cant find the leak

What you need is something that you can trace. Why don't you look on the market for a low power, short lived, radioactive source. Mix it in water and flush it into the situation. At least you'll be able to track it no matter where it goes... And being short lived, it won't leave anything bad behind. That's how the big-boys do things.

How about adding some dye at the meter? Make sure it's food quality!

Why food quality? I'm just wondering where we're going here?

I'm assuming this leak is in a domestic water supply. I wouldn't want to put poison there!

If he could not find the water, how would he be able to find the colored water?

Assuming there is some natural moisture, and depending on the quantity of the leak, the water alone might not be obvious, while a change in color should be.

A whole lot depends on the kind of soil and/or subsoil/bedrock. If this is a significant leak, then the soil must be rather porous to some depth.

Based on your description I would say you have (only) the sound of (what could be) a leak, but not the substance of a leak. You said (or implied) that your test revealed loudest sound near an exterior wall (and foundation footing), but no unusual soil moisture exterior to the foundation footings when you bored down outside. You said the area your test localized was close to the utility entrance line to the house. So let us re-think your situation based on what you know so far. The "test sound" being close to a wall, there's a good chance this might be the location of a branch distributing water to different locations in the house. Or perhaps to a sprinkler grid?) One thing you might have tried to get more information from your test would have been to open and close faucets, one by one, throughout the house (and outside faucets as well) to see what effect it had on the sounds detected by your test apparatus (or ears). It could be that there is really no leak at all, only a partially open valve somewhere...such as in a water closet. In other words, you need to eliminate possibilities other than a leak into the soil under the slab. Please note that a leak does not have to be in close proximity to the sound of running water; in fact it can be all the way on the other side of the house.

You could also try boring away from the test sound in a generally down hill direction (in Dallas, most likely towards the curb). Stratification and cracks in the "white rock" common to Dallas can provide a path way for water to flow undetected to a distant point where it might emerge (something like an artesian well), or build up pressure (and cause soil/curb/pavement lifting) or cause subsidance (soil/curb/pavement sinking). Sometimes the direction of escaping underground water (if there truly is any) can be sensed as overmoist soil at a distance from the actual leak.

Another suggeston would be to replace the line where the sound is detected...since you've already dug down. It could simply be that leakage is occurring all along the length of the line, but not fast enough to saturate the soil noticably in any location.

In summary, you first need to eliminate the possibility that there is not actually any leak.

Most water lines have a pressure limiter. Increase the pressure to the max or remove the pressure reducer totally and let the pressure from the systems pump show you where the leak is.

... and blow out the water heater! (unless there is an appropriate shut-off valve)