Merge experimental LZX compressor

[wimlib] / doc / imagex-capture.1.in

diff --git a/doc/imagex-capture.1.in b/doc/imagex-capture.1.in
index 18d07ed6c40831e5ca76d99ea8833d80c006ad12..08ac4226735bb10a22328ab3ef8fc0340b51c93a 100644 (file)
--- a/doc/imagex-capture.1.in
+++ b/doc/imagex-capture.1.in
@@ -184,7 +184,7 @@ for \fB@IMAGEX_PROGNAME@ capture\fR, since the compression mode for
 is automatically set as such).
 \fITYPE\fR may be "none", "fast", or "maximum".  By default, it is "maximum".
 This default behavior is different from Microsoft's ImageX, where the default is
-"fast". \fB@IMAGEX_PROGNAME@ capture\fR instead gives you the best compression
+"fast". \fB@IMAGEX_PROGNAME@ capture\fR instead gives you the better compression
 ratio by default and makes up for the slightly slower compression by being
 faster than Microsoft's software in the first place and using multiple CPUs when
 available.
@@ -192,6 +192,15 @@ available.
 You may also specify the actual names of the compression algorithms, "XPRESS"
 and "LZX", instead of "fast" and "maximum", respectively.
 .TP
+\fB--compress-slow\fR
+Spend even more time compressing the data in order to achieve a higher
+compression ratio.  Currently, this only has an effect with LZX ("maximum")
+compression.  Depending on the data, compressing with this option will take
+around 10 to 15 times longer and produce a LZX-compressed WIM about 1% to 5%
+smaller than one produced with the default LZX compressor.  Depending on the
+data, the resulting WIM may be approximately the same size (typically no more
+than 0.4% different) as a LZX-compressed WIM produced by WIMGAPI.
+.TP
 \fB--threads\fR=\fINUM_THREADS\fR
 Number of threads to use for compressing data.  Default: autodetect (number of
 available CPUs).