Exploring the visual soundscape of SS Rajamouli’s ‘Baahubali – The Beginning’

Baahubali-Index
Socialize it

The date 10 July 2015 will forever be remembered as the day when India witnessed the release of its biggest motion picture release till date in SS Rajamouli’s ‘Baahubali – The Beginning’ .

There has been much written about the visual extravaganza that Rajamouli has managed to bring to the big screen, TheGraphicSlate.com also did an exhaustive interview with the visual effects supervisor Srinivas Mohan to better understand how did the team manage to create the magnum opus over the last 3 years in production.

In this study, we now look at how Dolby Atmos was used for the mixing of the sound in the feature film and also understand the approach that was used by sound designers PM Satheesh and Manoj Goswami along with the sound mixer Juztin Jose, who had a discussion about the mix of the film with a couple of months before the actual mix started and then began their adventure of being a part of the biggest film in the history of Indian cinema – ‘Bahubali – The Beginning’.

Immediately after finishing the mix of Anurag Kashyap’s Bombay Velvet,  Juztin Jose started the mix of  Bahubali and the next 45 days were the most challenging and the most memorable days of his life.

In this case study Juztin shares the mix approach and the thought process behind a few important sequences in ‘Bahubali – The Beginning’ exclusively with TheGraphicSlate.com.

The Approach

It was very important for me and my mix team to understand the vision of the director and what the sound designer and his team expected. Along with that it was crucial for me to be a part of the story and emotion. I saw the Telugu and Tamil  version  of  the  movie. After that, through many discussions with the director SS Rajamouli and music director MM Keeravani, it was clear that the sound of the film should enhance the mood and culture of the scene conceived. To achieve this it was very critical that the sound design, the music and the mix find equilibrium so that the sound and the visuals complement each other in conveying the mood and grandeur of the film.

Juztin-Jose-1

We decided to mix the film in native Dolby Atmos. There is no other format that allows true immersive sound and so much resolution in panning and positioning that it brings the audience to be a part of the story. To eloborate, when you’re in the sea or in the forest, and you hear the waves or insects, you feel like you’re in that space. It suddenly becomes very real. The goal was to extend the film off the screen and into the theatre and make the audience part of the experience. We didn’t want the audience to sit back and watch the movie. We wanted them to experience Bahubali.

The Signal Flow

As I mentioned, the film was natively mixed in Dolby Atmos with 9.1 bed and 110 objects on the maximum side. We averaged around 60 objects per reel. The mix setup involved 3 Pro Tools | HD workstations running simultaneously handling total average of 700 voices in every reel. We split the sessions as mentioned below:

– Workstation 1 – Pro Tools | HDX2 – Handling Dialogue, FX, Foley and Crowd.

– Workstation 2 – Pro Tools | HD Native – Handling the full open tracks of background score and songs.

– Workstation 3 – Pro Tools | HD2 – For video playback and recording 7.1/5.1 and stereo downmix.

The Pro Tools session was set in 32bit float/48kHz/broadcast wave format in Pro Tools HD Version 10.3.10. We used Avid’s D – command  control surface for the mix.

As the background and music tracks were on the HD Native workstation, the bed and object outputs were assigned to HDX2 inputs. The output of HDX2 workstation were assigned to the Dolby RMU for monitoring.

baahubali-sound-1-Signal flow in the Mix Stage

FILM MIX AND PREMIX STAGES

Dialogues

We knew the film would go loud due to it’s high intensity action sequences and the drama and we needed to control it. For this, one of the techniques was to make the dialogues smooth and rounded as it defines the nature of the mix. Sharper dialogue would eventually result in tiring of ears and mind.

The film had been shot outdoors hence there was not much play of reverb other than a few slap back delays for dialogue on reflective hilly areas. A cave had been shown as a meeting place of the rebel group and for recreating the presence of the dialogues and their reflections, we used the reverb as objects and panned overhead using the cave presets on Altiverb 7.0. Revibe and R360 were also used wherever it was required.

The beauty in this was that it was never static and we were able to get it as close as it would be in the real world of that era. There is something special about those times. There were no traffic and so the density of the atmosphere would have been thinner. The slapbacks would have to be brighter and cleaner. One thing that was interesting was the dialogues were pitched higher. This was also because language as a culture was not something that would be below certain octaves. If you consider the musical elements and instruments of that era, they were mostly in the upper mid range. So, any song or communication had to be above that. This thinking eventually led to the fact that the reverbs should not be as dark as what we have today.

Even the slap backs of that era were different because the walls were of different texture. If you go to any historic places you will notice the reflections are bright. One play we had was the brightness of the reflections were lesser when it was a common man’s place like the street because the materials they would have used to construct would be more absorbant unlike the rich. That was one subtle way we used to show the class divide that existed then.

Crowd

There are massive crowd sequences throughout the film. Spanner, a plugin made by Cargo Cult was used extensively along with Atmos panner to achieve the desired impact. A wide range of reverbs were used to smoothen the sharpness, that  helped  in  enhancing  the  crowd  sequences.  Reverbs  are  what  gave  the bigness of the scenes as our mind will immediately take over the hugeness by then.

Background Score and Songs

MM Keeravani has composed the background score and music for the film, which I feel is one of the finest in the recent times. It helped in elevating the emotional content of the film as well as escalating the grandeur of many scenes, including the war sequences. Usually, in films background score comes in a limited number of tracks. Yet, for Bahubali we received around 60 stereo tracks.

In a war oriented film like Bahubali the rhythm pattern mix is very critical. On many occasions, the rhythm tracks have been used as an object to bring in more width and separation, especially in the Hi-Mid frequency. We also experimented with having objects in screen wide surrounds instead of the screen channels. This technique helped a lot in getting more separation and less clumsiness in the 7.1 & 5.1 downmix. Wide pan technique was used on double bass, violins, horns etc.

Personally, I don’t see any problem with object panning in music tracks, provided the height and width parameters of the panner are properly managed. Overhead channels were used for reverb of the score which mainly included horns and the strings section. This helped in enlarging the total sound spectrum and making it more immersive for the audience.

To play out the drama, vocal FX, reverbs of aalaps etc were used as overhead. There were separate 7.1 Bus masters for Rhythm, Horns, Strings, Chorus and Music on HD Native workstations which were routed to a common input on the HDX2 workstation. The main reverbs we used were Phoenixverb from Exponential Audio. One of the beauty of this reverb is that it was created for music with lush silky reverb tails and has one of the smoothest sonic textures I know.

Part of the Background Score Session for waterfall Sequence

FX, Foley and Ambience

This was the most challenging part of the movie. The film being an epic with some jaw dropping war sequences and waterfall shots were just a joy for any sound person. The way the tracks were laid out and designed were very clean and hassle free. All the credits for this is to PM Satheesh and Manoj Goswami, the sound designers.

We used Objects and Spanner extensively during the FX mix for smooth as well as efficient panning of individual elements. The movement of horses, chariots, arrows, chains and sound FX like swishes, whooshes etc were so precise and natural that it brings the audience into the warfield or the world of Bahubali. Spanner was used to pan the bed channels, especially in the water fall sequences and on additional crowd FX in the war, as the sound would move complimenting the visuals on screen.

Earlier on, even during Bombay Velvet, I realised that there are a few ways to bring dynamics into the film.

They are: Level, Pan and Frequency

I followed these ideas to get what you see on screen. Basically, the frequency bit was done by the designers and the score. I took care of the Levels and Pans. This is what can change the audience from being a third person with safe zone, to first person and being part of the real experience. This was how I was able to get the audience to the root of the culture and the characters that they experience. If I can’t convince myself of this, then the audience won’t be.

This was why the movie was challenging for me and my associate Sarath Mohan. We spent more than 700 hours in mixing 4 languages of the film and we were adamant that whatever language the audience chooses to see, they should have the same experience. All of these inspired me to share in the following lines, some of the techniques we used and what we managed to do in Bahubali.

The Waterfall Sequence

The first 30 minutes of the film revolves around many shots with a huge waterfall. The director was very specific that he needed the sound of this waterfall and it’s power to be felt by the audience to experience it’s grandeur. It was a beautifully shot scene with so many visually arresting shots.

One of the techniques we did with the sound FX layer of waterfall, was we assigned it proportionately to the Bed as well as the object tracks, and the low frequency was treated separately to enhance the low rumble.

Baahubali-Shivudu

LoAir and RBass were the plugins used to enhance waterfall sound with parameters set such that it doesn’t clash with the low frequency of the score. It was very important that we don’t lose the scale of the waterfall yet have the emotion and narrative through the score. We made good use of the height channels to envelope the audience to make them feel as if they are there.

The biggest challenge in this scene was that because it was water and it had a wide frequency spectrum, there was enough to clash with dialogues or score. The two ways I used to avoid the clash yet have the feel was by placement and EQ. There was another interesting technique that we were fortunate about. If you think, towards the beginning of any movie, the theater would be colder than what it would be because there were lesser people and it would take time for the temperature to register.

Since the waterfall happens in the beginning portions and the colour tone of the sequence was primarily cool, it allowed us to have more higher frequency spectrum to get the splash of water and the drops. This added to the experience. The grandeur was controlled by the low frequency and the score. We didn’t ride the score to create space for the dialogues. Instead the approach we took was to let everything have its place in the mix. We achieved this by positioning and EQ. This was yet another advantage of Dolby Atmos. Because I always feel that the audience wants an experience in whatever theater they see the film and it is my responsibility to provide that to them.

The War Sequence

The crux of Baahubali – The  Beginning is its 30 minute long war sequence, which is one of the best choreographed fight sequences in Indian Films. The main challenge in the mix of the war sequence was to manage the balance between music , crowd and sound FX. It was very important to have equal weightage to both the score and the effects.

As the sequence starts, we had to maintain the rhythm of the soldier footsteps. This was treated as a 7.1 bed. When Kalakeya and his army charges towards Mahishmati, the crowd, sound fx and foley sounds of their movements were panned shot by shot using Spanner. When we cut to the conversation of the queen and the principal characters , the movement of the opposing army were kept as a distant sound from surround using Reverbs and low end roll off.

Baahubali-Rana

There is a shot of a metallic wall being formed to fend of the enemy army. The sounds for these and the pans were done using atmos objects to give the resolution of the movement. We wanted the audience to be in the wall with Bahubali’s army.

The arrow fight sequence was a mix of beds and objects; with object tracks used mostly to define the height and  trajectory of the arrow. But the most interesting sequence was the one in which there was a tactical move by heavy cloth being targeted towards the charging enemy group setting it on fire with a flaming arrow to burn the enemies out. I had to create the astonishment and the wonder of this in the mix.

The flying of cloth was mostly stage and bed oriented to give enough  density  for  the  cloth.  After  that,  when  the  cloth  catches  fire,  a combination of bed and object tracks were used to enhance the whole scene. The audience doesn’t know what to expect there, so I had the full freedom to responsibly choose what they had to hear. The score that rises later was balanced that it rises as the audience realises what happens. This was important because it was a commercial film and we wanted the audience to react by cheering or whistling and still drive the score through it without pushing levels.

As the war progress the objects were reduced and FX were bed oriented to avoid distraction and also to give enough space for the score. But for individual shots of stabs, arrow movements, fight reactions involving movement etc, object tracks were used. Horses and chariot movements were either panned using Atmos panner or used as bed, depending on the shot.

There was also a sequence treated in a way that the main villain throws soldiers towards the hero. Here, we created silence during the time of the soldiers being flown towards the hero, and once their body touches hero’s body fx charges in.

Again when everyone fall down as a pile, it is pin drop silence (reactions) for a while and the moment everybody in the pile starts flying in all different directions the sound effects and BGM returns with full bang with LFE pumped up to give the full impact of the hero rising. Honestly, this was a hero sequence and we wanted it to have a full masala feel to it in all aspects.

Part of the FX session for the War Sequence
Part of the FX session for the War Sequence

Final Notes

We followed a strategy in the mix of Bahubali – The Beginning. During the initial portions, including the title, the mix was 7.1 bed oriented and as the story progressed Atmos Objects started becoming prominent and towards the end Objects content were gradually reduced and mix became more 7.1 bed oriented. The main highlight of using Atmos Panner is that there is nothing else that will deliver  the  precision  of  pan  expecially  when  sound  travels  across  to  the audience.

Overhead direct output was used mainly for assigning music reverbs only. Most sound effects that need to be sounded from above the head were achieved the positioning with the help of Atmos panner ;also considering for the down mixes.

Also, in the score, the pans were as follows:

1. Double Bass Front Speakers Left, Center, Right

2. Cellos Panned away from the stage towards the first few speakers

3. Low, Mid and High Strings more back and within the space to give separation.

Atmos Monitor of a sequence
Atmos Monitor of a sequence

It was crucial that the balance of this stays and also gives us the lushness that the score originally intended to create. The score has a purpose and I wanted to make sure it was translated across to the audience. We wanted to achieve the emotion that a symphony would give us. But to do this, we also had to change placements so as to not lose the story which is the main idea. Another technique we used was to also correct the pitch of the characters dialogue to match the age and attitude and also fit with the score.

A crucial aspect to achieve the mix of 4 languages in 700 hours was the disciplined methodology followed. What we did was to mix the Telugu version first. The Hindi is a dub based on the Telugu video. The Tamil version was shot differently and it was impossible to do two mixes. Instead, we got the sound editor of the movie, Gokul KR to edit the Tamil version to match the Telugu mix. Then, any changes and smoothing would be adjusted accordingly, and the Malayalam mix would be built from that.

None of this would have been possible if it wasn’t on Pro Tools. I cannot imagine adjusting automation on a million systems to get this. Mixing in the box, as many call it, for me has only advantages. We save time, effort and mostly have a way to achieve a creative output in a shorter time. The stability Avid offers on their system and the way it can handle so much load consistently over and over is what is unique about it. Add to this the various plugins and processing that we have and there is hands down nothing to compare it with in my opinion. Not only did we complete it in 700 hours (which is not less), I would say this is the fastest way I have achieved this with such a complicated movie. I know that it would have been way more than that if I had even attempted to mix otherwise.

Juztin-Jose-2

One very important thing to understand is that the pan that the Dolby Atmos panner generates in a down mix is completely different from the conventional pan laws that are used. This means that there is no way to create a mix that we can make with this format. But to me what is more important is that even in the down mix, the audience would get to hear a very close experience of what the Atmos mix sounds like. That would have been impossible for me to achieve otherwise. So I have heard conversations of saying that the film doesn’t require Atmos. My take on it is if you pan anything in the surround channels, you need Atmos. Its not the overheads, but the resolution and the amazing down mix that defines the format.

At the end of the day, the audience wants an experience of Baahubali, Mahishmati, the grandeur and the story. They must connect with the characters and the culture. They shouldn’t just see the movie. They should be a part of it. And as a team, we were able to work in an invisible medium that delivered all of this. There is only one other art form that deals with the invisible. Magic. Are we magicians? Maybe not. But we definitely are crafting the illusion.

Baahubali


Add
itional Mix notes

Avantika Introduction: Avanthika’s  introduction  chase  and  fight  sequence  has swords, bows and arrows. The sword and the arrow sounds were panned using the atmos panner and every arrow swoosh was panned according to the shot. It helped bring the audience into the action in the middle of the jungle with arrows flying all around them. The shot in which the camera pans from the mountain to deep into the forest through the top of trees, treatment was given in a way that the ambience sounds in LCR of 7.1 bed moves to the overhead using atmos panner in sync with the camera movement.
Kattappa Introduction: FX for Kattappa’s introduction scene is treated as Bed with wide panning which takes the audience to the location. The movements of fire sparks that are generated during the fight between Kattappa and Aslaam Khan were panned using Atmos panner to compliment the scene.
Shiva – Avantika Confrontation:In the first confrontation scene of Shiva and Avantika, he is attacked by an arrow shot by her. In the scene, the trajectory of arrow’s swishes was muted and there was a deliberate extra impact put into the arrow as it hits the back of the tree to give a disruptive break to the scene.
Avalanche scene: There is a sequence with a heavy avalanche eruption and big snow breaks. The avalanche sound is mostly placed in front bed output, and we created the perspective of sound according to the shots by adjusting the low end frequencies of the avalanche and hence avoiding monotony.Also in the scene we see the hero and heroine escape riding in a snow cart from the avalanche. The Score which is of run of the mill type genre was treated in such a way that least music was pumped to LFE to avoid LFE being over burdened with low frequency sounds of the score as well as the low end of avalanche sound FX.
Bhadra Killing: The shot begins with the hero slowly gaining consciousness. We rolled off the high frequency to slowly bring it in as the hero comes to his senses. In the fight sequence following it. Later, when hero chases the Badhra and chops off his head, the whole scene was treated as silence, once the head is chopped off. It was the Director’s intent to symbolise that at that point, nature too stood still.
You Might Also Like
Gadgets
Zion Xtreme DDR4 RAM Review: Matches up to its international counterparts
By Vivek Kapadia | 1 year ago
Games
Review – iBall Compbook Excelance – A laptop under 10k
By Vivek Kapadia | 1 year ago
Animation
Review- Finding Dory; more about getting lost and discovering yourself than finding others
By Ronald Rodrigues | 1 year ago
Gadgets
Review: BIOSTAR H110MD Pro D4 – A decent Skylake experience for the budget price
By Vivek Kapadia | 1 year ago
Gadgets
Review Asus Zenfone Max (2016) – A phone in the garb of a power bank
By Vivek Kapadia | 1 year ago
VFX / SFX
Review: Alice Through the Looking Glass; storytelling and creativity keeps you hooked on ‘Time’
By Ronald Rodrigues | 1 year ago

ad-970×90, Around the web, Footer Links