http://www.k-wave.org/forum/topic/changing-spatial-resolution-gives-strange-results Support for the k-Wave MATLAB toolbox en-US Wed, 13 May 2026 00:21:25 +0000 http://bbpress.org/?v=1.0.2 q http://www.k-wave.org/forum/search.php http://www.k-wave.org/forum/topic/changing-spatial-resolution-gives-strange-results#post-7955 Thu, 26 Nov 2020 14:21:11 +0000 DNF 7955@http://www.k-wave.org/forum/ <p>There's a time-of-arrival difference for the second pulse that bounces off the interface, though it may be related to the same thing.</p> <p>Thanks for the explanation. I'm used to only considering band-limited sources, and should probably not be messing around with initial value problems at this point, since I'm missing some intuition, apparently. </p> http://www.k-wave.org/forum/topic/changing-spatial-resolution-gives-strange-results#post-7954 Thu, 26 Nov 2020 14:07:13 +0000 Bradley Treeby 7954@http://www.k-wave.org/forum/ <p>In the limit of infinite bandwidth, the response should be an n-wave. This is described in the series of seminal papers about photoacoustic waves by Diebold (e.g., <a href="https://journals.aps.org/prl/abstract/10.1103/PhysRevLett.67.3384">this one</a>). The waves in the lower resolution simulation look "smoother" because of the band-limiting.</p> <p>Not sure what you mean by the timing being off? </p> http://www.k-wave.org/forum/topic/changing-spatial-resolution-gives-strange-results#post-7948 Thu, 26 Nov 2020 12:22:19 +0000 DNF 7948@http://www.k-wave.org/forum/ <p>I'm coming back to k-wave after a pause of several years, and ran into a problem that may be trivial (hopefully). I used the example_ivp_axisymmetric_simulation.m script, and wanted to just double the spatial resolution in each direction:<br /> <pre><code>Nup = 2; % upsampling factor Nx = 128 * Nup; % number of grid points in the axial (x) direction Ny = 64 * Nup; % number of grid points in the radial (y) direction dx = 0.1e-3/Nup; % grid point spacing in the axial (x) direction [m] dy = 0.1e-3/Nup; % grid point spacing in the radial (y) direction [m] kgrid = kWaveGrid(Nx, dx, Ny, dy); % define the properties of the propagation medium medium.sound_speed = 1500 * ones(Nx, Ny); % [m/s] medium.sound_speed(Nx/2:end, :) = 1800; % [m/s] medium.density = 1000 * ones(Nx, Ny); % [kg/m^3] medium.density(Nx/2:end, :) = 1200; % [kg/m^3] % create initial pressure distribution in the shape of a disc - this is % generated on a 2D grid that is doubled in size in the radial (y) % direction, and then trimmed so that only half the disc is retained source.p0 = 10 * makeDisc(Nx, 2 * Ny, Nx/4 + 8, Ny + 1, 5*Nup); source.p0 = source.p0(:, Ny + 1:end); % define a Cartesian sensor mask with points in the shape of a circle sensor.mask = makeCartCircle(40 * dx * Nup, 50); % remove points from sensor mask where y &lt; 0 sensor.mask(:, sensor.mask(2, :) &lt; 0) = []; % run the simulation using the axisymmetric code sensor_data = kspaceFirstOrderAS(kgrid, medium, source, sensor);</code></pre> <p>The output waveform is changed quite significantly, into something closer to a 'sawtooth' wave, and the timing of the arrivals also seem to be 'off'. Perhaps I need to tweak the time-step or something, but should this not happen automatically? </p>