What is the purpose of this project?

We are taking a fresh look at the humpback whales’ vocalizations, both song and non-song. During Phase 1, just like with all other projects (Manatee Chat and Dolphin Chat), we will be simply identifying whether or not humpback whale vocalizations are present in the file. The next Phase will be classification of calls and the final phase will be matching calls to see if we can find individually recognizable calls. Additionally, we will be using both historical data (from way back, 1960s and up) and current data to look how much change (or consistency) we can discover.

When and where were all of these recordings made?

We are starting with data from a number of publically available databases and will focus first on the Atlantic Ocean humpback whale vocalization recordings.

Is this project similar to Dolphin Chat or Manatee Chat?

Curiously, this project is more similar to Manatee Chat, because we will be focusing on individual calls. But it also has similarities to Dolphin Chat, as we will be looking to identify other acoustic events, like boat noise, since these sounds are nearly always present in many habitats of the Atlantic humpback whales.

What is the “Other” category and what should we mark as “Other?”

Humpback whales are known to create sounds by breaching, splashing their tails or flippers. These sounds are not produced the same way as song vocalizations, or non-song calls, but are still thought to be a part of humpback whale communication repertoire.
Additionally, there could also be some other interesting sounds present, since these recordings were made in the open ocean.

What is the final count of subjects?

This is a long-term and we will be adding datasets as we go along. We estimate that the total number of subjects will be around 20000-25000 files.

Are there breathing sounds present in the recordings?

No, because just like dolphins and manatees, humpback whales do not breathe under water.

Some calls are very faint; what should I do if I encounter those?

We recommend to be conservative in your classifications, so if the vocalizations are faint, but you can still hear /see them on the spectrogram, you can mark it as a humpback whale vocalization.

There are many overlapping vocalizations in some files; why is that?

Humpback whales are known to chorus often when they sing their songs, so these overlapping vocalizations are from several humpback whales chorusing together. Some vocalizations could be louder/better quality than others in the chorus, meaning that the humpback whale was closer to the hydrophone.

Why are there snapping shrimp sounds in Whale Chat files? Aren’t these recordings made offshore in deeper waters?

Surprisingly, sometimes you will hear snapping shrimp in the recordings. This happens because recordings were made in reef areas or near the shore where snapping shrimps were abundant. However, compared to Dolphin Chat recordings, snapping shrimp sounds will be less frequent in Whale Chat recordings.