Arsthbt

I study in Berlin and I've noticed that long vowels preceding an R have a very different quality. In specific, they are consistently more open, sometimes mixing with their "short vowel" equivalent.
For example, the starting vowel of the diphthongs in:

Wer ([ve̞ɐ] or even [vɛɐ]), der (de̞ɐ, can go as low as [dɐ]!), Leer

doesn't sound at all like the vowel in

Tee [tʰe:], Bete [be:tə]

Same can be said for the other vowels, but "e" and "o" seem to be affected the most.

Is this considered standard German? Has anyone written anything about it (I love reading books about phonology/phonetics)
Here is a 4 seconds recording of Helmut Krauss on Audible:
https://drive.google.com/open?id=14oobrQXIRCIQmyv4GnR7MNj0wQyRB4zJ

In Northern Germany, the R is pronounced as a gliding ɐ even after short vowels. Can we say that in Northern Germany there is no contrast between short and long vowels before R?

I'm inclined to answer negatively as far as Standard German is concerned. I know you mentioned Northern Germany, but then again your recording is definitely Standard German.

German vowel contrasts combine length and tenseness: [i: ɪ] [y: ʏ] [u: ʊ] [o: ɔ] and finally [e: ɛ], [ø: œ].*

Closed tensed long [e:] and open untensed short [ɛ] are clearly distinct, even if [ɐ] follows: lehrt, kehrt, mehrt, sehr, entbehrt – lernt, merkt, nervt, bergt. They remain fully distinct even when [e:] is shortened, for instance when it is prosodically unstressed (e.g. sie lehrt jetzt in TÜbingen).

Some words such as der, wir, ein have weak forms. So if you point to the word and ask someone to pronounce it they will say [de:ɐ̯ vi:ɐ̯ aɪ̯n] but in a sentence, they might be realised as [deɐ̯] > [dɛɐ̯] > [dɐ], [viɐ̯] > [vɪɐ̯] > [vɐ], [n]. This is completely normal and part of Standard German. However, these sort of reductions are not accepted for words such as leer, sehr, etc.

Caveat: Lack of stress and speed of articulation will automatically lead to tensed vowels such as [e] being realised as less tense, i.e. they will become more open. But even then, the minimal pairs mentioned in the paragraph above should remain distinct. (Also note that nothing I have said strictly precludes that R realised as [ɐ] does have an influence on the preceding vowel.)

*_{Note that the situation for e-like vowels is more complicated than for the others, because [ɛ:] also exists, but not for all speakers.}