Alesis QS 8.1 Electronic Synthesizer
Transcription
Alesis QS 8.1 Electronic Synthesizer
Alesis QS 8.1 Electronic Synthesizer Interface Analysis and Design Recommendations Based on Principles of Cognitive Ergonomics IOE 536 – Cognitive Ergonomics December 13, 2004 Written By Raul Baez-Toro Jason Clark Tom Ferris Shameem Hameed Il-hwan Kim IOE 536 Cognitive Ergonomics 1.0 Introduction The lights are bright and the sound is almost deafening. As the keyboardist in the band, your solo is coming up soon. Right before the solo, you attempt to punch in the button combination of the one specific sound patch you will need. Somewhere in the middle of the five-button sequence, you forget that you also need to deactivate the transpose function (which has shifted the pitch of the sound output). With so many changes to be done in a short period of time, you almost miss your cue. Then you wish you had missed it when the first notes of the solo are output in the wrong musical key. Our team used a systematic approach to evaluate an Alesis QS8.1 64-Voice Expandable Synthesizer interface in a representative user domain. Real life experiences similar to the previous example have motivated us to analyze this model and suggest improvements that could be made to the interface. Modeling the Cognitive Triad, we evaluated the user (agent), the synthesizer system, in a live performance environment (world), through the interface (representation), and how the interactions between the elements affected performance. The functionality of this model is outstanding, with a large variety of complex sound output modulations, but the model fails to provide a user-centered interface that is suitable for a domain where real-time performance is expected and errors are discouraged (e.g. a live performance in a bar). The first step in our analysis was to interview users that were familiar with a live musical performance domain, with varying levels of skills and experience using electronic synthesizers. We used this information to define a persona and target user scenario, in order to narrow the scope of our heuristic evaluation. We defined a set of phenotypic errors and problems (domain-specific descriptions, i.e. “transpose icon is easily missed”) that we could then translate to genotypic issues (cognitive language, i.e. visibility issue), to be analyzed in our heuristic evaluation. Through the use of Jacob Nielsen’s heuristic evaluation and severity ratings, we narrowed our analysis to focus on the most critical issues of the interface for the user in the target environment. We used the genotypic descriptions of these critical issues, and came up with cognitive language solutions (i.e. “increase visibility of the system status”). Then, by shifting to a top-down abstraction hierarchy (translating functional requirements to physical form), we propose physical design solutions. 1.1 Brief Description of the System A music synthesizer is an electronic device that can reproduce sounds from many musical instruments, and modulate those sounds in multiple dimensions. It may be thought of as the “Swiss Army Knife” of musical instruments. The process of condensing a multitude of functionality into a single instrument necessitates a complex interface. The following section 1 IOE 536 Cognitive Ergonomics describes key characteristics of the interface that are relevant to our evaluation. For a more comprehensive system description, see Appendix A. LCD Display This display is the only reliable source of visual feedback available to the synthesizer operator. Figure 1 indicates the primary areas interest for our analysis. Figure 1: Areas of Interest on the LCD Display Table 1: Description of LCD Display Components Highlighted in Figure 1 1 The large numbers on the left side of the LCD displays the Program or Mix code. The first (two) digit(s) represent the instrument group (in this case <00>), and the last digit represents the instrument subgroup, (in this case <1>). 2 The top line of the display shows the name of the particular sound patch that is active. This patch is unique to the current Bank, numerical code (the large numbers), and Mode (see the Sound Patch Selection Buttons explanation below). 3 TRN: An up or down arrow will appear in this area of the LCD if transpose is active. ABCD: The four vertical bar graphs represent the level of the Multi-Function Controllers AD. 4 This displays the Bank number. There are five Banks for each Program or Mix code. 5 This displays the current mode (Mix, Program, or Edit). * Refer to Appendix A, Figure A2 for illustrations of the following buttons/controls Sound Patch Selection Buttons The process of sound patch selection is characterized by four menu choices, illustrated by the flowchart in Figure 2. 2 IOE 536 Cognitive Ergonomics Figure 2: Sound Patch Selection 1. First select either Program or Mix play mode by pressing the corresponding button. Program Mode is characterized by sound patches that replicate single instruments or synthesized sounds. Mix Mode combines multiple Program Mode patches into single selections by either overlaying the sounds or dedicating separate key ranges to output the different sounds. 2. Next, make choices from the following three menus (instrument group, instrument subgroup, and Bank). These choices may be done in any order, and the changing of one menu does not affect the constancy of the other two. A) Instrument Group. Press one of the thirteen buttons in the top row of the code selection area. Each button corresponds to sound patches that are based on representations of a certain instrument; the buttons range from [00] (Piano) to [120] (Drums/Percussion). Note in Mix Mode, only buttons [00] to [90] are active. B) Instrument Subgroup. Press one of the ten buttons ([0] through [9]). The sub grouping is arbitrary, as this value does not convey any information about the modulation of the sound. The Instrument Group and Subgroup make up the numerical code displayed on the LCD. C) Bank. Each of the five banks for a numerical code represents a separate sound patch. Pressing either [◄ Bank] or [Bank ►] will cycle through Preset1 through Preset4 and the User Bank, which is defined through Edit Mode or defaults to a 5th unique preset. With 13 Instrument Groups in “Program Mode”, 10 in “Mix Mode”, and different sound patches for every combination of Instrument Subgroup and Bank, there are over 1000 possible sound patches accessible on the QS8.1. Switching between patches that have none of the above menu choices in common can require up to five button presses (Mode, Instrument Group, Instrument Subgroup, and cycling two Banks away). Volume and Multi-Function Controllers The use of sliders allows naturalistically mapped analog level control (sliding up increases the level) and can be assessed at a glance. Volume is controlled with the furthest left slider in Figure A2 and the other four sliders represent Controllers A,B,C, and D, which modify different non-volume parameters (i.e., sound “fullness”, reverberation, distortion) of a sound patch. A user must refer to the LCD display to observe the actual output level of the 3 IOE 536 Cognitive Ergonomics Controllers, any changes in mode/patch and the controller parameters are reset to zero and placed in different states of activation. Transpose Function The transpose function shifts the musical key of the output sound. When the interface is transposed, pressing a white or black key on the keyboard will output the pitch that is normally associated with a key that is a certain number of half-steps up or down from the pressed key. The LCD will display an up or down arrow to indicate that this function is active, and the direction of the transposition. The transpose function is activated by holding down the transpose button in Figure A2 and pressing the key of the desired output (this key’s pitch will be output when middle C (C3) is pressed), and deactivated by holding down the button and pressing C3. 2.0 Analysis Methods and Results Our team incorporated a number of analysis methods to evaluate the QS8.1 synthesizer. First, we conducted interviews with experienced users to determine common problems with the interface and to gain some insight into the user environment. Using the information from the interviews, we were able to develop a user environment scenario and a representative persona. Then, we utilized this persona in the scenario to conduct a heuristic evaluation, based on Jakob Nielsen’s Usability Principles. Finally, we isolated the most critical issues from the heuristic evaluation using Nielsen’s Severity Ratings 2.1 Open-ended interviews The first step was to conduct open-ended interviews with users that were experienced with this interface or the interfaces of similar synthesizer models. Through these interviews, we were able to compile a list of common difficulties with operating the QS8.1 interface. We were also able to understand the live performance environments where these users interacted with the synthesizer. For a summary of interview questions and representative responses, refer to Appendix B. 2.2 User Environment and Persona The information from the interviews was used to define a scenario that represents the user environment and to define a persona, which is an artificial identity that represents a target user in that environment. In the subsequent evaluation, team members attempted to analyze the interface from the perspective of the persona in the defined environment. 4 IOE 536 Cognitive Ergonomics For our heuristic evaluation, the design principles are evaluated within the context of live performance and preparation for the performance, of a keyboardist in a rock band (the target user persona). The following assumptions summarize the target user persona and the environment: • Scenario Ambience: In the live performance scenario the auditory modality is saturated by the music produced from the various instruments and the noise of the crowd. The ambient illumination is low, and there are intermittent flashing lights, leading to a compromised visual modality. Any work done in preparation for the performance is done with negligible environmental distraction. • User Expertise: The keyboardist is considered to be an expert in this domain; therefore the manipulation of the piano keys may be done in a skill-based manner. The cognitive processing required for this is assumed to be negligible. • Distributed Attention: The user’s attentional resources are distributed between the continuous skill-based task of playing music and coordination with the other band members. This coordination is done with visual and auditory cues from the band members and their instruments. The attentional resources required to operate the interface will compete with the attention on the band members. • Time Pressure and Error Criticality: The performance scenario is real-time, so there is little opportunity for recovery from errors. The user interacts with the interface to change sound patches or parameters of the sound at different instances throughout the performance. Each instance may be characterized by time pressure, because such changes commonly need to be made without disturbing the musical flow or rhythm of the piece. Also, errors in button-press combinations can have serious consequences, in that a completely unexpected sound may be output, interrupting the entire music sequence. 2.3 Heuristic Evaluation and Severity Ratings Our team next conducted a basic heuristic evaluation of the QS8.1 interface based on Jacob Nielsen’s Usability Principles. We primarily focused on the gulfs of execution and evaluation for the persona in the live performance environment. After analyzing each usability principle, we assigned severity ratings according to Nielsen’s method. Severity ratings range from zero to four, with four being reserved for the issues that require the most urgent attention. Table 2 includes a brief description of the usability issues that received the highest severity ratings. For the complete heuristic evaluation and more detailed explanation of each issue, refer to Appendix C. Table 2: Summary of Usability Principles and Issues 5 IOE 536 Cognitive Ergonomics Usability Principle Visibility of the System Status Specific Issues 1. 2. 3. Match Between System and Real World 1. 2. 3. Support Recognition Rather than Recall 1. 2. Recognition, Diagnosis, and Recovery from Errors 1. 2. Severity Rating The display is spatially displaced from the majority of interface controls, forcing the user to shift visual attention from the controls to get action feedback. The active status of the multifunction controllers for each sound patch is not clearly communicated, and can only be discovered by trial and error. Also, the particular sound parameter that is modified by each controller is not displayed. The transpose function icon on the LCD is not salient, and serious errors can result if the user is unaware that the transpose function is active. 4 Sound patch codes are arbitrary and convey no information about patches with similar codes. Incremental changes within Instrument Subgroups or cycling through Banks represent substitutive, not additive changes. The physical state of the multi-function controllers does not always match the actual level of the represented sound parameters. The particular sound parameter that is modified is not constant across sound patches; therefore expected modulation may not match output. The transpose function representation is an up/down arrow on the display, which is a mismatch to the corresponding left/right shift in key output. 3 The user is forced to recall button sequences for specific sound patches. The user may have great difficulty remembering multiple patch button sequences that must be entered within short periods of time. The transpose function display does not indicate the particular musical key of the transposition. The user must remember this information. 3 The QS8.1 does not have an online help system. The only source of information about how to handle/avoid particular errors is the Alesis manual. There are few available escape routines or other error management implementations. A user is not able to “undo” an accidental function command or return to a previous state. 3 3.0 Design Improvement Suggestions This section is concerned with design improvements that address each of the most severe usability issues. This is not a comprehensive list; given more time we could report many more improvement suggestions. For illustrations of many of these design suggestions, refer to Appendix D. 3.1 Visibility of the System Status Illuminate Sound Patch Buttons (Figure D1) When entering a sound patch sequence, the user is unable to tell which patch is active without shifting visual attention to the LCD display. The user will encounter difficulty in bridging the gulf of execution, as he will not immediately be informed that the system accepted his command. To address this problem, a general solution is to co-locate visual 6 IOE 536 Cognitive Ergonomics feedback with the control. Therefore, we recommend illuminating the sound patch selection buttons as they are pressed. This will allow the user closure in confirming that buttons were selected. One problem that may result from this is that feedback is now distributed in different locations, instead of locating it solely on the LCD screen. Show Controller Status and Label Functions (Figure D2) It is difficult to determine the status (active/inactive) of the multifunction controllers in the current interface. This is a problem with perceived affordances, and again, in bridging of the gulf of evaluation. To address these problems, the system should be modeled as a reliable direct manipulation interface (DMI), where the affordances are clearly communicated. Therefore, we suggest light indicators below each controller to show which controllers are active for the selected sound patch. Another problem is that the controller labels (A,B,C,D) are not informative, as they do not indicate the underlying function, which changes depending on the active sound patch. Symbols on interfaces should allow a user to intuitively understand represented variables, and these variables should be consistent. To address this, appropriately label the DMI controls, and force them to manipulate consistent variables. We suggest limiting each controller to adjust only one specific sound parameter, and labeling the controllers with the particular parameters they modify. Integrating these changes will communicate system affordances to the user through a direct manipulation interface in a clear and consistent manner. One issue that will need to be further addressed is where to position the labels as to not be visually obstructed. Salient Transpose Representation (Figure D3) The current transpose function icon is not salient on the LCD display. Since a user’s attention is divided across modalities, the ability to focus visual attention on displays representing specific functions is hindered, and this can result in errors of omission. To address this, increase the salience of feedback and display it in such a way to accommodate pre-attentive reference. We suggest illuminating the transpose button when the function is active. By displaying a salient light signal, the user can pre-attentively realize the activity of the transpose function, and therefore will be less likely to commit errors of omission without distracting visual attention. In addition, by illuminating this feature, the operator can bridge the gulf of evaluation with simple binary feedback. 3.2 Match Between System and Real World Force System level to match Controller level 7 IOE 536 Cognitive Ergonomics There is a possibility of mismatch between the physical controller level and the internal system level of a sound parameter, in that the system level resets when switching sound patches. Multiple inconsistent representations can result in user models that do not match system models. Having one consistent direct manipulation interface will simplify such a system. We suggest that the system level be forced to match the physical controller level at all times. The current LCD indication system for these controllers should be removed to maintain consistency. This change would preserve the naturalness principle and provide a better match between the user mental model and system model. One problem with this change is that now each controller must be manually reset when changing sound patches if the current parameter modifications are not desired. Additive Sound Patch Changes Within an Instrument Group, the Instrument Subgroup and Bank of a particular sound patch seem to be arbitrary, in that there is no information conveyed from these entries. This requires a user to discover system state by data-driven procedures (trial and error). By accommodating knowledge-driven search, a desired system state may be found more readily. We suggest that incremental changes in either the Subgroup or Bank should represent sound parameter values on a continuum. For example, if code 50 represented the piano group then code 51 may be made to represent a piano sound with an incremented amount of reverberation, and the amount of reverberation could be even more finely tuned by cycling through the Bank positions at this code. Allowing the representation to be made more additive as opposed to substitutive, and providing the opportunity to develop the appropriate mental model will reduce the requirement for exploration by the user in searching for a desired sound patch. It should be noted that there is no universal standard to quantify parameter increase for some instruments, and this could be an issue in implementing this design suggestion. Display Transpose Output Key (Figure D4) The transpose function display icon mismatches the keyboard layout in that the up/down arrow represents a right/left shift on the keyboard. This violates the naturalness principle. Changes should be made to reflect the actual state of the system with clear symbols understood by all users. We suggest that the pitch (the particular key on the keyboard, for example, Bb3) of the transposed output be displayed in place of an arrow to avoid this mismatch. 3.3 Support Recognition Rather than Recall Hotkeys (Figure D5) 8 IOE 536 Cognitive Ergonomics It is difficult for the user to remember a number of sound patch button sequences, and be forced to quickly enter them in a small window of time during a performance. Since the operator bears ultimate responsibility for the operation of the system, building intelligent system elements to allow the user to use shortcuts will decrease errors of commission and remove the mental workload burden of recalling some procedures. Our suggestion is to add shortcut buttons (hotkeys) to the interface. These will be pre-programmable buttons which will provide one-touch access to specific sound patches and reduce the burden of memory recall and the shift in attention to focus on the interface. The user should be aware that the addition of this automated function allows for latent errors to be introduced, if the hotkeys are not programmed properly before the performance. Display Transpose Output Key (Figure D4) Since the current transpose display shows the direction, but not the magnitude of transposition, the user is forced to remember the particular key of the output. This is an example of the keyhole effect, where only a restricted amount of information about the system is available. The user is not allowed any additional peripheral information. In order to bridge the gulf of evaluation, the display should provide more information about the system state. Our solution is to display the output key in place of the arrow symbol, which is more informative to the user as it displays the magnitude of transposition. 3.4 Recognition, Diagnosis, and Recovery from Errors Help Button (Figure D6) Online help is not available through the QS8.1 interface. The number of mistakes can be reduced by supporting robust user mental models. We suggest the provision of a help button, which when pressed in conjunction with any other button would display a short description about the function of that button. Pressing help again would return to the normal display. This function assists in knowledge-based behavior mode and may help the user to generate more robust mental models. Undo Button (Figure D6) The QS8.1 interface does not support error management. A necessity in supporting error management is an escape routine, which allows a user to mitigate the consequences of erroneous actions. We suggest implementing an “undo” button that allows a user to revert back to the state of the system before the last function/step performed. This would allow an 9 IOE 536 Cognitive Ergonomics escape routine for when the user makes an error or is confused by interactions with the interface. 4.0 Next Steps Our analysis and proposed solutions were limited by time and resources. Given more of either of these, we would continue the effort in the following sequence: 1. Perform an iterative design process that includes building prototypes of suggested designs, then critically evaluating them within our group. 2. Simulate the environment and perform user testing with musicians to measure performance improvements of the redesign and identify potential problems. 3. Implement and test the most successful redesign solutions in the domain scenario of a live performance. 4. Propose improvements for production by performing various analyses. (e.g. cost/benefit analysis). 10 IOE 536 Cognitive Ergonomics Appendix A: Comprehensive System Description The synthesizer can be divided into two main areas based on functionality. 1 1) Device Controls (Buttons that allow functional control of the synthesizer) 2 2) Music Interface (The black & white keys that control musical output) Our evaluation was limited to analysis of the Figure A1 Top view of Alesis QS 8.1 Device Control Interface, which is detailed in figure A2. Figure A2: Device Control Interface Components Interface Power Switch Function Power plug and power on/off I/O Interface The output devices are connected to the keyboard via this interface. Hardware add-ons, such as preamplifiers and various foot pedals, are also connected here Memory & Function Expansion This interface allows for expansion of system memory, additional sound libraries, etc. 11 IOE 536 Cognitive Ergonomics Pitch Bend/Modulation Wheels The Pitch Bend Wheel is used to expressively bend the pitch of the synthesized sound. The Modulation Wheel is used to create interesting sonic changes in the current Program or Mix. [Alesis Manual] Volume & MultiFunction Controllers The volume control slider is used to adjust the sound output level of the synthesizer. The controllers A, B, C, and D are programmable and can be used to obtain a hands-on control of many different parameters. [Alesis Manual] Edit Mode Controls [▲ Value] / [▼ Value] - When in edit mode, these buttons increment/decrement the selected value. [Edit Select] - This button activates Edit Mode. Pressing either <Mix> or <Program> cancels Edit Mode to return to normal play. [Store] - It is involved in saving and loading both User and Card Banks, in copying sound patches, and when initializing individual patches within Program Mode. [Page ◄] / [Page ►] - In Edit Mode, these buttons cycle backward / forward through the available “pages” for the current parameter (page indication is in the upper right of the LCD). Seq Select & Transpose Seq Select: Pressing this button activates Sequence Playback Mode. Transpose: This activates the transpose mode for the synthesizer output. Holding the Transpose button and pressing the third C key from the left (also known as C3) resets this mode. Sound Patch Controls These 27 buttons grouped together at the right side of the front panel, are used to select a particular sound patch within the current mode. LCD Display Figure A3: LCD Display Components 1) The large numbers on the left side of the LCD displays the Program or Mix code. The first (two) digit(s) represent the instrument group (the top row of buttons in the sound patch controls – in this case <00>), and the last digit represents the instrument subgroup, (the bottom row in the sound patch controls – in this case <1>). 2) The top line of the display shows the name of the particular sound patch that is active. This patch is unique to the current Bank, numerical code (the large numbers), and Mode. This location shows the name of the selected Function in Edit Mode. 12 IOE 536 Cognitive Ergonomics 3) The icons displayed at these labels represent the activity of several functions. [CLP] In Program or Mix Modes, an exclamation point (!) will appear in this area of the LCD if the signal clips internally. [SEQ] A blinking arrow will appear in this area of the LCD if a card sequence is triggered. [TRN] An up or down arrow will appear in this area of the LCD if transpose is active. [ABCD] The four vertical bar graphs represent the Controller A-D slider positions. 4) In Play Mode the middle line displays the bank number. In Edit Mode it displays the name of the parameter being edited. In Mix Program Select mode, it displays the Program and its MIDI channel. 5) This displays the current mode (Mix, Program, or Edit). 6) The bottom line displays the MIDI channel numbers 13 IOE 536 Cognitive Ergonomics Appendix B: Interview Summary and Representative Responses Our team conducted unstructured interviews with three keyboard musicians who had varying levels of experience with the QS8.1 interface. The following are representative questions and responses during those interviews. Q: Describe a scenario that you have been in where you have played this or a similar keyboard instrument and used some of the specialized functionality of that instrument. R: Playing in a bar, with a band. I am paying attention to the other band members to know when to take cues, either from the music or from visual communication with the other members. It’s very loud and I can’t see some things because of the lights. I may get nervous and mess up. Also I might be a little intoxicated if there’s free beer for the band. Q: What are some common problems you have within these scenarios when interacting with the interface of the QS8.1 or a similar instrument? R: I might want to switch between two or three sound patches that require long sequences of button presses. This is difficult to do if I only have a few measures to switch. Also, switching between some sound patches drastically changes the output volume, so I have to adjust that with each switch. R: Sometimes I use the transpose function for a particular song, and then I start the next song and I forget to turn off transpose. This causes a horrible dissonance with the other instruments until it can be fixed. It’s easy to forget that transpose is on because the synthesizer does not clearly display that the function is active. R: I really have no idea what’s going on with the edit functions for this. Whenever I accidentally or purposely enter Edit Mode, I can’t ever get out of it and so I have to either overwrite something or unplug the synthesizer. Q: What are some other improvements you might suggest for the QS8.1 interface? R: I never know what the controller sliders change specifically. They should either always change the same parameters, or have a way to label them. R: I could really use some kind of shortcut buttons so I don’t have to always remember which codes I need to enter and try to enter them quickly in a very small amount of time. 14 IOE 536 Cognitive Ergonomics Appendix C: Complete Heuristic Evaluation Heuristic Category Visibility of the System Status Focus Severity The system should always keep users informed about what is going on, through appropriate feedback within reasonable time. 4 The selection of features on the QS8.1 is dependent upon layers of options and menus that provide feedback through a 1.5x3 inch LCD display. Due to the limited space available on the display, symbols displayed can have multiple meanings dependent upon the current mode. The lack of description for these symbols can lead to errors in the user environment. For example: 1. The symbol for transpose function is a small up/down arrow ( ↑ ↓ ), which indicates transpose activation and direction (to a higher or lower pitch). However, it does not indicate the magnitude of the transposition, and therefore the particular musical key of the output cannot be inferred directly. Also, due to the diminutive size of the indicator on the LCD, it is not salient. 2. The display is spatially displaced from the majority of interface controls. This necessitates a reorientation of visual attention for closure/feedback after performing an action with the interface controls. 3. The physical state of the multi-function controllers does not always match the actual level of the represented sound parameters. Also, the particular controllers that are active depend on the particular sound patch. To determine which controllers are active and at what level, one must find the representation of the level on the LCD display (see **figure A3** of Appendix A). Although the LCD indicates which controllers are active (for the current sound patch), it will not identify the particular sound parameter they modify, nor the parameter modified by the modulation wheel control. There is an inconsistency in that the level of modulation according to the modulation wheel is determined only by the physical position of the wheel. 4. The status of I/O devices, such as the 3 control pedals and amplifier is not indicated on the LCD display. Therefore, appropriate settings for these devices can only be found through trial and error. This can be a problem if, for example, no sound is heard when a key is pressed: this may be due to sound patch or volume settings, or a number of settings or malfunctions in the peripheral devices or their connections. 5. Volume settings do not produce consistent sound output volume. As the operator changes sound patches, the intensity of the sound can increase or decrease without 15 IOE 536 Cognitive Ergonomics adjustment of the main volume. There is neither an indicator nor a control for this condition. The overall lack of representation of system settings can have implications in a domain where error tolerance is imperative. Since the system is “dumb” it cannot determine the difference between error and desired performance. The system operator then becomes the only measure of defense. Without the appropriate information, errors can occur, even for an experienced user. Heuristic Category Match between System and Real World Focus Severity The system should speak the users’ language, with words, phrases and concepts familiar to the user, rather than system-oriented terms. Follow real-world conversations, making information appear in a natural and logical order. 3 The QS8.1 displays settings with a complex array of symbols, and the user is forced to decode/interpret them. There are multiple examples of potential display/real world mismatches that may occur when attempting to interpret the system settings: 1. The numerical value of the subgroup within an instrument type seems to be completely arbitrary, in that it does not convey any information about similarity to sounds with nearby values. There are also substitutive changes between sound patches in Banks of the same subgroup. The user is then forced to randomly select sounds until they find the desired output. Desired sound patches could be found much more quickly if numerical codes and incremental changes of Bank represented additive changes in the sound parameters. 2. The modulator and multifunction controllers give no indication as to the particular sound parameters they address, and therefore the expected modulation may not match what is output when switching sound patches. Trial and error is the only user approach for discovering represented controller parameters. 3. The transpose function representation is an arrow pointing either up or down on the display, but the transposed output key is physically located left or right of the activated key (for example, if the output is transposed to the key of D, the arrow should point to the right instead of up, because the key D is located to the right of middle C, which is the normal output key). When the user is in a live performance scenario, these mismatches can result in errors. 16 IOE 536 Cognitive Ergonomics Heuristic Category User Control and Freedom Focus Users often choose system functions by mistake and will need a clearly marked “emergency exit” to leave the unwanted state without having to go through an extended dialogue. Support undo and redo Severity 2 In the defined user environment, monitoring occurs primarily through audition. Change in pitch, key, tone, sound type, etc. are typically realized by the user without any need for visual indication. However, if the change is unsolicited or unplanned, the user’s mental model is disrupted. They must then diagnose the change and adjust the settings to meet operation preference. Issues that illustrate limitations of the QS8.1 in terms of user control and freedom are: 1. The interface lacks an expedient method to reverse a change (i.e., an ‘undo’ button), often forcing the operator into a taxing knowledge driven environment that detracts from their ability to maintain contiguous play. 2. The complexity in manipulating controls for certain functions can limit the types of changes that can occur during a live performance. Situations exist where a user must avoid switching to certain sound patches because the button sequence is too long to be completed in a small window of time. Heuristic Category Consistency and Standards Focus User should not have to wonder whether different words, situations, or actions mean the same thing. Follow platform convention. Severity 4 The QS8.1 has multiple modes, Mixes, Programs, I/O options, etc. Depending on the configuration, different buttons, sliding controllers, modulators, and even keys can have different functions. With respect to the LCD display, the only features that are relatively consistent are the large numbers indicating instrument group and subgroup. The meaning of the remaining display elements will vary depending on the current mode. With this in mind we have identified the following violations: 1. There are much inconsistent functionality for the <Mix> and <Program> buttons between operation modes. In addition to switching to either Mix or Program mode in normal operation, these buttons perform the following actions: • <Program> or <Mix> must be pressed to exit Edit mode. • <Mix> must be pressed simultaneously with an instrument subgroup button to enter Demo mode, and <Mix> must be pressed to exit this mode. 17 IOE 536 Cognitive Ergonomics • Confirmation of an edited sound patch in Program Mode is done with the <Program> button, and cancellation is achieved with <Mix>. These examples illustrate the wide range of non-intuitive actions that are completed through these controls. 2. The level of the multi-function controllers resets each time the user changes sound banks, even though the physical state of the controllers remains constant. This forces the user to re-tune the modulation by manipulating each controller prior to playing or the desired sound may not be achieved. 3. The transpose icon indicator on the LCD does not provide information about the transposition magnitude, resulting in possible error generation. Heuristic Category Error Prevention Focus Even better than good error messages is a careful design which prevents a problem from occurring in the first place. Severity 2 The QS8.1 system is “dumb” in that it cannot determine the difference between error and desired performance. The system operator then becomes the only measure of defense. There are no security measures to promote error management, such as an “undo” button. For the most part, this is not a critical issue in the user scenario, since error correction does not require a process that is too complicated. However: 1. Having to key in long sequences of buttons to reach desired sound patches increases the likelihood of an error in live performance. 2. If Edit mode is purposely or inadvertently entered, it is very difficult to exit without turning off the power. Since edit mode has the potential to make permanent changes to the sound patches, there should be a more available exit/undo routine. 3. There is a lack of display of the states of peripheral devices. Therefore, if there is an errant setting in any of these devices, the user will not be aware until sound feedback makes the error apparent. For example, sustain pedal settings may be inverted depending on its state when the power switch is turned on (depressed pedal should sustain the note, while releasing the pedal stops it, but the reverse occurs). This is not communicated to the user until sound is produced. 18 IOE 536 Cognitive Ergonomics Heuristic Category Focus Severity Recognition rather then Recall Make objects, actions, and options visible. The user should not have to remember information from one part of the dialogue to another. Instructions for use of the system should be visible or easily retrievable whenever appropriate. 3 With the large functionality of the QS8.1, it is nearly impossible for a user to keep track of all active sound patches, settings, and active devices. Therefore, it is imperative that all state settings are displayed clearly and efficiently to promote user recognition at a glance. Some examples of the interface not effectively supporting recognition are: 1. The LCD display on the synthesizer is too small to clearly communicate the large amount of information about the state of the system. Due to the size limitations, some modes will require the user to page through multiple displays to achieve the desired goal. 2. The user is forced to recall button sequences for specific sound patches. If multiple patches are to be utilized in a short period of time during performance, the user may have great difficulty remembering all of the patch button sequences. A preprogrammable shortcut button is desired. 3. The transpose function display does not indicate the particular musical key of the transposition. Therefore the user must remember this information, or discover it through trial and error. Heuristic Category Flexibility and Efficiency of Use Focus Accelerators -- unseen by the novice user -- may often speed up the interaction for the expert user such that the system can cater to both inexperienced and experienced users. Allow users to tailor frequent actions. Severity 1 The QS8.1 is quite flexible in the enormous range of output sounds, and modulations on the sounds that it may produce. However, in the live performance scenario, the efficiency of the interface has much room for improvement. 1. Frequent button sequences or combinations that include more than two or three button presses could be simplified by allowing programmable macro buttons. This would increase efficiency by representing a complicated action with a single control. 19 IOE 536 Cognitive Ergonomics Heuristic Category Aesthetic and Minimalist Design Focus Severity Dialogues should not contain information which is irrelevant or rarely needed. Every extra unit of information in a dialogue competes with the relevant units of information and diminishes their relative visibility. 0 A satisfactory product is one that not only serves function, but has a desirable form. The aesthetic value of an interface is largely subjective, and so it is difficult to define aesthetic issues for the QS8.1. Assuming the perspective of the user persona for this analysis, we are able to define the following issues: 1. The desired size of the LCD display, in aesthetic terms, depends on the environment. A smaller display creates an overall look closer to a traditional piano. In the live performance scenario, the user prefers to have a larger screen to assist visibility, and also to communicate to the audience the large amount of functionality of the synthesizer. 2. The information labels surrounding the LCD display are too small and thus difficult to read, especially in the low lighting of the environment. 3. The physical weight of the synthesizer can be overwhelming when regularly moving it for performance setup. The QS8.1 is a particularly heavy model. 4. The controls visible in the interface are numerous and in some cases, poorly grouped according to Gestalt principles (like controls should be co-located and oriented appropriately). It is difficult to achieve a minimalist design in the controls while still maintaining the functionality and level of efficiency of this model. Heuristic Category Focus Severity Help Users Recognize, Diagnose, and Recover from Errors Error messages should be expressed in plain language (no codes), precisely indicate the problem, and constructively suggest a solution. 4 Help should be accessible from any point in the system, whether in response to an error or to explain functionality. However, the QS8.1 does not have an online help system. The only way for users to find the information they seek is to refer to the Alesis manual. This can cause serious problems to the users, especially during live performance. 20 IOE 536 Cognitive Ergonomics Heuristic Category Help and Documentation Focus Even though it is better if the system can be used without documentation, it may be necessary to provide help and documentation. Any such information should be easy to search, focused on the user's task, list concrete steps to be carried out, and not be too large. Severity 3 Documentation should clearly explain the functions of a system and how to use them in common tasks, and should do so in an organized manner that facilitates efficient search. This can be achieved through a good keyword index and through tabs, headers, and a table of contents that also shows how functions are grouped. Overall, the Alesis manual for the QS8.1 meets these specifications. It provides detailed step-by-step help focused on familiarizing the user with the functionality of the synthesizer. However: 1. The manual seems to be written for a target user that is using the synthesizer in a highly technological studio environment. This includes complex descriptions of sound output characteristics and interface capability for other studio devices. Therefore, the user who wishes only to use high level functions for purposes of live performance must consciously filter the manual content to find the documentation of use. 2. Some of the language is not plain enough for the target user in our evaluation. Again, the manual seems to primarily address studio professionals. 21 IOE 536 Cognitive Ergonomics Appendix D: Redesign Suggestion Figures Figure D1: Illuminate Sound Patch Buttons Figure D2: Show Controller Status and Label Functions Figure D3: Salient Transpose Representation 22 IOE 536 Cognitive Ergonomics Figure D4: Display Transpose Output Key Figure D5: Hotkeys Figure D6: Help Button and Undo Button 23