Synchronization between locations is an important factor for enabling remote shared experiences. Still, experimental data on what is the acceptable synchronization level is scarce. This paper discusses the synchronization requirements for watching online videos together — a popular set of services that recreate the shared experience of watching TV together by offering tools to communicate while watching. It studies the noticeability and annoyance of synchronization differences of the video being watched, as well as the impact on users’ feelings of togetherness, both for voice chat and text chat. Results of an experiment with 36 participants show that when using voice chat, users notice synchronization differences sooner, are more annoyed and feel more together than when using text chat. However, users with high text chat activity notice synchronization differences similar to participants using voice chat.