Scrolling capture fails for many web pages, PDF docs etc.
I'm often asked why scrolling capture fails for many web pages, so decided to post this and the possible workarounds:
Scrolling capture relies on capturing the visible part of a page, scrolling it a little bit, capturing again and comparing the two to find how much the contents has moved. If your page has contents that animates quickly, or elements that don’t scroll, just stay in place, any such comparison will fail and I cannot fix it.
Workarounds include using “scroll region” function and carefully outlining what you want to capture with scrolling, avoiding non-scrolling elements (e.g. navigation) or fast animated ad columns.
Another workaround would be to use Firefox browser and our plugin for it, click HyperSnap plugin icon in the toolbar and choose “Capture entire page”. This avoids any scrolling, just makes the browser render the entire page to a bitmap. Also: scrolling capture will never work for the new Microsoft Edge web browser.
Capturing PDF documents with scrolling fails
The reason you sometimes cannot capture PDF documents with scrolling is due to the background behind the pages in a PDF viewer, which has some kind of pattern or shading, but does not move as you scroll the pages. If you capture something that is not uniform color and does not scroll with the rest of the window, auto-scroll will not work. The reason is that HyperSnap has to compare the images before and after each scroll, to find identical parts and what was added at the bottom, to correctly stitch the images. If some part of the window moves, and another (background) does not move, such comparison will fail, as there is no identical parts before and after the scroll. The problem is similar if you have some part that are animated fast or change the image before and after scroll.
The work around would be to either set a uniform color background in your PDF viewer, or make PDF viewer narrower so that the background on each side of the page is not showing, and setting the viewer to show continuous pages (so that no background is shown between the pages). That way the images before and after each scroll will have identical large parts and can be correctly stitched. Here are some more suggestion by Thomas "The Lesser Admin":
Here is one partial solution to the scrolling problem:
If you have a webpage, that can't be auto-scrolled, look at the content in the first half of your visible screen. More often than not, you will find all movable objects there, like banners, clips, videos.
Below a certain point you only have the desired content.
So, if you scroll the page down until banners etc. are above your visible screen and NOW start auto-scrolling, it might very well work.
I tested it with several pages that could not be autoscrolled when starting on top, but auto-scrolled when starting below the "banner section".
Another method is scrolling with strg-shift-G to select an area without p.e. an adjoining vertical banner (tested with computerbase.de). The banner follows the scrolling down normally preventing autoscroll. But with selecting the content area without the banner (to the right), HS scrolls down.
post edited by Admin - 2016/08/20 12:54:30