In an earlier post where I consider piano music by Messiaen, I wrote:

I want to explore an aesthetic of music that is 'kinaesthetic'. Kinaesthetic aesthetics is constructed from the sonic gestures of a piece experienced within time, leading to a perceived architecture outside of linear chronology.   This would be a holographic image, combining many variant data elements into a very simple insight – perhaps into a statement like “this piece is Unfolding', or 'here the piece is Rotating'. (I cannot rule out the possibility that for some music, the simplest statement of form is that “Here there be Dragons”.)

We all the time set aside linear, chronological and time based thinking when we talk or think about music. It's a significant conceptual leap. If we talked about our lives with the same language that we use to talk about a piece of just-heard music, we would emerge from a flatland of perception into a very full and harmonic awareness of what it means to be a human being.

Gestural Spectral composition brings shape and architecture to the foreground, and is an evolution of this earlier analytical writing.  The spectrogram is a specific tool that directly modifies the rules and logic of compostion and perception.    The form of the piece is paramount, and the perception of the form is an essential part of understanding the music.  A visual representation  should guide the listener towards a perception of the form.  

This is not a casual undertaking:  it is learning to apprehend sound just a step outside of time.   This requires some basic building blocks and some agreed upon usage for fairly common terms.  The building blocks are Technical and Conceptual.

Technical 
How to 'draw' shapes in sound.  I begin with squares, circles and lines.  Blocks of static and sine waves. Position on the spectrogram modifies both played frequency and temporal relationship.  Duration is length, Frequency is Height.  Amplitude is depth

Conceptual
The spectrogram extracts sound from time (just a step).  The Sonic Object can be recognized, manipulated, perceived.  I reduce the sonic palette to static and sine waves.  If I can describe sound as a simple shape, then I can eventually describe complex and irregular shapes like  a human voice, or a recording of Mahler's Urlicht. 

The sample video here shows 8 shapes.

1.  A blast of static.  A block across the full screen.
2.  A sine wave at 440Hz. A single line
3.  A Filled Square
4.  A Filled Circle
5. A Square Outline
6.  A Circle Outline
7.  Interlocking Squares
8.  Interlocking Circles. 

 The video is a screen recording from Sonic Visualiser.  The Green shape below the spectrogram is the wave form of the sound.  A waveform represents the change in amplitude over time.  The Spectrogram shows changes in Frequency. 

You can see that the circles are not precisely circular and that the squares are not perfectly squares.  This is because at each step of the way, I am required to manipulate tools, to set configuration and to manipulate output. It is difficult to maintain the pixel to pixel relationship between X and Y axis as the file transfgers between software.  For example, when I import the image file to Photosounder, I have to establish a duration in seconds that is equivalent to the pixel width of the file.   Understanding that the limitations and features of the tools I'm using will define the output -  this is no different than a composer recognizing that the construction of a piano will yield different results than an oboe. 

A workaround is to create a blank 'template' from photsounder that I know is equal to 30 seconds, then import that in to my graphics tool, do my business, and then reload in to photosounder.  I do this for larger pieces, but didn't have time for this sample.  Trust me that it has nothing to do with the underlying sound and everything to do with the display settings of the tools I'm using.  

I'm belaboring some very obvious points, and will continue to do so over more posts.   I want to be clear on the conceptual and technical language in spectral composition.  If I can understand the idea of a sonic 'Cube' as an object that can be considered outside of the linear chronological unfolding of the sound, then I can begin to consider complex forms in the same way.