SOLARIA (Synthesizer V Studio)

SOLARIA	SynthV Studio	SynthV Studio 2	|	Originals	|	Gallery

SOLARIA (Synthesizer V Studio)
Product Information
Company	Eclipsed Sounds, LLC
Release Date(s)	January 19, 2022
Status	Released
System Requirements	Windows, MacOS, Linux
Base Language(s)	English
Type	Synthesizer V Studio (AI)
Licensing	Proprietary
Affiliations	Dreamtonics Co., Ltd.
Product Page	Physical copies: Eclipsed Sounds (Limited stock) Digital copies: Eclipsed Sounds Dreamtonics (International) Dreamtonics (Intl.) (Bundle Set) Dreamtonics (Mainland China) Dreamtonics (Mainland China) (Bundle Set) Lite versions: Eclipsed Sounds

This article is about the Synthesizer V Studio software known as a voice database. If you are looking for the Synthesizer V character then click here.

History

2018

Production for Eclipsed Sounds, LLC's first vocal synthesis project began in late 2018. There have been discussions between the team members ranging from several hours at a time to a few times a week, with countless hours outside of the meetings, to try to produce a virtual vocalist while wondering if they can really produce one. CEO Elijah Damon (a.k.a. "khanadee") met singer-songwriter Emma Rowley, the virtual vocalist's eventual voice provider, at a local recording studio they both worked at by chance, where Elijah was cited to be assisting. He had pitched the concept of voicing a database to her, to which she became interested immediately and was excited about the concept of vocal synths, but was too busy to officially sign on and begin work. Elijah had also sent his first semi-informal inquiry about the project to Dreamtonics Co., Ltd. around the same time. By then, the vocalist did not have a final name or design yet, but the initial premise was to have Emma Rowley be the voice provider. Very loose concept sketches were illustrated, and by the end of the year, the project was only a vague idea.

2019

The project began in earnest half a year later. By that time, there was possibly a voice provider, a basic character idea, and communications with Dreamtonics, with the possibility of developing for Synthesizer V, which was still a very new voice synthesizer. In July, Elijah invited Niamh Easter (a.k.a. "niamhsboba"), Taylor Hennessy (a.k.a. "celestrai"), and Irunabara, to whom he explained the state of the project. It was decided to re-register Elijah's LLC "Eclipsed Sounds", which was originally a company that created authorized, fan-produced merchandise for other vocal synthesizers, for the business side of the project and focus on logistics planning. Taylor was decided to be the lead developer and outreach manager, Niamh was the lead designer, and IrunaBara was internal consulting. The team committed time to research how to best meet the vocal synthesis community's needs.

On November 26, there were hopes that the vocalist would make reference to the company's "eclipse" theme and base her on the sun. During this time, the company discussed long-term business plans and crowdfunding as a possible way of completing the voice database. While the project was based on the goal of securing Emma Rowley as the voice provider, she was still busy, thus affecting the plans to ensure that the project's timeline would work with her schedule. Ultimately, she was chosen due to Eclipsed Sounds' desire to ensure the highest quality vocal they could create. In what later became of Synthesizer V AI, her voice would result in a very natural singer with her professional training and experience.It was noted that Emma Rowley didn't mind that the vocalist would sound like herself and hoped that others would use vocal synths. In addition, Irunabara sketched an initial concept for the character.

In December, Emma Rowley's schedule was still booked. The company instead focused on further developing the character design, with some refinements by Niamh. The team was aware that in vocal synthesis spaces, character designs for vocals were a sort of tradition and helped imply that the software was a vocalist. The appearance at the time became more of a "sun goddess".

2020

In February 2020, the Eclipsed Sounds team decided on crowdfunding as the way to procure funding for the vocal database production. It was considered to be a more challenging option than the alternatives, but with a great deal of research and preparation, the team believed it was possible. They studied successful crowdfunding of the past, including fan projects within the vocal synthesis community, other software crowdfund, and other projects that stood out. In addition, there were discussions of creating a test voice on a free vocal synthesis engine for the crowdfund, but the team decided against it to avoid misconceptions during a critical phase.

In March, the COVID-19 pandemic affected the United States, which slowed down production as the team sorted out their personal lives and living situations. During this time, the character mascot concepts were further developed by Niahm and Taylor, and it was decided to change the name of the project as the internal names were not final. The project's name was changed to "SOLARIS" after Niamh shared a list of suggested names.

In April, Emma Rowley reached back out to Eclipsed Sounds, confirming her participation as the voice provider. However, she was new to the concept of vocal synthesizers and was unsure of the recording process, though these did not deter her enthusiasm. By this time, Eclipsed Sounds' status was reinstated as an LLC within the state registry. The crowdfunding plan was being improved upon and the design concepts were further developed again. During character-intensive meetings, it was decided that many aspects of the character's appearance would be based on Emma Rowley. Canon character information was in consideration but was scrapped to avoid the core of the project's flexibility for the users. At this time, the company's logo was being worked upon. Near the end of the month, after deciding on the character concept, Eclipsed Sounds searched for a potential illustrator. They wanted an art style that would blend realism with fantasy and exaggerated elements to reflect how SOLARIA would blend the realism of an AI vocal with "the endless possibility" and user control of Synthesizer V. With this in mind, they looked for active voice synthesizer community members who may suit this concept, resulting in possibly selecting Superstellar as the illustrator. She was noted to have a high level of technical skill, artistic flexibility, and a natural stylistic fit.

In May, the team was able to plan the timeline and business strategy and was meeting as much as possible to work on the project, often around three to four times a week for three or more hours. While waiting for information and correspondence from Dreamtonics, the meetings were used to make decisions about business logistics, and create and transfer financial accounts, with most of the team contributing their own funds to establish a starting asset base. Toward the middle of the month, the team received information about the second release of Synthesizer V (known as Synthesizer V Studio), including an early look at the plans for the new version. The team also learned that they missed the first wave of vocal production and that the project would not progress over the time period that was initially expected. This gave the team an opportunity to continue to work out every aspect of the project before launching. With information about the cost of production, the team was able to plan in greater detail, including the creation of their first full draft of the crowdfunding campaign and sending the detailed pitch about the project (character mascot ideas, voice provider samples, 60-day crowdfund promotion timeline, and initial analysis of the market) to Dreamtonics.

In June, after the pitch was shared and reviewed by Dreamtonics, Eclipsed Sounds worked on expanding it and researching further logistics. Around the end of the month, Dreamtonics officially announced Synthesizer V Studio, and Eclipsed Souns had their first live meeting with Dreamtonics founder Kanru Hua to refine the pitch. Superstellar, the vocalist's key illustrator, also officially signed onto the project, and the near-final version of the design was completed.

In July and August, the crowdfunding pitch was further expanded to include details about the company's plans. The company logo was also refined. This period was cited to be a highly productive period for internal planning. In early July, the "Market Analysis and Marketing Plan" document was completed with five drafted pages of outline and detailed financial market analysis of the English vocal synthesizer user base and potential user base.

In September, the character mascot concept was finalized and the reference sheets were sent to Superstellar for the official artwork. By this time, the research and planning for the Market Analysis document was completed, which grew to a seventeen-page document with every consideration made in the process of creating the project, in addition to a spreadsheet with the current public interest in synthesized English vocals to pinpoint the best marketing strategy and vocal tone to aim for. The Eclipsed Sounds website and additional social media accounts were created along, with the team devising social media strategy documents. Small character merchandise exclusive to the crowdfunding campaign was also decided on, with the team carefully choosing the artists they wanted to work with based on the company's public image goals. The crowdfund planning document was also revised.

In October, it was decided to move forward with Dreamtonics' suggestion of creating a Synthesizer V test voice before proceeding with the project to ensure the team was getting the vocal tone they intended. At the time, the plan was to have a Standard style voice database, so the test voice would be a small subset of the number of recordings required for the full version. The test was to purely judge if Emma Rowley's voice would be processed well for the engine. The test was slated to begin later in December as Emma Rowley was to be married on December 5. Toward the end of the month, sketch drafts of the official illustration were given to the team by Superstellar. On October 30, Synthesizer V AI was introduced to the public.

In November, the team decided to make an AI voice database the primary goal rather than have it as a stretch goal after they had first heard about its production. Emma Rowley had responded to the team about the test voice library schedule and was able to set aside time for the recordings. At the end of November, the recording session began and the strings were recorded at A3 as it was about the average pitch of Emma Rowley's original song "Miles", which was expected to be used as the crowdfund launch's demo song. Unexpectedly, the draft campaign on IndieGoGo was slated for deletion due to not launching within a certain number of days since creation, which the team had hoped to use it to fine-tune the crowdfund over time. Everything was moved to a document to save the work.

The final set of recordings for the test library was delivered to Dreamtonics about one week after Emma Rowley's wedding. There was a total of six recorded takes of the test voice script. Further discussions about choosing the AI type of voice database over Standard commenced, with talks about how AI was more cost-effective. Standard voice databases must be recorded in a recording studio to maintain quality, while an AI voice database could be recorded from home with sufficient noise processing. Superstellar delivered the nearly completed key artwork around that time.

2021

Q1

In January 2021, Dreamtonics was handling the configuration for the test voice database while Superstellar was completing the key artwork. Eclipsed Sounds, in the meantime, had been working on the website, logo, and crowdfunding plan. On January 21, the test voice database was delivered to Eclipsed Sounds, which was only to be used for internal evaluation and was designed specifically for "Miles", with notes that proper pronunciation was not guaranteed for lyrics and musical contexts outside of the song. Eclipsed Sounds evaluated the quality of the sound and determined that tone was good enough to proceed. They later contacted MystSaphyr, Edlinklover, and staySOCKY to inquire about illustrating for crowdfunding reward merchandise.

In February, the decision to create an AI voice database over Standard was finalized. The team also decided to use the test database for the crowdfund's promotional purposes. The first draft of the crowdfund launch video was created and the first drafts of the reward merchandise were also delivered. February was spent fine-tuning the promotional timeline, which was used as a "master document" for creating a social media strategy and planned promotional works. There was also progress in updating the logos and section banners for the crowdfunding page. Taylor was cited to be the designer for SOLARIS's logo.

In March, Superstellar updated SOLARIS's facial proportions to better suit the voice after listening to a sample of the test database. staySOCKY had completed their acrylic charm design while Emma Rowley provided the instrumental for "Miles" for its use as a demo. The Synthesizer V project file for the song was also completed. The Eclipsed Sounds team then created stickers as a crowdfunding reward, then wrote up the tier descriptions for the crowdfunding plan. In mid-March, an in-depth meeting with Kanru Hua commended with discussions about the crowdfunding plan, which was found to be productive and clarified the plans. The crowdfund was launch date in mid-May, which was later than the initial plans, though the team saw it as extra time to polish the campaign and promotional material.

Q2

In April 2021, the team decided to create multiple short demos with the test database, with intention of posting them during the crowdfunding period on TikTok. Many of the demos were never published as too many demos of the test database may lead to misconceptions about the final voice. By the end of the month, initial teaser posts were ready for the primary social media account.

In May, Eclipsed Sounds renewed the website's domain name and spent time reviewing a near-final draft of the campaign video. On May 8, Niamh conducted a promotional interview with Emma Rowley, which was to be posted on YouTube, and final campaign edits were made. On May 10, Eclipsed Sounds officially revealed a small sample of the upcoming female Synthesizer V Studio vocal initially introduced as "SOLARIS". On her webpage, the voice provider and illustrator were revealed to the public. According to the company, the demo used a test library based on the Standard recording method and was created with a very limited recording list, meaning that it would only give listeners an idea of what SOLARIS would be like. The crowdfunding campaign was confirmed to begin on May 14. This campaign would allow the library to be released using the AI synthesis method, resulting in SOLARIS becoming a voice database for Synthesizer V AI. The company's goal with this voice database was to have SOLARIS sound closer to her voice provider in singing style, skill, and tone. This especially would be made possible with the capabilities of AI training.

In response to fans' questions about the production, Eclipsed Sounds noted that the main focus was to create an AI voice database for SOLARIS. Depending on how her crowdfunding campaign proceeded, a standard voice database could be a possible option for a stretch goal depending on the demand. When asked about a Lite version, it was noted that there were no plans. They stated that sales from SOLARIS' completed voice database would help them do things they were hoping for with no clarity of what this meant. SOLARIS was confirmed that she would be using the ARPAsing method like Eleanor Forte as this was the standard for Synthesizer V English. In regards to physical copies, it was confirmed that a limited edition physical box would be made available as a crowdfund backer reward. As Eclipsed Sounds was a small company, there was a possibility that they may not have the resources to create boxes for long-term. On May 11, a long version of "Miles" was uploaded and SOLARIS' design was revealed. At that time, Circus-P and GHOST signed on to create demos. It was noted in the production diary that the team had wanted to invite producers for demos earlier on, but were concerned about the possibility of leaks. On May 13, the funds recipient page for the crowdfund, which was indicated as completed days earlier, showed an error in authentication which could take several days to re-process. Hours later, Elijah and Taylor managed to fix the issue in time for the launch of the campaign.

On May 14, SOLARIS' crowdfunding campaign officially started and more information about the project was revealed. It was revealed that up until the campaign's launch, everything for the SOLARIS project was funded out of pocket by the team members. In order for completion, funding from outside sources was a necessity. The initial goal was shown to be $40,000, which would allow the team to create the AI voice database for SOLARIS in conjunction with Dreamtonics, payment toward the voice provider, create detailed key artwork with Superstellar, produce the crowdfund rewards, pay the guest artists (Edlinklover, MystSaphyr, and staySOCKY), and commission original demo songs from producers such as Circus-P and GHOST. The pledge rewards included:

• Key art rewards: digital wallpapers, a button badge, and an acrylic stand by Superstellar
• SOLARIS team rewards: PDFs of the concept art and stickers illustrated by the Eclipsed Sounds team
• Guest artist rewards: a button badge by Edlinklover and acrylic charms by MystSaphyr and staySOCKY
• SOLARIS Synthesizer V Library Rewards: the digital edition or the limited physical edition
• Limited item rewards: a tote bag and a t-shirt by Superstellar
• Add-on item rewards: a production diary, a custom voice line (similar to exVOICEs), and an autograph by the voice provider

Stretch goals included improving the quality of SOLARIS' singing capabilities with more AI Training data and additional key artworks. Should the campaign result in incomplete funding, the team would attempt to gather the remaining necessary funds, which would result in delaying the rewards and library completion. If this could not be done within six months from the final day of the crowdfund, pledges would be refunded.

A projected timeline was also made available. The illustrations of the rewards from the guest artists would be revealed later on May 16, 18, and 21 respectively. A Q&A stream would also be launched on May 21. A second Q&A livestream would be available on June 5 and a third on June 28, which was when the crowdfunding campaign was scheduled to end. The digital rewards (excluding the voice database) would be distributed in August and the physical merchandise would be distributed in December. The digital software rewards would be distributed in January 2022 and the physical edition rewards would be distributed in February.

Regarding the idea of switching to a Standard voice database in the event of incomplete funding, Eclipsed Sounds confirmed that there was no intention. They explained that for Standard voice databases to work, voice providers needed to fly to Japan to record at the Dreamtonics Studio, which was not an ideal option. In regards to additional samples with the test voice database, it was noted that there would be no further full demos using it. Genres for future demos would depend on how the AI voice database would turn out, though the company hoped to have a variety of examples. Regarding the terms of the license, users can use the voice database freely for commercial use, but for character use, the user should contact Eclipsed Sounds if the profits exceeded $1,000. When asked about other languages for SOLARIS, Eclipsed Sounds stressed that there were no plans. Due to the desire of having SOLARIS be as high quality as possible, in addition to Emma Rowley only knowing English, the company did not want there to be any mistakes in pronunciation and risk the quality. There were no current plans for a starter pack as this would need to be negotiated with Dreamtonics. SOLARIS would be released with Synthesizer V plug-ins for the Pro studio. Within the first day of the campaign, the project raised 25% of the goal. On May 21, Meltberry was signed on as a demo producer.

On May 23, Eclipsed Sounds asked fans for ideas about rewards for a higher tier due to popular demand. New low and high campaign tiers and the physical box were confirmed to be revealed in the second livestream. An interview with Emma Rowley was released on May 30, giving insight about her life as a musician and detailing her thoughts on the concept of voice synthesizers.

In June, Flanger Moose signed on as a demo producer. On June 12, two new tiers for the crowdfund were added: the "SOLARIS Ultrafan" and "Exclusive Merchandise tiers. The physical box was also revealed. For users who were unable to contribute to Indiegogo, a PayPal contribution link was opened.

On June 24, SOLARIS was fully funded with 2.5% of the proceeds coming from PayPal, 82% from Indiegogo, and 15% from a private investor. On June 27, the Indiegogo crowdfund reached 100% of its initial goal, bringing the total received funding to 120% A special "Final Synthesizer V SOLARIS Livestream!" was held following the successful crowdfund of which several announcements were made. The first was the announcement of the producers who would be writing demo songs which included: Meltberry, Flanger Moose, Circus-P and GHOST. Towards the end of the stream, Eclipsed Sounds announced that SOLARIS' name was changed to SOLARIA. This was due to a variety of reasons. The primary reason was because of an "abundance of other characters, IPs, and products named "SOLARIS", and to quickly evoke the idea of a singer on name alone ("SOLARIA" being a portmanteau of "solar" and "aria")".

Q3

Sometime in July, the team had a glimpse at the first draft of Flanger Moose's demo and its video illustrations by irarugii. GHOST delivered their first work-in-progress. After many requests from fans, Superstellar was commissioned for a character reference sheet, which was made available to the public on July 29..

In August, the recording process was scheduled to begin but was delayed due to Emma Rowley contracting COVID-19, which did not allow her to record for a large portion of the month as she needed to recover. The team received the final DVD case guidelines for SOLARIA's physical copy, allowing them to move forward with producing the design for the DVD insert, but was put on hold until they knew the capabilities of the final voice database.

On August 20, a user asked about the possibility of a Standard voice database after the crowdfund finished. Eclipsed Sounds responded that they were focusing solely on AI and may consider Standard if there was a high demand in the future. On August 30, Eclipsed Sounds posted an update on the crowdfunding page that the digital supporter rewards were being distributed. They also confirmed that the data collection phase was completed in Mid-August and that they were currently processing the recordings sourced from Emma Rowley. They noted that this may take some time to be completed.

Q4

In September, the recordings went through an extremely thorough noise removal and audio clean-up process led by Taylor. This took a lot of time due to the number and length of the samples. Eclipsed Sounds met with Emma Rowley's agent to discuss the final terms for the database and have a formal contract written out for future work, such as her signatures on the physical editions of the voice database.

On October 7, it was announced that there were two updates with the physical backer rewards. The small physical rewards were noted to be progressing smoothly and would likely be shipped in December. After receiving multiple requests from backers, Eclipsed Sounds decided to update the design to use an extra-large DVD case to match physical editions of other Synthesizer V software (particularly those distributed by AH-Software Co. Ltd.). As for the voice library itself, it was noted that the audio clean-up phase was nearly completed in mid-October, with the team ready for the next stage at the end of the month. Demonstration songs were also progressing in development and were expected to release before the end of the year. Eclipsed Sounds hoped to share the results sometime soon. Shortly after, Circus-P showed his first work-in-progress for his demo. Meltberry's demo was also shown to the team for the first time.

On November 16, Eclipsed Sounds confirmed that development had been continuing smoothly over the previous weeks. They noted that this may be the final time people would hear from them before the samples arrived. The work had been split as evenly as possible for each team member and team meetings were frequently held to write and revise configuration reference documents. Every sample was configured and then checked by another team member, which, at the end, Taylor would check for consistency before they were delivered for AI processing. The voice database sample set-up was completed by the end of November and was sent to Dreamtonics to initiate AI processing. The timeline was updated accordingly:

• December: Expected to have small physical rewards shipped and coupon codes to be distributed. Physical backer rewards may have shipments delivered in January or February 2022. Most of the demonstration songs were expected to be posted.
• January: All software backers to receive their copies of SOLARIA. Both digital and physical supporters would receive digital copies and activation codes. SOLARIA was expected to be available on the company website, marking this as her expected overall release date.
• February: Final physical packages to be mailed for physical supporters along with the entirety of their rewards.

On December 7, the team heard the first sample of the completed AI database and began testing it to ensure it met their standards. It was noted that the default Synthesizer V English dictionary did not perfectly fit the way Emma Rowley would naturally sing certain words, thus initiating the need to produce a custom English dictionary for the voice database. The team manually checked the 3,000 most common English words for any stand-out issues to increase the coverage of the dictionary. On December 8, SOLARIA's AI voice database was shown to the public for the first time. Further demo songs were expected soon afterward. On December 9, it was announced that the website underwent an update, now featuring a new Timeline page and an FAQ page. According to the FAQ, the team was still determining the final price for SOLARIA, but confirmed she would cost less than $100. Regarding if there would be any more Synthesizer V voice libraries after her, Eclipsed Sounds expressed that they would love to do so, though it would depend on how well SOLARIA performed in sales.

On December 12, Emma Rowley was able to listen to the initial AI sample and was impressed with the results. A few days were spent revising the initial decision and confirming the creation of SOLARIA's Lite version.

On December 15, it was announced that SOLARIA would be released on January 19, 2022, and would be purchasable through Eclipsed Sounds' website. Digital backers would receive her at least one week prior. Commercial sales would be digital-only, with no plans to sell more physicals at that time. A Lite version would be made available sometime after the commercial release, which was decided after much deliberation and many requests from the community and backers. A week later, the team was made aware of the Cross-lingual Singing Synthesis update for Synthesizer V Studio and was opted in to have SOLARIA be able to make use of the new function, which would allow her to not only sing in English, but also in Japanese and Mandarin Chinese. This update was officially introduced by Dreamtonics via Synthesizer V Studio version 1.5.0 on December 23. On December 30, it was announced that SOLARIA would be available for pre-order on January 1, 2022.

2022

On January 2, 2022, it was announced that crowdfund backers who pledged for SOLARIA's voice database would be receiving their digital downloads and activation codes on January 12 at 12 AM EST. When asked about the delivery of the Lite version, it was later clarified that it would be released sometime after the physical crowdfund rewards have been fulfilled. Reason was cited that it would take more time to be completed. Around this time, Taylor was finishing the first versions of the scripts to be included with the voice database, including the revised dictionary and a Spanish pronunciation dictionary using Japanese phonetics to further expand SOLARIA's language capabilities.

Pre-orders were closed on January 18 to prepare the file distribution for all pre-order customers, who were to receive their copies a day early. Preparations for the site to switch to full commercial release were underway. SOLARIA was released on January 19 as intended.

On February 18, Eclipsed Sounds showed the physical editions which were nearly ready to be shipped in the following week. The autographed DVD inserts were still in production while the discs were still on the way due to a snowstorm delaying their production. On March 5, it was announced that the first wave of physical crowdfund packages was ready to ship on March 7, though delays in fulfillments were expected due to the number of varied items within the Superfan and Ultrafan tiers. The physical copies of SOLARIA were expected to ship in the coming week as well. The remaining digital crowdfund rewards were expected to be delivered on March 10. SOLARIA Lite released on March 26.

On May 6, when asked about adding Vocal Modes to SOLARIA, Eclipsed Sounds confirmed that it was currently in progress and that they hoped it would be completed soon. On May 24, it was announced that the Vocal Mode feature for SOLARIA was nearly ready. Two were revealed and confirmed to be "Soft" and "Solid". On July 2, it was confirmed that SOLARIA would have seven Vocal Modes in total, with the remaining ones confirmed to be "Airy", "Power", "Clear", "Passionate", and "Light". Her Vocal Mode update was expected to release within the next two weeks. On July 5, Eclipsed Sounds announced that SOLARIA's Vocal Mode update would be available on the following day, July 6.

On July 21, SOLARIA was one of the AI voice databases to receive an update for HDVM compatibility (a feature introduced in Synthesizer V Studio Pro 1.7.0, also released the same day).. However, some users complained about unstable voice-quality results when re-rendering in 1.7.0 projects saved under earlier versions of the editor that made use of her voice database (after updating it to v106). Eclipsed Sounds suggested users to finish their songs with her previous version before updating to the new version; they also offered installers for her previous versions (through the setup PDF document included with SOLARIA) in case users were unsatisfied with the new update.

On July 22, Dreamtonics explained that by default, the AI Retakes feature favors "rich timbre and expressions over smooth and stable vocalization", resulting in inconsistent behavior relative to previous versions; as a workaround to stabilize the voice output, they offered these steps below:

• Empty the selection.
• Open AI Retakes panel, go to Timbre tab.
• Under Takes -> Global Settings, reduce Expressiveness.

Responding to requests from users, Dreamtonics made available for download that same day rolled-back versions of SOLARIA (as well as other affected AI voice databases) by logging in to their store download website.

On July 31, it was announced that SOLARIA was made available on the official Dreamtonics International store which was helpful to customers who preferred Google Pay. A bundle pack with the Synthesizer V Studio Pro editor was also made available through this platform. She was later made available on the Dreamtonics Taobao shop on October 28 with the same options as the International store.

On November 17, it was announced that a limited stock of 45 SOLARIA physical copies was set to be available starting November 18. These were leftovers from her crowdfund and were to be sold after extreme demand from the user base and were expected to ship in February 2023. It was confirmed that this edition would not be reprinted in order to respect its exclusivity.

On March 1, 2023, Eclipsed Sounds announced that the leftover SOLARIA physical boxes would begin shipping on March 8.

Updates

SOLARIA AI v101 - January 19, 2022 

Adjusted clarifications to the license agreement.

SOLARIA AI v103 - July 6, 2022 

A free update for SOLARIA that enables her to use the Vocal Mode function. Seven variations were included: "Clear", "Soft", "Airy", "Power", "Passionate", "Solid", and "Light".

SOLARIA AI v104 - July 12, 2022 

A free update for SOLARIA that improves tone consistency with v101 when not using Vocal Mode and general sound quality improvements over v103 based on feedback from users.

SOLARIA AI v106 - July 21, 2022 

A free update for SOLARIA was released following the Synthesizer V Studio 1.7.0 update. This enables her to use the new High Dynamics Voice Models and AI Retakes features.

SOLARIA AI v108b1 - November 10, 2022 

A free update for SOLARIA was released following the Synthesizer V Studio 1.8.0b1 update. This enables her to use the new Diffusion Probabilistic Models (DPM) feature.

SOLARIA AI v108 - November 24, 2022 

Final version update compatible for Synthesizer V Studio 1.8.0 which enables SOLARIA to use the new Diffusion Probabilistic Models (DPM) feature if the beta version was not installed previously.

SOLARIA AI v109b1 - June 21, 2023 

A free update for SOLARIA was released following the Synthesizer V Studio 1.9.0 update. This implemented the new rap feature available for the English and Mandarin languages. Cantonese Chinese was added to the Cross-lingual Singing Synthesis feature as well as general improvements to the overall function, resulting in more accurate pronunciation that sounds closer to a native speaker when singing foreign words.

For SOLARIA specifically:

• Her Vocal Modes' strength and realism were increased and normalization of Vocal Mode audio were improved.
• Her base tone became somewhat stronger.
• An improved expression in her low range, now reaching D3 without quality loss.

SOLARIA AI v109 - July 6, 2023 

Final version update compatible for Synthesizer V Studio 1.9.0 which enables SOLARIA to use the rap feature (English and Mandarin Chinese only), allows for Cantonese Chinese Cross-lingual Singing Synthesis, and improves her Vocal Modes, tone, and range (as listed in v109b1's notes) if the beta version was not installed previously.

SOLARIA AI v110b1 - August 2, 2023 

A free update for SOLARIA was released following the Synthesizer V Studio 1.10.0b1 update. This enhanced pitch generation by utilizing Reinforcement Learning from Human Feedback (RLHF), translating to fewer off-key errors, making better use of vibrato, and creating more expressive transitions between notes. This also allowed for the use of the new "Enhancement" slider in the AI Retakes panel, which controlled the level of changes introduced by RLHF. For Synthesizer V Studio Pro users, feedback submission was enabled, allowing users to consensually share their preferred takes with only Dreamtonics for research and development purposes, where models trained with this feedback would be deployed in future voice database updates. Users may also opt out of this feature if undesired.

SOLARIA AI v110b2 - September 28, 2023 

A free update for SOLARIA was released following the Synthesizer V Studio 1.10.0b2 update, which was the second beta version following version 1.10.0b1 that introduced enhancements to voice databases and integrated the latest user feedback via Reinforcement Learning from Human Feedback (RLHF). Regarding the editor, numerous reported issues were addressed for a more fluid experience. A new generation of pitch models were trained using the extensive feedback from users over the past month, and the new models offered greater accuracy in voice timbre reproduction, as well as a reduction in mispronunciations. Dreamtonics also focused on issues from the previous beta release that made Vocal Modes seem less distinct on certain vocals, ensuring clearer differentiation in the 1.10.0b2 update. The latest update also introduced a new feature in the Note Properties Panel that allowed for even division of a note into multiple syllables, proving to be a useful tool for songs with faster tempos.

SOLARIA AI v110 - October 19, 2023 

Final version update compatible for Synthesizer V Studio 1.10.0 which enables SOLARIA to use Reinforcement Learning from Human Feedback (RLHF) as well as other enhancements (as listed in v110b1 and v110b2's notes) if the beta version was not installed previously.

SOLARIA AI v111 - October 26, 2023 

A free update for SOLARIA was released that added support for the Japanese language to the rap feature.

SOLARIA AI v112b1 - December 18, 2023 

A free update for SOLARIA was released following the Synthesizer V Studio 1.11.0b2 update. This implemented support for the Spanish language in the Cross-lingual Singing Synthesis feature.

SOLARIA AI v112 - April 4, 2024 

Final version update compatible for Synthesizer V Studio 1.11.0 which enables SOLARIA to sing in Spanish through Cross-lingual Singing Synthesis if the beta version was not installed previously.

Voice Database Information

Demonstrations

Synthesizer V Studio Standard

Note: It was confirmed that this voice database may not be released.

Demonstrations
Miles - Short Version	YouTube
Miles - Full Version	YouTube
From This Moment On ft. Emma Rowley - Short Version	TikTok Twitter

Synthesizer V AI

Demonstrations
Miles - Short Version	YouTube Twitter
Dawn - Short Version	YouTube bilibili
untied laces ft. Meltberry - Short Version	YouTube Twitter
Aura - Short Version	YouTube
Our Universe - Short Version	YouTube
White Winter Hymnal - Japanese & Mandarin Chinese Cross-lingual Singing Synthesis	YouTube
Our Universe - Full Version	YouTube
untied laces ft. Meltberry - Full Version	YouTube Bandcamp
Aura - Full Version	YouTube Bandcamp
Dawn - Full Version	NicoNico YouTube bilibili SoundCloud Bandcamp
Miles - Full Version	YouTube
Miles - Short Acapella Version; "Soft" & "Solid" Vocal Modes	YouTube
CYCLES - Short Version	YouTube
Erode - Short Version; "Power", "Airy", and "Soft" Vocal Modes	YouTube
If You Are An Echo ft. ASTERIAN	YouTube

Eclipsed Sounds Demos
Dawn	eclipsedsounds.com
If You Are An Echo	eclipsedsounds.com
Aura	eclipsedsounds.com

Voice Databases

• Standard (Unreleased)
• AI

Product Information

Optimum

  Range: A3

Package details as noted:

SOLARIA's test voice database used the Standard recording method. This voice database was only created to showcase what SOLARIA would potentially sound like. As Eclipsed Sounds favored the production of AI, the Standard voice database was only a possibility based on high demand or may not release at all.

Vocal traits as noted:

This library was recorded in a lower register at A3, which was selected because it was the average pitch of Emma Rowley's original song "Miles".

Phonetic notes as noted:

The test voice database used a very limited recording list. Because of this, it was not even the size of a Standard Lite library.

Product Information

Optimum

  Tempo: 80 ~ 180 BPM  Range: Sing mode: D3 ~ G5; Rap mode: G3 ~ F4

Versions

  Lite Version/FLT Available?: Yes (Lite)  Starter Pack Available?: Yes  Web Version Available?: No

Package details as noted:

SOLARIA is a feminine English voice database based on the singing style of Emma Rowley. This would produce a natural, realistic, and more human-like voice. According to Eclipsed Sounds, SOLARIA is noted to have "warm low notes, a powerful core range, and a clear high register". She is also noted to be versatile and expressive, suiting a wide variety of genres and applications.

Included in the package are the following items:

• Bonus images: 2 key artwork and reference sheet by Superstellar
• Example files: edited and unedited versions of the 4 original songs ("Dawn", "Untied Laces", "Aura", and "Our Universe") and the "White Winter Hymnal" cover.
  • Files for "CYCLES" and "Erode" are expected to be released in the future.

Vocal traits as noted:

• Before the production of the AI voice library, Eclipsed Sounds expected to have SOLARIA sound closer to her voice provider, Emma Rowley, in singing style, skill, and tone.
• SOLARIA maintains the test "Standard" library's traits in her lower notes, though clearer and stronger. With parameter adjustments, it is possible to potentially emulate the original timbre across her whole range.
• GHOST, one of the demo producers, noted SOLARIA's quality to be impressive.
• unit.0, another demo producer, praised SOLARIA's versatility, range, and dynamics.
• As per the Synthesizer V Studio 1.7.0 update launched on July 21, 2022, through High Dynamics Voice Models (HDVM), which later migrated to Diffusion Probabilistic Models (DPM) per the 1.8.0b1 update on November 10, AI voice databases may use AI Retakes to create multiple "takes" of a sung section with variations in pitch, dynamics, and timbre. This replicates the randomness of human vocalists, adds a more realistic feel, and allows users to find an ideal vocal without the need for "laborious tuning." This feature is only available in the Synthesizer V Studio Pro Editor.
  • This does not apply to the Lite version.

Phonetic notes as noted:

• Phoneme format: ARPABET
• Eclipsed Sounds confirmed that SOLARIA would occasionally vocal fry given a certain input, which is done using the "stop" [cl] phoneme, which typically performs best between vowels and low ranges. It should be noted that the [cl] phoneme would not always be vocal fry. The library performs a fry situationally based on what was learned from the singer's input.
• Through Cross-lingual Singing Synthesis, SOLARIA AI is able to sing in not only English but also in Japanese and Mandarin Chinese. Since the release of her v109b1 update on June 21, 2023, she can also sing in Cantonese Chinese with this feature; since her v112b1 update on December 18, she can also sing in Spanish with this feature. Cross-lingual Singing Synthesis is only available in the Synthesizer V Studio Pro Editor.
  • This does not apply to the Lite version, thus restricting the voice database to English only.
• A total of 34 voice lines and 227 breath samples were included with SOLARIA's voice database.
• Bonus dictionaries include the 1.0 dictionary, which fine-tunes SOLARIA's pronunciation and was tested with 3,000 English words, and for Synthesizer V Studio Pro, there is a dictionary that allows SOLARIA's Japanese (through cross-lingual singing synthesis) to pronounce Spanish words. Dictionaries to support German and Italian via English phonetics were added on December 31, 2022, though these were noted to be created programmatically and would not be perfect in every circumstance.
• Bonus scripts include 4 tools for Synthesizer V Studio Pro users: changing track color, English phoneme reference, phoneme length & strength mass editor, and mass phoneme replacer.
  • On August 4, 2023, Eclipsed Sounds acknowledged that the scripts were not compatible with macOS Ventura (version 13+). A replacement would be investigated as soon as possible.
  • With the inclusion of track colors as of Synthesizer V Studio 1.10.0b1, Eclipsed Sounds announced that they would discontinue their own track color script and remove it from download once the final version of 1.10.0 was made available.
  • New bonus files that were planned for distribution with SAROS would be added to SOLARIA's and ASTERIAN's distributions.

Vocal Modes as noted (Synthesizer V Studio Pro Editor only):

• As of July 6, 2022, the paid version of SOLARIA became compatible with the Vocal Mode function introduced in Synthesizer V Studio 1.6.0. Her v103 update included seven variations:
  • "Clear": Provides a bright voice that helps SOLARIA's clarity in mixing. When mixed with other modes or her default voice, it can be used to naturally raise the brightness of her tone without relying on effects. Suggested combinations include "Power" to reduce noise, "Passionate" to provide a boost, and "Solid" to create a crisp, strong sound.
  • "Soft": A gentle and mellow tone that strips out the natural power in her voice and gives the users discrete control over her expression. Also has added breathiness and darkness to her voice. Suggested combinations include "Airy" for the softest combination and "Passionate" to enhance the natural tone while pulling back from purely soft.
  • "Airy": A whispery mode that provides a unique tone. While similar to "Soft", it removes the support from her voice and leaves only a light and gentle quality, creating unique effects as backing vocals or a new layer of expression for more atmospheric vocals. Suggested combinations include "Soft" for the softest combination and "Clear" to create a bright and whispery tone, adding additional texture when used for harmonies.
  • "Power": Brings out her belting ability directly, easily lending a strong and emotional voice. Suggested combinations include "Solid" for the strongest sound and "Clear" or "Passionate" to reduce noise.
  • "Passionate": Is closer to her default tone and is based on the most emotional lines of the songs in her data, increasing her natural emotional expression. Suggested combinations include "Solid" to create a very strong and emotive tone, and a combo of "Soft" and "Airy" to bring her emotive sound to a very soft tone.
  • "Solid": Packs energy into every syllable and helps bring out a different side of her strong tones. When mixed with other modes or her default voice, it can add a base of support to add more "punch". Suggested combinations include "Power" for the strongest combination and "Passionate" to enhance tone while pulling back from purely soft.
  • "Light": Based on the upper part of her range, typically described as the "head voice". Provides a unique tone across the range with bright high notes and solid low notes. Best used for songs in the higher register. Suggested combinations include "Airy" to create a believable falsetto tone, "Soft" to create a more ethereal sound, and "Clear" to make a very clear and present sound.
• On July 6, regarding the v103 update, users noted that the sound quality overall seemed to have decreased. The "Power" Vocal Mode was also noted to have a reduction in volume. Eclipsed Sounds acknowledged this and noted they were undergoing an investigation regarding the change. In regards to "Power", they were looking to improve this Vocal Mode to require less user adjustment in the future.
  • The v104 update released on July 12 improved upon the tone consistency with v101 when not using Vocal Modes and included general sound quality improvements over v103. At the time, Eclipsed Sounds planned to address Vocal Mode volume changes in a future update, stating that they could be circumvented in-editor by using the track volume slider, loudness parameter, or mixing with other Vocal Modes.
• For rap mode, Eclipsed Sounds suggests users to use her "Solid" vocal mode in low amounts. Increasing the "Tension" parameter can also make her rapping tone more believable.

System requirements as noted:

• OS: Windows 11/10/8.1 or later, Mac OS X 10.11 or later, Ubuntu 18.04 or later (64bit)
• RAM: 2GB or higher
• CPU: Intel Core i5 or equivalent AMD processor
  • For editing AI in high quality mode: 4th Generation Intel Core i5 series or later, AMD Athlon X4 845 or later, including Ryzen

License agreements as noted:

• Synthesizer V SOLARIA License Agreement - English; as of July 2, 2022
• Regarding the Lite version:
  • SOLARIA Lite may not be used commercially in any way
  • In uploaded works using the Lite version, the user must indicate in the upload title or prominently in the description or associated text that they used "SOLARIA Lite"
  • Users may not attribute this vocal to any other source other than "SOLARIA Lite" or "Synthesizer V SOLARIA Lite"

Voice database sample

• SOLARIA AI Solfège:
  
  • Vocal Mode Solfèges:
    • "Clear":
      
    • "Soft":
      
    • "Airy":
      
    • "Power":
      
    • "Passionate":
      
    • "Solid":
      
    • "Light":
      
• SOLARIA AI Lite Solfège: