Flags

Definition

Flags are an editing feature in UTAU that helps change the tonal quality of notes. Flags include Y and H (clarity of sound), BRE (breathiness), an g (gender: + for lower, - for higher).

Table of Flags and Their Functions

General Flags: The following flags are used in most resamplers.

Flag Base Value Setting Range Feature
g 0 -100…+100 A flag to control the effect offered by the formant (the voice quality determined by the structure of the throat or mouth). This flag value is set with the + or - symbols appended, like e.g. g+10, g-10. No symbol will be interpreted as +. When setting positive values, the voice becomes deeper, more mature and masculine (large values, usually from +15 to +30, can cause a female voice can become male.) When setting negative values, the voice becomes thinner, more childlike and feminine (large values, usually from -15 to -30, can cause a male voice can become female.)
B 50 0…100 Alters breathiness as adjusting BRE in the note properties.
N none none As checking the "no formant filter" box in the note properties.
t 0 -9…+9 Flag to adjust the pitch in 10 cents (1/10th of semitone) units. This flag value is set with the + or - symbols appended, like e.g. t+5, t-5. No symbol will be interpreted as +.
Y 100 0…100 The part outside of the fixed range in consonants is called breathiness. (Details are omitted.) By specifying a small value like e.g. Y0, the breathiness part of consonants becomes relatively stronger, and the articulation is considered to be better. (As a side effect, noise appears that makes high notes sounding metallic, thus increase the flag value, or adjust simultaneously the H low-pass filter described below.)
H 0 0…99 A low-pass filter to emphasize the bass and cut the treble. (When using together the C, D, E low-pass filters described below, they produce the same effect.) It has the effect of mitigating the metallic noise on high notes, but as a side effect the sound becomes muffled.
h 0 0…99 A low-pass filter operating outside of the breath component of consonant (breathiness). As it emphasizes the high frequencies of consonant components, it is unsuitable to sound sources where the consonant component is unstable. Note: If set too strong, voices becomes hoarse even with sound sources in which the consonant component is stable, and you need to reduce the value of the Y flag.
F 3 0…unspecified This adjusts the stength of the formant filter. The formant filter depends on the frequency defined by "source frequency * specified value". It is generally better not to touch it, but when noise appears in low tones, you can suppress it by specifying values around F4 .. F7. This flag is valid for Resampler development versions too, but changes are not as big as in the default generic Resampler version.
L none 0…unspecified A fixed frequency for the "F" flag above. The formant filter depends then on the frequency defined by "170Hz * specified value". When used simultaneously with F, this value takes precedence.

Resampler (generic) Flags

Flag Base Value Setting Range Feature
b 0 0…+100 BRE adjustment after the formant filter. BRE changes are loosened when its pitch is very different from the pitch of the primary sound, and the voice becomes unpleasantly rough. In addition, because it is not influenced by the low-pass type filters (C, D, E, H and h), the sound coming from BRE can be bad and muffled.
C 0 0…100 A low-pass filter especially reducing the high-frequencies. When set to 100, the volume is 100% at 0kHz, 50% at 11kHz, and 0% at 22kHz.
D 0 0…100 A low-pass filter cutting the midrange. When set to 100, the volume is 100% at 0kHz, 0% at 11kHz, and 100% at 22kHz.
E 0 0…100 A low-pass filter cutting the bass and treble. When set to 100, the volume is 100% at 0kHz, 0% at 7.1kHz, 100% at 11kHz, and 0% at 22kHz.
c 50 0…100 The value of the C flag before the formant filter adaptation.
h 0 0…99 A low-pass filter operating outside of the breath component of consonant (breathiness). As it emphasizes the high frequencies of consonant components, it is unsuitable to sound sources where the consonant component is unstable. Note: If set too strong, voices becomes hoarse even with sound sources in which the consonant component is stable, and you need to reduce the value of the Y flag.
P 86 0…100 Peak compressor. Align the peak volume of the primary sound. (The volume setting and the envelope changes are applied separately.) There is zero variation with a value of 100. When set to 99 or less, the variation produced is proportional to the volume of the primary sound and to the parameter value. Because only the peak volume of the primary sound is aligned, a sense of instability in the volume will remain in sound sources with unstable volume changes, even if it is set to 100.
W unspecified unspecified Produces a robotic voice. This flag is highly experimental, and is generally not used.
x unspecified unspecified unknown (to be filled in later)

Each sampling engine may have its own specific flag sets.

See Also

Genderbent/Genderbend

_f
Unless otherwise stated, the content of this page is licensed under Creative Commons Attribution-NonCommercial-NoDerivs 3.0 License