Problem
Large still images consume excessive memory because they are imported at full camera resolution for small on-screen use.
Solution
Root Cause / Diagnostic:
Importing unscaled 24–60 megapixel digital camera raw/JPEG stills (e.g., 8000x6000 px) forces the graphic rasterizer to allocate hundreds of megabytes of RAM per image frame buffer. When used as small on-screen picture-in-picture or thumbnail assets, this excessive resolution wastes memory and slows UI caching.
Actionable Fix:
1. Batch downscale high-resolution still images to 200% of delivery resolution (e.g., max 3840x2160 for 1080p, 7680x4320 for 4K) using Photoshop or Squoosh.
2. Convert still images to uncompressed PNG or high-quality TIFF formats to avoid repetitive JPEG decompression cycles during timeline scrubbing.
3. Check NLE RAM consumption in Activity Monitor / Task Manager to confirm system memory allocation drops back into standard operational ranges.
Pro Tip:
Pre-scaling oversized still images to double your project canvas resolution affords ample room for dynamic digital pans and zooms (Ken Burns effect) without choking system memory.
Importing unscaled 24–60 megapixel digital camera raw/JPEG stills (e.g., 8000x6000 px) forces the graphic rasterizer to allocate hundreds of megabytes of RAM per image frame buffer. When used as small on-screen picture-in-picture or thumbnail assets, this excessive resolution wastes memory and slows UI caching.
Actionable Fix:
1. Batch downscale high-resolution still images to 200% of delivery resolution (e.g., max 3840x2160 for 1080p, 7680x4320 for 4K) using Photoshop or Squoosh.
2. Convert still images to uncompressed PNG or high-quality TIFF formats to avoid repetitive JPEG decompression cycles during timeline scrubbing.
3. Check NLE RAM consumption in Activity Monitor / Task Manager to confirm system memory allocation drops back into standard operational ranges.
Pro Tip:
Pre-scaling oversized still images to double your project canvas resolution affords ample room for dynamic digital pans and zooms (Ken Burns effect) without choking system memory.