Sir,

We thank Drs Atan, Foy and Scanlon for their interest in our study. Our work was retrospectively done on digital image bank photographs for which we wish to define the effect of compression on detection and not the effect of initial image size on detection. In spite of this our conclusion still shows first, that 1.26 MB to 118 KB compression (1:11) remains adequate (which is in accord with the English National Screening Committee's recommendation upon compression ratios), and second that larger image sizes than those we used must be tested clinically. We also think that ‘bigger is better’ concerning retinal images.