Mercurial > hg > svgui
changeset 1094:8a815776151c spectrogram-minor-refactor
Split out cache rendering functions and some fixes to calculations
author | Chris Cannam |
---|---|
date | Thu, 07 Jul 2016 19:18:31 +0100 |
parents | cd22f74dc159 |
children | ba62684a4512 |
files | layer/Colour3DPlotRenderer.cpp layer/Colour3DPlotRenderer.h |
diffstat | 2 files changed, 205 insertions(+), 136 deletions(-) [+] |
line wrap: on
line diff
--- a/layer/Colour3DPlotRenderer.cpp Thu Jul 07 19:18:01 2016 +0100 +++ b/layer/Colour3DPlotRenderer.cpp Thu Jul 07 19:18:31 2016 +0100 @@ -25,6 +25,8 @@ #include <vector> +//#define DEBUG_SPECTROGRAM_REPAINT 1 + using namespace std; Colour3DPlotRenderer::RenderResult @@ -55,13 +57,23 @@ m_cache.setZoomLevel(v->getZoomLevel()); cerr << "cache start " << m_cache.getStartFrame() - << " view start " << startFrame << " valid left " << m_cache.getValidLeft() << " valid right " << m_cache.getValidRight() + << endl; + cerr << " view start " << startFrame << " x0 " << x0 << " x1 " << x1 << endl; + bool bufferIsBinResolution = useBinResolutionForDrawBuffer(v); + + if (bufferIsBinResolution) { + // Rendering should be fast in this situation because we are + // quite well zoomed-in, and the sums are easier this + // way. Calculating boundaries later will be fiddly for + // partial paints otherwise. + timeConstrained = false; + } if (m_cache.isValid()) { // some part of the cache is valid @@ -140,7 +152,18 @@ rightToLeft = isLeftOfValidArea; } - renderToCache(v, x0, x1 - x0, rightToLeft, timeConstrained); + // Note, we always paint the full height. Smaller heights can be + // used when painting direct from cache (outside this function), + // but we want to ensure the cache is coherent without having to + // worry about vertical matching of required and valid areas as + // well as horizontal. That's why this function didn't take any + // y/height parameters. + + if (bufferIsBinResolution) { + renderToCacheBinResolution(v, x0, x1 - x0); + } else { + renderToCachePixelResolution(v, x0, x1 - x0, rightToLeft, timeConstrained); + } QRect pr = rect & m_cache.getValidArea(); paint.drawImage(pr.x(), pr.y(), m_cache.getImage(), @@ -175,113 +198,69 @@ //!!! should we own the Dense3DModelPeakCache here? or should it persist } +bool +Colour3DPlotRenderer::useBinResolutionForDrawBuffer(LayerGeometryProvider *v) const +{ + DenseThreeDimensionalModel *model = m_sources.source; + if (!model) return false; + int binResolution = model->getResolution(); + int zoomLevel = v->getZoomLevel(); + return (binResolution > zoomLevel); +} + void -Colour3DPlotRenderer::renderToCache(LayerGeometryProvider *v, - int x0, int repaintWidth, - bool rightToLeft, bool timeConstrained) +Colour3DPlotRenderer::renderToCachePixelResolution(LayerGeometryProvider *v, + int x0, int repaintWidth, + bool rightToLeft, + bool timeConstrained) { - // Draw to the draw buffer, and then scale-copy from there. + cerr << "renderToCachePixelResolution" << endl; + + // Draw to the draw buffer, and then copy from there. The draw + // buffer is at the same resolution as the target in the cache, so + // no extra scaling needed. DenseThreeDimensionalModel *model = m_sources.source; if (!model || !model->isOK() || !model->isReady()) { throw std::logic_error("no source model provided, or model not ready"); } - // The draw buffer contains a fragment at either our pixel - // resolution (if there is more than one time-bin per pixel) or - // time-bin resolution (if a time-bin spans more than one pixel). - // We need to ensure that it starts and ends at points where a - // time-bin boundary occurs at an exact pixel boundary, and with a - // certain amount of overlap across existing pixels so that we can - // scale and draw from it without smoothing errors at the edges. - - // If (getFrameForX(x) / increment) * increment == - // getFrameForX(x), then x is a time-bin boundary. We want two - // such boundaries at either side of the draw buffer -- one which - // we draw up to, and one which we subsequently crop at. - - bool bufferIsBinResolution = false; - int binResolution = model->getResolution(); - int zoomLevel = v->getZoomLevel(); - if (binResolution > zoomLevel) bufferIsBinResolution = true; - - sv_frame_t leftBoundaryFrame = -1, leftCropFrame = -1; - sv_frame_t rightBoundaryFrame = -1, rightCropFrame = -1; - - int drawWidth; - - if (bufferIsBinResolution) { - for (int x = x0; ; --x) { - sv_frame_t f = v->getFrameForX(x); - if ((f / binResolution) * binResolution == f) { - if (leftCropFrame == -1) leftCropFrame = f; - else if (x < x0 - 2) { - leftBoundaryFrame = f; - break; - } - } - } - for (int x = x0 + repaintWidth; ; ++x) { - sv_frame_t f = v->getFrameForX(x); - if ((f / binResolution) * binResolution == f) { - if (rightCropFrame == -1) rightCropFrame = f; - else if (x > x0 + repaintWidth + 2) { - rightBoundaryFrame = f; - break; - } - } - } - drawWidth = int((rightBoundaryFrame - leftBoundaryFrame) / binResolution); - } else { - drawWidth = repaintWidth; - } - - // We always paint the full height. Smaller heights can be used - // when painting direct from cache (outside this function), but we - // want to ensure the cache is coherent without having to worry - // about vertical matching of required and valid areas as well as - // horizontal. That's why this function didn't take any y/height - // parameters. int h = v->getPaintHeight(); - clearDrawBuffer(drawWidth, h); + clearDrawBuffer(repaintWidth, h); - vector<int> binforx(drawWidth); + vector<int> binforx(repaintWidth); vector<double> binfory(h); bool usePeaksCache = false; int binsPerPeak = 1; + int zoomLevel = v->getZoomLevel(); + int binResolution = model->getResolution(); - if (bufferIsBinResolution) { + for (int x = 0; x < repaintWidth; ++x) { + sv_frame_t f0 = v->getFrameForX(x0 + x); + double s0 = double(f0 - model->getStartFrame()) / binResolution; + binforx[x] = int(s0 + 0.0001); + } - for (int x = 0; x < drawWidth; ++x) { - binforx[x] = int(leftBoundaryFrame / binResolution) + x; - } + if (m_sources.peaks) { // peaks cache exists - // calculating boundaries later will be too fiddly for partial - // paints, and painting should be fast anyway when this is the - // case because it means we're well zoomed in - timeConstrained = false; - - } else { - for (int x = 0; x < drawWidth; ++x) { - sv_frame_t f0 = v->getFrameForX(x0 + x); - double s0 = double(f0 - model->getStartFrame()) / binResolution; - binforx[x] = int(s0 + 0.0001); - } - - if (m_sources.peaks) { // peaks cache exists - - binsPerPeak = m_sources.peaks->getColumnsPerPeak(); - usePeaksCache = (binResolution * binsPerPeak) < zoomLevel; - - if (m_params.colourScale.getScale() == - ColourScale::PhaseColourScale) { - usePeaksCache = false; - } + binsPerPeak = m_sources.peaks->getColumnsPerPeak(); + usePeaksCache = (binResolution * binsPerPeak) < zoomLevel; + + if (m_params.colourScale.getScale() == + ColourScale::PhaseColourScale) { + usePeaksCache = false; } } + cerr << "[PIX] zoomLevel = " << zoomLevel + << ", binResolution " << binResolution + << ", binsPerPeak " << binsPerPeak + << ", peak cache " << m_sources.peaks + << ", usePeaksCache = " << usePeaksCache + << endl; + for (int y = 0; y < h; ++y) { binfory[y] = m_sources.verticalBinLayer->getBinForY(v, h - y - 1); } @@ -294,57 +273,139 @@ rightToLeft, timeConstrained); - //!!! now scale-copy to cache + if (attainedWidth == 0) return; - if (attainedWidth == 0) return; + // draw buffer is pixel resolution, no scaling factors or padding involved int paintedLeft = x0; if (rightToLeft) { paintedLeft += (repaintWidth - attainedWidth); } - if (bufferIsBinResolution) { + m_cache.drawImage(paintedLeft, attainedWidth, + m_drawBuffer, + paintedLeft - x0, attainedWidth); +} - int scaledLeft = v->getXForFrame(leftBoundaryFrame); - int scaledRight = v->getXForFrame(rightBoundaryFrame); +void +Colour3DPlotRenderer::renderToCacheBinResolution(LayerGeometryProvider *v, + int x0, int repaintWidth) +{ + cerr << "renderToCacheBinResolution" << endl; + + // Draw to the draw buffer, and then scale-copy from there. Draw + // buffer is at bin resolution, i.e. buffer x == source column + // number. We use toolkit smooth scaling for interpolation. - QImage scaled = m_drawBuffer.scaled - (scaledRight - scaledLeft, h, - Qt::IgnoreAspectRatio, (m_params.interpolate ? - Qt::SmoothTransformation : - Qt::FastTransformation)); + DenseThreeDimensionalModel *model = m_sources.source; + if (!model || !model->isOK() || !model->isReady()) { + throw std::logic_error("no source model provided, or model not ready"); + } + + // The draw buffer will contain a fragment at bin resolution. We + // need to ensure that it starts and ends at points where a + // time-bin boundary occurs at an exact pixel boundary, and with a + // certain amount of overlap across existing pixels so that we can + // scale and draw from it without smoothing errors at the edges. + + // If (getFrameForX(x) / increment) * increment == + // getFrameForX(x), then x is a time-bin boundary. We want two + // such boundaries at either side of the draw buffer -- one which + // we draw up to, and one which we subsequently crop at. + + sv_frame_t leftBoundaryFrame = -1, leftCropFrame = -1; + sv_frame_t rightBoundaryFrame = -1, rightCropFrame = -1; + + int drawBufferWidth; + int binResolution = model->getResolution(); + + for (int x = x0; ; --x) { + sv_frame_t f = v->getFrameForX(x); + if ((f / binResolution) * binResolution == f) { + if (leftCropFrame == -1) leftCropFrame = f; + else if (x < x0 - 2) { + leftBoundaryFrame = f; + break; + } + } + } + for (int x = x0 + repaintWidth; ; ++x) { + sv_frame_t f = v->getFrameForX(x); + if ((f / binResolution) * binResolution == f) { + if (rightCropFrame == -1) rightCropFrame = f; + else if (x > x0 + repaintWidth + 2) { + rightBoundaryFrame = f; + break; + } + } + } + drawBufferWidth = int + ((rightBoundaryFrame - leftBoundaryFrame) / binResolution); + + int h = v->getPaintHeight(); + + clearDrawBuffer(drawBufferWidth, h); + + vector<int> binforx(drawBufferWidth); + vector<double> binfory(h); + + for (int x = 0; x < drawBufferWidth; ++x) { + binforx[x] = int(leftBoundaryFrame / binResolution) + x; + } + + cerr << "[BIN] binResolution " << binResolution + << endl; + + for (int y = 0; y < h; ++y) { + binfory[y] = m_sources.verticalBinLayer->getBinForY(v, h - y - 1); + } + + int attainedWidth = renderDrawBuffer(drawBufferWidth, + h, + binforx, + binfory, + false, + false, + false); + + if (attainedWidth == 0) return; + + int scaledLeft = v->getXForFrame(leftBoundaryFrame); + int scaledRight = v->getXForFrame(rightBoundaryFrame); + + QImage scaled = m_drawBuffer.scaled + (scaledRight - scaledLeft, h, + Qt::IgnoreAspectRatio, (m_params.interpolate ? + Qt::SmoothTransformation : + Qt::FastTransformation)); - int scaledLeftCrop = v->getXForFrame(leftCropFrame); - int scaledRightCrop = v->getXForFrame(rightCropFrame); - - int targetLeft = scaledLeftCrop; - if (targetLeft < 0) { - targetLeft = 0; - } - - int targetWidth = scaledRightCrop - targetLeft; - if (targetLeft + targetWidth > m_cache.getSize().width()) { - targetWidth = m_cache.getSize().width() - targetLeft; - } - - int sourceLeft = targetLeft - scaledLeft; - if (sourceLeft < 0) { - sourceLeft = 0; - } - - int sourceWidth = targetWidth; - - if (targetWidth > 0) { - m_cache.drawImage(targetLeft, targetWidth, - scaled, - sourceLeft, sourceWidth); - } - - } else { - - m_cache.drawImage(paintedLeft, attainedWidth, - m_drawBuffer, - paintedLeft - x0, attainedWidth); + int scaledLeftCrop = v->getXForFrame(leftCropFrame); + int scaledRightCrop = v->getXForFrame(rightCropFrame); + + int targetLeft = scaledLeftCrop; + if (targetLeft < 0) { + targetLeft = 0; + } + + int targetWidth = scaledRightCrop - targetLeft; + if (targetLeft + targetWidth > m_cache.getSize().width()) { + targetWidth = m_cache.getSize().width() - targetLeft; + } + + int sourceLeft = targetLeft - scaledLeft; + if (sourceLeft < 0) { + sourceLeft = 0; + } + + int sourceWidth = targetWidth; + + cerr << "repaintWidth = " << repaintWidth + << ", targetWidth = " << targetWidth << endl; + + if (targetWidth > 0) { + m_cache.drawImage(targetLeft, targetWidth, + scaled, + sourceLeft, sourceWidth); } } @@ -391,6 +452,9 @@ int columnCount = 0; vector<float> preparedColumn; + + int modelWidth = sourceModel->getWidth(); + cerr << "modelWidth " << modelWidth << endl; for (int x = start; x != finish; x += step) { @@ -413,10 +477,10 @@ for (int sx = sx0; sx < sx1; ++sx) { #ifdef DEBUG_SPECTROGRAM_REPAINT -// cerr << "sx = " << sx << endl; + cerr << "sx = " << sx << endl; #endif - if (sx < 0 || sx >= sourceModel->getWidth()) { + if (sx < 0 || sx >= modelWidth) { continue; } @@ -429,8 +493,8 @@ ColumnOp::Column fullColumn = sourceModel->getColumn(sx); - cerr << "x " << x << ", sx " << sx << ", col height " << fullColumn.size() - << ", minbin " << minbin << ", maxbin " << maxbin << endl; +// cerr << "x " << x << ", sx " << sx << ", col height " << fullColumn.size() +// << ", minbin " << minbin << ", maxbin " << maxbin << endl; ColumnOp::Column column = vector<float>(fullColumn.data() + minbin,
--- a/layer/Colour3DPlotRenderer.h Thu Jul 07 19:18:01 2016 +0100 +++ b/layer/Colour3DPlotRenderer.h Thu Jul 07 19:18:31 2016 +0100 @@ -63,7 +63,7 @@ binDisplay(AllBins), binScale(LinearBinScale), alwaysOpaque(false), - interpolate(false), + interpolate(false), //!!! separate out x-interpolate and y-interpolate? the spectrogram actually does (or used to) invertVertical(false) { } ColourScale colourScale; // complete ColourScale object by value @@ -167,10 +167,15 @@ // then repainting from cache to the requested painter. ScrollableImageCache m_cache; + bool useBinResolutionForDrawBuffer(LayerGeometryProvider *) const; + RenderResult render(LayerGeometryProvider *v, QPainter &paint, QRect rect, bool timeConstrained); - void renderToCache(LayerGeometryProvider *v, int x0, int repaintWidth, - bool rightToLeft, bool timeConstrained); + void renderToCachePixelResolution(LayerGeometryProvider *v, int x0, + int repaintWidth, bool rightToLeft, + bool timeConstrained); + void renderToCacheBinResolution(LayerGeometryProvider *v, int x0, + int repaintWidth); int renderDrawBuffer(int w, int h, const std::vector<int> &binforx, const std::vector<double> &binfory,