton Active Member

    i hate to be noobie on that; but can you explain that in simple terms, so i understand those concepts?
  4. kmetal

    kmetal Kyle P. Gushue Well-Known Member

    Boston, Massachusetts
    Home Page:
    nobody starts off without being a noob, no shame hamillton. hope i can help.
    when you listen to your reference mixes, and what your recording/mixing, there is a 'volume' level where people are reguraded as having the most 'accurate/flat' hearing, while minimizing hearing degradation. check outFletcher. it relates test results, how accurate is our hearing, depending on playback volume.? the flattest line is theoretically the best.

    so, when you are starting to record say, a mic'ed guitar w/ a stock pre-amp for an interface digitally, you have ta work a 'trim/or gain' knob for whatever you plugged in. lets just say your micing a guitar, so you put the mic in front of the amp/or acoustic's body maybe a fist width away.

    lets plug into channel 1. after you assign the in/outs you record enable the software, usually by clicking an 'R' on the new track.

    Now it's "gain staging" this means ya get the most workable 'volume.level'. usually start w/ any faders at 0, (or unity) use thr 'trim' knob while someone is playing, and watch the meters/listeten. when the meters are around half way your in the area for digtal. yellow is the new red. they'll likely read -20 to - 15 on their number markings. do this every channel.

    so when your mixing the stuff, the almost inevitable enhancement of sonic quality (of ruining thereoff lol) will boost the levels you hear thru your main output. this shouldn't usually hit 'the red" in digital, because you still have to make a master copy peeking at yellow is more than enough.

    that'll setup the levels that your 'system is hearing'. now ya gotta hear that from the speakers.

    so turn up the master, or main, output on your interface/consonle. Use your phone app, or radio shack 'DB Meter' and put it on the desk, or hold it near your ears, keep the levels between 80-90 and you'll be roughly average.

    turn up loud to hear bass, and very importantly turn very low to hear overall balance.
  5. bouldersound

    bouldersound Real guitars are for old people. Well-Known Member

    Boulder, Colorado
    When he says "85dB" he means 85dB SPL, probably C weighted, which is the acoustic loudness in the room. That's your standard listening volume, but you want to check mixes loud and soft as well.

    The other numbers are in dBFS, the level in the digital recording software. The dBFS number is always negative or zero, as it refers to the highest level possible in digital recording, 0dBFS. You always want to leave some space between the peaks of your audio and 0dBFS. Keep your peak levels around -12dBFS when you record. When mixing keep the overall mix peaks from getting too close to 0dBFS. If you're doing a rock mix and starting with the kick drum you might want to set it to peak around -16dBFS or so to allow for the buildup of level as you add tracks.

    Later, after the mix is done, you can process the stereo mix for overall loudness. Don't try to do that in the mix stage.
  6. DonnyThompson

    DonnyThompson Distinguished Member

    Akron/Cleveland, OH
    Home Page:
    ...and, if you are going to have it professionally mastered at some point, avoid using any hardcore dynamics processing and allow around -6d of headroom for the M.E. to work with. If you send them a master that has already been limited to -0.5, there's no room for them to work.

    For that matter, avoid strapping any processing to the 2 Mix until you know what you are doing. ;)

  7. mberry593

    mberry593 Active Member

    Silver Spring, Maryland
    Home Page:
    What are you doing? The answers are different. If you are doing music only, the comments above are appropriate. I would only add that it would be good to try to get the analog Radio Shack meter rather than the digital one. It is both less expensive and, IMO, easier to use.

    But when you use the term "post production," I must assume that you have something to do with broadcast. In that case, things are much more difficult.

    In Europe---
    You should conform with EBU R 128
    EBU Technology & Innovation - Loudness
    EBU Technology & Innovation - Loudness
    EBU Technology & Innovation - Loudness
    EBU Technology & Innovation - Loudness

    You may find this to be of interest too...
    EBU Technology & Innovation - Loudness

    The good news is that the Hofa company in Germany have a free plug-in that will allow you to measure your material for compliance. It is the model 4U meter, fader & pan.

    I have played with this and find it to be perfectly adequate. The fader & pan are sort of irrevalent to the purpose of the metering but they don't get in the way.

    In The US---
    The Hofa plugin isn't for the US. It is one dB off for the US standard. -23 vs -24

    Here are the US standards.
    Don't worry too much about the metadata, networks have been stripping it off.
    Also the epicenter of unpleasantness is right here in Silver Spring, Maryland where The Discovery channel has been irritating people by rejecting programs that don't adhere to their own standards.


    Notice that they expect measurements from a Dolby LM100. That's a $3,200 device. Even the Dolby Media meter plugin is $795.

    I have been happily using the Waves WLM.

    Broadcast Loudness Meter Plugin - WLM | Waves

    It is $400 but Waves is almost always running some kind of sale. I was able to purchase mine for much less by combining a WUP voucher with a sale.


    EDIT: The Hofa link is broken. Cut & paste this....
  8. bishopdante

    bishopdante Guest

    I can't work out from your question if you're talking about monitoring level or meter level?

    Are you asking about meter level of stereo files sent off for mastering? If so, you know who you should ask about that? Your mastering guy.

    There are various standards out there. If you send stuff to the BBC, they want -24dB if you're using 24 bit, and will send them back if they're all-over considerably louder. That's conservative, and many people complain about such stringency. However, that's not too bad an idea, since each bit accounts for approximately 6dB. They're suggesting that you send them 20 bit files plus some headroom. In contemporary music, the standard is to supply peaks somewhere between -6dB and -12dB meter level.


    Most mastering people will appreciate it if you don't overcompress your mix, and leave that to them. You might not agree about that, though. Some people will prefer stems, and have you go in there and do a final master mix with them. It's not cutting a vinyl stamper anymore, a master, so it's basically an extension of the mixing process, there's no exact rules in that respect. In today's sense it's just the finishing of a mix. Although you might be cutting vinyl, in which case you need to be far more careful about how you mix.

    Regarding monitor level, that shouldn't really be changing how you mix. You should probably get a bit more technical than that.

    Also, part of mixing is using all sorts of different monitor levels from whisper quiet barely audible to rip-yer-face-off. Put it through a ghetto blaster. Take it to a club and put it through a PA. Listen in the car. Listen in a friend's living room. Put it through a bunch of different output systems. In my opinion it's best not to use just one monitoring level, and just one set of monitors.
  9. RemyRAD

    RemyRAD Member

    Your description only defines your naïveté. Because if it's too loud, it's too loud. If it sounds like mud. You're listening to your recording technique and not the equipment. If it's too low? You don't know how to get the best bits out of the machine? I mean why record everything as the least significant bit? And of course you don't want to be looking at red peak meters because peak does not mean it is at its best. An easy misinterpretation to make. Good engineering really doesn't need any Mastering. Lousy engineering requires this magic pill that people know how to deliver as Mastering Engineers. Because really good mixes are simply really good mixes and they sound like it regardless of level. The loudest CD on the block simply means that they traded off enough fidelity with just the right amount of distortion. Because you're not going to have the loudest CD on the block unless you limit, clipboard truncate all of your peaks and transients. And then what happens to your music and the way it sounds? There's your mud. Rightly so.

    So really the answer to your problem is just learning how to engineer better. All of your assumptions are really quite way off the mark. Not even close. If you feel I'm being a bit hard on you and verbose, I apologize. But that's really the name of that tune. Why does everybody think that making a good recording is a one click procedure? It isn't. It might be if you were a kid playing with an iPod? But real engineers like being real engineers and not playing with toys. So ya have to make a fundamental decision. Do you want to be an engineer or do you want to play with toys? Now this is not saying that entry-level equipment are all toys. They're not. They are 100% adequate tools to use. And learning how to use all of the beginner stuff first to obtain good recordings is usually the proper procedure to take. I mean you don't go to college first and then enter elementary school do you? No. Of coarse not. I mean perhaps so if you were a wunderkind? But there are very few like that. I don't think you're one of them? I know I'm not. Well maybe I am? Yeah, I am.

    So you've gotten a lot of good feedback here from a lot of very knowledgeable people, including one of my supervisors from over 30 years ago. And you can tell he was a supervisor over it NBC-TV and Fox by all of that wonderful techno-speak and qualifiable specifications. Which generally doesn't mean much to the home recording enthusiast. But it's all part of getting all of your ducks in a row... boat. And then trying to show them how to use the oars with those little flat duck beaks. What a hassle. And their little wings go flapping about in that you have to duct tape them to the bench in the boat. And you don't want to have to go through anything like that right? And you know that ducks don't have any sphincter muscles? And that doesn't usually bode well by the client either. So forget about getting all of those ducks in a row... boat.

    So all you're really complaining about is the fact that you really don't know how to set your record levels properly. This setting of levels varies greatly depending upon your signal source. Legato musical lines make it fairly straightforward. Highly percussive transients make the job a little more challenging. Understanding that you want to record levels just shy of making your peak lights, light, is where you basically want to be. Different meters have different kinds of visual ballistics. Old-fashioned mechanical VU meters, we used since the beginning of time, never indicated any peaks. You had to know that whatever levels you were looking at on the VU meter, that your peaks were going to be +10-+20 D. be beyond what you could see on the meter. An analog did not clip on the tape recorders quite like digital even when that happened. So today it's a little bit more of a juggling act. Bad microphone technique and placement, makes it more difficult to establish proper and consistent recording levels. Well let comments from microphone placement as well. So it's not just from not tweaking the gain trim properly.

    For folks like yourself that have trouble setting gain trims properly, at least you can take advantage of higher resolution recording. Recording at 24 bit at any sample rate, will provide for some improvement. Though it is still restricted by the analog front end. And if you don't know how to set the preamp gain trim properly, you will always have these problems.

    In other aspects of prior recording history and something that many of us still do, is that your preamp from the microphone gets routed to a compressor/limiter. This dynamic range modification can make it much more easy to squeeze 10 pounds of sound into a 1 pound disk drive. So some folks opt for an external microphone preamp/EQ/compressor gizmo as a specialized feeder for important or problematic sources. And a little goes a long way. Only drawback is, if you don't do it right, you got something worse than if you had done it wrong without it. So then you can only think about that great line that Clint Eastwood made in the dirty Harry movies and that's... Are you feeling lucky today... punk?

    So if you're recording in the studio, this should not be an issue at all. If you're trying to do impromptu location recordings, they can get a little more dicey.

    Since I came from the birth of digital recording back in 1983, 16 bit at 44.1 kHz, was all there was. And while it offered performance in certain areas that far exceeded any analog recorders, there were also trade-offs. Because you don't get something for nothing. There's always a price to be paid. And in many ways, sloppy level setting became even more apparent in the digital age. Digital clipping can be downright awful. Though minor transient clips frequently go unnoticed. Anything that's legato in nature will suffer greatly when clipped. So when we recorded at 16 bit, 44.1 kHz, we wanted our recording levels should be as high as possible to take full advantage of as much resolution as we could squeeze out of 16 bit, 44.1 kHz. Because with digital, lower levels mean lower resolution. With analog, it simply meant more noise or hiss. So our multitrack analog recorders actually sound like a bunch of angry snakes if you didn't know how to charm it. So we used additional noise reduction devices and downward expanders and gates to help keep that hiss, in check. And we don't have to deal with that at all even with 16 bit, 44.1 kHz digital, today. So anything that's 24-bit and higher in sample resolution is virtually overkill. It's not overkill if you're recording some famous rock band and have been contracted for $50,000 for a year-long project, to want to record at the deepest bit depth and highest sample rate possible. For everything else... there's MasterCard. I mean we did not use 30 IPS on our 24 track analog machines on all projects. Frequently, we just utilized 15 IPS. What's that tell ya? It still comes out is a beautiful professional product. Nothing has been lost by recording at 15 as opposed to 30. Though there is a different sonic signature. Did that mean 15 was not adequate for professional work? Hell no. In fact sometimes you'd want 15 over 30. And I look at our bit depth and sample rates much like that. I really don't think that rock 'n roll needs resolution like symphonic work does? A basic element of rock 'n roll is distortion. Not true for a symphony. Soda 16 versus 24 and 44.1 versus 96 make any difference? Only if somebody is paying you and requests that difference and of course anybody that wants to sell you something. I mean some engineers know what they're doing and other engineers cannot remove a childproof cap from a bottle of aspirin. Thank God otherwise they might get poisoned? Because they've heard you need a lot of aspirin to make a hit record. But you're not supposed to take it all at once. I mean for all the crappy rock bands I've recorded... I almost wanted to eat an entire bottle of aspirin at once. Just so I could puke blood on their recording.

    Don't try to swallow a tampon because you could choke.
    Mx. Remy Ann David
  10. GZsound

    GZsound Active Member

    Near Portland, Oregon
    Home Page:
    I have a way of setting levels that seems to work without being very complicated.

    I have a group of songs from artists in the same genre as the music I am producing. They are all national artists with big dollar studio recordings that have been nationally released.

    I find three or four songs that match the style of the song I am producing. Say a ballad, or a fast tempo song, etc.

    I set my Radio Shack DB meter up eight feet from my monitors. I get db readings from the national recordings. I then check and set levels on the song I am producing to make sure it matches national release levels.

    I don't pay attention to db except to make sure my song averages the same level as the national recordings. I use Waves L1 to set levels properly without distortion.

    One of my previous recordings was chosen for inclusion on a national release compilation CD. We had decided to go for quality instead of making a two by four of the song. Big mistake.

    When played in context with the other songs, mine sounded worse than the others because the level didn't match the rest of the CD. It was too quiet. Never again.
  11. bouldersound

    bouldersound Real guitars are for old people. Well-Known Member

    Boulder, Colorado
    I open the file in Sound Forge and run the analysis tool. It shows the RMS value of the audio. But I always use my ears to match levels subjectively.

    So they just threw all the songs together without properly remastering? They should have gotten the unmastered 24/48 mix files (or whatever was the best available) and finalized all the audio to be compatible. That's their failure, not yours.

    I had the same thing happen on a podcast. The song I mixed and mastered was quiet compared to the others because the amateur podcaster just played the files directly without any level adjustments between songs. That's not how a real radio station (the model for podcasting) does things. My response was the same, if that's how they operate I'll have to slam my stuff harder to stand up against other songs since you can't expect the people playing it to make the adjustment like you could with a real radio station.
  12. RemyRAD

    RemyRAD Member

    This kind of stuff has been going on since the beginning of time. I had a similar instance happened to me back in 1978. A song I recorded ended up on the local rock station's compilation album. Our competition was going to " Master ", the album. So I'm glad I went to the mastering session. Sheffield Recordings Ltd. felt I had too much low-end in my recording. So they world my low-end off by 9 DB! Which was way way too much. Finally got them to split the difference instead. Then it was still weak on the low-end but then so was all of their stuff coming through their lousy MCI JH 416 console and to their KLIPSCH Monitor speakers. And their MCI console couldn't hold a candle to my custom console.

    C'est la vie
    Mx. Remy Ann David

Share This Page