Video transmission is a very challenging task in opportunistic networks, due to their coarse characteristics of frequent disruptions and uncertain data delivery. In a previous work [1], we presented promising results of scalable video transmission in opportunistic networks. Users’ viewing experience was taken into account, by the means of playout delay and perceived quality. However, there was still a space for the results to be bettered, and we suggested to change the transmission unit from SVC layers to their composing NAL parts. In this paper, after reviewing the last work and results, we implement the new idea, outline its pros and cons and propose a new integrated solution. The new proposal is a network density aware packetization technique, which dynamically reforms the transmission units into larger chunks with a higher probability of successful delivery. Both proposals are experimentally evaluated using the opportunistic networks environment simulator