Loudness Standards

On my video project, there was intention of uploading to YouTube. YouTube has a normalization feature on the site so that when playback moves from one video to another, there isn’t a massive change in volume, leaving the user to change levels when listening.
When YouTube applied this feature, a guideline was left. Without going into detail, we can summarise things by saying that everything must be under -13LUFS (Loudness Units Full Scale) or they will be compressed, but if the video is too quiet, they will be brought up to blend with other videos.
Keeping this in mind, it can be very easy to stay under this ‘limit’ but mixing and mastering to it can be an experience if you are new to this side of things, which I am. So, when I was approaching this task, I did what made sense to me. I found a LUFS meter, applied it to my master track and set it up accordingly;

What we are looking at is a graph that just tells us the perceptive hearing levels (yellow line). the true peak levels (blue line), the attenuated peak levels (green line) and the perceptive peak levels (blue shade). On the right is a drop down menu box that allows us to set the target LUFS limit, I’ve set it to -13LUFS for the sake of this picture but I set it to -14LUFS when bouncing out for the sake of head room with peaks that may occur between sample readings, (as I am not sure of how this demo version of the meter counts for that.) underneath that box is what looks like a fader, this is a built-in gain meter that allows us to alter it as we need, according to our project. The ‘adjust to’ button above takes in the sampled information and automatically adjusts it accordingly to the work but I personally didn’t use this feature and wouldn’t really recommend it as it is too easy to make mistakes and miss them later on.

Before bouncing out my video project, I just made sure that the louder moments stayed under -13LUFS on the yellow line, because that’s what YouTube cares about when it comes to LUFS standardisation and that was that.

Calibrating Studio Monitors

Calibrating your audio work space is a very important step into the world of mixing and mastering, it allows us to correctly hear what our speakers are putting out and it allows our speakers to push information out correctly, so that we can be aware of what we are and more importantly aren’t hearing.

There are a few things you will need to start this process;
1. SPL (Sound Pressure Level) Meter
2. Studio Monitors (Speakers)
3. Measuring Tape
4. Tone Generator (Pink Noise)

Before we get started on this we need to adjust our monitor and listening positions.
Using our measuring tape, we need to make sure that we are seated in an equilateral triangle from our ears, to the left and right speakers.
There is also some great info on that here:
http://ehomerecordingstudio.com/studio-monitor-positioning/

Now, the SPL Meter, this should be positioned to where our head would be, at ear level.
Whilst being careful with our hearing protection we want to start generating some pink noise for our meter to pick up on, to know what we want to read, and how to get that, it is a little situational, so I’m going to leave another link. This one below may even be better than what I’m typing here.
https://www.youtube.com/watch?v=idGvZnSnPhs

I WAS going to get into detail about volumes for calibration but the video link above talks about it pretty openly. To summarise, we need to make sure that the SPL meter shows the same results for each speaker and that we don’t move it whilst calibrating. Depending on the room we are working with we could be aiming for different decibel(dB)/SPL levels. In a small to mid-sized room we should be aiming for around 70-75dB/SPL from each speaker, remember, this is loud pink noise for what should be an extended period of time, please use hearing protection, preferably ear muffs, but over-ear headphones should do fine.

Elastic Audio With Dialogue

Elastic audio allows you to time-stretch samples within a Pro Tools session, it is a feature that is built into the DAW. For my sound replacement project, because it is in another language, the overdub work wasn’t needed to be perfectly lip-synced because that is just impossible, but, with elastic audio, we can me it sound closer to what we are seeing, just to create an easier listening experience.
The reason we use elastic audio instead of just simple time-stretching is due to a few or more advantages, elastic audio can  work in multiple reading algorithms at once and allow us to select what process it uses before we bounce the track, what this means is, to put it simply, we can either time-stretch with the famous ‘Alvin and The Chipmunks’ effect where squishing the sample (or stretching out) alters the length of the waveforms, ultimately affecting the pitch, or, we can have Pro Tools only stretch out the sample whilst attempting to reproduce the waveforms that would be lost, this way it can sound as if the sound is just made for a longer period. Be wary though, depending on the quality of the sample, using this method can create many digital artifacts appear in the sound, and if undesired, can change your work completely.

Anyway, let’s get on with how we use it. Below is a couple images showing that there is a tab to the left of each audio track that has a small triangle shape with a small line coming out of the top, almost like an oil can. If we click on that, we get a drop down menu. For the work I am trying to make, we click Polyphonic mode, then, above that is another tab that should say Waveform, we click for the drop down menu and get two new options, ‘Analysis’ and ‘Warp’.

In the image below we can see that once selecting ‘Analysis’ we get some vertical lines on our track, these are markers, showing where Pro Tools has recognized prominent transients within the audio information. In this mode, with our grab tool we can make alterations to the makers, remove markers and create markers. These markers are important to be aware of when we get into our next mode.

Below we see that I am now selecting ‘Warp’ mode. This mode allows us to start time-stretching our samples according to the ‘Polyphonic’ sampling mode and the markers we made with the ‘Analysis’ mode tool.

Below we have two images, one after placing markers, two of which are barrier markers and one is an alteration marker. I included the track above the one we are editing for the sake of comparison, we are looking at the differences in the purple track over the exact same section of sound.

With the grab tool, I created many markers by simply double clicking and adjusting their position in ‘Analysis’ mode, then in ‘Warp’ mode, I grab a marker and move it to alter the length of the clip.

To go into further detail on what I meant when I said ‘barrier markers’ and an ‘alteration marker’, what we need to do the take further control of what we are doing here is; firstly imagine that we have one sound that we want to edit, in that sound there are two syllables, an “Ah” and an “Oh”, if we imagine that the “Ah” is too long and the “Oh” is half a second too short, we would create a marker on the first transient of the “Ah” sound, another marker at the end of the sound, and a final marker between the change of syllables (from “Ah” to “Oh”). Now that the ‘barrier’ markers are set and we have our middle marker, we can move the point at which the syllables change in time, by simply grabbing the marker to moving it from left to right or vice versa, according to our desired result.

Correct Use of Contracts and Agreements

In this industry, a lot of us will be running into projects where we end up talking about agreements on paper, just so no one gets hurt in the long run and such.
A lot of this can be very daunting, especially for someone like me that hasn’t had much experience with signing papers.

I was looking up some details just to make sure I didn’t post anything incorrect and I came across this website called artslaw.com.au.

What this site lists is a boatload of information on making agreements within different creative industries and works and a series of templates for specific contracts that you can buy to use for your own work.

This last trimester I worked on two projects, one for sound replacement and one for game composition. In the sound replacement, I was the producing engineer and I brought on two actors. In the game, I was approached by the lead designer and asked to write one song to be listed in among the works by a peer of mine that was leading the audio team. Neither of these projects were planned for profit or payments to anyone, just some people helping each other out.

In the sound replacement project, what I needed was a ‘Collaboration Agreement’. These kinds of agreements are made to ensure that all members are informed on what is needed, by whom and for whom, and to try to ensure that the project isn’t dropped without reason, leaving people in the dark with an end result of providing nothing without some sort of recognition. On the subject of recognition, these agreements are also likely to address how people will be credited and how their efforts will be implemented, and if not, what would happen in that situation.

In the game project, what would be needed is either a ‘Game Development Services Agreement’ or a ‘Music Commission Agreement For Games’, depending on the situation and circumstances, for my experience personally, it would be the Game Development Services Agreement, this is because I was approached and asked to create some work with the intent of a collaborative effort. Although, I was only asked to create one piece of music, this would typically be decided upon using the Commission Agreement as it is more aligned with a proper contract for providing work but, the big difference in this project is that I was asked as a peer, not as an employee and the big key here; I was asked to create a track set with specific parameters, according to the game’s design.
If I was asked to create a song so that the designers could simply place it in there as a song, the story would be different but, because I was asked to create the work as a background setting of music, this is what makes it line up with the Game Development Services agreement.

Reflecting On Stalker Project

In the original piece there is a constant flow of water making sounds that we cannot see, I didn’t want to place this in the project for a couple reasons, one was that it would be just this constant noise taking up space and I initially worried about it cluttering up the listening experience for the viewer, the second was that I didn’t want to take samples from online nor was there a way for me to record this sound on my own, so I decided it wasn’t needed.
I feel now that it was a pretty poor choice to do so. I could have easily used a freely available sample of a waterfall or drainpipe flowing and that would have been that, but now the space feels empty, boring, and the overall scene was incredibly difficult to fill out the sounds without it. Not only that but, for the aesthetic implementation, it felt like 3 people were sitting around underground, I didn’t capture the tension in the original, nor did I bring in anything new. I feel as if the constant water flow sound would have made that for me.

I also feel that the reverb on the second role ‘Stalker’ was badly implemented, I do still think that I went the right way about using it in comparison to ‘Writer’ and his reverb sounds but I personally just did a bad job at mixing it in correctly. I should have turned down the aux send controlling the values of the reverb a bit and made a cleaner frequency roll-off.
Other than that, I feel that the first clip we hear from ‘Writer’ came out better than I was expecting, I think that this is mostly due to the microphone used and it’s natural frequency response. It sounds more ‘in place’ than it would have if I used a dry signal from say, the C414 microphone I planned to use in the beginning.

Once uploaded, I will embed the video here.

When comparing the sound of my sound replacement to the sound replacement on other videos (on YouTube), I think I had made a fair approach but I definitely hear the differences in quality, it might be due to deadlines with time constraints but I’m also aware of my troubles that I had during recording sessions and my constant feeling of content with my own work. I tend to think that I’ve done a great job until it comes time to the end of the work and I compare it to something that wasn’t my reference material.
I feel that in this project, I used the best of what I got to the best of my abilities but if I made the time to do so, I should have set up a much better recording session and made sure to record better samples, rather than to record to what’s on screen and accepting it as good enough. It’s not just an ability to record a clear recording, but I need to develop the ability to make good calls on what I have just made myself, I can only do this by making more content that I’m unhappy with really, but that’s what we all do.

Removing Noise Floor

For my Stalker sound replacement project, I mentioned in an earlier blog post that my dialogue recordings didn’t go as I had hoped for. With the recordings that I did end up getting, there was quite a lot of ‘noise’ coming through to a noticeable point, I’m going to explain what I did and what the product I used says it does.
Unfortunately Izotope won’t share their secrets so until I can work out more, this is what we will be working with.

Izotope’s RX Dialogue De-noise is both a stand-alone application and  a plugin designed to recognise the differences between noise and sound. It does this by analysing the loudest frequencies in a given moment and compares them to the loudest frequencies in another, it attempts to figure out what is being ‘un-used’ or is undesired and allows us to manipulate from there.
As you can see in the picture above, there are three tabs at the top; ‘Manual’, ‘Auto’ and ‘Learn’.
Manual mode lets us play a clip from the track the plugin is assigned to, we then click Learn and the plugin takes a snapshot of that moment’s information, from there we mess about with the threshold and reduction amount, according to the sound we recieve. Auto mode, however, attempts to analyse the frequency information on its own, whilst attenuating the ‘noise’ over a constant time-lapse.

In an attempt to make things  little easier to understand, I have some more pictures showing how I used the program in my latest project.

As you can see at the top of this image, I have a channel named ‘WriterDialogue’ and on it is the plugin RX De-noise,

To the right of that are two short clips of dialogue. I simply played one of these clips on loop at a time (so that the program wouldn’t define the silence in between the clips as noise floor), set the plugin to ‘manual mode’ and as the clip was playing, I clicked ‘learn’. I then made changes accordingly with the threshold and reduction metres. Examples below;

Stalker – General Update 2

Just due to time constraints and other mishaps in people’s personal lives, some of the roles have been moved around and some of the equipment used was changed (mostly for worse but easily acquired).

The roles are as follows:
Recording and Mixing Engineer: Christopher Gleeson
Writer’s Dialogue Performer: Cameron B Clayton
Stalker’s Dialogue Performer: Jonathon Bonham

As for equipment, the only changes are with the dialogue microphones and locations for both Stalker and Writer. Instead of using  a tuned studio with 2x AKG C414 microphonesm we ended up needing to use what was at our disposal, for Stalker we used an RODE NT-USB microphone, I was pretty unhappy with both the recording due to the sound we got and because I couldn’t be there for the recording. I became inquisitive upon receiving the assets for Stalker and asked how the recording went, the performer told me that the microphone was placed next to a wall with the back facing the wall about 10cm’s away from it and about 2 metre’s away from a window to the side, the wall reflecting the sound from behind the microphone is fairly obvious on the dry signal so I’ve tried my best to mask it without ruining the mix of the scene on screen.

As for the recording of Writer’s part, well, this is even worse but the location was clearly thought out as it is fairly silent and no reflections got its way into the recording. The microphone was your standard iPhone microphone. Though I’m not sure which model, it came out better than the other recording just due to the preperation before-hand and because of this comparison to Stalker’s part, I can’t complain too much. It’s just unfortunate that all members involved had some interuptions that couldn’t be cast aside.
A finished product that followed through to the end is what I’m happy about though, I just wonder how I can learn from unexpected mishaps.

Strat And Slash Game Songs

I am just posting to say I have uploaded the two ‘Strat And Slash’ songs up to Soundcloud.
Check out my previous blog on them here!

I’m not sure if I will get to share the game around but if I can, I will post it here in the future. If I can’t, I’ll probably just remove this paragraph.

Case Study – ‘When No One Walks With You’ by Vildhjarta. 2011.

https://youtu.be/akvZW_G6RgA

In the broad genre of ‘metal’, there are some similar expectations that are needed across the varying ocean of hard hitting sounds. Though, many of these techniques that go into the pre- and post-production stages go unnoticed by different producers and this is where the majority of genre differences come into play but for now, I want to put some light on what is almost absolutely required for the majority of these styles.
The guitars are practically never DI’d, one or two dynamic microphones up against the amplifier cabinet is what will give the instrumentalists the tone they expect and what the listeners want.
Multiple guitar tracks playing the same rhythm parts are expected with hard left and right panning, this is to create tension in dynamic, loud surrounding guitars are part of the metal aesthetic.
Heavy amounts of equalisation on every track is a must. The slightest movement of a microphone can change everything when working with loud sound sources and clear end results, not only does microphone placement matter on its own but in metal, it plays an important role because the listeners expect the perfect, clear blend of aggressive guitar distortion (yes, clear distortion is what I said) and timbre.
Metal instrumentation can become very technical and demanding so filter automation is key, especially when moving into new ‘scenes’ or riffs of a composition. Automation also includes time-alignment of percussion, strings and vocals. ‘Metal’ is such a broad term in the age of internet discovery as experimentation is a constant, so automation means everything.

In 2009, Swedish band, Vildhjarta would take inspirations from bands like Meshuggah, using down-tuned, extended range guitars with heavy use of polyrythms and single note emphasis to release their first EP ‘Omnislash’. This would excite fans all over the world as it took a distinct sound to new places for a specific community to find entertainment and inspiration, along with this release and all of Vildhjarta’s online posts was left with the term ‘Thall’. The meaningless word was nothing but a slurred rendition of an enemy in the video game ‘World Of Warcraft’ that the band members found funny, this term would excite fans as they had no idea what it meant or if it was the title of Vildhjarta’s upcoming album. Through this excitement, fans would spread the word all over the internet, making up different definitions of the word, ending up using it for a general description of the sound most affilliated with ‘Djent’, the high-tone, scratchy palm-muted sounds of an electric guitar technique that would differ from the common, more expected sound but ‘Thall’ was to describe that mixed with long reverb effected guitars and a dynamic use of high and low notes. With all of this fan hype, Vildhjarta would release their first album ‘Måsstaden’ in 2011 which took the young, online community in new directions, bands would eventually take this specific sound and merge with other progressive metal styles like ‘Protest The Hero’ and the creation of most notable bands like ‘Periphery’ and ‘Animals As Leaders’ would spark into the genre.

Below is an upload of Vildjharta’s song When No One Walks With You, with a simple analysis contained within Variations Audio Timeliner.
https://drive.google.com/open?id=0B-nAPcFDeF1QeVdVRkJjelZKQWM

Stalker – Dialogue Specific Update

This is a ‘What I did.’ blog for what happened when I lost my voice actors. It might not be the best thing but at least some insight might make its way in.

So, I’ve been working on a sound replacement project where I am taking a scene from an older Russian film and giving it an English overdub whilst also making the sound effects sound ‘cleaner’ but keeping it relatable to the quality of the film.
The major part on this piece is dialogue from one character with one line in order to have this character’s response, so this is absolute key.
From day one, my goal was to get a good voice actor on board, for unlucky project managers, learn from me.

I originally had put out the project idea onto collaboration pages on Facebook and some forums around the web. I got a lot of responses and was feeling confident that everything would fall into place, unfortunately I had so many people on board that I couldn’t nail it down to two actors and kind of kept everyone in the know.
This ended up being a pretty bad move on my part because lots of actors lost interest and when my main bets were not committed, I no longer had the many actors at my disposal, I was left with three or four people who were still interested but time was running out, fast. The first actor that I got into recording sessions gave me about a minute (for a five minute scene) of demo material whilst we were still fleshing things out, we decided to get back together at a later date once we were both completely prepared, as time went on I got ready and back in contact with this actor, unfortunately he had made plans for other work and had to cancel, even more unfortunate, he took the second role with him.
By this time, one of my two backup plans had lost interest and I had lost contact, so for the last contact that was still happy to help, we had practiced the lines and made preparations, when we got into the studio for recording though, this was another story completely.

We had everything ready and he said he was good to go, I got the ‘record-arm enabled’ going and my last backup plan got the shivers, he told me it was nerve racking and needed to quickly go to the restroom. Ten minutes later I got a text message from him saying that he “Wasn’t up to it.” and I just didn’t hear back from him.

With only days left I was more or less stranded, stuck between a bad place and… well, another bad place, things were bad.
I just went back to that first FaceBook page and posted a distress call. Luckily two more actors showed their interest, one confirming his experience in the field to me. We are set to practice and record in the next few days, so here’s hoping.

The lesson here is, contracts are absolute, words are not.
Wish me luck!