Is that stainless?. It looks as if the amps are too high so you are having to travel faster to keep control, because the gas shield is moving at the same speed as the torch its moved away before the metal has cooled and allowing oxidation of the hot metal, also if you haven't purged the inside you will be dragging oxidation up from the inside of the joint,
What amps, tungsten, and gas flow are you using.
Make sure you run 3-4 cm beads at a time let it cool with enough post-flow of argon and then continue. Maby you have an argon gas coverage issue ? your torch and argon setup ?
I would use a post flow of about 4 to 5 seconds, I think your amps are a little high, 30 - 35 would be more like it, 6lpm gas flow minimum, a 1.6 tungsten ground to a point, with no more than 10mm stickout, 1mm filler rod, no. 7 or 8 ceramic, good joint fit up is essential as well, when welding make sure that the hot end of the filler rod stays in the gas shield at all times,
To add to what's already been said arc length is very important. Keep it short, ideally around 1, 1.5mm. From the pics it looks like you're losing the torch angle as you progress round the joint which'll mess with shielding quality and increase the arc length
Here is a good shot showing How important is Gas coverage.
After i was seeing black-ish weld i switched to a bigger alumina cup.
Of course 5-10 amperes can make a difference.
Look at the characteristic line of grey at my picture, where the post flow shields and where its not.
Although this oxidation goes away with a brush...
A POINT OF NOTICE IS TO Look Behind your pipe for underhangs. Goops of black metal protruding to your pipe is more serious than a grey weld on top....
For the thicknesses you are welding I would set up at around 35A as a starting point, a couple of amps either way can make a world of difference on thin stuff. Filler would be 1mm and a 1.6mm sharp tungsten and a close arc, post flow no less than 5 seconds. The biggest problem welding small diameter tube is torch angle, as it varies quite quickly as you progress around the circumference. If you allow the torch to lean back then arc length and heat input will increase. Shielding gas is a problem also as it tends to flow around small tubes, leaning the torch back increases the problem by drawing air from behind the cup, this is another reason why maintaining the correct torch angle is critical. Shielding can be improved by using a gas lens and a large cup. As already mentioned, back purge will also prevent the inside of the tube from coking up and being drawn through the pool.