Troubleshooting FrameMaker Content Conversion in RoboHelp – Part 2: Images


I hope my first post of the series was useful. In the second post of the series, I am going to discuss the quality of images generated by RoboHelp when FrameMaker content is brought (import/link) in RoboHelp. I will address the common misconceptions that occurred in this area by recommending proper workflows. I will also explain what’s going on behind the scenes for those having geeky brains.

Why are my images still dull after improvements in TCS 3?
Adobe Technical Communication Suite 3 boasts about improved image conversion from FrameMaker to RoboHelp. However, many users are still not able to find any improvement and their images still appear to be of degraded quality.

The fact is, the image conversion process has been greatly enhanced in TCS 3 to achieve much better quality of images as compared to previous versions of TCS. Let’s see how it works.

TCS 3 now provides a new route for image conversion where an anchored frame inside a FrameMaker document is converted to an image through Distiller. In order to turn on this route, RoboHelp provides a special checkbox under image settings.

In order to turn on/off this setting:

  1. Go to File->Project Settings
  2. Go to Import tab
  3. Click Edit button for FrameMaker conversion settings to open Conversion Settings dialog
  4. Go to Image settings
  5. Check/Uncheck the check box for Use Distiller to Generate Image

To use the new conversion route, turn on the setting. The check box is checked by default in a newly created RoboHelp project. Please verify the setting in an upgraded project.

The image quality obtained should be of high quality. However, one may still argue that if the distiller route is used why is the image quality not same as PDF generated through Acrobat Distiller from FrameMaker and may it still find it of degraded quality.

Here is the reason why? And how can it be made to match the PDF quality?

For those who may not be interested in exploring the answer to “Why?”, here are the quick tricks that may be applied.

  1. Prepare images as per the media(web or print). For example, if you wish to have good quality images on web and are not concerned about the print quality on paper when printed from FrameMaker, import images at 96DPI (or your screen DPI) inside FrameMaker.
  2. Set distiller PDF Job options setting to “High Quality” inside FrameMaker.
  3. Place a marker in the anchored frame.


For Geeks:
Here is the explanation for those who are interested to know things going on behind the scenes. Further reading may be skipped by those who are looking for a quick solution which is described above.

Every display device (e.g. computer screen) has a DPI (Dots per inch or Pixel per inch) setting. This DPI setting is determined from the Screen physical dimension and the display resolution. For example, if the screen size is 17” (diagonal 17”, W X H = 13.27” X 10.62”), and the resolution is 1280X1024, then the DPI can be roughly calculated as 1280/13.27 = 96 DPI (approx)

Now an image is defined in terms of pixel dimensions. It has no DPI setting. DPI setting of an image is a meaningless notion. Each image can be viewed at different DPI settings. There could be a recommend DPI for an image at which it should be displayed on a particular medium or device. Different medium or display devices have different requirements for DPI settings for displaying images.

For example, image resolutions requirements for printing on paper is quite high(or the order of 200-300 dpi) as compared to requirements for Web (normally 96 dpi). Even on web image requirements could be different for mobile devices and desktop.

When an image is imported in FrameMaker, FrameMaker asks for DPI setting which defines the resolution at which the image should be displayed when published. In other words, it asks the user to provide the setting to render each pixel of the image on the computer screen. If the screen resolution is 96DPI and the image scaling setting is also 96DPI, one pixel of the image is mapped to one pixel on the screen. If the image scaling setting is set to 192dpi on the same screen, 2 pixels of the image is mapped to one pixel on the screen thus scaling down the image. When this image is converted to PDF through Acrobat Distiller, the entire image is scaled down to display it at a smaller size. However, there is no loss of pixel information. The entire pixel information is preserved so that when the PDF document is zoomed in, the pixel information is used to scale up the image thus maintaining the quality of the image. A high DPI value (say 200DPI or 300DPI) is recommended for good print quality on paper.

However, a similar functionality is not available in the online(web) world when the HTML is viewed in a browser. Browsers do not do a very good job of scaling images. RoboHelp understands this limitation of browsers and recreates the image at the resolution mentioned for the image in FrameMaker while importing the image. The recreated image does not contain extra pixel values as that present in the original image. Thus, when the image at normal size in a browser is compared to the scaled up in an image in a PDF, it looks degraded in quality though they more or less look same when viewed at the same zoom value.

This can be explained through the following image snapshots. The images look more or less same in the three cases.

The difference appears when the image is zoomed.

The solution to this problem is to avoid scaling of an image on a particular medium and prepare image as per the medium. This can be ensured if the DPI mentioned during the import of images in FrameMaker is same as the screen DPI at that moment. Normally, today’s screens display at 96 DPI. Hence, this should be the preferred DPI setting to be specified while importing images in FrameMaker. Following this guideline, should greatly improve the image quality.

NOTE: However, there is precaution to be followed. As explained above there is a difference in the print and web paradigm. Where print needs images placed at a very high DPI value, web needs images to be inserted at screen DPI value. So, if your images in FrameMaker are going to be printed on paper, do on apply the rule of 96 DPI. Otherwise, the images in print may get distorted.

Another trick that can be applied is to set the distiller PDF Job options setting to “High Quality” inside FrameMaker. This sets the distiller to generate a high-quality image which is used by RoboHelp during image conversion.

Lastly, a marker can be placed in a text box inside the anchored frame so as to ensure that RoboHelp follows the distiller route and does not try to intelligently recreate the image without distiller.

All these tricks should help achieve a high-quality image with acceptable results. I hope this explains the image conversion route followed by RoboHelp and you find it useful to resolve the image conversion difficulties faced by you.

Mayank Agrawal, RoboHelp Engineering Team

FrameMaker, Products, RoboHelp, Support

Posted on 06-07-2011


  • By tabassum - 2:25 PM on October 3, 2016   Reply

    I am trying to cross reference a part of an image to a table present in the same topic.
    Please let me know how to do it.

    • By Stefan Gentz - 8:07 PM on October 3, 2016   Reply

      Hi Tabassum,

      please check the answer, that my colleague Manminder Singh gave on your question on YouTube for the video “RoboHelp 10: Image Map – hyperlink sections of screen captures”:

      This should answer your question.

      Kind regards,
      Stefan Gentz

  • By feline1 - 11:28 AM on September 27, 2013   Reply

    I’m not finding these blog suggestions work reliably for me (using TCS4). Mayank, when a raster image is imported into FrameMaker (into an anchored frame), and one looks at the “Object Properties” dialog, there is a panel called “Scaling” which has *two* fields. One is the DPI your blog talks about, the other is “scaling” (a percentage value). Can you explain how these interact and affect the import in Robohelp?
    For instance, typically, if I grab a raw screenshot and import it to Frame at 96dpi, it will be much too big. I don’t want to throw away pixels and resample it in another image editor before importing, because that will generally give terrible result (especially with 1-pixel wide text etc). So instead, traditionally I would also use the “scaling” parameter (say, 60%) to make the image an appropriate size in Frame.
    However this often (but, confusingly, not always) seems to mean the DPI parameter changes (sometimes to another value, sometimes to “Unknown”). Are they just linked? And this % x DPI is passed to Robohelp?
    Generally, what most users want to happen (I believe!) is for, at default zoom, their image to be the same size in Robohelp as it is in Frame, relative to the paragraphs of text. I appreciate that Robohelp may need to downsample the image to achieve this… but it keeps seeming to change the size of the image, either making it enormous, or much smaller than it was in Frame. It’s very unpredictable.
    The “Distiller” route seems broken too. I can see what job option settings are specified in Distiller when it does the conversion (no downsampling, lossless compression etc), but the output in Robohelp bear no relation to this – they’re usually heavily JPEG compressed and resize/resampled. What’s going on? Is RoboHelp doing further processing *after* Distiller?

  • By Marina Sheiner - 8:31 AM on January 31, 2013   Reply

    Has anyone managed to achieve acceptable image quality with any of
    these workarounds? I’ve tried all four: using 96 dpi images, selecting HighQuality
    settings in Distiller, putting a marker inside anchored frame and copying all images into RoboHelp Image folders. Unfortunately, none of these worked for me.

  • By Joseph Lorenzini - 2:57 PM on May 11, 2012   Reply

    There’s another workaround that you have missed. If the images from the FrameMaker folder are copied into the RoboHelp Images folder (or folders), then the really bad images are replaced with pristine, high quality images. This process could be scripted. 

  • By Diana - 10:30 PM on July 18, 2011   Reply

    Mark – I am so close to nirvana with my FM->RH but still run into some issues. One of them is image conversion. I’ve tried the tips above and have not had consistent success. There is one exception — I have not tried the marker because I don’t know what text to put in the marker tag or which marker tag (graphicscale?) to use. With WebWorks, we used the GraphicScale but not sure what you meant by ensuring RH uses Distiller by placing a tag. Can you please elaborate?

Join the discussion