|🛠||This subject is work in progress.
Please bear with us while improvements are being made and assume good faith until the edits are complete.
For information on how to help, see the guidelines. More subjects categorized here.
|📰||This subject requires intervention. What is being worked on: Please note the details of this software are subject to change without notice.
For information on how to help, see the guidelines. More subjects categorized here.
VOCALOID5 (commonly shortened to "V5") is the current successor to the VOCALOID4 engine by the company YAMAHA Corporation.
The software was released on July 12, 2018.
The first hint that VOCALOID5 was about to be released appeared in late June 2018 when the announcement of several VOCALOID retirements were being made at the VOCALOID SHOP. Many took this as speculation as to preparation for a new upcoming version of the engine, as this happened when VOCALOID3 was replaced with VOCALOID4, and several things occurred including a website overhaul.
The first official "leak" also appeared in the form of Haruno Sora, a VOCALOID/VOICEROID by AH-Software Co. Ltd., whose Amazon description referred to her as a "VOCALOID5" vocal. The release date was noted to be July 26, 2018, indicating that VOCALOID5 may be released before or on the same day.
VOCALOID5 was later released on July 12, 2018, hours after both the website updated and Sora was leaked. This new version introduced 4 new vocals: Amy, Chris, Kaori, and Ken. VOCALOID5 was confirmed to release in two versions: Standard and Premium. Standard included the new 4 vocals while Premium included the same vocals in addition to CYBER DIVA II, CYBER SONGMAN II, VY1, and VY2. Amy, Chris, Kaori, and Ken replace the VY series and act as the new Yamaha standard vocals.
An upgrade offer was allowed for all versions of VOCALOID3 and VOCALOID4 to the new VOCALOID5 version, including VOCALOID NEO and VOCALOID Editor for Cubase. In the past offers had only been given to direct upgrades from each engine version to the next.
- Ver.5.0.2 was a general stability fix as well as fixing a few bugs that either cause crashing or glitches.
- Ver 5.0.3 was a fix for stability and performance improvements as well as for correcting crashing issues.
- Ver.5.2.0 was a major update to the engine. The most notable updates were the introduction of multicore / multithreading part rendering and improved processing skill of Singing Skill. It also introduced new features to the media parts including saving the key and tempo, saving .WAV files as media, and transposing non-wav media files into different keys, adding media by double clicking and sampling it by single clicking. It is also possible to download new media if available from the program's startup. Some phrases were also renamed. It improved tag narrowing for searching and added a reset button to clear tag navigation. Functionality for lengthening and shortening phrases and parts by either double or half tempo were added. For the VST/AU ReWire version for DAWs it added support for changing song position, turning on/off loop, and adjusting the loop range. 
- Ver.5.2.1 provided general improvements to stability.
- Ver.5.3.1 gave some supplementary abilities while editing and stability improvements.
- Ver 5.4.0 brought some improvements to existing functions and to stability, as well as stability improvements for the 4.5 Editor for Cubase.
- Ver 5.4.1 acknowledges the engine is now compatible with macOS Catalina and brings some more stability improvements.
- Ver.5.5.0 added a Part/Note Duplication function, improved keyboard note selection and editing scrolling process, revised the activation process and fixed some stability improvements
- Ver 5.6.0 improved playback processing, and made it possible to gradually listen to parts from the beginning while rendering; also brought some stability improvements. 
- Ver 5.6.1 fixed a bug that was introduced on the previous update, which caused some Spanish phonetic symbols not to be converted properly. 
- Ver 5.6.2 fixed a bug of sound cutting off at the end of the first part when VOCALOID part was used continuously over two parts in the same track.
Windows 7 Support End
On the 1st Jan 2020, Yamaha announced Windows 7 support would end. This impacts users with that version of the Windows OS.
- OS: Windows 10, Windows 8/Windows 8.1, Windows 7, OS X 10.13, macOS 10.12, macOS 10.11 (64bit only)
- CPU: Intel multi-core processor
- RAM: 8 GB recommended (minimum 4 GB)
- Standard: At least 20 GB
- Premium: At least 35 GB
- Other: Audio device, internet connection (for authentication, deauthentication, software updates, etc.)
- Monitor size:
- Minimum operating environment: 1366 x 768
- Recommended environment: 1920 x 1080
Additional installation notes
Note these requirements only apply to the VOCALOID5 engine itself. Users need to check each VOCALOID package and its requirements before installation as some variation may occur between releases. When used as a VST/AU together with a DAW, the operating requirements of the DAW must also be met at the same time.
The license, much like in VOCALOID3 + NEO and VOCALOID4, still allows the user to install on either Windows or Mac, though they can only have one or the other. The software must be uninstalled from its current OS to be able to install on the opposing one. However, note that regardless 1 license of VOCALOID5 is given upon purchase, therefore VOCALOID5 cannot be installed on multiple computers at the same time. This is the same for all voicebanks as well as the engine so while the user can have the installation on either Windows or Mac, they cannot install it on both or multiple computers.
VOCALOID5 installs according to the version of the OS, meaning an English OS will result in the VOCALOID5 interface being in English when installed. Users wanting the software in any other language have to change their language OS to force VOCALOID5 to be installed in that language. This is different to VOCALOID3 and VOCALOID4 where the language was selected upon installation.
Ken is the default vocal for VOCALOID5 and the engine will always revert back to his voicebank if an issue occurs, though this can be changed to other vocalists if the user doesn't want him to be the default vocalist.
Announced Vocals for the VOCALOID software. Names and images presented are placeholders and may not reflect the final product when complete.
We urge readers be considerate and careful when supporting unconfirmed projects, especially if they ask for financial endorsement.
In October 2018, a new offer was put up for "Vocaloid5 Plus". This was a 5% discount for either version of VOCALOID5 plus one of the 70 voicebanks offered for sale on the VOCALOID SHOP from the VOCALOID3,VOCALOID4 and VOCALOID5 range. This did not include upgrade offers and was available only to the Japanese version of the site. Any Vocaloid bought this way is considered a "Starter Pack" release.
On the English version of the VOCALOID SHOP, there are starter packs available with either the Standard or Premium version of VOCALOID5 included. These starter packs can be bought with voicebanks such as Fukase, Macne Nana English, and Megpoid English.
A upgrading campaign was announced in April 2019. This allowed anyone who had a 16 digital serial code to get VOCALOID5 cheaper. The campaign allowed any VOCALOID2, VOCALOID3 and VOCALOID4 product to get the discount, and applied for even voicebanks Yamaha often did not normally attach promotions to such as Sweet ANN or Prima.
VOCALOID PHRASE PACK SERIES
In December 2018, Vol.1 of a new series called VOCALOID PHRASE PACK SERIES was released. This expanded the phrases VOCALOID5 contained within its library. Providing the user had the VOCALOID5 engine, these would be released for free. They are similar to the past VOCALOID-P data series and like the previous series are supportive examples of Vocaloid capabilities.
A second pack was released on the 22nd of April 2019.
Examples of usage
To contribute an example- see this blog entry to download the VSQx.
|CYBER DIVA II||CYBER SONGMAN II|
|Haruno Sora (Natural)||MEIKA Hime|
For a list of VOCALOID5s new parameters see Parameters
As mentioned, all new purchases of VOCALOID5 come with 4 vocals: Amy, Chris, Kaori and Ken. This is a first for any new version of the engine. Outside of "Starter Pack" versions, Vocaloid has always come with no vocals included with the standard editor. Therefore, any new VOCALOID user can get started without having to purchase additional vocals. This also means every user who purchases VOCALOID has access to both English and Japanese voicebanks.
In addition, the VOCALOID software now comes with 1,000 lines of pre-set text for both Japanese and English voicebanks and 1,000 audio samples. In terms of the sample text, the text given will be dependent on the vocalist. For example if Amy is selected then only the English phrases will appear. But if Ken is selected, only the Japanese ones will appear. These are mostly used to easily create loops.
10 pre-set "singing styles" are able to be used and so are 11 audio effects. When choosing the singer, a list of pre-set "types" (including "Breath", "Looped", "Robotic", etc) appears. The "Types" can then be coloured to give an expressive tone (this function is similar to the "colour" adjustments made via E.V.E.C.), such as adding Growl, Brightness, etc. This allows all vocalists within VOCALOID5 to produce a unique tone suitable to the user's needs. Note that as with any such function (see E.V.E.C.), some vocals do not handle some combinations; one may handle "Robotic" better than another. Trial and error is involved like anything else and recommended tempo/vocal range still play a part in results. Many of these are similar or reused from the VOCALOID NET service.
ASIO can now be used as well as an audio stream source, if the audio device supports it. ASIO will result in lower playback latency and smoother audio playback. MIDI Keyboards can be used for real-time recording, though this was part of VOCALOID Editor for Cubase it had not been a feature of VOCALOID itself.
It is AU (Audio Units) Compatible.
Additional vocals are no longer referred to as "Plug-in Libraries Databases" (VOCALOID2 and VOCALOID3), "VOCALOID libraries" (VOCALOID4), and other names but are now known as "Add-on Voicebanks". A minor change of classification, though the use of "voicebank" is notable. Previously "voicebank" had been used in English speaking communities as a shorthand or common term to refer Vocaloid library database as a simplification. Later the term had slowly seen an increase in usage over time among studios. As noted by Gizmodo unlike all past VOCALOID versions, VOCALOID5 has the engine and VOCALOID Library are now more interlocked with each other, which is why it is now referred to as a "voicebank".
A new "Vocaloid4.5" version was offered within VOCALOID5 to VOCALOID Editor for Cubase users which continued the use of the VOCALOID4 Editor for Cubase. This version lets users import VOCALOID5 vocals into the editor. Both versions can be used for the same project.
The GUI was overhauled and performance improved during rendering. It is now easier to both adjust perimeters and each has example presets for ease of use. Several of the perimeters have expanded functions for finer adjusting and control.
VSQ and VSQX files are replaced with VPR file types. This engine can read VPR, VSQX, MIDI (SMF), and WAV files.
The engine now supports up to 32-bit equalising (VOCALOID4 only supporting 16-bit) and up to 192 kHz rates while VOCALIOID4 supported up to 90KHz.
32 tracks can be supported, this is double the amount of all past versions of Vocaloid which only supported 16.
ReWire returns, this was absent since VOCALOID2 as a capability of Vocaloid, being excluded from both VOCALOID3 and VOCALOID4. It is, however, unsupported and may encounter issues. In addition VST (VST2) support returns also having been removed from VOCALOID3 and VOCALOID4. However, VOCALOID5 cannot be used as both a standalone or VSTi, therefore to switch between the two users have to close their current version. VST 3rd party plug-ins are unsupported.
DJTECHTOOLS also noted in their review that VOCALOID5 is much more realistic sounding then VOCALOID4.
VOCALOID5 is better then VOCALOID4 at adjusting the wave form and is suitable for more delicate adjustments. This is especially useful for English Vocaloid voicebanks. Consonants can be checked for their timing and locating problem phonetics is much easier. 
VOCALOID2 vocals are unsupported in this version entirely and a direct V2 → V5 import is impossible.
VOCALOID5 does not support XSY. While few VOCALOID are impacted by this greatly, the ARSLOID package was left with two voicebanks "bright" and "soft" unable to access XSY. The result is the pair are left only able to act as low quality voicebanks. The XSY function got replaced with "Styles/colours" which create more stable vocals. While XSY allowed two vocals to be combined to create new ones, depending on the combination, the new vocals were entirely different from the vocals used. Style/colours is more reliable and stable, but does not allow the same level of vocal creation. This mostly impacted those who used XSY for new vocal recreation, but also impacts even those who use XSY for its original intended function, which was allowing smooth transaction between voicebanks.
Some issues were found upon initial release including a bug that caused the engine to crash upon exporting and a comparatively longer loading times between editing the file and playback.
An issue with the Ver.5.0.2 update was when dragging the drawn parameters caused the engine to possibly crash. This was fixed in update 5.0.3.
Piapro Studio still works with VOCALOID5 but there are currently concerns such as Piapro Studio being unable to load the new save file type (.vpr files). MIDI and VSQX files made in Piapro Studio are unusable in VOCALOID5. E.V.E.C. must be used within VOCALOID5 as if the user wrote manually in VOCALOID4. Updates will continue on Piapro Studio in regards to the release of VOCALOID5. Currently the most impacted V4x release is Megurine Luka V4X due to its heavy reliance on Piapro Studio and EVEC. Voicebanks intended for VOCALOID5 currently do not work with Piapro Studio. As a result, users who rely on the software and which to move into the VOCALOID5 generation cannot and are restricted on what how they can use the software. An update to allow Piapro Studio to work with VOCALOID5 is confirmed not going to happen as Piapro Studio is being developed into its own software independent of Vocaloid.
Japanese vocals in VOCALOID5 are largely the same as VOCALOID3 and VOCALOID4 vocals. So far, no noted changes between VOCALOID4 and VOCALOID5 have been reported other then basic increases of quality other then natural progression of technology and developer experience. English vocals so far have reported to use the same methods as developed in VOCALOID4. There is no information on the situation with other languages. When reporting on why they were not updating to V5, Beijing Photek S&T Development Co., Ltd. noted this among the reasons for not updating to V5, feeling that not much change happened to the engine and that the Japanese trial had no noticeable differences. While comparison of files between V4 and V5 do show differences in sympathised results when subject to user experimentation, the issue has existed since VOCALOID3, with VOCALOID3 having been the last major update to Japanese voicebanks and how they work.
VOCALOID5 is the first engine to have issues with long standing problems that have existed within the franchise across the past versions came to a head. This lead to the engine being the first one to see a mixed response which includes some heavy negative criticism as well as long term issues become a folly for the franchises progression. While negative criticisms have always been within the franchise, the consequences of those negative responses had never hugely impacted support for the engine by 3rd party studios.
In comparisons to the VOCALOID3 to VOCALOID4 update, the VOCALOID5 engine has had a major overhaul and many new features have been presented making it a very different software to past version. The initial first-day reaction was mostly positive for this reason. Upon reviewing it, Gizmodo noted the engine felt much more modern.One of the problems with it was the overall over realming lack of response to the update compared to past versions.
However, the higher price tag for the engine and the higher computer specification requirement have also been cited as a negative point. In regards of the price, this issue is attenuated if you own one of the previous versions of VOCALOID due to the price to upgrade being much lower than the price to buy VOCALOID5 brand new. Other issues occurred with 2 of the 8 vocals in the Premium version offer, namely VY1 and VY2. Their newer VOCALOID5 versions do not contain the addition vocals that were supplied with VY2v3 and VY1v4.
- Note; if producers can afford the full premium version it is recommended over the standard version due to its value for money. Currently, to buy VY1 and VY2 in VOCALOID5 brand new are $90 each which means buying both VY1 and VY2 alone would cost $180. However, to buy them in the Premium form, it only costs an extra $129 on top of the price of the Standard version, with the added bonus of gaining access to CYBER DIVA II and CYBER SONGMAN II who are also normally sold at the same price of $90 each. In addition, buying two different English vocals and two different Japanese vocals would cost more than this, making the current Premium version one of the cheapest multipack offers for different vocalists and languages on sale for VOCALOID. However, if you already own any members of the 4 VY series vocals, please note there is not much difference between them and older versions in VY1v4 and VY2v3 have more voicebanks.
The VOCALOID5 editor was, according to one website, quite easy to edit vocals with and get an array of tones even with elder VOCALOID3 or VOCAOID4 vocals. VOCALOID LINK, however, stated that producers may simply be happy with updates to GUMI, Hatsune Miku, Otomachi Una, or to the Chinese vocaloids (which at this point were as popular as the Japanese Vocaloids) in VOCALOID4, although the last few Chinese Vocaloids have been released to a much less excited response. VOCALOID LINK expressed that there was a feeling of Vocaloid coming to an end because of this slower release trend and many feel it is the end of an era.
The lack of change for Japanese voicebanks was cited for the first time as a negative response, Beijing Photek S&T Development Co., Ltd noted in their first impression responses for why they did not continue with the VOCALOID5 software. Criticism was already beginning to be noted by VOCALOID fans over time as releases such as Gackpoid V4 and Rana V4 were not much different to older versions V3 Gackpoid and Rana.
3rd Party Developer response
VOCALOID5 was the first engine reported to have a sense of disdain which was unaffected by the inclusion of 4 vocals within the engine. Part of this was to do with the lack of large announcement in comparison to VOCALOID3 nor a rush of new exciting releases for VOCALOID4, with AH-Software Co. Ltd. being the only 3rd party to release a VOCALOID5 vocal by January 2019. AH-Software reported growing sales globally and was expanding on its successful campaign in Germany.
However, despite AH-Software's response, one of the growing issues with VOCALOID5 was the number of departing studios from the software:
- Upon the release of VOCALOID5, sales of VOCALOID4 ceased as per normal for the updating of a newer engine. Beijing Photek S&T Development Co., Ltd. reported being unable to order more VOCALOID4 sales and after contacting Shanghai HENIAN Information Technology Co. Ltd., they found that SHN had no stock either. This was cited as among the many reasons they shifted to Synthesizer V as it allowed more freedom, although it was normal for YAMAHA to cut off sales of older engines when a newer one is released. However, this the first time it was reported as a problem.
- PowerFX Systems AB. also left, though reasons are currently unconfirmed.
- Internet Co., Ltd. stated they were not producing a VOCALOID5 vocal. They have not yet stated if they are completely out of the franchise. A later response noted that Internet co., Ltd was not switching engines and may produce some products later on.
- Long term veteran, and the most successful company, Crypton Future Media, Inc. cited they would be releasing their own synthesizer based on Piapro Studio. This marked Hatsune Miku's departure from VOCALOID for the first time since her initially release on August 31, 2007.. This included the departures of MEIKO, KAITO, Kagamine Rin & Len, and Megurine Luka.
- In July 2020, Zero-G Limited responded to a fan confirming that they are not currently making any new VOCALOIDs.
- Other VOCALOIDs, such as flower and Yuzuki Yukari, opted for new releases on CeVIO, although their departure from VOCALOID is not confirmed as the new releases may coexist with their VOCALOID versions.
- On December 25, 2020, VocaTone announced that they were working in collaboration with Misbah Studios to produce a new voice synthesizer: Maghni AI. Their first two vocalists, Audine and Aurum were announced. Aurum was formerly known as "ausgris", a project that was initially introduced as working VOCALOID voicebank in 2016, but did not have a finalized name at the time. For years, ausgris had little information about the progress of his VOCALOID production, and for three years, VocaTone was dedicating time to creating Maghni AI. However, departure from VOCALOID was not explicitly stated.
- In December 2021, when asked about considerations for migration to another engine, Vsinger's official Weibo account answered that they had been long studying and made several voicebank demos with tech providers. Different engines are based on different technologies. VOCALOID was based on sampling, which better reproduced the original voice texture, but was less natural and may be harder to use. AI voicebanks were easier for beginners and sing more naturally, but do not sufficiently reproduce the original voice texture and the voice may also sound as though it was synthesized. Vsinger noted that the voice is the most important part of a virtual singer and was trying to find ways to preserve and reproduce the characters' original voices, wishing to deliver more satisfactory products.
At the time of its release, there was no environment set up for developers, so supported 3rd party features and functions did not start appearing until the following year in 2019.
The reference manual for this program can be found here.