Now its time to bring this project to a close. Below you can see the completed 3 tracks of ‘The Reserve Lounge Collection’, ready for academic submission before the deadline on the 8th May. Unfortunately I have recently been informed that my client John has been called away on business for 10 days so unfortunately I could not meet with him face to face. However I did email the tracks to his personal email address so he could listen to them when he had some free time on his trip. After a few days I received a reply, filled with positive comments and excitement. It seemed John was delighted with the final product’s professional sound and careful production, and was excited to play it to his customers when he got back into the country. He also asked me to drop into the club when I had finished my university course to discuss some possible business for future music production for him, which I delightedly accepted. My full personal reflection on this and the project as a whole can be seen in the ‘Personal Reflection’ tab. Anyway, here are the final tracks:
Year: 2014
Product Testing
Now that the main bulk of the product is finished I felt I needed to test its effectiveness based on my research into creating the songs. My product testing is be based on previous research into emotional responses to music as you can see in an earlier blog post. The ‘circumplex model of emotion is a theory that splits recognisable human emotions into 4 sections and how they occur, which formed the main basis of my product testing. I began by inviting groups of friends to my flat and asking them to each fill in a brief survey in which they would mark their particular emotional states out of 10, 0 being not feeling that emotion at all, 10 representing feeling that emotion to the highest degree. The emotions the subjects rated are the ones that are featured on the circumplex model of emotion. By doing this I was able to fit each individual into the models 4 emotional states: High arousal/positive impact, High arousal/negative impact, low arousal/positive impact and low arousal/negative impact. I then asked the groups to sit through and listen closely to my product in silence so each of their emotional states would only be affected by the music. Afterwards I would simply ask them to fill out the same survey. By doing this I was able to analyse the emotional state of the test subjects before and after listening to my product. If the majority of the subjects showed emotional change towards the positive impact sections of the circumplex model like my product was intended to do, then I could deem the product a success. Below is a screenshot of the survey the subjects were asked to complete. You can find the full version as a link in my supporting materials section of this blog.
For my research I collected 20 sets of results from 20 different friends and members of family to ensure I encompassed a wide range of material and was able to make solid judgement on my final results. I then plotted my results on a radar graph showing the survey before listening and the survey after in different colours so it was easy to differentiate between the two. As you can see by the examples below, the blue shows the survey results before listening to the product and the red after.
You can see an obvious shift towards the positive emotional sections of the circumplex model from blue to red. I found that out of 20 test subjects, 17 of them showed a shift into a more positive emotional state where as the remaining 3 either remained in a similar emotional state or moved towards the negative emotional sections of the model. This could be down to a variety of factors from entering the test in a particularly bad mood, to personally disliking the style of music they were listening to. However as 85% of test subjects showed a positive emotional response to my product, I would like to deem my research and product as successfully. Based on this, I feel that my work is now ready to be submitted ready for my academic deadline and my clients deadline as well.
Mastering in Practice
Now comes the time for the finishing touches, the glue that binds the song together, the process of making your music sound polished and professional. Mastering is the final process of the practical side of my audio project and as discussed in my previous post, must undergo careful consideration. My choice was to master my songs with dynamics in mind rather than loudness as the Reserve Lounges music system, Kaleidovision, features automatic gain control, in which all songs are played back at a similar consistent volume. However at the same time I must bare in mind that these songs could end up being used in a commercial environment, meaning that although they must be dynamic, they cant be too quiet either. In this post I shall explain my personal mastering process and the techniques I have employed to help achieve a dynamic and professional sounding product. Below is the original final mix of my 3rd song without any additional processing, basically the song at the stage where its ready to be mastered and finalised:
So at first listen you can tell there were some equalisation issues, it lacks top end and the bass isn’t as deep and defined as it is intended to be for music of this style. To combat this issue I began by using the Waves Q-Clone EQ plug-in, the reason for this choice is that it has a fantastic visual representation of the frequency spectrum and what you are doing to edit this. I began by adding a shelf at 15 kHz by a few decibels to try and bring out the top end a little more, however any more than 5 decibels it started to make the track sound thin and weak, I made the concious decision to try to avoid that happening and moved onto the bass. For this I boosted around 70 Hz to 140 Hz in an attempt to enhance the 909 kick and the bass frequencies that filled out the track. Although this did help i still wasn’t happy with how the bass line was coming through, it still sounded too thin and powerless. Based on this I decided to utilise the MaxxBass plug-in by Waves which works by adding artificial bass frequencies to your current bass tone to thicken it up and make it sound more powerful. The ‘medium’ pre-set worked perfectly for this as any more would have caused the bass to sound too prominent when played on club speakers. As a final stage of the ‘fixing section’ of the mastering process, I decided to add some tape emulation using the Krammer Tape plug-in from Waves. The subtle use of tape emulation helped bring some warmth and low end saturation to my track, a desirable trait in typical Down-Tempo music. Most processes so far remain as subtle adjustments that only a trained musical ear can register. However it’s lots of little changes that make a big difference.
The next stage of my mastering chain was the enhancement process, the stage where the track is made to sound polished and professional. To achieve this I began by using a stereo enhancement plug-in by Waves which allowed me to widen the stereo field in which the track sat. By adding some subtle stereo widening, the track was able to reach the far sides of the left and right stereo field, making the track sound bigger and more lively. This was the first technique that really made a big difference to the sound of the track. Next came this use of multi-band compression, which allowed me to lightly compress each frequency band individually. Bass and treble frequencies tend to stand out more in any track and therefore tend to require slightly more compression, whereas too much compression on the mid-range can make the track sound flat and dull, making multi-band compression a popular choice among mastering engineers. Once again I used a Waves version of a multi-band compressor, set the ratio’s at a gentle 2:1 and then adjusted the threshold to taste. It was important that I did not over-compress the track as the dynamics played an important role in effecting the songs listeners in a positive way, and therefore shouldn’t be squashed. Finally I used the L3 Multi-maximiser to adjust the level to an appropriate commercial standard in comparison to the reference material given at the beginning of the project. This was to ensure that my work wasn’t considerably quieter than other commercially released material should it be played outside of the Reserve Lounge.
After I felt the mastering stage was near completion I ran the track through the TT dynamic meter plug-in by the Pleasurize Music Foundation as a way of measuring dynamic range in music. I found that its highest DR (Dynamic Range) value was DR18 and its lowest was DR7, which by today’s standards makes for an extremely dynamic piece of music. Upon completion of the song, the dynamic range meter calculated an average DR value of 10, which is considered by many audio professionals as the ideal balance between wide dynamics and loudness. Below you can hear the final mastered and completed track:
I felt this was the best example to use as an example of mastering practice as it showed the largest difference between the original mix and the final master. A similar approach was adopted to the other two songs produced but obviously certain parameters of the plug-ins were adjusted to suit the particular song. My personal feelings on how I think the mastering stage went can be seen on the ‘Personal Reflection’ page of this blog.
Mastering Research & The Loudness War
Upon reaching the mastering stage of my audio project I came to the realisation that more consideration would have to be put into the process than first thought. Through research into mastering I found that I would have to make a choice between mastering my songs loud to compete with other commercial music in the Reserve Lounge at the sacrifice of dynamic range, or master the songs to feature more dynamic qualities yet risk customers loosing interest due to my work being perceived as quieter than other releases. To understand the reasoning behind this we must first understand exactly what mastering is and what this entails.
“Mastering is the last creative step in the audio production process, the bridge between mixing and replication (or distribution). It is the last opportunity to enhance sound or repair problems within an acoustically designed room, under an audio microscope.” (Katz, 2007. p12)
Mastering, often referred to as ‘The Black Art’ in simple terms is the process that gives recordings a polished, cohesive and professional sound, best described as “Photoshop for audio” by mastering engineer Ian Shepherd (2010). Mastering audio commonly consists of 3 different creative techniques to help create a professional sounding track: EQ, compression and limiting, when each process is used subtlety they tend to make a huge overall difference, but at the same time the overuse of just one of these can have a detrimental effect on the audio. The most commonly abused technique is compression. Compression is a technique that takes the loudest and quietest parts of a song and reduces the difference in dynamic range between them. This technique can be utilised in a varied number of different circumstances, from smoothing out the peaks and raising the quieter parts of a vocal take so that it sits consistently within the mix and remains audible, to adding extra punch to a kick drum or bass guitar or even in the mastering process to give the perception of power and punch. Like anything in the audio processing world the sensible and subtle use of this technique can have some extremely positive and rewarding effects on a final mix or master. Unfortunately this process can also warrant some negative implications if used incorrectly. The term hyper-compression is used to describe the excessive use of this tool and how it is utilised to push the waveforms of music closer to the 0dBFS (decibels to full scale) limit to make the song sound louder. Unfortunately through doing this you sacrifice dynamic range, therefore erasing the light and shade of the song. Everything tends to hit you at the same level in a wall of sound throughout the song rather than alternating between loud and quiet which in turn, can cause a fatiguing effect on the listener. The main reason for the use of hyper-compression is a desperate attempt for the artist’s track to be louder than their competitors. This all stems from the ‘Louder is better’ theory, in a way that the louder something is, the more frequencies our ears can detect within that audio, giving an improved listening experience. However loudness is a completely subjective effect, in other words, the loudness of a sound is completely dependant on the person experiencing it. Unfortunately not everyone believes this, Record labels and artists have been pressuring mastering engineers into adopting the ‘louder is better’ concept.
“Every Year CD’s seemed to be getting hotter. When a new loud benchmark was reached, somebody else figured out a way to go even louder. It was nothing less than a Loudness War.” (Milner, 2010, p. 243)
The concept of the Loudness War has a dramatic effect on the way music is mastered in the modern age, and now i must face the same choice. Due to the vast dynamic qualities of the music I have written for my client and the way it is designed to effect its listeners emotionally and subconsciously, It is important that it is mastered to enhance these dynamic qualities, rather than crushing them. Mastering these tracks too loud could cause a fatiguing effect on its listeners and cause the Reserve Lounge to loose customers, the exact opposite of what the project was originally intended to do. However at the same time, for music to be considered quieter than its competitors is considered commercial death, meaning my product will have to be perceived at the same loudness as the other commercial releases in the Reserve Lounge’s playlist. Luckily, upon investigation of the Kaleidovision music playback system that the rooms audio runs from, I found that it featured a loudness normalisation option, similar to the iTunes soundcheck function. Apple’s popular media playing software ‘iTunes’ introduced the Sound Check feature in 2002 within its software and hardware devices. The utilisation of this facility was to bridge the gap between volume changes when listening to tracks in shuffle mode.
“The idea behind Sound Check is essentially avoiding the annoying (and potentially even dangerous) volume changes that can occur if you play list moves from, say, a quiet jazz ballad into a full on screaming metal track.” (Robjohns, 2014, p. 23)
Sound Check mode uses a loudness normalisation algorithm to calculate the louder songs and compare them to the quieter songs and brings them to a similar playback volume. The reason this feature closes the question of how to master my work for the Reserve Lounge is because it eliminates the need to hyper-compress commercial music as the iTunes Sound Check feature will use its algorithms to adjust the playback volume accordingly anyway. “All music will be standardised to an average playback level, so the “loudness” of your music will have no effect on its playback volume” (Shepherd, 2012). This also means that no matter how dynamic or potentially quiet my product is mastered, it will still be played at the same level as the other commercial releases within the Reserve Lounge playlist. My next post will include a walkthrough of how I will master my music and will include comparable audio samples. Sources: Katz, R. (2007) Mastering audio. 2nd edition. Amsterdam: Elsevier/Focal Press. Shepherd, I. (2010). What is Mastering?. [online] Available at: http://productionadvice.co.uk/mastering-as-photoshop/ [Accessed: 19 April 2014]. Milner, G. (2010) Perfecting Sound Forever. First edition.London: Granta. Robjohns, H. (2014) The End Of The Loudness Wars. Sound On Sound, Volume 29, (Issue 4) Shepherd, I. (2012). The best loudness metering plugins money can buy – and my favourite alternative (for only 9 dollars !). [online] Available at: http://productionadvice.co.uk/the-best-loudness-metering-plugins-money-can-buy/ [Accessed: 19 Apr 2014].
Client Meeting 4 – Monday 31st March 2014
Today marked the 4th meeting with my client, John O’donohuge. The meeting gave me an opportunity to play John the 3 songs I had been working on for his VIP room. The songs that were played were production demo’s of the final product, meaning some finishing touches and mastering were still yet to be applied. Once again we used the Reserve Lounge itself to play my work so far in an attempt to gauge the acoustic qualities of the room and apply them in the mastering process. For example, if the bass bins are perceived louder or more powerful compared to the overhead speakers, I will need to balance this in the mastering stage by knocking a few decibels off the lower frequencies of the tracks. As well as this we played the songs in comparison with the other material the music interface had entered in to the playlist to grasp an idea of how much louder commercial tracks were compared to mine. The reason for this was that many average listeners perceive louder music as being better due to the increased frequencies that our ears detect from louder music, so to have your music as the quietest song in the playlist would be considered commercial death. The theory of the Loudness Wars will be covered in detail in a later blog post when deciding on mastering strategies for the songs. While listening through my work in comparison to other commercial releases I realised that the playback volume was roughly the same throughout and upon further investigation into the Kaleidovision music system I found that it utilises automatic gain control (AGC). This is an application that uses loudness normalisation algorithms to analyse songs in the playlist and play all of the music back at a similar volume through measuring the average peaks throughout songs and bringing them to the same decibel value. The point of loudness normalisation within commercial spaces like the Reserve Lounge is to create a consistent and pleasurable listening experience. This meant that in the mastering stage I would be able to master the songs to feature a wide dynamic range, rather than sacrificing dynamic range for loudness. Once again this is something that will be covered in more detail later on.
With regards to Johns feelings on my work so far, he was extremely impressed. His only negative comment was that the songs didn’t sound quite as big and professional as other commercial releases featured on the playlist. As John wasn’t musically minded I felt it was necessary to explain the next steps of the production of the songs and the mastering stage. I explained that the point of mastering was to make these songs sound big, polished and professional and that the next time he heard my work he would be amazed at how much of a difference the mastering stage makes. The next stage of production was adding some finishing touches and final mixing of the songs and then finally the mastering stage to complete the product. Following this will be the testing stage, where the final product will be played to selected ‘average listeners’ and they will be asked to document their emotions before and after the music has been played. Following this will be my final client meeting and delivery of the product to my client.