Describir: Speech recognition threshold in slightly and fully modulated noise for hearing-impaired subjects.