Vergence-accommodation conflict: Difference between revisions - VR & AR Wiki

Line 1:

'''Vergence-accommodation conflict''' ('''VAC'''), also known as '''accommodation-vergence conflict''' or sometimes '''accommodation–vergence mismatch''', is a visual and perceptual phenomenon that occurs when the [[Brain|brain]] receives mismatching cues between the distance to which the eyes are pointed or converged ([[Vergence|vergence]]) and the distance at which the eyes' lenses are focused ([[Accommodation (eye)|accommodation]]).<ref name="Hoffman2008">{{cite web |last=Hoffman |first=D. M. |last2=Girshick |first2=A. R. |last3=Akeley |first3=K. |last4=Banks |first4=M. S. |title=Vergence–Accommodation Conflicts Hinder Visual Performance and Cause Visual Fatigue (Journal of Vision, Vol. 8, Issue 3, Article 33) |url=https://jov.arvojournals.org/article.aspx?articleid=2192424 |year=2008 ~~|doi=10.1167/8.3.33~~ |access-date= [Insert Access Date Here] }}</ref><ref name="Kreylos2014VAC">{{cite web |last=Kreylos |first=Oliver |title=Accommodation and Vergence in Head-mounted Displays |url=http://doc-ok.org/?p=1602 |website=Doc-Ok.org |date=2014-04-13 |access-date= [Insert Access Date Here] }}</ref> Because natural viewing conditions tightly couple these two mechanisms, breaking that link is a primary cause of visual discomfort and performance issues in modern [[Virtual reality|virtual reality]] (VR), [[Augmented reality|augmented reality]] (AR), and other [[Stereoscopy|stereoscopic]] 3-D displays, including nearly all mainstream [[Head-Mounted Display|head-mounted displays]] (HMDs).<ref name="Hoffman2008" />

'''Vergence-accommodation conflict''' ('''VAC'''), also known as '''accommodation-vergence conflict''' or sometimes '''accommodation–vergence mismatch''', is a visual and perceptual phenomenon that occurs when the [[Brain|brain]] receives mismatching cues between the distance to which the eyes are pointed or converged ([[Vergence|vergence]]) and the distance at which the eyes' lenses are focused ([[Accommodation (eye)|accommodation]]).<ref name="Hoffman2008">{{cite web |last=Hoffman |first=D. M. |last2=Girshick |first2=A. R. |last3=Akeley |first3=K. |last4=Banks |first4=M. S. |title=Vergence–Accommodation Conflicts Hinder Visual Performance and Cause Visual Fatigue (Journal of Vision, Vol. 8, Issue 3, Article 33) |url=https://jov.arvojournals.org/article.aspx?articleid=2192424 |year=2008 |access-date= [Insert Access Date Here] }}</ref><ref name="Kreylos2014VAC">{{cite web |last=Kreylos |first=Oliver |title=Accommodation and Vergence in Head-mounted Displays |url=http://doc-ok.org/?p=1602 |website=Doc-Ok.org |date=2014-04-13 |access-date= [Insert Access Date Here] }}</ref> Because natural viewing conditions tightly couple these two mechanisms, breaking that link is a primary cause of visual discomfort and performance issues in modern [[Virtual reality|virtual reality]] (VR), [[Augmented reality|augmented reality]] (AR), and other [[Stereoscopy|stereoscopic]] 3-D displays, including nearly all mainstream [[Head-Mounted Display|head-mounted displays]] (HMDs).<ref name="Hoffman2008" />

==Physiological Basis==

Line 6:

* '''[[Accommodation (eye)|Accommodation]]''': The [[Ciliary muscle|ciliary muscle]] adjusts the shape and thus the [[Optical power|optical power]] of the [[Crystalline lens|crystalline lens]] within each eye to bring the image of the target object into sharp focus on the [[Retina|retina]]. This response is primarily driven by retinal blur.

In natural vision, these two systems are tightly linked through fast, reciprocal neurological signals known as the [[Accommodation reflex|accommodation–vergence reflex]].<ref name="Kreylos2014VAC" /><ref name="Kramida2016">{{cite web |last=Kramida |first=G. |title=Resolving the Vergence–Accommodation Conflict in Head-Mounted Displays (IEEE Transactions on Visualization and Computer Graphics, Vol. 22, Issue 7) |url=https://~~doi~~.org/~~10.1109~~/~~TVCG.2015.2473859~~ |year=2016 ~~|doi=10.1109/TVCG.2015.2473859~~ |access-date= [Insert Access Date Here] }}</ref> This coupling ensures that the eyes focus at the same distance they are pointed, allowing for clear, comfortable, and efficient vision. Stereoscopic displays disrupt this natural coupling because binocular disparity cues drive the vergence system to the ''simulated'' depth of a virtual object, while the accommodation system is driven by blur cues to focus on the ''physical'' display surface, which is typically at a fixed optical distance.<ref name="Kramida2016Resolving">{{cite web |last=Kramida |first=Gregory |last2=Varshney |first2=Amitabh |title=Resolving the Vergence-Accommodation Conflict in Head Mounted Displays |url=https://www.cs.umd.edu/sites/default/files/scholarly_papers/Kramidarev.pdf |website=Department of Computer Science, University of Maryland |year=2016 |access-date= [Insert Access Date Here] }}</ref>

In natural vision, these two systems are tightly linked through fast, reciprocal neurological signals known as the [[Accommodation reflex|accommodation–vergence reflex]].<ref name="Kreylos2014VAC" /><ref name="Kramida2016">{{cite web |last=Kramida |first=G. |title=Resolving the Vergence–Accommodation Conflict in Head-Mounted Displays (IEEE Transactions on Visualization and Computer Graphics, Vol. 22, Issue 7) |url=https://ieeexplore.ieee.org/document/7296633 |year=2016 |access-date= [Insert Access Date Here] }}</ref> This coupling ensures that the eyes focus at the same distance they are pointed, allowing for clear, comfortable, and efficient vision. Stereoscopic displays disrupt this natural coupling because binocular disparity cues drive the vergence system to the ''simulated'' depth of a virtual object, while the accommodation system is driven by blur cues to focus on the ''physical'' display surface, which is typically at a fixed optical distance.<ref name="Kramida2016Resolving">{{cite web |last=Kramida |first=Gregory |last2=Varshney |first2=Amitabh |title=Resolving the Vergence-Accommodation Conflict in Head Mounted Displays |url=https://www.cs.umd.edu/sites/default/files/scholarly_papers/Kramidarev.pdf |website=Department of Computer Science, University of Maryland |year=2016 |access-date= [Insert Access Date Here] }}</ref>

==Causes / Occurrence in Display Technologies==

The vergence-accommodation conflict is inherent in display technologies where the perceived depth of content differs from the physical or optical distance of the display surface:

* '''Fixed-focus HMDs''': Nearly all consumer VR and many AR headsets use lenses to place a virtual image of the display screens at a fixed focal distance, typically between 1.3 and 2 meters (though this varies).<ref name="Kreylos2013HMD">{{cite web |last=Kreylos |first=Oliver |title=Head-mounted Displays and Lenses |url=http://doc-ok.org/?p=1360 |website=Doc-Ok.org |date=2013-07-24 |access-date= [Insert Access Date Here] }}</ref> Viewers must accommodate to this fixed plane to see a sharp image. However, stereoscopic rendering creates virtual objects that appear at various depths, requiring vergence changes. Objects rendered virtually nearer than the fixed focal plane induce a ''positive VAC'' (eyes converge more than they accommodate), while objects rendered virtually farther induce a ''negative VAC'' (eyes converge less than they accommodate).<ref name="Shibata2011">{{cite web |last=Shibata |first=T. |last2=Kim |first2=J. |last3=Hoffman |first3=D. M. |last4=Banks |first4=M. S. |title=The Zone of Comfort: Predicting Visual Discomfort With Stereo Displays (Journal of Vision, Vol. 11, Issue 8, Article 11) |url=https://jov.arvojournals.org/article.aspx?articleid=2192969 |year=2011 ~~|doi=10.1167/11.8.11~~ |access-date= [Insert Access Date Here] }}</ref>

* '''Fixed-focus HMDs''': Nearly all consumer VR and many AR headsets use lenses to place a virtual image of the display screens at a fixed focal distance, typically between 1.3 and 2 meters (though this varies).<ref name="Kreylos2013HMD">{{cite web |last=Kreylos |first=Oliver |title=Head-mounted Displays and Lenses |url=http://doc-ok.org/?p=1360 |website=Doc-Ok.org |date=2013-07-24 |access-date= [Insert Access Date Here] }}</ref> Viewers must accommodate to this fixed plane to see a sharp image. However, stereoscopic rendering creates virtual objects that appear at various depths, requiring vergence changes. Objects rendered virtually nearer than the fixed focal plane induce a ''positive VAC'' (eyes converge more than they accommodate), while objects rendered virtually farther induce a ''negative VAC'' (eyes converge less than they accommodate).<ref name="Shibata2011">{{cite web |last=Shibata |first=T. |last2=Kim |first2=J. |last3=Hoffman |first3=D. M. |last4=Banks |first4=M. S. |title=The Zone of Comfort: Predicting Visual Discomfort With Stereo Displays (Journal of Vision, Vol. 11, Issue 8, Article 11) |url=https://jov.arvojournals.org/article.aspx?articleid=2192969 |year=2011 |access-date= [Insert Access Date Here] }}</ref>

* '''[[3D television|3D Cinema and Television]]''': VAC also occurs here, but symptoms are often milder. The screen is typically farther away, the [[Field of view|field of view]] is smaller, and content creators can limit disparities to keep virtual objects within a "zone of comfort" relative to the screen distance.<ref name="ISO2015">ISO 9241-392:2015, Ergonomics of human-system interaction — Part 392: Ergonomic requirements for the reduction of visual fatigue from stereoscopic images. International Organization for Standardization, 2015.</ref>

* '''[[Optical see-through display|Optical See-Through (OST) AR]]''': In OST AR glasses, virtual images (often at a fixed focus) are overlaid onto the real world. This creates a conflict not only between vergence and accommodation for virtual objects but also a potential mismatch between focusing on real-world objects at various distances and the fixed focus of the virtual overlay. This can introduce depth discontinuities, reduce the perceived registration accuracy of virtual objects, and cause discomfort.<ref name="Zhou2021">{{cite web |last=Zhou |first=Y. |last2=Li |first2=X. |last3=Yuan |first3=C. |title=Vergence-Accommodation Conflict in Optical See-Through Display: Review and Prospect (Results in Optics, Vol. 5) |url=https://~~doi~~.~~org~~/~~10.1016~~/~~j.rio.2021.100160~~ |year=2021 ~~|doi=10.1016/j.rio.2021.100160~~ |access-date= [Insert Access Date Here] }}</ref>

* '''[[Optical see-through display|Optical See-Through (OST) AR]]''': In OST AR glasses, virtual images (often at a fixed focus) are overlaid onto the real world. This creates a conflict not only between vergence and accommodation for virtual objects but also a potential mismatch between focusing on real-world objects at various distances and the fixed focus of the virtual overlay. This can introduce depth discontinuities, reduce the perceived registration accuracy of virtual objects, and cause discomfort.<ref name="Zhou2021">{{cite web |last=Zhou |first=Y. |last2=Li |first2=X. |last3=Yuan |first3=C. |title=Vergence-Accommodation Conflict in Optical See-Through Display: Review and Prospect (Results in Optics, Vol. 5) |url=https://www.sciencedirect.com/science/article/pii/S266717002100160X |year=2021 |access-date= [Insert Access Date Here] }}</ref>

==Effects and Symptoms==

Line 23:

* '''Focusing Problems''': Difficulty rapidly refocusing between virtual objects at different apparent depths because the natural reflex is disrupted. Users may also experience lingering focus issues or unusual visual sensations after removing the HMD.

* '''[[Virtual Reality Sickness|VR Sickness]] / Discomfort''': VAC is considered a significant contributor to symptoms like nausea, dizziness, and general discomfort associated with VR/AR use.

* '''Reduced Visual Performance''': Measurable degradation in tasks requiring fine depth judgments, reduced reading speed, slower visuomotor reaction times, and increased time required to fuse binocular images.<ref name="Hoffman2008" /><ref name="Lin2022">{{cite web |last=Lin |first=C-J. |last2=Chi |first2=C-F. |last3=Lin |first3=C-K. |last4=Chang |first4=E-C. |title=Effects of Virtual Target Size, Position and Parallax on Vergence–Accommodation Conflict as Estimated by Actual Gaze (Scientific Reports, Vol. 12, Article 20100) |url=https://~~doi~~.~~org~~/~~10.1038~~/s41598-022-24450-9 |year=2022 ~~|doi=10.1038/s41598-022-24450-9~~ |access-date= [Insert Access Date Here] }}</ref>

* '''Reduced Visual Performance''': Measurable degradation in tasks requiring fine depth judgments, reduced reading speed, slower visuomotor reaction times, and increased time required to fuse binocular images.<ref name="Hoffman2008" /><ref name="Lin2022">{{cite web |last=Lin |first=C-J. |last2=Chi |first2=C-F. |last3=Lin |first3=C-K. |last4=Chang |first4=E-C. |title=Effects of Virtual Target Size, Position and Parallax on Vergence–Accommodation Conflict as Estimated by Actual Gaze (Scientific Reports, Vol. 12, Article 20100) |url=https://www.nature.com/articles/s41598-022-24450-9 |year=2022 |access-date= [Insert Access Date Here] }}</ref>

* '''[[Focal Rivalry]]''': Particularly in AR, the conflict between focusing on a real-world object and a virtual object projected at a different focal distance can make it difficult or impossible to see both sharply simultaneously.

Line 42:

# '''Avoid Rapid Depth Changes''': Avoid sudden disparity jumps (> 1 D) or rapid oscillations in depth for prominent objects. Allow the visual system time (at least 500 ms) to adjust to significant depth changes.<ref name="Kramida2016" />

# '''Optimize UI Placement''': Present user interface elements, text, and critical information at or slightly behind the display's native focal plane where VAC is zero or negative (which is generally better tolerated).<ref name="Shibata2011" />

# '''Simulate Blur''': When hardware cannot provide correct focus cues, incorporate [[Gaze-contingent display|gaze-contingent]] [[Depth of field|depth-of-field]] rendering (simulating blur for objects not being looked at) to provide [[Monocular cues|monocular]] depth information that aligns better with vergence, potentially reducing cue conflicts.<ref name="Koulieris2017">{{cite web |last=Koulieris |first=G-A. |last2=Buhler |first2=K. |last3=Drettakis |first3=G. |last4=Banks |first4=M. S. |title=Accommodation and Comfort in Head-Mounted Displays (ACM SIGGRAPH 2017 Courses) |url=https://~~doi~~.org/10.1145/3084873.3084901 |year=2017 ~~|doi=10.1145/3084873.3084901~~ |access-date= [Insert Access Date Here] }}</ref>

# '''Simulate Blur''': When hardware cannot provide correct focus cues, incorporate [[Gaze-contingent display|gaze-contingent]] [[Depth of field|depth-of-field]] rendering (simulating blur for objects not being looked at) to provide [[Monocular cues|monocular]] depth information that aligns better with vergence, potentially reducing cue conflicts.<ref name="Koulieris2017">{{cite web |last=Koulieris |first=G-A. |last2=Buhler |first2=K. |last3=Drettakis |first3=G. |last4=Banks |first4=M. S. |title=Accommodation and Comfort in Head-Mounted Displays (ACM SIGGRAPH 2017 Courses) |url=https://dl.acm.org/doi/10.1145/3084873.3084901 |year=2017 |access-date= [Insert Access Date Here] }}</ref>

===Technological Solutions===

Line 50:

|-

! [[Varifocal display|Varifocal]]

| [[Eye tracking|Eye-tracking]] determines the user's gaze depth, and the display system adjusts a single focal plane to match that depth using [[Tunable lens|tunable lenses]] (e.g., liquid crystal, liquid lens, Alvarez) or mechanically moving components (screen or lens). | Meta Reality Labs Butterscotch Varifocal (2023);<ref name="DisplayDaily2023">{{cite web |title=Meta’s Going to SIGGRAPH 2023 and Showing Flamera and Butterscotch VR Technologies |url=https://displaydaily.com/metas-going-to-siggraph-2023-and-showing-flamera-and-butterscotch-vr-technologies/ |website=Display Daily |date=2023-08-04 |access-date=[Insert Access Date Here]}}</ref> UNC Wide-FOV deformable-mirror NED.<ref name="Dunn2017">{{cite web |last=Dunn |first=D. |last2=Tippets |first2=C. |last3=Torell |first3=K. |last4=Kellnhofer |first4=P. |last5=Akşit |first5=K. |last6=Didyk |first6=P. |last7=Myszkowski |first7=K. |last8=Luebke |first8=D. |last9=Fuchs |first9=H. |title=Wide Field-of-View Varifocal Near-Eye Display Using See-Through Deformable Membrane Mirrors (IEEE Transactions on Visualization and Computer Graphics (TVCG), Vol. 23, Issue 4) |url=https://~~doi~~.org/~~10.1109~~/~~TVCG.2017.2657058~~ |year=2017 ~~|doi=10.1109/TVCG.2017.2657058~~ |access-date=[Insert Access Date Here]}}</ref> | Delivers correct focus cue at the depth of fixation. Challenges include eye-tracking latency and accuracy, depth switching speed, limited depth range, and potentially incorrect blur cues for objects not at the fixation depth.<ref name="UNC2019">{{cite web |title=Dynamic Focus Augmented Reality Display |url=https://telepresence.web.unc.edu/research/dynamic-focus-augmented-reality-display/ |website=UNC Graphics and Virtual Reality Group |year=2019 |access-date=[Insert Access Date Here]}}</ref>

| [[Eye tracking|Eye-tracking]] determines the user's gaze depth, and the display system adjusts a single focal plane to match that depth using [[Tunable lens|tunable lenses]] (e.g., liquid crystal, liquid lens, Alvarez) or mechanically moving components (screen or lens). | Meta Reality Labs Butterscotch Varifocal (2023);<ref name="DisplayDaily2023">{{cite web |title=Meta’s Going to SIGGRAPH 2023 and Showing Flamera and Butterscotch VR Technologies |url=https://displaydaily.com/metas-going-to-siggraph-2023-and-showing-flamera-and-butterscotch-vr-technologies/ |website=Display Daily |date=2023-08-04 |access-date=[Insert Access Date Here]}}</ref> UNC Wide-FOV deformable-mirror NED.<ref name="Dunn2017">{{cite web |last=Dunn |first=D. |last2=Tippets |first2=C. |last3=Torell |first3=K. |last4=Kellnhofer |first4=P. |last5=Akşit |first5=K. |last6=Didyk |first6=P. |last7=Myszkowski |first7=K. |last8=Luebke |first8=D. |last9=Fuchs |first9=H. |title=Wide Field-of-View Varifocal Near-Eye Display Using See-Through Deformable Membrane Mirrors (IEEE Transactions on Visualization and Computer Graphics (TVCG), Vol. 23, Issue 4) |url=https://ieeexplore.ieee.org/document/7850947 |year=2017 |access-date=[Insert Access Date Here]}}</ref> | Delivers correct focus cue at the depth of fixation. Challenges include eye-tracking latency and accuracy, depth switching speed, limited depth range, and potentially incorrect blur cues for objects not at the fixation depth.<ref name="UNC2019">{{cite web |title=Dynamic Focus Augmented Reality Display |url=https://telepresence.web.unc.edu/research/dynamic-focus-augmented-reality-display/ |website=UNC Graphics and Virtual Reality Group |year=2019 |access-date=[Insert Access Date Here]}}</ref>

|-

! [[Multifocal display|Multifocal / Multiplane]]

Line 65:

|-

! Emerging Optics

| Novel optical components like Alvarez freeform lenses,<ref name="Liu2024">{{cite web |last=Liu |first=Y. |last2=Cheng |first2=D. |last3=Wang |first3=Y. |last4=Hua |first4=H. |title=A Varifocal Augmented-Reality Head-Up Display Using Alvarez Freeform Lenses (Journal of the Society for Information Display, Vol. 32, Issue 4) |url=https://doi~~.org~~/10.1002/jsid.1286 |year=2024 ~~|doi=10.1002/jsid.1286~~ |access-date=[Insert Access Date Here]}}</ref> tunable fluidic lenses, and deformable membranes are being explored for compact, low-power varifocal or multifocal elements. | Primarily research stage. | Aim for integration into smaller form factors. Manufacturing challenges, response time, optical quality, and control complexity remain active research areas.

| Novel optical components like Alvarez freeform lenses,<ref name="Liu2024">{{cite web |last=Liu |first=Y. |last2=Cheng |first2=D. |last3=Wang |first3=Y. |last4=Hua |first4=H. |title=A Varifocal Augmented-Reality Head-Up Display Using Alvarez Freeform Lenses (Journal of the Society for Information Display, Vol. 32, Issue 4) |url=https://onlinelibrary.wiley.com/doi/abs/10.1002/jsid.1286 |year=2024 |access-date=[Insert Access Date Here]}}</ref> tunable fluidic lenses, and deformable membranes are being explored for compact, low-power varifocal or multifocal elements. | Primarily research stage. | Aim for integration into smaller form factors. Manufacturing challenges, response time, optical quality, and control complexity remain active research areas.

|}