Icon = Sign that relates to an object through its similatity

Index = ‘‘ throug direct causal relationship (e.g. smoke relates to fire)

Symbol = ‘‘ through an abritory or conventional relationship (e.g. letter to a sound)

Carrier : physical Sign displaying the symbol

Meaning : What is the message?

Object : what is depicted?

earcon: abstract, synthetic tone -> create soundmessages -> UI, learning

auditory icon: everiday sound, e.g. recorded in nature -> anology with events -> functional feedback, get attention

Earcons

Pro: Can be used to design and control the desired effect

Con: No association with everyday events/ needs to be learned

Microtypography

Small scale text composition (Serifs, Ligation, Font, hight)

Macrotypography

composition of text on paper/screen

proportional: every character has an individual width

non-prop.: each character has the same width (e.g. curier)

this is inhomogenic and distroys the attention of the reader

Kerning is the adjustment of space between individual letter pairs to create visually balanced and pleasing text.

Line spacing is the vertical space between lines of text

1. Segmentation (Analysis):

   • The speech signal is segmented into short overlapping frames around pitch marks (periodic instants like glottal pulses in voiced sounds).

   • These frames are usually centered around pitch periods to maintain the natural periodicity of the speech.

2. Modification:

   • To change pitch, the algorithm shifts the spacing between segments.

   • To change duration, it either repeats or skips certain segments.

3. Reconstruction (Synthesis):

   • The modified frames are overlapped and added together, aligning them at pitch-synchronous points to produce a smooth signal.

Due to manipulation the result can sound a bit unnatural -> longer and more units used

1. Parametric Synthesis: Generates speech by manipulating parameters (like frequency, pitch, and formants) using a mathematical model, often resulting in robotic or unnatural-sounding speech.

2. Concatenative Synthesis: Constructs speech by piecing together small recorded segments of human speech (units), aiming for more natural-sounding output.

3. Unit-Selection Synthesis: A type of concatenative synthesis that selects the best-matching units from a large speech database based on target and concatenation costs for high naturalness.

4. Statistical Synthesis (e.g., HMM-based): Uses statistical models trained on speech data to generate speech waveforms, balancing flexibility and intelligibility, though often with smoother but less expressive output.