A possible origin of big size files

[Gianni Becattini]

After having received so much from this forum, I try to carry on a little contribution. I do not pretend that what I say is completely right... it is more a suggestion to the investigation of a possible bug.

One of the my biggest complains was the pdf file size, and I tried in every way to reduce it. 

I noted that sometimes Publisher wrongly extends the layers of an AP file outside a picture frame.See 

Well, this possible bug could have another kind of manifesting itself. I try to better explain: if I have a picture and some adjustment layers UNDER IT (hierarchically), all is OK. If I place these layer ABOVE the picture, the effect seems to be the same, but in the pdf file we have more copies of the picture one over the other, one for each adjustment layer.

Because my book contains about 700 large AP files with 2-3 or more adjustment layers, I asked myself if the effect could be catastrophic. So I took courage with all my hands and corrected all the photos where the layers were ABOVE the picture. In this way, the size reduced from more than 9G to 3.5G and I have no more duplicated images in the pdf. I noted also that, if I have two identical image duplicated in the AP file, both are transferred to the pdf (even if the image below is not visible). And the export flag "include layers" not always has the expected result.

I underline: it is just a suggestion, after one month of continuous working maybe I am confused... but it seems so. If not, forgive me and forget it.

 

[NathanC]

Hi @Gianni Becattini,

15 hours ago, Gianni Becattini said:

Well, this possible bug could have another kind of manifesting itself. I try to better explain: if I have a picture and some adjustment layers UNDER IT (hierarchically), all is OK. If I place these layer ABOVE the picture, the effect seems to be the same, but in the pdf file we have more copies of the picture one over the other, one for each adjustment layer.

Are these adjustment layers nested/clipped onto the images within the .afpub document, or are they in an .afphoto file that has been placed as a resource in your .afpub file? I'm guessing that you're referring to the images being nested/clipped onto the image file when you say they're 'under it'? As otherwise the adjustment layers would have no affect on the image if they were lower in the layer stack/hierarchy than the actual pixel image.

Without looking at your source .afpub document and images/afphoto files it's difficult to say exactly what's happening or if there's a bug. Could you possibly send me a copy of your .afpub file, along with a few of the .afphoto resources so we can look into this further? It doesn't have to be the entire file, just a cut down version or a singular page where we can observe what's happening with the file size when exporting to PDF dependant on the adjustment layer position. I've included a private upload link below.

https://www.dropbox.com/request/kxNTWTPqRvRbxtLwZBbj

Many thanks 🙂

[Gianni Becattini]

You are right, I was not precise. When I say "below", I say "hierarchically", so this way (this is OK):

When I say above, I mean thi s way:

and this construction produces multiple images. I can upload something if you like, but you can do the same as above. The problem seems not to be linked with a particular file.

All that is INSIDE AP file, not inside Abu.

[NathanC]

Hi @Gianni Becattini,

Thanks for following up and clarifying on my queries. I've just made up a few test documents for this scenario. I opened the same image in Photo and applied the same  adjustments used in your screenshot, with the only difference being the position of those adjustments in the layer stack, one file had them above the pixel layer, and the second file had the adjustments nested into the image as child layers. After saving them out both .afphoto files had a file size of 268MB.

I then created a new afpub doc and placed the first .afphoto file with the adjustments above it in the stack as a linked resource, I then exported to PDF using the (For export) preset. I then replaced that .afphoto file with my second one which had the adjustments nested inside the pixel layer and repeated the export using the same (for export) preset After comparing the file sizes of the two PDF's they were identical at 235MB, even down to the KB. 

It could be that there are other variables here not taken into account in my test documents, so if you could upload an .afphoto file and .afpub file where you're encountered this problem I can then confirm if the problem can be reproduced, and in addition if you could send me a screenshot of your PDF export settings for consistency that would be great.

Thanks!

[Gianni Becattini]

Yes, I do but for that I need my backups (now all my photos have been modified) - I can do in the weekend when I am home.

The pronblem arose because the Editor checked my pdfs with some tool and he found many superimposed photos.

[thomaso]

18 hours ago, Gianni Becattini said:

I noted also that, if I have two identical image duplicated in the AP file, both are transferred to the pdf

Not directly the mentioned situations but it may matter for the OP's thoughts that APub doesn't export a copy of a linked image resource as referenced element inside a PDF but as an independent, separate item. Different to ID for instance which seems to work with referenced elements and thus results in smaller PDF file size. This was discussed a few years ago and confirmed by a Serif moderator, unfortunately I can't find the thread any more. (and my used English term "referenced" may be wrong / imprecise).

Attached two sample PDF + export options.
(apub: 10.3 MB – id: 3.3 MB – placed jpg: 2.7 MB)

apub image dup.pdf -  id image dup.pdf  - 

-

[Gianni Becattini]

Sorry for having disappeared; I tried to rebuild the previous situation but I went into troubles with the perfidious iCloud and then gave up. In this moment it is difficult for me to replicate the previous situation.

I am however rather sure of what I said. The problem has been detected by my editor, using a tool called PitStop Pro. The company is Enfocus: https://www.enfocus.com/en/pitstop-pro

It allowed to detect the duplicate images, that, from my side, I removed with the different layer organisation, as told in the beginning.

I am trying to recover one of the old files (those with the duplicate images).