Filter entries in an aggregation by wildcard string?

[very_sneaky]

hey all, I'm having trouble formulating a query, and I'm hoping someone can help. This is the format of my data:

{
    "srcIp": "192.168.0.54",
    "dstIp": "74.125.195.147",
    "srcMac": ["ec:f4:bb:4f:b0:96", "0a:f7:bc:44:c0:d7"],
    "dstMac": ["78:54:2e:9f:10:28"],
}

My objective is to return a list of unique srcIp's, and count the number of srcMac's for each unique srcIp that have the prefix ec:f4:bb.

The query I have at the moment is:

{
    "size": 0,
    "aggs" : {
        "unique_ips": {
            "terms": {
                "field": "srcIp",
            },
            "aggs": {
                "num_macs": {
                    "cardinality": {
                        "field": "srcMac"
                    }
                }
            }
        }
    }
}

And this works great; it returns the number of srcMac elements for each unique srcIp. However, it doesn't match the prefix string. To be explicit, for the provided sample data, i want it to return srcIp: 192.168.0.54 and num_macs: 1, as only the first element of the srcMac list matches the prefix ec:f4:bb.

I've tried a number of combinations of filter, filters, nested, match, and wildcard directives, but haven't been able to get it to do what I want. For some of them, it looks like they may not work with aggregations, and others it's probably likely that I just don't know what I'm doing.

Can anyone point me in the right direction/give me an example of how I can get what I want out of this? Any answers are appreciated.

Cheers.