The Televes noise may be close to 3 dB as described in this Solid Signal review:
“Televes’ mast-mounted amplifiers are a little different from other amplifiers you’ll find on the market. They’re sort of a mix between preamplifier and distribution amplifier — the power levels are extremely high, but then you have the ability to combine multiple signals and, at least according to the manufacturer, there’s a higher noise figure than you normally find in a true preamp.
The company is really addressing the way amplifiers are used in the 21st century. When all TV broadcasting was analog, the preamp itself was more important for pulling in distant signals, and the noise figure was something that was absolutely all-important. After all, it was possible to pull in a very weak signal and amplify it so it would be stronger, but you certainly didn’t want to add a lot of noise as well.
Today, the main reason for amplifying is to overcome long cable runs or distribute the signal to multiple TVs. Televes’ engineers built their amplifiers for that purpose, making them super-powerful, and also adding the ability to power antennas that, themselves, are already amplified.
In real world testing, gain from this amplifier was extraordinarily high, averaging 26dB throughout the usable range. As you can see from the chart, the strongest gain was in the lower end of the UHF spectrum which is perfect for most broadcast applications as the majority of broadcasts are in that range.
Various documents available on the web put the noise figure of these amps between 7.5 and 10 dB. That seemed very high even for a distribution amplifier, which is more of what this amp is. Without a full laboratory to test, our labs tried two different methods to derive the real-world effect of noise on the amplifier.
Using mathematics and analyzing the difference between the signal-to-noise ratios of unamplified and amplified modes, the data suggest that the amplifiers add about 2.75dB of noise when operating, far less than the rated number.
Using a secondary methodology, the antenna’s output was “padded” by 26dB to remove the increase in gain. Therefore, the difference between the signal-to-noise ratio of an unamplified antenna and the signal-to-noise ratio of an amplifed antenna should be the amount of noise inherent in the amplifier. The result, 3.15dB, is very consistent with the mathematical result and the difference between the two numbers could be attributed to minor differences in reception as the tests were taking place.
As a control test, the same measurements and methods were applied to a Winegard LNA100 amplifier, which has a published noise figure of 1dB. The Winegard amp’s number, using those methods, was 1.15dB, which suggests that the calculations are reasonably accurate.
When it comes to Televes’ published specs, they seem largely attributable to extreme honesty and scrutiny in the testing process and real-world results for this amp are far more consistent with other amplifiers with similar power levels.”