Speech Graphics in 2024
A Year of Innovation, Recognition, and Pioneering AI Technology
By Michael Berger (CTO) and Gregor Hofer (CEO), Co-founders of Speech Graphics
As the year comes to a close, we would like to take some time to reflect on a year of product developments, industry recognition, and dynamic growth at Speech Graphics. We continue to lead the field in audio-driven animation with the expansion of our nonverbal animation tech including breath, laughter, and improved real-time responses; we’ve improved our lip-sync and multi-lingual capabilities; and we’ve been featured in many exciting games! Below you can find out more about Speech Graphics in 2024, including news about our new platform, Rapport.
Starting the Year with a Bang
At the start of the year, we were thrilled to win the Dragon Award for Pioneering Spirit at the China-Scotland Business Awards, and in March we were named by the Financial Times as one of the fastest-growing companies in Europe! We’ve seen continued growth in 2024, particularly in our Singapore team which supports our clients in Asia.
In Q1 we also released our Remnant 2 Case Study where SGX was used to animate 14 characters, including a 14-foot tall, Root-infested Wolf creature, complete with 10 eyes, featured in the video below!
Chris Mead, Animation Director at Gunfire Games, said: “This software (SGX) not only elevated our facial animations to match our exceptional voice talent, but it also empowered us to scale seamlessly.”
We are very proud to get testimonials like this from our clients, and watching back how they have used SGX to bring their projects to life is a joy for us!
What’s new in SGX?
This year saw the release of SGX 4.2 and 4.3 with a range of exciting new features. We extended our Unreal integration in SGX Studio UE5, exposed a lot more metadata for editing, added metadata export, tightened up file management and user experience, and improved our lip sync as well as nonverbal behaviors. The Speech Graphics Knowledge Base also continues to expand with new articles and helpful videos, plus a version in Mandarin, with Japanese and other languages coming soon!
Language Modules
In order to continue providing the best multi-lingual pronunciation accuracy in the market, we've released several language updates, including:
Mandarin 2.1 - Improved pronunciation
Korean 2.0 - Improved pronunciation, plus support for mixed writing systems in transcripts
French 1.3 - Improved pronunciation of nasal vowels
Breathing, Laughter, and Effort Sounds
At Speech Graphics we aim to detect and animate every sound that can be made by a human - not just speech. These non-verbal nuances are essential to creating immersive character experiences in games.
Breath: SGX 4.3 features animation of the breath cycle, automatically synchronized with inhales and exhales detected in the audio and weaved into pauses in the speech. The intensity, speed, and rate of breath are all determined by the vocal performance.
Laughter: SGX 4.3 adds a new auto mode: Laughter. This is triggered whenever laughter is detected in the audio. The behavior of the new auto mode is defined by mapping it to a behavior mode with various laugh-like expressions.
With SGX 4.3, we continue to lead the way in the generation of high-quality animation at scale and with creative control. These new features empower creators to produce realistic, expressive, and engaging characters more quickly and efficiently than ever before.
Find out more about how it works and other features of SGX 4.3 here.
What’s new in SG Com?
SG Com is our real-time SDK that converts audio into facial animation with a latency of only 50 milliseconds. As with the rest of SGX we have added new abilities to detect and animate behaviors like laughter and breath – all in real-time! In addition, we improved nonverbal behavior by lower response times and more natural motion. We expanded live behavior controls and metadata output. This was all included in the release of SG Com 5.1.
We previewed SG Com 5.1 at GDC in San Francisco with a popular interactive demo. In this demo, attendees were invited to speak into a microphone, and their voices drove both a speaking character and a reacting character at the same time, both from the same voice. The characters' emotional responses to the voice could be changed according to the desired interaction. We demonstrated this with both anime characters and MetaHumans. You can see some clips in the video below.
Find out more about what's new in SG Com 5.1 here.
Rapport: Democratizing Digital Character Animation
With the rapid growth of the audio-driven animation market, we’ve identified the need to provide a more accessible animation platform for customers wanting to create engaging digital experiences. This is where Rapport comes in. Rapport provides a complete tool suite for anyone to create, animate, and deploy real-time digital character experiences.
Rapport has really taken off in 2024. We’ve seen international collaborations with celebrities and multinational companies as well as the launch of our self-service product (which you can try for free here). We’ve explored how gaming companies can reuse their high-quality assets to interact with fans and communities, and we’ve broken into new sectors. You can read our full review of Rapport in 2024 here.
Awards and recognition
As the year came to a close, we were honored to be nominated for our third TIGA Award, having won the last two years running. It’s a testament to the passion and dedication of our team, and we couldn’t be prouder. On top of this, games we’ve had the privilege to work on have been nominated for 16 awards across 10 categories at The Game Awards. It’s an amazing recognition of the creativity and innovation that drives the projects we’re a part of.
Summary and Plans for 2025
Speech Graphics has had a groundbreaking 2024, pushing the frontiers of AI-based character animation with never-before-seen features like audio-driven breath and laughter, while continuing to improve integrations, control, and core lip sync quality. We’re committed to intensive, customer-driven R&D and to providing value in new sectors through Rapport.
We have lots in store for 2025, including an extended SGX Unreal integration, new features to give users more control over the character setup process, a revolutionary new Animation Styles selector, and new emotional AI.
- Gregor & Michael
