People make gas discharge tubes like neon lights for example. There are also many demonstrations of gas glowing tubes. Each gas has a different color and some are very close to white. I know it may produce some UV, but I know that the quartz bulbs on mercury vapor prevent most UV.
These demonstration tubes I see on videos are often very dim. Is there not a way to make a bright light with just gas in a tube? What's preventing more voltage?

I'm asking this cause I want to try to make a gas tube myself, but I want it to be bright like a lightbulb, not like a glow tube you see in demonstrations.

Neon signs?Those are good and bright.Good ones for light collections are beer signs since they are common.Since they can be more than one color there could be diffrent gases or phosphors in the tubes.

There are few major problems:
Most gasses emit majority of their photons out of the visible spectrum, only small part is visible.
If you mix gasses, most often the gas with the lowest energy takes over.
And there are only few gasses that are inert enough even in the plasma environment to not corrode the tube components in short time.

So the most efficient and reasonably lasting way to make tubes of any color is the use of UV radiating mercury (with some noble gas as a buffer) and then a phospor converting that to the desired color.

There are few major problems:
Most gasses emit majority of their photons out of the visible spectrum, only small part is visible.
If you mix gasses, most often the gas with the lowest energy takes over.
And there are only few gasses that are inert enough even in the plasma environment to not corrode the tube components in short time.

So the most efficient and reasonably lasting way to make tubes of any color is the use of UV radiating mercury (with some noble gas as a buffer) and then a phospor converting that to the desired color.

Thanks. I wonder why do fluorescent lights run such a low voltage compared to neon? I don't think it's the neon gas itself cause it seems like when they use other gases in neon tubes, it's still in the thousands of volts. Maybe it's pressure?

I'm going to try to get some neon electrodes and a tube. I already have a vacuum pump and some gas. I never welded glass before, so it will be interesting. I want to try to make an actual light bulb with it.
I read neon lights are more efficient (maybe it's not true). A phospor coated neon running at super high voltage maybe is more efficient than fluorescent then.

Good questions! I am curious too, so this is just speculation. It does seem if run at lower voltage neon is dim, as in a starting SOX lamp despite higher current. I’ve seen them burn in neon tubes at extra high voltage and they’re almost blinding bright and the electrodes glow red like an HID! This was on how it’s made and frightening. I think in vapor lamps, that metal vapor adds so much more hunger for current (hence needing a ballast to prevent run away conductivity and exploding arc tube). I think maybe it’s actually impossible to feed a vapor lamp thousands of volts especially once the metal is melted and vaporized at full warm up as the more voltage would push more current to overload. This is only speculation though.

Second thought here, kind of reminds me of LED’s!! You can have a direct chip and a direct gas but neither makes a comfortable living area light or enough. So both instances they take it the route of an exciter color that’s efficient for the chip or gas and let the phosphors take over from there just make different phosphor recipes! Guess that doesn’t cover metal halide where a mix of vapors does the trick right off the plasma.

Thanks. I wonder why do fluorescent lights run such a low voltage compared to neon? I don't think it's the neon gas itself cause it seems like when they use other gases in neon tubes, it's still in the thousands of volts. Maybe it's pressure?

It is the gas composition, pressure and tube shape (length and thickness).

I'm going to try to get some neon electrodes and a tube. I already have a vacuum pump and some gas. I never welded glass before, so it will be interesting. I want to try to make an actual light bulb with it.
I read neon lights are more efficient (maybe it's not true). A phospor coated neon running at super high voltage maybe is more efficient than fluorescent then.

First there is a big technical difference between a true Neon discharge and what the generic term "neon tubes" uses to be used for.
The true Neon tubes are practically always reddish in color, as that is what the Ne is radiating. These tubes then use to be clear.
The colorful gliwing tubes, which use to be white substance coated on the inside (assume the eventual outer surface paint removed) are practically all mercury fluorescent tubes, so the main radiating component in the plasma is Hg, generating UV which is then exciting the phosphor layer and that generates the final light.

Now when speaking about efficacy, low current density thin narrow tubes are indeed favorable design for high efficacy, but the signage business needs extra properties: Not that high brightness and instant, harmless (not imposing any extra wear) starting and simple connection of multiple tubes to a common power source and a long service life (in the 100khour ballpark). Plus it should be compatible with not that (chemically) clean and accurate hand manufacture of the final tubes (emission coating on hot cathodes is just extremely easy to poison), allowing reasonable storage time for the components (completed hot cathode assemblies have short shelf life, hot cathodes usually are sealed into the final product minutes after being made, while signage cathode assemblies are lying in a shelf for weeks or recently even years before being used for final tube making, yet still are able to yield the needed tube life).
That led to a cold cathode (no warmup time, no reactive chemicals, so easy to hand assemble with simple heat degassing cleaning), low current high voltage tubes. The low current is there essential to keep the cathode losses acceptable, as cold cathodes tend to have really high nonluminous voltage drop between the metal vs plasma (in the 100'sV, against about 10..15V for their hot, thermionic emission based counterparts).
And last thing the nature how the lamps need to be driven means the ballasts used to feed them use to have really high losses, compare to the ballasts for hot electrode lamps.

Even with the low current and high voltage, the efficacy is rather poor (in the lower 10's lm/W), compare to a standard (e.g. 36W; ~80 lm/W when including ballast losses) fluorescents. However the efficacy of the CCFL does stay even when scaling down the power as the efficacy of the hot electrode lamps does (heating power does not scale downthat much), so with power levels of 1..5W the CCFL could be more efficient (include the ballast losses; it retains its 30lm/W ballpark, but the hot electrode fluorescents fall bellow 20 lm/W at these power levels). Together with their thin tube format nature, this made them an ideal LCD backlight source (till LEDs).