DOC PREVIEW
MIT 6 111 - Music Transcriber

This preview shows page 1-2 out of 6 pages.

Save
View full document
View full document
Premium Document
Do you want full access? Go Premium and unlock all 6 pages.
Access to all documents
Download any document
Ad free experience
View full document
Premium Document
Do you want full access? Go Premium and unlock all 6 pages.
Access to all documents
Download any document
Ad free experience
Premium Document
Do you want full access? Go Premium and unlock all 6 pages.
Access to all documents
Download any document
Ad free experience

Unformatted text preview:

Roberto Carli 6.111 Project Proposal Alessandro Yamhure 11/04/2005 1 MUSIC TRANSCRIBER Overall System Description The aim of this digital system is to convert music played into the correct sheet music. We are basically implementing a music transcriber that transcribes sound inputs into visual sheet music on a computer monitor. The system is divided into three main modules: The Note Recognizer is responsible for analyzing the sound input of the system and outputting its note name (F, C#, etc) and rhythmical duration. It takes in one input (the audio) and gives out four outputs; note (the name of the note), duration (how many beats it lasts), new note a pulse that signals the arrival of a new note to the Video Display module and beat, a pulse emitted at the tempo of the music. The Metronome is responsible for converting a pulse signal beat into a rhythmical “click” audio sound, which will be fed to the audio codec. The Video Display module is responsible for taking the note and its duration when new note goes high and converting it into the appropriate video display of staves and notes on the computer screen.Roberto Carli 6.111 Project Proposal Alessandro Yamhure 11/04/2005 2 Note Recognition Module (Alessandro Yamhure) The main input is an audio input that comes into the module from a microphone or amplified instrument and passes through a codec module which amplifies the audio and converts it from analog to digital. PITCH DETECTION: The digital audio data is then temporarily stored in RAM before serving as the input to the FFT submodule. The FFT submodule takes the Fast Fourier Transform of this data and temporarily places its results into RAM memory, before it is used by the Peak Detector. The result of the FFT is basically a frequency plot of the audio input. The Peak Detector is responsible for using the frequency plot in RAM to detect the peak frequency of the Frequency Data Out and in this way detects the pitch of the input. The Look-up Table submodule uses the peak frequency to decide the actual musical name for that frequency (F#, Bb, etc). The Note serves as one of the outputs of the Note Recognition Module but also serves as the input to the compare submodule. RHYTHM DETECTION: The compare submodule compares the current note to the old note that was output during the previous processing cycle by the look-up table. If they are equal, change signal is low. If the two values are different, then change is high. If change is high we have a “new note” and the signal New Note, which is an output of the Note Recognition module, goes high for one clock cycle. It also serves as an input to the counter submodule. The counter submodule is a counter that gets incremented every time it is sent a pulse by the Tempo Clock. When change is high, counter outputs the value up to which it has counted as the duration of the note which is another output of the Note Recognition module. Tempo Clock appropriately divides the system clock (using a counter) into the beats of the music (like a metronome). This is important in order to measure the duration of a note, which will allow us to have a rhythmical component in our transcriber.Roberto Carli 6.111 Project Proposal Alessandro Yamhure 11/04/2005 3 Video Display (Roberto Carli) The video display module will convert the digital signals giving information about notes and durations into an actual music sheet, which will be displayed. This module will take three inputs from the note recognizer: the value of the note, its duration and the newnote signal (display another note). The way I am planning to tackle the display is by using an object-oriented visualization style. The duration of the note will determine its shape (full, quarter etc…) while the value of the note will determine its position on the stave. The video display will have several submodules: Tracker: The tracker submodule will be an FSM whose function is keeping track of the position of the latest note and, when a new note has to be displayed, to determine what position it will occupy on the screen. Tracker will take inputs note and newnote, and will output the screen coordinates cx and cy where the note shall be displayed. XVGA: The XVGA module will take the clock input and generate the appropriate signals for video display hcount, vcount, hsync and vsync. It will work for a resolution of 1024x768 and a refresh rate of 60Hz. Finally, it will take the pixel input from the Video module in order to generate the image. ROM: The ROM will store bit values representing the image (pixels ON and pixels OFF) of notes of different duration, which can be requested by inputting the appropriate address (the duration signal itself will work). The notes will be approximately 10x20 pixels large. They will be drawn in such a way that they will all align correctly on a stave with respect to their upper-left pixel, so that with same coordinates cx cy, every note will be drawn on the same line of the stave. Video: The core of the module. Video will take in the coordinates cx and cy of the note to display from the tracker, the color for the display, the duration of the note and the appropriate video signals, and will generate the pixel outputs for the screen. At reset, Video will rely on two sprite submodules which will generate pixels for the stave and the treble clefs. When a start signal is given from the Tracker, Video will ask the ROM to generate the appropriate note figure, and will feed it to a logic module. The logic will generate pixels for that figure having coordinates cx, cy as reference point for the hcount and vcount of the upper-left pixel. Finally, Video will OR the various pixel signals so as to emit a positive pixel (color specified by the color input) if any submodule (stave, clef, note) emits a positive pixel. Metronome (Roberto Carli) Finally, there will be a metronome which will emit a “click” audio signal, so as to let the musician hear the tempo he specified. The Metronome will take as input the signal beat, a pulse signal given by the note recognizer representing the tempo of the music, and it will output a “click” audio signal to_codec which will go to the audio codec and subsequently to speakers.Roberto Carli 6.111 Project Proposal Alessandro Yamhure 11/04/2005 4 Note RecognizerVideo DisplayMetronomebeat to_codecnote 5duration 2newnoteaudio inputto_monitor 5Overall Block DiagramRoberto Carli 6.111


View Full Document

MIT 6 111 - Music Transcriber

Documents in this Course
Verilog

Verilog

21 pages

Video

Video

28 pages

Bass Hero

Bass Hero

17 pages

Deep 3D

Deep 3D

12 pages

SERPENT

SERPENT

8 pages

Vertex

Vertex

92 pages

Vertex

Vertex

4 pages

Snapshot

Snapshot

15 pages

Memories

Memories

42 pages

Deep3D

Deep3D

60 pages

Design

Design

2 pages

Frogger

Frogger

11 pages

SkiFree

SkiFree

81 pages

Vertex

Vertex

10 pages

EXPRESS

EXPRESS

2 pages

Labyrinth

Labyrinth

81 pages

Load more
Download Music Transcriber
Our administrator received your request to download this document. We will send you the file to your email shortly.
Loading Unlocking...
Login

Join to view Music Transcriber and access 3M+ class-specific study document.

or
We will never post anything without your permission.
Don't have an account?
Sign Up

Join to view Music Transcriber 2 2 and access 3M+ class-specific study document.

or

By creating an account you agree to our Privacy Policy and Terms Of Use

Already a member?