Cedar Audio Retouch: Like Audio Cough Syrup . . . Or Explosion Syrup
We take a closer look at some modules used within the Cedar Audio Cambridge System at the 125th AES Convention, and in this video, we focus on Retouch 4. Cedar Audio not only provides audio retouching, but forensics teams benefit from Cedar's systems as well.
Retouch can remove individual sounds from a mix, whether it be an unwanted background noise that bled into your condenser mic or loudness obscuring a wire-tapped conversation.
GORDON REID: I'm Gordon Reid. I'm the president of Cedar Audio Limited, a company based in Cambridge in the UK, and what I'm going to show you very, very quickly is the Cedar Cambridge System, which is our flagship product. It's the one on which we developed most of our new products and new algorithms and bring new ideas to the marketplace.
What we've got here is a Cedar Cambridge System with a wide variety of modules running on it. Some of these are concerned specifically with audio restoration for the traditional pro audio markets, for CD mastering, for DVD mastering, and so forth. Some of them are little more esoteric; for example, we've got a full set of filtering systems which are used for forensic, that's police, and security work which will help to dig speech out of background noise and make things intelligible hopefully to stop bad things from happening. Then we've got other tools which are specifically designed for use in film production, soundtrack cleaning up and so forth, and the one I'm specifically going to show you now is a product called Retouch, which was designed for removing or manipulating individual sounds within an audio signal, not necessarily a musical signal, not necessarily just a speech signal but any signal at all.
So, for example, in the music industry, let's say you have that perfect take in a live concert and somebody coughs. Well, you can't go back and recreate the evening. Wouldn't it be great if you could just remove the cough without affecting the desired audio at all. Or for example, if we were talking in the security industry, you may have the situation where somebody says something just as something very loud happens, hopefully not a gunshot or an explosion, let's be kind and say just a door slamming, and just at that moment somebody said something quite quietly and you want to remove the very loud sound so that you can uncover the quiet one. So, this technology allows you to manipulate individual specific sounds.
So, if we turn to the screen, what we have here looks very much like a piece of you running a bunch of software modules, and at the most basic level that's what it is, although it's a very specific PC, server-grade PC with sonme other hardware, and what I've got here is an audio signal of some sort, and I'm gona invoke the Retouch process by clicking on this icon, and what we get is what's called a spectrogram, and along the X-axis here you're looking at time. Here, you're looking at frequency going up the Y-axis, and then the various colors represent the loudness of the signal at any given moment in time at any given frequency. And we can display this in numerous different ways to identify most clearly the signal content. We can zoom in on particular things like so, and there are other display methods and so on that we can use to make the representation carry as much information as possible so that we can identify the sounds and manipulate them. Having done that, we can then do interesting things with the signal.
Let's say this sound here is a sort of an unpleasant fluttering high-frequency sound of some sort. I'm not sure which because I haven’t listened to this particular signal, but let's say that we want to remove this part of it and leave the background untouched. You can't do that with a conventional filtering system, but with Retouch what we can do is zoom in even further and we can say this bit of signal here for example, in the most very basic level I want to take that and get rid of those frequency components, so I just need to adjust the system a little. And what I've done there is I've marked the actual bit we want to change. I'm identifying some signal with which I want to remodel that part of the audio and then I just hit the retouch button and you can see that those sounds have been completely eliminated without effecting anything else either in time or frequency around them, and we can use that to, as I said, we can remove pops, door slams, gunshots.
We can remove individual notes from the mix. Let's say that the orchestra is playing and the clarinet hits B-flat when everybody else is playing B, well you can see those harmonics completely clearly and you can either remove them individually or even, if you're very, very careful and you've got some time to do it, you can take those and you can shift them to the right frequency and actually move the note into the right position harmonic by harmonic, and some of our classical mix engineers have been using this system to do that for some years.




Post new comment