Update img2nef.md

diff --git a/img2nef.md b/img2nef.md
index 8ba0a593bf3b388506e1b8af86f299d0e65a4a7c..628fd94a63ddc6356ac8237604a1fd9d5497dfa5 100644 (file)
--- a/img2nef.md
+++ b/img2nef.md
@@ -218,7 +218,7 @@ Because you'll likely be developing your generated raws in either raw-processing
 ### Shoot a highly-detailed scene with good exposure\r
 Since img2nef doesn't use the raw image data from template NEFs, the content of the scene you shoot doesn't directly matter. However, img2nef does have sensitivity to scene content for a different reason - embedded jpgs. For maximum compatibility, img2nef replaces the embedded jpgs of the template NEF when generating the output NEF by overwriting them the jpgs "in-place", meaning it doesn't try to rearrange the EXIF sections. This means the embedded jpgs in your template NEF need to be large enough to accommodate the embedded jpgs img2nef creates from your source image, otherwise img2nef is forced to increase the jpg compression level to make sure the embedded jpg fits. In the worst case scenario, img2nef will be unable to fit the embedded jpg(s) even when using maximum compression, meaning the generated NEF will have the original template embedded jpg.\r
 \r
-To prevent this, shoot your template NEF with a scene with lots of detail and well-exposed. This will create the maximum embedded jpg size possible, leaving enough room for the high-quality embedded jpgs img2nef will create. The template NEFs I included img2nef were shot using a noise pattern displayed on a computer monitor - you can download that image [here](https://photos.smugmug.com/photos/i-Lv7XTmQ/0/MPXGZcjMCj3DKHrnsR4nHPRMMm53PXSb8dLKXvfL3/O/i-Lv7XTmQ.png).\r
+To prevent this, shoot your template NEF with a scene with lots of detail and well-exposed. This will create the maximum embedded jpg size possible, leaving enough room for the high-quality embedded jpgs img2nef will create. The template NEFs I included img2nef were shot using a noise pattern displayed on a computer monitor - you can download that image [here](doc/img2nef_screen_noise_pattern.png).\r
 \r
 ### Shoot with lens electrically disconnected\r
 Nikon Z cameras embed lens correction profiles into raw files. These profiles are automatically applied by both raw image processors like Photoshop and by the camera itself (in-camera multiple exposure and retouching features). These profiles are intended to correct geometric distortions from the projection of the lens, along with various aberrations and vignetting.\r
@@ -240,3 +240,4 @@ The next best option is to shoot with a lens that has very little optical distor
 img2nef is written primarily in Python, with the NEF encoder logic written in C for performance. A pre-compiled version of the encoder is included in the Windows, Mac, and Linux directories, each of which is automatically loaded based on the detected platform. Before loading the platform-specific library img2nef will first check for the library in its base directory - this allows you to compile/use the NEF encoding module for non-standard platforms as well. The encoder module should compile without modification on any platform - it's coded to work with various native integer sizes and on both small and big endian configurations. You can compile the module via the following gcc invocation:\r
 \r
 `gcc -shared -O2 -o nefencode.so nefencode.c -fPIC`  \r
+\r