"A Heuristic and GOMS Analysis

of the Trim Edit Function

in Apple Computers' Final Cut Pro 3.0"

 

by David Reiss

3/12/03

 

 

 

 

 

Introduction

The subject for this interface analysis is the trim edit function in Apple Computers'  Final Cut Pro (FCP), version 3.0. This software allows a single user to edit full motion, high-resolution video, stills, and audio into a finished video program. Like many of the major players in the desktop video editing field, FCP is based on a timeline editing model, built around a visual representation of a timeline presenting clips strung out from left to right in a sequence based on time.  System status is visible by an array of tools such as time code, clip frames, and audio and video tracks that help the user to visually gather needed information about the status of the timeline at any given moment.

In most editing software the timeline is a key component of the interface. It is usually integrated as part of all the various modes of editing. That is to say that the timeline is not only a visual guide, but is a component used for user input and actions in all modes of editing. This timeline based editing model is widely considered intuitive. Most users have been working on something like this for well over a decade, thus making the interface design habitual and easily recognized. This consistency of design is why Apple's software looks like most other editing software, taking the standard interface layout of source (viewer) and record (canvas) monitors over a timeline, as seen above.

In most desktop video editors, when in editing modes such as Edit, Trim, or Effect, the timeline provides visual feedback and user control, as well as input of any number of elementary operations at any given time. In FCP, the critical function of Trim, which provides the ability to add/delete frames from an outgoing/incoming clip at a transition, disables the timeline and opts for a floating Trim window over the existing interface. This is a move away from the traditional editing workflow model, and Trim being a critical function, I argue adds a layer of complexity while sacrificing functionality and affordance.  For this study I will concentrate on using the GIDs and forgo the extensive array of keyboard commands built into the software.

            Having worked on other desktop video editing systems for many years, I can attest to the differences between other products and FCP. Other software editors recognize the importance of the timeline and build their interface and functionality around it, so the timeline is the locus of attention not only for visual cues as to what is activated or selected for most editing functions, but also has the ability to control most user modes and functions.  Not so with Final Cut Pro 3.0, as we'll discover in the following heuristic analysis.

            In the basic interface layout as shown above, there are four key components to this GUI. The source (viewer) monitor, record/timeline (canvas) monitor, timeline and the tool palette. Let's run through a sequence of actions using the Trim feature of the program. The Trim function can involve moving an individual clip's In or Out points, changing the location of an edit point between two clips, or simply moving a clip to a new location. There are two ways to Trim in FCP.

 

1. Timeline Trim:

First the user selects the Zoom In tool from the tool palette, then positions the mouse over the section of the timeline they wish to enlarge and clicks until the tracks are large enough to be visually distinctive in size and content.

Then the user selects from one of the two Trim tool modes on the tool palette, either Ripple Edit, or Roll Edit. For our purposes we'll use the Roll Edit function, which slips a clip's start and end points without changing its placement in a timeline.

Once the Trim function mode is selected, the cursor is then placed onto a cut between two clips, selecting the edit and providing visibility by the Trim icon cursor. The user click/holds and then drags the cursor and the source monitor changes to show the incoming and outgoing frames of the edit.

Visually the user gets feedback as to how many frames the edit is being slid down the timeline by a yellow window with a time/frame count next to the edit. The two new smaller video screens update themselves as the edit is manipulated as well. This is the typical way a timeline Trim feature is implemented, but usually with more functionality built into the timeline.

 

2. Dedicated Floating Trim Window:

With the cursor back to pointer mode, move the mouse over the timeline between the two clips.

The cursor changes from a point to a double line visually cueing an activated feature.

At this point clicking and dragging will only shorten the end of the first clip, or shorten the start of the second clip, leaving a hole in the timeline. Double clicking, however, will invoke the dedicated floating trim window function.

This new window highlighted in red above, is layered over the basic edit source and record windows, and gives the user a larger video window to view the clip(s) modification of start and end frame, and allows for more precise control via the dedicated Trim Frame buttons,

.

The timeline at this point is no longer enabled, and in fact if you click on either clip or the cut in the timeline, this new trim window will be removed. The only way to control the Trim function at this point is to click the Trim Frame buttons in the center of the trim window, or by entering numbers via the keyboard.

            In this instance, the weakest component of this interface is the timeline. While in the trim window, the timeline is not grayed out visually, but should it be, since it does not allow the user to do anything except to exit the trim window mode. This is not intuitive, as most users are used to having timeline functionality across all editing modes. FCP is inconsistent as it deviates from the standard practice of timeline user interactivity, with no benefit added that I can see. If the software grayed out the timeline function so the user’s locus of attention was shifted to the active trim window, at least that would give the user some affordance of the dedicated mode and functionality of the trim window.

The other weak component is the tool palette, which creates the need to move the cursor around constantly to change pointer function. It seems like Apple has taken its cue from graphic programs like Adobe Photoshop, which require the cursor to have multifunctional modes. In a video editing environment, the combination of disabling the timeline for some functions, and relying on a multi-mode cursor for most others is a weak design. This will become clear as we go onto the GOMS analysis, followed up by two alternative designs.

 

GOMS Analysis of Trim in FCP 3.0

Legend:   m = mental preparation   h = homing   p = pointing   k = keying

1.

Select zoom in tool

Cursor icon change

Position, Click and Zoom in timeline

 

zoom in:: MHP K

system response:: R

select timeline clips:: MHP KKK

2.

Select cursor or trim tool

Cursor icon change

 

click button:: MHP K

system response:: R

3.

Select trim edit in timeline

 

double click button:: MHP KK

4.

New trim edit window over interface

 

system response:: R

5.

Select trim function from buttons

(-5, -1, +1, +5)

 

click button:: MHP KKK

system response:: RRR

6.

Select timeline to close trim Window

Or click close window button

 

click button:: MHP K

system response:: R

7.

Reset cursor tool

 

 

click button:: MHP K

system response:: R

 

            While the heuristic analysis clearly shows how the timeline's lack of integration in the design is a problem, the GOMS analysis shows that the user spends a lot of time navigating the interface and changing cursor modes via the tool palette.  If this function was something that was used infrequently, I’d be hard pressed to criticize the interface Apple designed for FCP. But the ability to trim is an essential mode in video editing, and having such a cumbersome design, hinged on the multiple functions of a tool palette is just poor design of an otherwise good execution of their GUI.

 

 

Alternatives

1. New design w/ dedicated function buttons in record/timeline monitor.

I would remove the tool palette from the GUI completely, and assign dedicated buttons to the record/timeline video window frame. This is nothing new, as most editing software has dedicated functions mapped in between the video windows, or onto the timeline frame itself. The lack of any text explaining button functionality also shows how the interface relies on the user's ability to recall rather than recognize button functions. Here the buttons are placed in logical groupings by their modes of Edit, Trim, and Effect, clearly labeled by text headings.

Also, this is a good opportunity to build in intelligence to create an attentive environment within the timeline itself and the interface as a whole. When the timeline cursor is close to a cut, and the Trim function is activated, the timeline anticipates the users need to trim an edit by highlighting that specific area between two clips in the timeline. This sequence of events, making the jump to the closest transition, causing the edit area to be indicated, selected, and activated without the user having to do anything else is a good example of an attentive environment and transparent computing. The burden here is shifted from the user to the computer.

            Another use of an attentive environment is the use of double clicking to magnify the timeline when in the basic Edit mode. This function is one of the most used, and there is no need to have a dedicated cursor tool for it. Simply double click in the timeline area to magnify it by 50%. Option -double click could shrink the timeline by 50%.  By anticipating the need to expand and contract the timeline constantly with a simple dedicated mode that addresses this specific task would create a more transparent, user friendly interface.

 

2. New design w/ dedicated Slip-Left and Slip-Right buttons.

            The ability to slip frames in a clip — to have any given shot in the timeline start sooner or later but remain in the same spot in the sequence — is a key function of Trim and general editing that is very useful. Dedicated buttons with keyboard equivalents make this adjustment faster and more user-friendly. These new buttons highlighted above are for Slip -Left 10 frames, -Left 1 frame, -Right 1 frame, and -Right 10 frames. This is regarded as a Clip function and labeled as such. A user can, in a single mouse click, accomplish what would take several steps: changing the cursor tool; clicking on desired clip; click, hold and dragging in the timeline to slide an edit; and then resetting the cursor tool.

 

Summary

In looking at what most users will need in a computer video editing environment, Apple has developed an editing system that favors those graphic software users who are used to a tool palette model for multiple cursor functions. In my model, removing the cursor/tool palette design and opting for more dedicated GID functions is a stronger approach, given what users do with this software, which is primarily add, delete and manipulate video and sound clips into a single timeline.

Focusing on the trim feature, we saw how the dedicated floating trim window causes confusion due to the disabled timeline, without any indication or visual clue as to what is not active in the GUI. This design feature works against the user by disrupting workflow, lacking consistency, and causing confusion on the part of the user.

Interesting to note, counter to Jef Raskin’s idea that creating content and controlling systems are two separate functions, in desktop video editing, the creation of content and control of the system are fused together by the nature of media form. When editing media, you are creating content by controlling the system and manipulating video and audio. Editing clips into a timeline is similar to typing words into a document, although a keyboard is not necessary for the former.

 

SOURCES:

1. Jef Raskin. "The Humane Interface"  ACM Press, 2000.

 

2. Jennifer Preece, Yvonne Rogers, Helen Sharp. "Interaction design, Beyond Human-Computer Interaction"  John Wiley and Sons, 2002.

 

3. Frank Ritter, Gordon Baxter, Gary Jones, Richard Young. "Supporting Cognitive Models as Users"  ACM Transactions of Computer-Human Interaction, Vol. 7, June 2000. Pages 141-173.

 

4. Ann Marie Seward Barry. "Visual Intelligence: Perception, Image and Manipulation in Visual Communication"  State University of New York Press, Albany, 1997. Pages 15-190.

 

5. Nathan Shedroff. "Experience Design" New Riders Publishing, April 18, 2001.

 

6. Apple Computer. “Final Cut Pro 3.0 Users Manual, Volume 2:Editorial” Apple Computer, 2001.