Talk:Noise floor

Factual accuracy

I didn't tag this one as being factual inaccurate, but I also believe that it has problems. For example, the figure at the bottom right says that signals below the noise floor cannot be measured. This is not always true, as GPS signals lie well below the noise floor but can be measued by integrating them over time. Someone much more knowledgable than I on this topic needs to do some work on this article. Thanks, --CapitalR 19:34, 22 June 2007 (UTC)[reply]

GPS signals are far above the noise floor: they have noise of a few meters measuring the distance between receivers and satellites. Averaging does help with GPS signals, but can't get past the noise floor limit. — Laura Scudder ☎ 20:58, 22 June 2007 (UTC)[reply]

I don't believe this article is inaccurate, I am removing the tag. However, I will also weaken the absolute statement at the end, because there are situations where signals below the noise floor can be extracted. For example, when using SDR for EME, DSP algorithms can remove much of the man made noise while looking for CW signals. This is surprisingly effective, but requires A/D converters with enough bits to overcome the noise floor, predictable noise, and a predictable signal. --ssd 14:25, 6 August 2007 (UTC)[reply]

Please refer to spread spectrum techniques and theory. 20080326

Not to mention that the picture in this article actually contradicts the claim that signals below the noise floor can be measured... Vedant (talk) 12:12, 28 October 2009 (UTC)[reply]