Some.... trivia:
Ideograms in viewing refer to a subconsciously-triggered response from the nervous system, an "ideomotor" musculature response in the hand in this case, producing a so-called "ideogram."
The word has a history and technically is something that is a whole word or meaning in a single symbol (e.g. chinese characters are ideograms).
They are not intended to include or involve visuals as a viewer experience. In fact the goal is to have the target "impact" the viewer prior to their having time for anything else at all.
There are four* ways in which ideograms are supposed to help and/or improve the viewer or view:
1 - They function as a self-created 'witness' (as dowsing calls it); essentially recording the 'energy' of the target into something physical, which the viewer may touch/feel for a sense. Some viewers do this well, some don't at all.
2 - Practiced consistently, ideally one tends to recognize certain gestalt-level target elements corresponding to the results. That is the "recognizing it by its shape" part, although "the feeling while making it" may be part of that. Some instructional coaches in viewing methods will have you practice this with various 'drills'. Some of them have limitations on the size. Some actually predefine the shape, although that contradicts the original idea of spontaneous free-response but that doesn't mean it's unworkable.
3 - Ideograms can also be pictograms, there's an overlap there, and in some cases may have some visual correlation with the actual target visual data (e.g. trees may have an id of a bunch of vertical lines).
4 - The 'process' of the ideogram, regardless of outcome (sometimes ignoring outcome), is sometimes used as a grounding or "re-anchoring" exercise by viewers at any point in the session.
Later approaches have not focused on the instant-autonomous-like-response and instead of made it more a brief stream of consciousness more like 'the flow of art,' and these tend to result in larger ids, which my buddy Ed the Toad once called them "Scribbliograms." I thought that worked perfectly. I've seen more pictographic correlation in these.
Other approaches make it far more visual but in this case, it shouldn't be called an ideogram anymore, at that point it's completely abandoned the reason the term was used in the first place. There are simply many ways to go about anything, and people come up with lots of them.
*As a fifth way that isn't all that different, but is best done after you are very familiar and practiced with ids, you can simply "imagine" your hand making the spontaneous shape. Even if you don't truly see anything (most people get concept-visuals more than literal-visuals), you still may have a pretty clear sense of the "feeling" you normally get from it, simply without the ink involved.
Visualization skills, or getting visuals, shouldn't be needed for ideograms. The "ideo" part is the "ideomotor response" -- that is not a conscious practice, its definition precludes that.
You can "stage 1" or use ideograms (in the classic CRV sense above) an entire session if you like, although the data will become more complex than 'gestalt' pretty fast of course.
Edited to add: And of course you can view without using them at all. They are just odd at first but with practice they can be very helpful.
Fwiw,
PJ