From 5a34e5bd8d52e8733f5a970b18c1f00136c3338e Mon Sep 17 00:00:00 2001 From: Keven Ates Date: Mon, 19 Sep 2022 14:08:09 -0400 Subject: [PATCH] Update ReadMe --- ReadMe.md | 30 ++++++++++++++---------------- 1 file changed, 14 insertions(+), 16 deletions(-) diff --git a/ReadMe.md b/ReadMe.md index a83d185..4e574a6 100644 --- a/ReadMe.md +++ b/ReadMe.md @@ -96,22 +96,20 @@ system installed version first. If not found, it uses the supplied version. 1. For ***Thumbs.db*** files, Windows(R)(TM) uses at least two format types to store these thumbnails: - 1. An **older** format that seems to consist of jpeg-like images that have been - reduce to a frame composition and image data. The standard JPEG header, - Define Quantization Table (DQT), and Define Huffman Table (DHT) are missing - (presumably, they are constant structures abstracted from the image). The - header, DQT, and DHT are supplies as constants during extraction. As such, - ***Vinetto may not reconstitute some Type 1 thumbnails correctly.*** - The PIL Image library is used to attempt proper reconstitution, but may fail - in certain circumstances. - - NOTE: Recent work has shown that the stored JPEG portions describe RGBA frames. - RGBA images are not a recognized format for JPEG. However, the image data identifies - as CMYK channel images. The color channels are stored "out of order" and the Key (K) - channel has been set to "full black" for all image generated thumbnails and - "selective" (presumably used as Alpha channel transparency parts for desktop display - over backgrounds) for all Windows OS constructed thumbnails generated (i.e., image - folder images). The extraction process reorders the CMY channels (C is Y and Y is C) - and sets the K Channel to "no black" for all extracted images. + 1. An **older** format that seems to consist of JPEG-like images that have been + reduce to a frame composition and image data. + - The standard JPEG header, Define Quantization Table (DQT), and Define + Huffman Table (DHT) are missing as, presumably, they are constants in + Windows(R)(TM) and were removed to reduce size. Substitute header, DQT, and + DHT are supplied as reasonable constants during extraction. As such, + ***Vinetto may differ slightly from the original display of Type 1 thumbnails***. + The PIL Image library is used to attempt proper reconstruction.

+ Recent work has shown that the stored JPEG portions describe CMY color + channels (no K channel) and an Alpha channel for transparency. The color + channels are stored "out of order" (C is Y and Y is C). Then, the images + prescribe to a "YMCA" image format (presumably as a joke). The extraction + process converts the image to a CMYK image (the K channel is "no black") with + no Alpha channel for all extracted images. 2. A **newer** format that is fully compliant with the JPEG format. Vinetto writes this type to file directly.