I calibrated my monitor, but how the hell do I calibrate cans?
You can't calibrate cans, that's just one of the reasons why cans are rarely used by by audio post professionals.
Editing 5 minutes of film might take anywhere from an hour to a day (obviously depending on what needs to be done) ...
Is that all, you must be really rushing it! Actually, it depends I what I'm doing. I get some rush jobs in sometimes, documentaries usually, where there's only 3 or 4 days for the whole thing. At the other end of the scale, you've got big budget features where an audio post team of 30 or more might be getting through 1 or 2 minutes of film a day! Even low budget features have an audio post team of a dozen or so and would usually take 12-16 weeks. So 5 minutes of film in an hour or even a day, sounds extremely rushed to me if it's just one person (you!) in the audio post team.
So, my hat is off to you guys that do this well, especially if you do it well with newbie equipment (like I got - zoom H6 + cheap condenser hyper pencil + audacity + sony vegas). I'm still trying to figure out how to normalize levels for proper loudness for broadcast TV without any test equipment, and that's an entire subject in itself.
There are many audio post tools available and to be honest quite a few of them you can do without, at a pinch as many of them are to aid workflow and speed up the process. This very important commercially though because clients always want miracles and what's more, they usually want them within a completely infeasible time scale. Some of the audio post tools are essentials though, loudness normalisation is one of them. There's no way to meet modern TV delivery specs without the right tools and more than a fair amount of knowledge. As you're discovering, it can be quite complex. If you've got any questions in this area, feel free to ask. Ah, you have got one:
doing a test on youtube, I found that the perceived loudness is wayyy too low. Not sure if I should worry about that, as youtube isn't the primary arena this is for. But I may have too much dynamic range for other formats.
Youtube is particularly problematic because there are no loudness specs and you find material of all loudness categories, from theatrical mixes (which are the quietest), all the way up to commercial CDs (which are the loudest) and the difference between these two extremes is vast; extremely roughly, about 30dB difference! Popular music CDs have very little dynamic range, whereas theatrical films tend to have a very large dynamic range. DVDs are closer to film than to CD and so is TV, although TV is generally a little louder than DVDs. What you put on Youtube depends on what you expect the majority of your audience to be watching on, for example, laptops and mobile devices need high levels and small dynamic ranges. In general, a theatrical feature usually ends up with at least 3 different mixes; One for cinema, one for DVD and one for TV. TV has very specific loudness specs, which include an integrated loudness (-24LKFS or -23LUFS in Europe), peak levels which are variable from broadcaster to broadcaster (generally -2dBTP and -1dBTP in Europe) and an LRA (loudness range) of no more than 20LUFS but this can also vary from broadcaster to broadcaster. Theatrical films do not have specific loudness specs as such, loudness is controlled by calibrating cinema sized dubbing theatres (mix stages) with the same calibration as the cinemas themselves, so the mix will sound the same volume in the cinema as it did on the mix stage. The problem here is that it's virtually impossible to calibrate a small room for theatrical mixing, so mixing in a small room, even one where calibration has been attempted, you never quite know how loud it will play in the cinema and therefore if some of your quiet details will be way too loud or inaudible.
You guys are either brilliant and patient, cyborgs or other artificial lifeforms, or absolutely insane. Maybe all of the above.
I would say mainly insane cyborgs with a bit of the others factors mixed in!
There is one small section with some wind noise that I can't seem to alleviate. Utterly no idea how to handle that ...
You'd think there's be a simple plugin to add & subtract harmonics on voice, either high, low, or in a range. But no idea where I'd find such a beast.
Wind noise is usually a serious problem because it doesn't usually exist within a frequency band but across the entire spectrum. This means that a fair bit of the wind noise usually "lives" in the same frequencies as the dialogue you want to keep and so there is no way to differentiate what you want to keep from what you want to remove. It's therefore usually impossible to remove all the wind noise and depending on it's severity, it's often impossible to remove enough of it for the dialogue to be usable, in which case ADR is the only alternative (if there aren't alt takes without wind noise). Using HP, LP and notch filters, along with say the CEDAR equipment Alcove mentioned, plus various other noise reduction plugins, it's usually possible to obtain usable dialogue, unless the wind noise damage is extreme. However, apart from the standard EQ filters, the good noise reduction plugins are very expensive and even then it's not just a case of hitting a noise reduction button! There are all kinds of parameters which need adjusting and may need constant re-adjusting (using automation) and exactly which plugins you use and in what order varies depending on the exact nature of the problem (all wind noise is not the same). So while a newbie with little high quality equipment/plugins or experience can often improve dialogue a fair bit, it's usually not enough to make it usable. There's no way around this and is one of the reasons why commercial audio post facilities and professionals still exist!
On the point of removing harmonics from dialogue. Are you still referring to wind noise? If so, then this isn't really applicable because noise by definition doesn't really have any harmonics but entire bands of frequencies. If you are referring to a hum though (you mentioned a fridge), this isn't noise per se because hums are specific frequencies rather than bands of frequencies. There are specific audio post tools to deal with hums but actually this is one of those areas where you can sometimes deal with the problem reasonably effectively without specialist tools, here's how:
Set up an EQ with a high (narrow) Q value and say 8dB (or so) of gain. Now sweep slowly through the frequency spectrum, starting at about 40Hz. You will notice a particularly big jump in volume at a particular point during the sweep. Set your EQ to the highest (narrowest) Q setting available and sweep more finely in the area where you noticed the volume jump, until you've narrowed it down to just a small number (preferably 1) of frequencies. If it were an earth loop hum for example, that frequency would be exactly 60Hz (in the US). Centre your EQ peak on that frequency (60Hz in our example) and turn the gain in the opposite direction, to it's maximum cut value. You've now identified and notched out the fundamental frequency of the hum. Now setup another EQ as before and continue on up through the spectrum, paying particular attention to where the harmonics of the fundamental you've already found should be. In our example (an earth loop hum) the first harmonic would be at 120Hz, then 180Hz, etc., creating a notch filter for each harmonic (or fundamental of a new hum) that you come across. Make sure you play a large portion of the take through each time you are trying to locate fundamentals and harmonics, as you don't want to start notching out the fundamentals or harmonics of the dialogue itself. The fundamentals and harmonics of the dialogue will usually change throughout a phrase, whereas hums will not, they will stay the same for the duration of the hum.
Professional tools for doing this job will setup the harmonic notches automatically and the notches themselves can be set much narrower (higher Q) than most EQs bundled with software and are therefore able to remove the offending frequencies more precisely (and therefore more transparently) than most basic EQ plugins. Be aware though that many mechanical hums may contain broadband noise in addition to the hum and/or may actually be comprised of several different hums, in which case it may be impractical to remove all the different fundamentals and their harmonics. In either case, the hum removal method I've described above may not improve the dialogue to the point of usability. Air con nearly always falls into this category for example and fridge hums often do. In these cases, notch filters may form just the first stage of the ultimate solution, if there is one!
One of the discouraging things about audio post is that the more you learn about it and the more effort you apply to it the more complex and difficult it appears to become. Don't let this put you off though! Remember, most no and very micro budget filmmakers get discouraged straight away or just aren't interested or can't be bothered in the first place, so you're already ahead of much of the competition to start with! Hums are actually a good example of this, quite a high proportion of takes contain some sort of hum but unless it's a really obvious hum most no budget filmmakers don't consciously notice them and so don't remove them. Try the method I've described above on a number of different sections of dialogue which you think currently sound OK. I'm sure you'll find some hums in some of those sections of dialogue which you didn't realise were there and you maybe surprised by the improvement from removing them.
I hope this helps?
G
