/x/ · velar fricative (j / ge / gi) Spanish j (and the g of ge / gi) is a voiceless velar fricative. The Mandarin h [x] is essentially this sound, so Mandarin speakers usually get this one right. What to watch: it is tighter than the Mandarin «hē», with more friction and further back. Spelling has three sources, and the x of México belongs here too.
Four steps to /x/
Confirm the gesture, then learn where it changes, then put it in words.
Start from the Mandarin h, then tighten
The initial of «hē» is already [x]. Spanish x has the same place with stronger friction. Start from h and raise the tongue back slightly until the friction is audible.
Learn the three spellings
j: jamón, trabajo.
g + e / i: gente, gira (but g + a/o/u is the g sound).
x: México, Oaxaca (place names, a handful of words).
Do not make it as light as the Mandarin h
Too little friction and it approaches zero — trabajo starts sounding like trabao. The friction has to be audible.
It weakens to h in many Latin American varieties
In the Caribbean and Chile x is often [h]. That is a dialect feature, not an error — but while learning, aim for the stronger [x].
Spelling: three spellings, one sound (x has two faces)
j and ge/gi are the rule; x is a historical leftover.
Word-initial and word-medial: always [x], no exceptions.
Note that g + a/o/u is [g] (gato, gusto). This is one of the most frequently tested spelling rules in Spanish.
Only in a handful of place names and their derivatives (mexicano). Everywhere else x is [ks]: taxi, texto, exacto.
Which one are you
The three most common readings.
My j is too light to hear
Not enough friction — trabajo turns into trabao. Raise the tongue back.
gente comes out with a hard g
The g of ge/gi is x, not g. No exceptions apart from loans.
I say México with ks
This x is x, not ks (place names only). But taxi and texto really do have ks.
Word list
11 words grouped by where the sound sits in the word. Click a word to hear it. These are recorded as separate words, not cut out of the sentences above — a word cut from a sentence carries its neighbours' transitions and always sounds muddier. All three groups use the same speaker, so what changes between them is position, not voice.
Word-initial3
Word-medial8
Corpus and acoustic measurements
Click any word in a sentence to play from that word. The pale band over the waveform marks the measured boundary of the target word.
Per-token data
Every number below is measured directly from the audio, not hand-annotated.
| Token | Accent | Sex | Target | Target word | /x/ duration | F0 | Pitch range | HNR |
|---|
Where these numbers come from
12 synthetic sentences from two Peninsular voices; phoneme boundaries from MMS_FA.
Corpus: 12 Spanish sentences whose target words contain x, with the target word never sentence-initial or sentence-final (those positions are distorted by pauses and sentence stress).
Boundaries: MMS_FA forced alignment; word and phoneme boundaries come from the same pass. The target phoneme is looked up in the word as declared by the page, never taken blindly from the onset.
Target duration: the column is the length of the target phoneme itself, not the whole word. The «target word» column is the word.
All 12 target words contain x, in onset (jamón) and medial (trabajo) position, across j, ge/gi and place-name x spellings. ★ This page reports duration only, not spectral centre of gravity: x (1.5–2.5 kHz) is separable from s (4–8 kHz), but with no s items here a single number would carry no information. A proper s/x comparison would need its own page.
What this page does not do: no paired differences (there is no pair to compare), and no verdict on whether you pronounce it correctly — the durations describe native speech, not a scoring standard.