At every stage of Sindarin’s conceptual development (including Noldorin and Gnomish), voiced stops became spirants after vowels: [Vb], [Vd], [Vg] > [Vv], [Vð], [Vɣ]. These same sound changes occurred in Welsh (WGHP/§103i, §104i). This phonetic development was foundational in Sindarin’s system of soft mutations whereby b, d, g became v, dh and ’ (indicating a vanished sound). Tolkien described this sound change in numerous places:

• Thus the changes whereby the distinction of the 3rd or a-series and the 4th or z-series was revived; and the letters r 4 f etc. were restored to the notation of voiced spirants — absent from ON but redeveloped, either at the very end of the pre-Exilic period or in early Exilic — were simply returns to the original Feanorian scheme (from notes on the Feanorian Alphabet from the 1930s, PE22/31).
• Since the voiced spirants v, ð and archaically ʒ, developed from ON b, d, g between vowels, became common in Exilic and were denoted as originally in the Feanorian system by r 4 f ... (ibid., PE22/32).
• DH represents the voiced (soft) th of English these clothes. It is usually related to d, as in S. galadh “tree” compared with Q. alda (LotR/1113).
• In Sindarin voiceless stops (i.e. p, t, k) before nasals became voiced > b, d, g, and then together with the original voiced stops in this position became nasals before homorganic nasals (tn, dn > nn; pm, bm > mm), but before other nasals became spirants as generally medially [emphasis added] (from notes on Elvish numerals from the late 1960s, VT42/26).

Examples of this sound change are numerous in both Sindarin and Noldorin. Here is a representative sample:

• ✶nebā > S. nef “on this side” (PE17/27).
• ✶edelō > S. Edhel “Elf” (PE17/141; WJ/364).
• ✶nāba-grota > nǭv-ʒrot > (North) S. Novrod “Hollowbold” (WJ/414).
• ᴹ✶ubrā > ofr (ovr) > N. ovor “abundant” (Ety/UB).
• ᴹ✶wedā > ON. weda > N. gwedh “bond” (Ety/WED).
• ᴹ✶Tūgore > tūghor > N. Tuor “strength-vigour” (Ety/GOR, TUG).

The sound change of [Vb] > [Vv] is sometimes obscured by the orthography of Sindarin and Noldorin, since Tolkien often used “f” to designate [v], especially at the end of words. Similarly, the sound change of [Vg] > [Vɣ] is obscured by the fact that [ɣ] later vanished, or it became a vowel before [l], [r], [m], [n] forming various diphthongs, though the exact details of these sound changes varied at different conceptual stages of the language.

Conceptual Development: The soft mutation systems of the Gnomish of the 1910s and Early Noldorin of the 1920s show the same lenitions of b, d, g to v, dh and ’ (PE11/7; PE13/120), which means they certainly had the same phonetic developments of voiced stops as in later Sindarin and Noldorin from the 1930s and forward. Analysis in this early period is complicated by the fact that voiced spirants were also part of the phonetic inventory of Primitive Elvish in the 1910s and 1920s, as pointed out by Roman Rausch in his Historical Phonology of Goldogrin (HSG/§2.3):

Note, however, that it is difficult – in fact practically impossible – to decide whether a root originally had a voiced spirant or a voiced stop, as both Qenya and Goldogrin turn them into spirants. For example, both *gweđ- and *gwed- would yield gwedh- in Goldogrin and ’wer- in Qenya (perhaps d > dh is a Common Eldarin change).

Fortunately, Tolkien himself sometimes indicated cases where a primitive form contained a stop (D) versus a spirant (Ð). This gives us a few clear examples of the spirantalization of stops in both Gnomish and Early Noldorin:

• ᴱ✶aduen- > G. adhwen “avenue” (PE13/108).
• ᴱ✶(s)wada > ᴱN. gwadh “bark, skin, peel” (PE13/146).
• ᴱ✶Eid(e)lā > ᴱN. Uidhel “elf, fairy” (PE13/155).
• ᴱ✶Ngol(o)dō- > ᴱN. Golodh “Gnome” (PE13/145).
• ᴱ✶labna > ᴱN. lafn “*animal” (PE14/70).

Clear examples of [Vg] > [Vɣ] are harder to find, but there is solid indirect evidence of this sound change in the early soft mutation systems (see above) and examples of early vocalizations of [g]:

• ᴱ✶dagla [probably > *daʒla] > ᴱN. dail “axe” (PE14/66).
• ᴱ✶dagnā́ [probably > *daʒna] > ᴱN. dain “high, noble” (PE13/141, PE13/161).

See the entry on how [ɣ] vocalized before [l], [r], [m], [n] for further discussion.

GS/§4.86 @@@

HSG/§2.3 @@@ in GLS