Re: [DCP-o-matic] dcp sound worries

Hello Nehtarios Clipping the audio can cause loud distortion, which could potentially damage loudspeakers and be very unpleasant for anyone in the auditorium. To amplify Carl’s point a little – if you will excuse the play on words – I’d like to walk through some details of his very nuanced first sentence. The physics of speakers requires the parts which move the air to be perfectly aligned within the magnetic channels that they float in. There are limits to the amount of fluctuation. When the limits are exceeded they cause damage to the parts in addition to creating a distorted sound field. (Translated: when the excursion of the parts hit their boundaries, it can tear the fine wires and light paper to shreds which fill the cavity and cause burning ooze on the metal magnet that turns to baked gunk that causes more problems.) Over time, a speaker's ability to fluctuate is reduced naturally, of course, but abusing the limits cause more dramatic problems much earlier. So, from the speaker component view, it is our responsibility as DCP creators to not allow bad decisions by the director to get into the field. The physics of the sound in the auditorium turning into audio in the human head takes the complication further. To begin, there are amazingly small hair follicles in the most inner part of the ears which can only tolerate so much movement and in humans (and other mammals), when over stressed they get damaged and eventually break. This can happen all at once or be a cumulative problem that manifests over time. So, we have the responsibility to the audience – who we presume are currently mostly mammals – to not cause any harm. (In all other vertebrates these hairs can grow back, so if your audience is amphibians and sharks, experiment as you wish.) Unfortunately, many cinemas use their components beyond the point of their peak efficiency. This will cause audio that sounds great on a set of brilliant new and well calibrated mix system to sound soft and un-dynamic on a beat up and uncalibrated system. These distortions annoy, and if the mix is loud, it annoys even more. Equally unfortunate, this is the same for humans as well. The ability that the majority of teenagers have to hear better and more, diminishes over time. And like any good distribution curve, there are many outliers, people who hear annoying sounds when others in their category hear as OK. As people age, or they return from military service with more damage in one ear than another, or listen to their ear pods at high levels for long periods of time, they join the group of outliers who hear different frequencies as distortion (and annoyance) at lower and lower levels. Unfortunately, we, as the final link in the professional chain, have to accommodate to a wide audience in a way that will present an appealing presentation to the widest audience. There is very little science for determining the exact what and who for these levels. Early numbers that governments used for work environments don’t correlate well for many reasons, first of which is that they were created during the dark ages when the determining factor was whether a person could still hear talking across a table – ignoring all other sounds and frequencies – as the determining factor of whether the sound level of years in the factory was unacceptable. On the other hand, there are no studies that say that a couple of hours of an intermittently loud movie will cause hearing damage. But we, as professionals, must not push the established envelop for many reasons, one which is that will invite authorities who are investigating audio loudness in dance clubs and sports arenas to also inappropriately include cinemas in the same studies and laws. Finally, there is a perceptual component in the human hearing system that ties this all together. We hear some frequencies better than others, and we hear these frequencies in different proportions at different levels. The theory goes that if the mix is loud and seems well balanced, when the level is dropped down it will seem out of balance. In practice, that happens with the result that the dialog will seem muffled and difficult to understand. See: Equal-loudness contour - Wikipedia <https://en.wikipedia.org/wiki/Equal-loudness_contour> The incorrect analysis of that situation is that the effects and music are too loud. Thus, we get a vicious circle going. When it is too loud the audience will ask for the level to be turned down, which will make the effects sound louder and the audience will ask for it to be turned down even more. With that, the artist’s mix is destroyed and the audience is thinking “I get better results at home.” Yet again, as DCP creators who hand the product to an ever more often untrained person who only knows ‘ingest’ and ‘play’, we should know and operate within the standards and recommended practices…and encourage others – mixers, directors and producers – to know and operate likewise. Directors who have heard their dialog being spoken for months may not be aware that others who hear it for the first time have no clue what is being said. (Will this be called the Christopher Nolan Effect?) Ask a movie mixer and they’ll tell you that the mixes are built around the dialog. But dialog mixers are no longer trained to look at a VU meter and keep things down in the -14 to -18 range – they only know to keep peaks out of the red. This won’t give a mix that reasonable effects and music can be built around. Mixers who are in the joy of creating on a wonderfully tuned audio system aren’t allowed to accommodate for anything other than the brilliance they hear. There are no absolutes that can be given about sound levels – each movie and presentation are different. It is like looking at a picture and asking if it is too bright or dark. Although there is an absolute that each projector should be emitting for pure white (48 candela per square meter), you would never have your movie only hitting close to that peak value. Logically, there is a balance of light and dark appropriate to the material. Fun stuff: This Artist Turns Every Frame in a Movie into a Single Photograph - Creators <https://creators.vice.com/en_us/article/jason-shulman-every-movie-frame-single-photograph> Visualizing A Movie's Color Footprint - Creators <https://creators.vice.com/en_us/article/visualizing-a-movies-color-footprint> There is no time unit on the picture that you sent, so it is difficult to judge. But I bet that if you watch some movies in a theater with a meter you will find that there is a lot more dynamics – the range from soft to loud – than the 10,14 LU (Loudness Units) window that your mix has. Also, it looks like you have 3 channels chosen but only two are represented. And while left and right might often diverge in a mix, these seem oddly different. Is your dialog track not separated from one of the Left or Right perhaps? LUFS as a measurement needs to be evolved for movies – there needs to be a 10 minute floating window for long duration materials and LUFS was created for short duration work) and so is not a perfect guide. Read this to get some data that might answer your question though. How Loud is a Movie? – a New Measurement Procedure By Ioan Allen … <http://isdcf.com/papers/ISDCF-Doc11-MovieLoudnessMeasurement20160315.pdf> Mr. Allen writes in terms of LEQm instead of LUFS, but the numbers can be generally used equivalently for learning purposes. Good luck to us all. C J Flynn Cinema Test Tools <http://cinematesttools.com/>