diff --git a/CHANGELOG.md b/CHANGELOG.md
index 9bddc15..d4404fa 100644
--- a/CHANGELOG.md
+++ b/CHANGELOG.md
@@ -21,6 +21,9 @@ All notable changes to this project will be documented in this file. It uses the
 ### 🐞 Bug Fixes
 
 *   `DateTime64` or `Time64` now defaults to millisecond precision ([#76]).
+*   Reading a ClickHouse string into a text column validates its bytes against
+    database encoding, raising an error instead of returning unreadable text.
+    Use `bytea` to preserve raw bytes ([#80]).
 *   Added the `date_time_output_format='iso'` setting to each chDB query to
     always format `timestamp` and `timestamptz` values in plain text formats
     with the ISO-8601 format in UTC, converting `timestamp` values from the
@@ -42,6 +45,7 @@ All notable changes to this project will be documented in this file. It uses the
   [v0.1.1]: https://github.com/clickhouse/pg_chdb/compare/v0.1.0...v0.1.1
   [#76]: https://github.com/clickhouse/pg_chdb/issues/76
   [#78]: https://github.com/clickhouse/pg_chdb/pulls/78
+  [#80]: https://github.com/clickhouse/pg_chdb/pulls/80
   [chdb_hook docs]: ./doc/chdb_hook.md
   [structure]: ./doc/chdb_hook.md#structure "chdb_hook Docs: structure"
 
diff --git a/doc/chdb.md b/doc/chdb.md
index 1622c68..74c9e82 100644
--- a/doc/chdb.md
+++ b/doc/chdb.md
@@ -82,6 +82,11 @@ Output:
 (5 rows)
 ```
 
+## Text Encoding
+
+See [Text Encoding in the `chdb_hook` documentation][text encoding] for how
+chDB byte strings map to Postgres `text`, `bytea`, and other types.
+
 ## Settings
 
 ### `chdb.max_memory`
@@ -181,4 +186,6 @@ Copyright (c) 2026, ClickHouse
     "ClickHouse Docs: max_parsing_threads session setting"
   [table functions]: https://clickhouse.com/docs/reference/functions/table-functions
     "ClickHouse Docs: Table Functions"
+  [text encoding]: ./chdb_hook.md#text-encoding
+    "chdb_hook Docs: Text Encoding"
   [semver]: https://semver.org/spec/v2.0.0.html "Semantic Versioning 2.0.0"
diff --git a/doc/chdb_hook.md b/doc/chdb_hook.md
index 4596241..81826fd 100644
--- a/doc/chdb_hook.md
+++ b/doc/chdb_hook.md
@@ -576,7 +576,7 @@ types:
 | IntervalSecond      | interval                    |                                  |
 | IntervalWeek        | interval                    |                                  |
 | IntervalYear        | interval                    |                                  |
-| JSON                | jsonb                       | Also reads into json             |
+| JSON                | jsonb                       |                                  |
 | LineString          | path                        |                                  |
 | LowCardinality(T)   | T                           |                                  |
 | Map(K,V)            | text[][]                    | One row of text items per pair   |
@@ -586,7 +586,7 @@ types:
 | Point               | point                       |                                  |
 | Polygon             | polygon[]                   |                                  |
 | Ring                | polygon                     |                                  |
-| String              | text                        | Also reads into bytea            |
+| String              | text                        |                                  |
 | Time                | time without time zone      |                                  |
 | Time64(P)           | time(P) without time zone   | P over 6 caps at 6               |
 | Tuple(...)          | text[]                      | Fields become text items         |
@@ -607,6 +607,31 @@ Postgres holds a narrower range than chDB in a few of these types; thus copy
 raises an error on a `Time` or `Time64` beyond 24 hours, and on a `Date32`
 outside the Postgres date range.
 
+### Text Encoding
+
+chDB reads `String`, `FixedString`, `Enum`, and `JSON` as bytes, with no
+guarantee of an encoding. Copying such a column into `text`, or into any other
+non-binary type, verifies bytes against database encoding and raises an error
+for data that cannot represent:
+
+```
+ERROR:  invalid byte sequence for encoding "UTF8": 0x00
+```
+
+Every encoding rejects NULs, which Postgres cannot store in `text`.
+
+Copy into `bytea` to keep bytes as chDB wrote them. Name such these, as
+[CREATE TABLE](#create-table-overloading) derives `text` for these types:
+
+```sql
+CREATE TABLE logs (id bigint, payload bytea) WITH (
+    copy_from = 's3://my-bucket/logs.parquet'
+);
+```
+
+`FixedString(N)` pads shorter values with NUL bytes. Copying into `text` drops
+trailing NULs, while `bytea` keeps all N bytes.
+
 ## Settings
 
 ### `chdb_hook.max_memory`
diff --git a/vendor/pg-clickhouse-c b/vendor/pg-clickhouse-c
index 236da5b..439b710 160000
--- a/vendor/pg-clickhouse-c
+++ b/vendor/pg-clickhouse-c
@@ -1 +1 @@
-Subproject commit 236da5b615c44b2b5b41dbee9972e71ea56b6e1a
+Subproject commit 439b71055777d7e8b03465d94cd86669f0af29f2