Hi,

We tested and we do not see the gray "amature" text shown in the picture you posted on the first page.

We are not aware of that font size is pt or px matters to the conversion, but line height matters. You can get text line being split in the middle if your text line height is too small.

Thanks!

Hi,

This makes sense. Internally we use pixel position to cut the page. The paging is done after layout, so for example if we have two lines of text with the following height:

Line 1: 50px
Line 2: 25px

And if the page is 60 px high then a page break will be inserted immediately after Line 1 since there isn't enough room to place line 2 on the same page. Page 1 in PDF will represent a rendering of from 0 to 50 px in the web page and clearly include line 1 and exclude line 2.

This would have problem if the line height is like this:

Line 1: 49.8px
Line 2: 25 px

In this case the paging will still be inserted at 50px because 49.8px is rounded to 50. In this case page 1 in PDF will still represent a rendering of from 0 to 50 px in the web page, however this same range now includes a tiny hair of line 2 since line 2 starts from 49.8px and the page breaks at 50 px. Visually this usually does not cause any problem because it can only be off half a pixel at most. However that is enough to trigger the underlying engine to include the ghost text due to the y position overlaps with the text line.

To complete resolves this issue, we need to systematically support sub pixel values (which the browser engine itself already support hence the issue you see) and this will be quite an undertaken. So I do not see this to be resolved right away. We will look into it to see if it is possible for us to support that.

Thanks!