Frame caption outside frame yet below it

LO 7.5.9 for Linux
.odt document

It appears to be unclear if frames can get caption. On one side Add/Insert Caption not available on Frames toolbar nor in toolbar customization dialog. On another hand if to select frame then do right-click the option Insert Caption… is available in context menu.
If to use it the caption is placed inside frame while the author here wants to have it below frame. No way was found to get it this way.

In this particular case frame is used to place two pictures side by side, then to decorate the pair of pictures with one single caption. Idk if doing in fashion as presented above is the optimal way.

I agree, nasty problem, write a bug report.
Insert image1 and apply caption. Anchor image1 As character and add another paragraph (in the frame). Check that image1 is the only content of the upper paragraph.
Then add image2 into the frame (upper paragraph) and anchor this one again As character.
Now you have two paragraphs in the frame, the upper one containing the images, the lower one containing text and field.

Thanks for hint. I personally prefer to have caption outside frame as its font size is expected to orientate always on document text font size. I can then scale picture by arbitrary factor but caption font size keeps be unchanged.

You must first understand what this caption feature provides. IMHO, it is a courtesy one-size-fits-all convenience offered for quick’n’dirty job. By this, I mean it is a compromise and may not fit exactly what you need or want, as exemplified by your question.

Insert>Caption creates a frame around the selected picture and adds a (styled) paragraph inside the frame above or below the image. You end up with a nested structure which is way too complicated in the majority of cases.

If i do it manually, I do it the other way round. I start by creating the caption paragraph (yes, it is more complicated because you must take care manually of the numbering by inserting the adequate field) because this paragraph has some relation with my surrounding text. Then I attach the image to the caption and I frame-style the image so that it has without any error all the desired wrap and position properties relative to the caption. In the end, my document has less frames and is more robust.

In your case, I’d anchor two images to the same caption paragraph and the frame style(s) would take care of the position. Above all, make sure that Allow overlap is not ticked. I think you need 2 frame styles because I seem to remember that the automatic conflict avoidance function repositions the image only in the vertical direction. Consequently, two style are probably needed to align images side by side, unless you opt for a table solution.

Remark: whatever you do, the Insert>Caption needs to anchor the outer frame to some element of your text. This ends up to some equivalent of my caption paragraph anchoring. You could object that the group caption+image is made independent from the text and this is preferable. However, Writer is lacking a “deferred” or “floating” anchor mode in which the frame would be flushed to next page if there is not enough space and text continues to flow up to the bottom of the page. Unfortunately, this is not possible presently and you have a rather high probability of annoying blank space at bottom of page.

In my former write-up I used captions mainly for tables and it work very well.

On another hand I learned from “Designing with LO” that for objects becoming inserted to document the Writer packs (rather always) them in a frame. Would it mean insertion of caption to receive an framed object on its input (in this particular case a picture) in order to pack it in an additional frame?
When used captions for tables in past it wasn’t visible in obvious way that caption insertion generates a frame.

If to practice it once (insert picture to document, then decorate it with one caption and do it using picture context menu > Insert Caption) one can see pretty well what is going on (view of formatting marks is ON). Indeed Insert>Caption conducts nesting of frames: picture frame inside caption frame. Well, it may look like be presented by Writer as frame nesting, however maybe it is more anchoring then nesting, idk.

It is different.

In this particular case frame is used to place two pictures side by side,…

In this case, I would merge the two images into one image in an external image editing program and only then in Writer.
You can then use the normal caption for images.

However, you have not mentioned how you want the images positioned and whether or not text should run around them.

Is my understanding right: frame style gets assigned to frame. Caption paragraph gets placed externally to the frame then for image which is located in frame internally it’s hard to anchor it to caption paragraph unless caption paragraph is inside frame too. I prefer personally to have caption outside frame as in this fashion it is more natural for me.

Me was lucky today to have two pictures side-by-side in a frame while using only one frame, the frame style and picture properties (maybe also paragraph alignment in DF, which eventually in the end was not necessary).
Frame style is positioning the frame on page/in document.
Picture properties position picture (in this case) in frame, or (in general case) in flow of document text and other objects.

Do you mean all that referring to picture made in remaining building blocks of document?

My feeling is in this fashion it won’t align well with the approach keep-it-simple. At arbitrary later point of time which maintainer of this document will remember that this particular spot uses a table to frame two visible pictures. It will require fair portion of attention to see it that point of time.

I see, although tables are also objects, they’re objects of different kind - hence “It is different” as you kindly mean. I agree it sounds reasonable, thank you.

New Point Frame/Image Wrapping: Parallel vs. Optimal
I still conducted zero experiments in this regard however by taking a look at descriptions of both I don’t see a difference to exist between these two. Well, maybe it’s just those 2 centimeters (which the Writer Guide is aware of (page 249, WG 7.5) and informs readers) which make the difference Optimal to Parallel.

Document author added one frame to page, frame is styled by custom frame style. Two pictures in frame, side by side. Each picture aligned as character and no overlap, picture frame style is custom style and child of Graphics frame style.
Right now is point of time when caption needs to be placed below frame. Custom paragraph style exists for caption - this was made to follow your method, I don’t however find field type referring to graphic captions to be placed in caption paragraph. How to approach this problem? In the end of day document will need to have index of pictures.
Frame encapsulating two pictures is anchored to caption paragraph, however both the frame as well as caption paragraph are close to end of page. The frame with graphics skips to next page (due to its size) but leaves caption paragraph to be on previous page. How to force graphic frame to be always before text paragraph it is anchored at?

Insert>Field>More Fields, Variables tab, Type Number range.
The Select column allows to choose among several counters. It is likely the counter is Figure. To make sure, righ-click on an existing number and Edit Fields. This will tell you which counter is used.

You can also create any custom counter.

Table of figures is added like a TOC. Where you want this table:

  1. Insert>TOC & Index>TOC, Index or Bibliography
  2. Type: Table of Figures
  3. Create from Captions, category: Figure (or your counter name)

Writer built-in captions seem to have more automated elements as just caption numbering. I mean the

  • head (seems to be same as caption category)
  • head character formatting
  • symbol used as deliminator: caption head to caption text

I wonder if it will result in any complications as of subsequent phases of writing in case all these elements to get just typed manually in custom caption paragraphs.

As I already mentioned, Insert>Caption is just a convenience feature. You can do the same or better manually. It does not really matter if all needed (desired?) information is entered automatically or manually. What is important is the result you want.

Doing things manually avoids unnecessary creation of frames (at least this is my goal when I choose the manual method, in addition of the freedom to sequence the caption bits how I like them).

Hi, Myself doesn’t use Insert > Caption. Rather I try to implement your method. I compare it to Writer built-in one because I see when doing captions manually certain levels of automation seem to become lost (it gets lost what on another side the built-in feature does). Well, the question of caption numbering has been clarified - this is however not the problem. The unclear points are those myself addressed in my previous post: caption line head, caption line deliminator. Built-in function seems to handle them by itself (from document author perspective it happens automatically). On another side if to compile caption lines manually, author needs to care for those as well.
The question is: if it will there will be troubles at later point of time due to this circumstance.

Not at all. Caption category and separator are literal text without any meaning. Type anything you like. They will be displayed just fine in the Table of Figures. Category and separator have no significance per se. They are used only once when Insert>Caption creates the caption. If you later modify these elements, existing captions are not updated, which clearly shows they play absolutely no role in the feature. As I said, this feature is just a convenience to avoid entering constant strings. Apart from frame addition, it is strictly equivalent to an AutoText entry. The only difference is you select an object and the caption (and extra structure such as frame) are added around the object.

Thanks for hint. Really? How will generator of TOF know for author to be using the colon, for instance, as separator?

I tried to implement your method in my document (caption line crafted manually) and run onto following effect. One page almost full, however little bit space is still free at bottom of page text area. For document content it is the point where a frame should follow yet be decorated by caption line. As next I place a line styled by custom caption paragraph style then insert a frame and anchor the frame at this caption paragraph. The frame height is to much in order to fit in page remaining free space. Hence Writer is shifting frame to next page but lets caption line (crafted manually) to be on previous page. This way page gets in its lower section following look: text body paragraph ends, frame caption line, blank space (a paragraph of low number of lines would fit here), space for page Footer, page break, frame on new page.

It does not need to know. It takes the full paragraph as the caption without slicing it into head, number, tail.

You must play with frame settings.

  • Position Vertical From top some distance in kbd>Paragraph text area
  • tick Keep inside text boundaries- in Wrap tab, Options, tick First paragraph, untick Allow overlap

If you still have caption and image on separate pages, change vertical position reference to Page text area, OK. Then restore Paragraph text area.

Myself created a paragraph style for caption paragraph to use it. Custom style is member of Caption paragraph styles family and was derived from some Writer stock caption style.

Frame decorated by a caption is present in document. Certain paragraph wants to make cross-reference to caption. To accomplish it the following steps are followed, Insert > Field > More Fields… > Type = custom-caption-category, Refer using = Category and Number, Selected the desired caption.

Cross-reference at source text is built from merely caption number, the caption category must be typed manually.

Myself didn’t mention following point before however should had it done. Sorry.
Document follows the convention regarding inserting images to document then captioning them. Top to down always this order vertically: first picture (anchored to paragraph), below the line of picture caption.

If objects of other kinds are used they follow same convention.

Writer to be placing picture on separate page than the caption is not the problem - it had very good reason in that particular case and author understands it.
Problematic are following points (convention presented above is used)

  • Writer moved picture to next page but left caption to be on previous one - my expectation (when mentioned convention is followed) when Writer decides to move the picture it is expected to do it also with caption and place them two on next page in their original order (as autor did it). It is bad if Writer inverts the vertical ordering picture to caption.
  • Writer decides to move picture to next page but undertakes no reordering of tail paragraph in order to avoid blank unused space at bottom of pages. Reordering the pair picture-caption is not the proper way because it breaks conventions. Putting frontal fractions of tail paragraph is acceptable.