jks

Like the first suggestion and I almost implemented the second on this last update.

I think I need to loosen up the criteria for "GPS is okay" so you don't have as much problem with the third issue. Same for "all channels in use".

The fourth issue of overlapping/obscured markers is definitely a problem. I like your idea of additional highlighting of the likely position marker.

And I agree about the button names..

Interesting idea. Remember that you can always double-click on the sampling station marker (blue) to open a regular Kiwi connection window in a new tab/window for monitoring. But having a little "snapshot" button on each line of the sampling station list that shows a short waterfall is a neat idea.
I'll look into it.

This has been discussed at length before, but I'll summarize.

The current FPGA logic behind the OV indicator is as pessimistic as it could be and almost certainly wrong. Right now if one single sample overflows the ADC the OV indicator will get lit and persist long enough to be seen. That doesn't really indicate a substantial overload condition that will lead to increased IMD products seen in the waterfall etc. The logic really needs to be a threshold/averaging function of some sort.

The +20 dB amplifier / matching buffer ahead of the ADC is a fixed gain block. So yes, you'd have to have new OV logic trigger a GPIO pin and control an external attenuator (with appropriate hysteresis so it isn't pumping on and off all the time).

Yes, this is an actual bug. I had an email about it earlier today and have a fix that should go out in tomorrow's update.

The area you've identified turns out to be exactly the transparent box that the audio FFT explanation text appears in. I forgot to enable mouse events for the box (HTML div actually) so that's why you get the dead zone.

Any difference between the waterfall carrier of a known reference station and the frequency scale comes down to frequency calibration of the ADC XO clock.

At z14 each waterfall FFT bin is roughly 0.9 Hz. But how this translates into what you see on your browser screen is a little muddled because those 2048 bins are "stretched" and anti-aliased to fit your actual visible screen width. But from experience the one above (not yours I know) looks fine for 5 MHz WWV.

Do you have GPS clock correction working? Without GPS correction XO temperature variation may make the frequency offset at z14 look much worse than what you have. To see the exact ppm and Hz difference use the manual clock correction procedure (see admin page "config" tab). The XO spec is +/- 50 ppm @ 66.67 MHz without GPS correction operating.