In youtube movies of argon-neon based HPSs startup i found that the mercury takes away the control of the discharge from the buffer gas very fast resulting is color shift from red to blue in less then 5 seconds from the ignition. This is not the case of xenon based HPS when the mercury start to take away the discharge only when the sodium start to vaporise (The lamp remains white for 45 second until the mercury and the sodium both takes away the control of the discharge). In simple words: mercury in argon-neon based HPSs vaporize much more fast and early (5 second after ignition) then xenon based HPSs.

Wait... red? I've never seen a red color given off by HPS lamps. I've always seen white, then orange, then orange-white during the warmup. That means I've only ever seen xenon buffer gassed lamps?

Well neon-filled retrofit lamps produce a red light shortly after turning on, but I've never seen red in standard HPS lamps (they start blue or white).

Edit: but I've seen some lamps which after reaching their final color are very reddish, not orange like normal HPS.

I know that argon-neon HPSs are retrofit for MVs, but in these lamps the color change to mercury much more early then standard xenon HPS after the ignition (Argon-neon: after 5 sec, xenon: after 45 sec [Or erlier if the lamp is older]).

I don't have any HPS retrofit lamps that use the Ne-Ar fill, but I have found that most HPS lamps I have start off white with a hint of blue (xenon discharge) and usually a bit of Na vaporizing at the arc tube ends (electrodes - or maybe this is due to incandescence of the electrodes?) - then, the arc tube gradually fills with more orange from the ends to the centre and becomes kind of a pinkish colour, then a very saturated yellow (low pressure sodium) which gradually gets more golden-white as the lamp heats up to full pressure & temperature.

Older, and higher-wattage lamps often stay white for a longer time at startup, and will show a mercury blue discharge, sometimes for quite a while, before the sodium starts to vaporize. But I've had some that never do this even when they start cycling at end of life. I wish they would! I do notice that lamps that have a pinkish colour when fully lit are most likely to have the cool colour changes on startup, and are also close to end of life, usually.

That colour change is what I like best about HPS lamps, as it was what got me interested in learning about HPS in the first place.

I've seen some lamps which after reaching their final color are very reddish, not orange like normal HPS.

I'm familiar with this effect, too. Some time ago I saw it discussed in a German forum, but no real explanation for it was found there. Any ideas what exactly causes such red hue in HPS lamps?

Since HPS lamps get most of their blue lines from mercury, isn't it possible that a reddening lamp would indicate lack of mercury?

Xytrell and kai: Usually the blue in the spectrum of HPS lamps is form molecular radiation of the sodium at high pressure. In a reddening HPS lamp, there are much lower sodium and the red light is because of the mercury. It relases from the HG-NA amalgam and its pressure increase and so its radiation. As a result the sodium line shrinks from the green wavelengths and broads to the red wavelength until its broads only to the red of the specturm (Like the sodium iodide in MH lamps). This extremely broadening of the sodium line only to the red and not also both the red and the green at equal levels is because of the very dominancy of the mercury and is the reason why sodium depleting HPS lamps look redded out.