Here, they describe the sound of the dragon’s Hidden World, and dive into the creative details of creating sounds for new dragons, such as the Light Fury and the Death Grippers, and how they created new, more emotive sounds for Toothless. They also look back at their now-classic sound work on the Dragon films:
Written by Jennifer Walden. Images courtesy of Dreamworks.
How to Train Your Dragon 2 was released half a decade ago, but that doesn’t matter for fans of the franchise who have been eagerly anticipating the final installment, How to Train Your Dragon: The Hidden World, since the first trailer was revealed back in June 2018. This time around, Toothless, Hiccup and their band of dragon riders must find a safe place for all of dragon-kind to live, a secret world hidden away from humans.
At the helm again is director Dean DeBlois, who has been with the film series since the start. Also returning is the Skywalker Sound team led by Oscar-winning supervising sound editor/sound designer/re-recording mixer Randy Thom and MPSE award-winning sound designer Al Nelson. Not only have Thom and Nelson worked on all three Dragon films, but Thom got started on the sound of Toothless even before the dragon was animated. Thom’s sounds helped to inspire the creation of the Toothless we all know and love.
Here, Thom and Nelson describe the sound of the dragon’s Hidden World. They dive into the details of creating sounds for new dragons, such as the Light Fury and the Death Grippers, and how they created new, more emotive sounds for Toothless. They also look back at their sound work on the Dragon films.
What does the ‘Hidden World’ of the dragons sound like? How did you help to define this space through sound?
Randy Thom (RT): The Hidden World is underground and underwater. You enter it through this giant circular waterfall as water from the ocean pours down into the hidden world. So water, of course, is one of the main sound elements. There are glowing crystal rock structures, stalagmites and stalactites, that you can fly by while riding on a dragon. Those make musical sounds. On top of that, you have thousands and thousands of dragons that are flying around and vocalizing, so it’s quite a world of sound down there.
For these thousands and thousands of dragon vocalizations, was this material that you were able to pull from the sounds you created but didn’t use for the first two How to Train Your Dragon films?
RT: We certainly used some of those and we made variations on those. We also came up with a bunch of new vocalizations for this. Both Al [Nelson] and I worked on creating new dragon vocalizations, as well as another sound designer we worked with named Jeremy Bowker.
It was really important to director Dean DeBlois that the dragons in this film have even more of a language than they had in the first two films
It was really important to director Dean DeBlois that the dragons in this film have even more of a language than they had in the first two films, so that you get the feeling that they are actually saying things to each other and not just squawking or roaring or grunting. We worked on that quite a bit, to make it seem like you are hearing the dragons talk to each other and say things in a dragon language to humans and respond to humans using dragon language.
Toothless needs to communicate some sophisticated emotions in this installment, both with Hiccup and with the Light Fury. How did you create the new, more emotive vocalizations for Toothless?
Maybe it’s a dog yawning and there is something about that that sounds halfway between an animal vocalization and human vocalization.
RT: Some of it was literally my voice, even though we always try to use animal vocalizations as much as we can as a starting point because it’s very difficult for human beings to make themselves believably sound like animals. There’s always something about it that tells the listener that it’s a person trying to sound like an animal and not a real animal. So we try to start with actual animal vocalizations, like elephants and whales and camels — every conceivable type of animal you can think of. We are constantly trawling the sonic universe trying to find animal vocalizations that we think will lend themselves to application for these How to Train Your Dragon films, especially animal sounds (even very common animals like cats and dogs) that seem to be like language, in the sense that they don’t sound like the standard bark or meow. Maybe it’s a dog yawning and there is something about that that sounds halfway between an animal vocalization and human vocalization. Whenever Al, Jeremy, or I happen upon a sound like that, we label it and store it away because it may be useful for some other scene in the movie.
When we can’t find an animal vocalization to do what we need it to do dramatically, we sometimes resort to using our own voices. We go into a dark room where nobody will be embarrassed by watching us and we do all kinds of strange sounds with our mouths. Usually the shooting ratio is very high. We have to perform/record typically a hundred of these vocalizations before we get one (usually by accident) that has some potential to be credible as an animal vocalization.
The next trick is to integrate all of those things, so that it doesn’t sound like you are going from an elephant to a goat to a human in the span of one phrase that the dragon might utter. There is some technical trickery that we use in an effort to change the source sounds quite a bit. For instance, we change the pitch as much as two or three octaves to make a dog sound like a creature that weighs 2,000 pounds.
Al Nelson (AN): One thing that I enjoy about the vocalizations that Randy described is that he’ll create such a vast palette, especially with his voice in combination with animal sounds, and the director might say, “Can it be more like an ‘oh?’” Or, “Can it be more like a ‘what?’” So the director will use our familiar language to explain what he would like Toothless to be doing. Then I or one of the other designers or editors will go through the palette that Randy has constructed and find the right inflection and continue to fine-tune this language. Sometimes we’re working with animation that isn’t complete, and then suddenly we’ll get the updated visual effects and Toothless’s mouth or eyes will be doing something that looks like surprise. So we go through our sounds and find some ‘surprise’ vocalizations. Randy has done a variety of emotive performances specifically to be able to construct this dragon language. It’s an enjoyable but challenging process.
RT: When we hear a sound that has potential, we will put it in our sound library and we’ll add metadata that describes its emotion, like ‘surprise,’ ‘fear,’ ‘sadness,’ or ‘what?’ or ‘wow!’ That makes it much easier for the editors to find potential candidates for a given sound in a given moment.
Any specific tools that help you to integrate the animal vocalizations with the human vocals?
AN: Of course our go-to is Serato’s Pitch n’ Time Pro because you can manipulate pitch in a graphical way. So if you need it to inflect up in the beginning and then down at the end you can draw that shape. You can also manipulate whether or not you want to change the length of it. Obviously, you can go too far and break a sound but Pitch n’ Time Pro is a good resource for pitch manipulation.
I like to use some of the Native Instruments products, like Kontakt. I sometimes use their sampler options. And then there are the basics like reversing a sound.
First and foremost, you have to start with the right material. Then from there, it’s just getting it to match — getting that goat to match that human and that camel. Maybe the goat is an octave too low and so we are going to pitch it up and then change the inflection of the human vocalization and then maybe smooth it out with a little bit of reverb. My advice is to not go too crazy with the processing because you always want the final sound to feel organic.
Popular on A Sound Effect right now - article continues below:
Latest releases:
-
20 %OFFEnds 1553986800 -
30 %OFFEnds 1553986800 -
Igniter is the new industry standard for creating any real-world or sci-fi vehicle and engine sound effects with ease and at unbeatable value. Whether you work in audio post or game audio, Igniter enables you to design, perform and automate any complex vehicle behaviour directly in your DAW – from sports cars, motorbikes, planes, helicopters, spacecraft and other engine sounds to moving ambiences, textures, Foley or whatever sparks your imagination.
Igniter is the only solution of its kind to come jam-packed with over 20 performable vehicles and 1,943 audio assets from vehicles including Aston Martin, Ferrari, Porsche, Tesla, Harley Davidson, Huey UH-1H, Agusta Westland 119x, CH-47D Chinook, Bombardier Challenger, Cessna 560XL, skids, sweeteners and much more. Recordings come from Watson Wu, The Recordist, Sounding Sweet, George Vlad, Echo Peak and Flysound.
For vehicle fanatics, Igniter Full Tank comes with 75GB of additional library content including onboard recordings, ramps & static rev bands, Foley sounds, skids, slides and much more!
-
Igniter is the new industry standard for creating any real-world or sci-fi vehicle and engine sound effects with ease. Whether you work in audio post or game audio, Igniter enables you to design, perform and automate any complex vehicle behaviour directly in your DAW – from sports cars, motorbikes, planes, helicopters, spacecraft and other engine sounds to moving ambiences, textures, Foley or whatever sparks your imagination.
Igniter is the only solution of its kind to come jam-packed with over 20 performable vehicles and 614 audio assets from vehicles including Aston Martin, Ferrari, Porsche, Tesla, Harley Davidson, Huey UH-1H, Agusta Westland 119x, CH-47D Chinook, Bombardier Challenger, Cessna 560XL, skids, sweeteners and much more. Recordings come from Watson Wu, The Recordist, Sounding Sweet, George Vlad, Echo Peak and Flysound.
-
31 %OFFEnds 1553814000
When advising young sound designers, I always tell them that it may be tempting to use all of these gadgets but they really should spend their time trying to find sounds in the real world that already have most of the characteristics that they’re looking for.
RT: One thing that gets sound designers into trouble most often is using too much processing and ending up with an artificial or electronic sounding result. That is fine for certain kinds of films but for creature vocalizations, if you process them too much, then they’ll sound too electronic. Sound designers have probably been reprimanded about that more often than anything else. Directors like creature vocals that sound natural and real and organic. They usually don’t want something that sounds processed. So when advising young sound designers, I always tell them that it may be tempting to use all of these gadgets but they really should spend their time trying to find sounds in the real world that already have most of the characteristics that they’re looking for. Then, they’ll only have to use technology 10% instead of trying to take a sound that doesn’t have that much potential and using technology 75% to make it work.
Did you do any fun field recordings for How to Train Your Dragon: The Hidden World?
Sound is very context dependent, I think. So everything about this new context that you are putting the sound into is going to make it feel like a unique sound.
RT: We recorded many potential sources for the dragon vocalizations, but we did not record ambiences. We mostly used material from our library for that. When we use things from the library, we are always careful to not overuse them. We don’t want to use a sound that is immediately identifiable as being used on some previous movie. But our library at Skywalker Sound is so huge that it’s really not difficult to find material in it that nobody has ever heard before, or that very few people have ever heard before. And very often, even if the sound has been used on another film, if you put it into a completely different context then it plays as a different sound. You’d have to go over it with a microscope to figure out that the sound had ever been used in any other way. Sound is very context dependent, I think. So everything about this new context that you are putting the sound into is going to make it feel like a unique sound.
With all that said, we do typically try to do as much recording as we can for every new film. On this one, though, we didn’t do a huge amount of ambience recording. Also, there isn’t as much fire in this film as there was in the previous two Dragon films. So, we didn’t do any new fire recordings for this film.
AN: There were two new things for the dragons though. Randy did design really cool sounds for the Light Fury. She has a very unique attack sequence that she uses. She’s very stealthy and cool, so her sounds are more musical and tonal. She does something much cooler than just blast fire. So a lot of Randy’s design was in that direction.
We did record a huge collection of fire for the previous film. One of Randy’s assistance Nia Hansen had a firebreather come out and do some awesome firebreathing tricks, like spitting fire and breathing fire. So, for these films we have a huge palette of fire sounds that we recorded. Most of the fire in this film is fire that is already associated with the Dragon franchise.
The new dragon sounds are more musical and tonal, and special. It’s not just breathing fire. The dragons also expel electricity and lightning and things like that. The filmmakers have taken this into more fantastical territory. These new dragons, especially the Light Fury, do even cooler stuff.
Some of the new bad dragons have this acid goop that they spew. Randy did some cool work on that and I think we did some new recordings for that as well.
It’s ‘been there and done that’ with the fire thing and so we went to new and different stuff which is what the filmmakers wanted.
RT: For the acid goop, one thing that I recorded was pouring liquid from one very hot pot or frying pan into another so that you get this dynamic sizzle. The main trick there is to choose raw material that has the most potential to make an interesting sound. And so since this acid goop is coming out of the mouths of the dragons, I found that using pots with a fairly narrow opening at the top (or metal containers with a fairly narrow opening at the top) more closely simulated the sound of a vocal cavity, i.e., a mouth. That way the sizzling sounds and the airy sounds of pouring liquid from one container to another became more musical and more dynamic and more like a vocal.
What were some of the sounds that you made for the Light Fury?
RT: She was a big challenge vocally, because we needed to suggest with her vocalizations that she was female. I’ve faced the challenge before of coming up with female-sounding creature sounds. It’s a bit of a conundrum because humans are not able to distinguish the difference between male animal vocalizations and female animal vocalizations. If you hear a dog bark, you really can’t tell if it was a male or female dog. It’s the same with any cat.
I’ve used various tricks to try to suggest the feminine aspect of the Light Fury. One (and you could quickly become mired in sexism with this) is that felines traditionally are associated with the female. So I did start with some cat purring and cat vocalizations that I think were useful. Believe it or not, I also used some of my own voice pitched up about three octaves for things that I just couldn’t do with cat vocalizations or with some other kind of animal.
Normally for dragons, I’m pitching my voice down. But this is one of the few cases where I found myself pitching it way up for the creature instead.
What about the Light Fury’s attack sounds? Did they differ from Toothless’s attack sounds?
RT: We came up with a distinctively different attack. Toothless has his signature sound when he dives and is about to attack. We came up with a different version of that for the Light Fury.
AN: Her sound is more musical and stealthier. Toothless’s attack sound was designed when we first met him and he was dangerous. So his attack sound is very aggressive. The Light Fury is much more elegant. We know Toothless by this third film, but the Light Fury is more mysterious. She’s not necessarily an adversary. She is kind of an enigma. She has this mysterious, musical/magical persona and her elements reflect that. They’re not necessarily dangerous; they’re just cool and stealthy. They still sound powerful but they don’t sound like the bad dragons. She sounds like a cool dragon.
Dozens, possibly hundreds, of dragons fly around the Viking ships.
What were some of the base elements that went into her sound? What was some of the processing tools that you used to manipulate them?
We have close to one million recordings of wind in our library, every conceivable kind of wind.
RT: All kinds of strange sounds went into it. It’s partly squeak and squealing and screaming and manipulating those sounds. There are things like firework rocket whistles and wind. We have close to one million recordings of wind in our library, every conceivable kind of wind. Sometimes when we’re doing a Toothless or Light Fury diving attack we will use a distinctive kind of whistling wind for one of the elements that you can either interpret as the sound of the dragon flying through the air or as an element of the vocalization — it doesn’t matter so much how you interpret it as long as it evokes the emotion you want it to evoke.
Did you have a favorite scene for sound?
It’s a special scene because it requires us as the sound designers to be the performers of these animated creatures.
AN: One of my favorite scenes is very simple and very sweet. It’s where Toothless first meets (or, directly engages with) the Light Fury. It’s very sweet and incredibly humorous. It’s all personality. So even though it is a simple scene — it’s really just two dragons vocalizing and communicating with each other — it relied on the vocalizations to get across the humor, the curiosity, and the frustration. There is no dialogue except dragon dialogue. It’s a special scene because it requires us as the sound designers to be the performers of these animated creatures.
RT: That’s my favorite scene too.
AN: That scene shows Randy’s creativity in a very special way.
Contrary to that scene, there are big hero moments that happen, and those are surprising. There is a rescue scene near the end and the audience is white-knuckled right down to the last moment. It has great sound. The villain, Grimmel (voiced by F. Murray Abraham) is fantastic. He gives a great performance. You really do think that our heroes won’t make it. The flying and attack sequence that happens required a lot of sound design and mixing work and editorial. That sequence reflects a lot of creativity.
There were new dragons that were the minions of Grimmel. Those were called Death Grippers. Their visual design gave us the opportunity to do something new and cool. Randy and Jeremy were involved in the initial vocalizations, and those were great. The director Dean then gave them the note that he wanted the dragons to do this clicking sound, which reminded me of the Predator alien sound. So Randy comes back with something a few hours later and it’s this really cool, liquidy, clicky, goopy sinister sound. I had the privilege of going back through and adding that in under the vocalizations. So when we first meet these evil dragons, the Death Grippers, they come walking past and you hear them coming in from the surrounds. It’s creepy! It’s really fun.
RT: The fun thing about these Dragon movies is that there are always new dragons that you haven’t seen before. Trying to come up with the sounds for the new dragons is a kick.
You’ve both been with the How to Train Your Dragon films from the start. Looking back, what’s been the most memorable moment in terms of your sound work on this franchise?
We like to get involved early enough on a project so that the sound ideas will influence the visual artists.
RT: In a way, my heart belongs with Toothless because he has been the central dragon in the series and he is the one that I started working on first, even before any animation had been done for the first film. I was asked to do some speculative sounds so I started playing with walrus sounds and elephants and whales. Those sounds influenced the way the animation was done a little bit, and that’s the way we like to work. We like to get involved early enough on a project so that the sound ideas will influence the visual artists. And that very much happened on How to Train Your Dragon.
If I had to name a character in the films that I feel the closest to, it would have to be Toothless.
By the way, Al has been giving me too much credit. He’s done a lot of creative sound design work on this series, including creating some of the dragon vocalizations for the Gronckles, and several of the other dragons are mostly Al’s work. It’s been such a pleasure to work with Al. As I’ve said many times before, creature vocals in my opinion are the hardest kind of sound design to do because they do have to sound organic but they have to be really evocative and interesting. Al is one of the few sound designers that I’ve run into who is good enough at his craft to do that work at the highest level. So I feel like I’ve been so lucky to work with him.
AN: That is very generous Randy, thank you.
Randy both advised and mentored me, and gave me incredible freedom to be creative and guide me through the process.
One of my favorite things about the franchise is that Randy kicked off the sound work before there was any animation, back on the first one. He asked me to come along and help with some of the design and that is one of the best points in my career. It was such a treat. I was young and eager. Randy both advised and mentored me, and gave me incredible freedom to be creative and guide me through the process. It was really an amazing opportunity and I’m glad that I was able to stick with it through the entire film series. It’s been great fun and a great learning experience.
The director Dean DeBlois has been the director on all three and he is one of the most genuine and creative and wonderful people in the industry, and in the world. It’s been such a great opportunity to work with him and I think we’re all better people for it.
RT: Many people in the movie business have very big egos and are very difficult to work with, but there isn’t a single person like that on the Dragon team. The editorial department is made up of wonderful people. Dean is a filmmaking saint, as far as we are concerned, because he really cares about sound. And, as Al said, he’s just a genuine and nice guy. We have all looked forward to working on these Dragon films, just for the camaraderie and for the opportunity to do new things.
A big thanks to Randy Thom and Al Nelson for giving us a look at the incredibly innovative sound of How to Train Your Dragon: The Hidden World – and to Jennifer Walden for the interview!
Please share this:
-
20 %OFFEnds 1553986800
-
30 %OFFEnds 1553986800
-
Igniter is the new industry standard for creating any real-world or sci-fi vehicle and engine sound effects with ease and at unbeatable value. Whether you work in audio post or game audio, Igniter enables you to design, perform and automate any complex vehicle behaviour directly in your DAW – from sports cars, motorbikes, planes, helicopters, spacecraft and other engine sounds to moving ambiences, textures, Foley or whatever sparks your imagination.
Igniter is the only solution of its kind to come jam-packed with over 20 performable vehicles and 1,943 audio assets from vehicles including Aston Martin, Ferrari, Porsche, Tesla, Harley Davidson, Huey UH-1H, Agusta Westland 119x, CH-47D Chinook, Bombardier Challenger, Cessna 560XL, skids, sweeteners and much more. Recordings come from Watson Wu, The Recordist, Sounding Sweet, George Vlad, Echo Peak and Flysound.
For vehicle fanatics, Igniter Full Tank comes with 75GB of additional library content including onboard recordings, ramps & static rev bands, Foley sounds, skids, slides and much more!
-
Igniter is the new industry standard for creating any real-world or sci-fi vehicle and engine sound effects with ease. Whether you work in audio post or game audio, Igniter enables you to design, perform and automate any complex vehicle behaviour directly in your DAW – from sports cars, motorbikes, planes, helicopters, spacecraft and other engine sounds to moving ambiences, textures, Foley or whatever sparks your imagination.
Igniter is the only solution of its kind to come jam-packed with over 20 performable vehicles and 614 audio assets from vehicles including Aston Martin, Ferrari, Porsche, Tesla, Harley Davidson, Huey UH-1H, Agusta Westland 119x, CH-47D Chinook, Bombardier Challenger, Cessna 560XL, skids, sweeteners and much more. Recordings come from Watson Wu, The Recordist, Sounding Sweet, George Vlad, Echo Peak and Flysound.
-
31 %OFFEnds 1553814000
+ free sounds with every issue:
🙏🙏🙏.