sane-project-website/old-archive/1998-04/0127.html

<!-- received="Mon Apr 13 15:17:10 1998 PDT" -->
<!-- sent="Tue, 14 Apr 1998 00:01:46 +0200" -->
<!-- name="Henning Meier-Geinitz" -->
<!-- email="henningmg@gmx.de" -->
<!-- subject="line-distance correction and mustek 600 II N" -->
<!-- id="" -->
<!-- inreplyto="" -->
<title>sane-devel: line-distance correction and mustek 600 II N</title>
<h1>line-distance correction and mustek 600 II N</h1>
<b>Henning Meier-Geinitz</b> (<a href="mailto:henningmg@gmx.de"><i>henningmg@gmx.de</i></a>)<br>
<i>Tue, 14 Apr 1998 00:01:46 +0200</i>
<p>
<ul>
<li> <b>Messages sorted by:</b> <a href="date.html#127">[ date ]</a><a href="index.html#127">[ thread ]</a><a href="subject.html#127">[ subject ]</a><a href="author.html#127">[ author ]</a>
<!-- next="start" -->
<li> <b>Next message:</b> <a href="0128.html">Henning Meier-Geinitz: "Performance mustek 600 II N"</a>
<li> <b>Previous message:</b> <a href="0126.html">Oliver.Rauch@Wolfsburg.DE: "Re: /proc/scsi/scsi"</a>
<!-- nextthread="start" -->
<!-- reply="end" -->
</ul>
<!-- body="start" -->
I looked into the line-distance correction code for the mustek 600 II<br>
N in mustek.c. Some details are explained in an earlier article. With<br>
the patch color works somewhat, but there are missing lines at the<br>
end. The problem is: The scanner sends data one line after the other<br>
in the following manner (Example):<br>
<p>
R-R-R-R-B-B-R-B-G-R-B-G ... R-G-B-R-G-B<br>
<p>
If n is the distance between two CCD lines, sane loses 3*n data lines<br>
at the end of the scan. The line-distance code in sane asumes the<br>
following: <br>
<p>
R-R-R-R-B-B-R-B-G-R-B-G ... R-G-B-G-B-G-B-G-G-G<br>
<p>
If I understood correctly, mustek.c would have to scan n more lines<br>
then specified to get the full sized image. I don't know most<br>
internals of mustek.c so I can't do this.<br>
<p>
By the way: My 600 II N doesn't need the line-distance-fix, the<br>
line-distance values which it sends are correct. I don't know, if this<br>
applies for all 600 II N.<br>
<p>
The following patch corrects the color scanning problem with the 600<br>
II N as far as I can do it. It will (hopefully) fix the problem with<br>
not ignoring empty lines in mustek.conf. And I put a usleep in<br>
sanei_ab306.c. This will allow to run sane without turning on<br>
debugging (which was the only way to get sane with the 600 II N to<br>
work for me). I don't know, why it works, but it does.<br>
<p>
<p>
diff -ru sane-0.72.orig/backend/mustek.c sane-0.72/backend/mustek.c<br>
--- sane-0.72.orig/backend/mustek.c	Tue Mar  3 04:02:43 1998<br>
+++ sane-0.72/backend/mustek.c	Mon Apr 13 21:25:53 1998<br>
@@ -1394,7 +1394,7 @@<br>
 	}<br>
     }<br>
 <br>
-  num_saved_lines = s-&gt;ld.index[0] - s-&gt;ld.index[2];<br>
+  num_saved_lines = s-&gt;ld.index[0] - s-&gt;ld.index[1];<br>
   if (num_saved_lines &gt; 0)<br>
     /* restore the previously saved lines: */<br>
     memcpy (out, s-&gt;ld.buf[0], num_saved_lines * bpl);<br>
@@ -1404,10 +1404,10 @@<br>
       if (++s-&gt;ld.lmod3 &gt;= 3)<br>
 	s-&gt;ld.lmod3 = 0;<br>
 <br>
-      c = color_seq[s-&gt;ld.lmod3];<br>
+      c = s-&gt;ld.lmod3;<br>
       if (s-&gt;ld.index[c] &lt; 0)<br>
 	++s-&gt;ld.index[c];<br>
-      else if (s-&gt;ld.index[c] &lt; s-&gt;params.lines)<br>
+      else /*if (s-&gt;ld.index[c] &lt; s-&gt;params.lines)*/<br>
 	{<br>
 	  s-&gt;ld.c[c].Qk += s-&gt;ld.peak_res;<br>
 	  if (s-&gt;ld.c[c].Qk &gt; s-&gt;ld.max_value)<br>
@@ -1421,17 +1421,17 @@<br>
 		  DBG (1, "fix_line_distance_pp: lmod3=%d, index=(%d,%d,%d)\n",<br>
 		       s-&gt;ld.lmod3,<br>
 		       s-&gt;ld.index[0], s-&gt;ld.index[1], s-&gt;ld.index[2]);<br>
-		  num_lines = s-&gt;ld.index[2] - s-&gt;ld.ld_line;<br>
+		  num_lines = s-&gt;ld.index[1] - s-&gt;ld.ld_line;<br>
 <br>
 		  /* copy away the lines with at least one missing<br>
 		     color component, so that we can interleave them<br>
 		     with new scan data on the next call */<br>
-		  num_saved_lines = s-&gt;ld.index[0] - s-&gt;ld.index[2];<br>
+		  num_saved_lines = s-&gt;ld.index[0] - s-&gt;ld.index[1];<br>
 		  memcpy (s-&gt;ld.buf[0], out + num_lines * bpl,<br>
 			  num_saved_lines * bpl);<br>
 <br>
 		  /* notice the number of lines we processed */<br>
-		  s-&gt;ld.ld_line = s-&gt;ld.index[2];<br>
+		  s-&gt;ld.ld_line = s-&gt;ld.index[1];<br>
 		  /* return number of complete (r+g+b) lines */<br>
 		  return num_lines;<br>
 		}<br>
@@ -2048,11 +2048,11 @@<br>
 	continue;<br>
 <br>
       len = strlen (cp);<br>
-      if (cp[len - 1] == '\n')<br>
-	cp[--len] = '\0';<br>
-<br>
       if (!len)<br>
 	continue;			/* ignore empty lines */<br>
+<br>
+      if (cp[len - 1] == '\n')<br>
+	cp[--len] = '\0';<br>
 <br>
       if (strncmp (cp, "option", 6) == 0 &amp;&amp; isspace (cp[6]))<br>
 	{<br>
diff -ru sane-0.72.orig/sanei/sanei_ab306.c sane-0.72/sanei/sanei_ab306.c<br>
--- sane-0.72.orig/sanei/sanei_ab306.c	Tue Mar  3 03:13:25 1998<br>
+++ sane-0.72/sanei/sanei_ab306.c	Mon Apr 13 15:58:02 1998<br>
@@ -196,6 +196,8 @@<br>
       ab306_inb (p, base + 1));<br>
   while ((ab306_inb (p, base + 1) &amp; 0x20) == 0);<br>
 <br>
+  usleep(100000);<br>
+  <br>
   DBG(4, "ab306_write: writing data\n");<br>
   for (i = 0; i &lt; len; ++i)<br>
     {<br>
<p>
<pre>
--
Source code, list archive, and docs: <a href="http://www.mostang.com/sane/">http://www.mostang.com/sane/</a>
To unsubscribe: echo unsubscribe sane-devel | mail <a href="mailto:majordomo@mostang.com">majordomo@mostang.com</a>
</pre>
<!-- body="end" -->
<p>
<ul>
<!-- next="start" -->
<li> <b>Next message:</b> <a href="0128.html">Henning Meier-Geinitz: "Performance mustek 600 II N"</a>
<li> <b>Previous message:</b> <a href="0126.html">Oliver.Rauch@Wolfsburg.DE: "Re: /proc/scsi/scsi"</a>
<!-- nextthread="start" -->
<!-- reply="end" -->
</ul>