> When terminated--as any mile-long line should be--it has a smooth rolloff, 3dB down at the top of the band, with no weird peaks or valleys.
I would not expect significant resonance on such a short line. And on a longer line, copper-loss tends to overwhelm any resonance.
Did you terminate BOTH ends in reasonable impedances?
Did you really connect test gear to a mile of telco wire withOUT isolation transformers??? Seems foolhardy to me.
5,000 feet of cable should be 150,000pFd or 0.15uFd capacitance. Using a too-simple R-C approximation, to be at -3dB at 15-20KC you must have driven with ~60-70Ω. This is now common practice in good broadcast DAs, but if you use a cheap mixer with low-price transformer you may not get that low.
Actually, unless you happen to have been put on a #20 or larger pair, your copper-loss will be greater than 100Ω. But this is distributed: you get a slow droop.
Drive with the lowest possible impedance: 600 is too high.
Hi-Z loading gives highest bass-mid output but the treble may droop 3dB at 15KHz. Loading-down to 100Ω hurts bass-mid more than treble and is "flatter" but everywhere lower than the hi-Z loading.
Frankly, for a fixed-price job, at a mile, just tell them to crank the treble knob.
At my first job, the studio in the academic building was 1,000 feet from the transmitter in the dorm. There were no handy conduits across campus. The previous engineer ordered a telco line, and because it was only 1,000 feet, did not specify any specs. Well, after a lot of terrible audio, he learned that the academic building was wired to the CO in City 2 miles one way, but the dorm was considered residential and was wired to the CO in Town which was 4 miles the other way. To get 1,000 feet, the telco had kindly given him 12 miles of wire jumpered through two COs.