ut 8, Hungary OFTILINK has developed and presented a polarization holographic reading writing system using 1-2 pm thick storage layer on a card form media [1,2,3]. Storage density of 1.3 bit/pm2 has been achieved without multiplexing with sparse code modulation and Fourier-filtering. Because of using thii storage layer the volume data density is extremely high in the system. Further enhancement of the areal density up to -100 bit/km2 can be achieved in an improved new polarization holographic system using a multilayer media instead of the thin film. A 6f system for reading with a suitable second spatial filter allows separation of the signal fiom different layers. This technique combines the advantage of the thin film holography with the idea of the confocal filtering.The thin-film polarization holographic memory card system of Optilink [1,2,3] has practical advantages, e.g. the card works in reflection mode, we do not use hardware servo, it is possible to read and write with different wavelengths, data encryption is possible, there is no problem with material shrinkage, etc. Using thin-film holography the possibilities of multiplexing are limited, but theoretical and experimental calculations show that with an accordingly optimized system (optical elements and codes) the data density can be increased without multiplexing.In optical memories data density is limited first of all by the wavelength and by the numerical aperture (NA) of the readinglwriting objectives. For data density optimization different custom Fourier objectives With high numerical aperture (2 0.74) were designed, fabricated and tested [2,6]. Theoretically in thin-film holographic memories the data density is inversely proportional to the hologram area. In OUT optical system the hologram s u e is controlled by a specially arranged Fourier filtering mirror [2,3]. This mirror cuts the higher diffraction orders. The role of cutting in the Fourier plane is twofold it controls the hologram size and prevents inter-page (hologram) cross-talk. Decreasing the Fourier aperture size causes failure of the optical system resolution which results in a higher inter pixel interference and an increasing bit error rate. -3 -a--= 3,OO 0 2 3 0 '5 2,oo 5 1,50 U 1,oo 2 0,50 0,oo 6% 15% 19% 25% 50% White rate Fig. 1. Calculated data density 0-7803-7379-0/02/$17.0002002 IEEE A well know method to avoid the effect of different error sources is the optimization of codes. Theoretical and experimental investigations show [ 1,4,5,7] that in our optical system the optimal code is a constantweight sparse modulation code. In this case we convert 8 bits ASCII characters into 4x5 binary matrices. The constant number of white pixels in every symbol is combined with disabling white neighbors of each white pixel. This code has advantageous effects to the noise reduction: the sparse code relives the effect of 2D interpixel interference arising from the low-pass nature (Fourier cutting) of the optical imaging system and the diffracted light intensity per pixel increases by using fewer...