VR Short Lesson 5 (Spanish)

Resolución en Realidad Virtual comparada con la vida real

http://www.roadtovr.com/michael-abrash-explores-next-5-years-vr-technology/

En función de las gafas de Realidad Virtual que usemos, la resolución varía drásticamente. Sin tener en cuenta las Google Cardboard, la resolución más baja entre los dispositivos de RV más punteros la apreciamos en las PlayStation VR, mientras que la mayor  se encuentra… ¡en las Samsung Gear VR! Efectivamente, ni Oculus ni HTC Vive. ¿Sorprendido/a?

	Oculus Rift (CV1)	HTC Vive	PlayStation VR	Samsung Gear VR	Google Cardboard
Compañía	Oculus (FB)	Valve + HTC	Sony	Samsung+Oculus (FB)	Google
Plataforma	PC	PC	PlayStation 4	Note4, S6	Android
FOV	110º	110º	100º	96º	80/90º
FPS	90 Hz	90 Hz	120/90 Hz	60 Hz	60 Hz
Resolución	1080x1200	1080x1200	960x1080	1280x1440	-
Tracking	Camara	Habitación	Camara	No	No
Wireless	No	No	No	Si	Si

Vale, genial, ¿pero a qué nos referimos exactamente con resolución? En realidad virtual tomamos como medida un parámetro denominado “pixeles por grado” (PPD por sus siglas en inglés). Este parámetro depende de los pixeles en pantalla y el campo de visión (FOV en inglés). Vamos a ver un ejemplo con las Gear VR:

* Visor: GearVR

* Resolución de pantalla: 2560x1440 pixeles, lo que supone 1280x1440 por cada ojo (la mitad de la pantalla para cada uno)

* FOV: 96º

* PPD horizontal: 1280 / 96 = 13.4

* PPD vertical: 1440 / 96 = 15

En otras palabras, esto significa que si giramos 360º sobre nosotros mismos, tenemos 360º x 13.4 PPD = 4824 pixeles a lo largo de nuestro campo de visión.

Moviendo la cabeza verticalmente abarcando 180º, tenemos un total de 180º x 15 PPD = 2700 pixeles.

En la realidad, nuestros ojos pueden apreciar 60 PPD con un campo de visión de 210ª/100º horizontales y verticales respectivamente. Esto implica tener 12,5600 x 6,000 pixeles ante nosotros.

Parece que estamos bastante lejos de alcanzar una resolución digital similar a la apreciada en vida real, ¿verdad?

El tipo de la imagen inferior es Michael Abrash, Chief Scientist en Oculus. Durante la presentación Oculus Connect 2016, Michael pronosticó que llegaremos a la resolución 4k por ojo en tan solo cinco años. Sáltate la introducción del video y ve directamente a la parte interesante a los 10:02 minutos.

Si no te ha convencido, echa un vistazo a la evolución de la resolución en móviles desde 1998 hasta 2016, ¡y saca tus propias conclusiones!

Si te ha gustado poder leer esta breve lección en español en lugar de en inglés, déjanos un comentario en este post o en nuestra página de Facebook.

¡No te pierdas la siguiente!

VR Short Lesson 5

VR Resolution vs. what we see in real life

  http://www.roadtovr.com/michael-abrash-explores-next-5-years-vr-technology/

Depending on the VR headset we refer to, resolution varies drastically. Without taking into account Google Cardboard, the lowest resolution among the top VR headsets is on Play Station VR, while the greatest is… on the Gear VR! Yes! Neither Oculus nor HTC Vive. Surprised?

	Oculus Rift (CV1)	HTC Vive	PlayStation VR	Samsung Gear VR	Google Cardboard
Company	Oculus (FB)	Valve + HTC	Sony	Samsung+Oculus (FB)	Google
Platform	PC	PC	PlayStation 4	Note4, S6	Android
FOV	110º	110º	100º	96º	80/90º
FPS	90 Hz	90 Hz	120/90 Hz	60 Hz	60 Hz
Resolution	1080x1200	1080x1200	960x1080	1280x1440	-
Tracking	Camera	Room	Camera	No	No
Wireless	No	No	No	Yes	Yes

Ok well, but how do we measure resolution? In virtual reality, we measure it based on PPD (pixels per degree). This parameter depends on both screen pixels and FOV (Field of View). Let’s take Gear VR as our illustrative example:

* Headset: GearVR
* Screen resolution: 2560x1440 pixels, thus 1280x1440 per eye (half of the screen for each)
* FOV: 96º

* Horizontal PPD: 1280 / 96 = 13.4 pixels per degree

* Vertical PPD: 1440 / 96 = 15.0 pixels per degree

This means that if we spin 360º horizontally, we can appreciate 360º x 13.4 PPD = 4824 pixels. Similarly, moving vertically our head up to 180º, we can see 180º x 15 PPD = 2700 pixels.

If we apply this formula to obtain the resolution in real life, our eyes shall capture 60 PPD with 210º/100º of FOV (horizontally and vertically respectively). This entails 12,600x6,000 pixels in view, nothing compared to 1,280x1,440 of GearVR.

It seems we are very far from achieving a digital resolution similar to the one in real life, isn’t it? Well, the guys in the picture below is Michael Abrash, Chief Scientist at Oculus, and he predicted 4K pixels per eye in solely 5 years during the annual presentation at Oculus Connect 2016. Skip the intro and jump into this part at 10:02

If he doesn't convince you, take a look here to the evolution of mobile screen resolution from 1998 till 2016, and then do the math.

Stay tuned for next VR Short Lesson and follow us on our FB site!

VR Short Lesson 4: Designing VR experiences

www.felixandpaul.com

How would you tell a story in VR?

The same story is told differently in each media, both its content and its intensity. The story of Batman can be experienced through books, comics, movies, videogames and who knows if also in theatre. And we experience it differently each time because each media has its own way of telling.

In VR, this is even most pronounced. The distinction between spectator and actor tends to zero. The role of the VR user is “active” in terms of interaction, opposing to “passive” roles assumed in classic narrative media. Now you are spectator and narrator. The story depends on you and is happening because you are there.

	Cinema	Literature	Theatre	Virtual Reality
Narrative representation	Visual	Mental	Visual	Visual
Presence	Not physical	Not physical	Physical	Immersive
Interaction	No	No	No	Yes

In cinema, the camera is under author’s control, but in VR the camera is under user’s control. What does this mean from author’s point of view? The standard rules of storytelling are no longer valid!

The complete scenario has 360º, that is clear. But as viewers, we can’t pay attention to all spaces at the same time. If we are watching a comedy, the main action should take place in a master scenario, avoiding have to move our head all the time form left to right, front to rear to follow the argument. Lateral secondary scenarios are being used to introduce new actors/characters, while rear scenario is used for minor actions.

These best practices apply for narrative stories, not for VR emotional experiences. In this kind of videos, action takes place at all scenarios at the same time, encouraging the viewer to pay attention to whatever he or she may prefer. Most likely, several views will be needed to enjoy the whole experience!

Felix and Paul Studios are considered by many the best cinematic VR directors. Check how they use both techniques in of the most famous 360º videos. VR features available only in Chrome or Firefox!

facebook.com/felixandpaul

These are only few brushstrokes about designing VR experiences, depending on what authors want to show, requirements and style change drastically. We will go further in future lessons, so stay tuned!

Are equal the two images we see in virtual reality?

Todays’ lecture is about introducing two important parameters that we need to take into account when buying a headset. We refer to the interpupillary distance (IPD) and parallax.

IPD is the distance between the center of your eyes’s pupils. Well, actually this is called real IPD. So as you may expect, this length should match the distance in between the center of the lenses of the head-mounted display. Otherwise, we could experience blurred vision or occasional double vision.

 developer3.oculus.com

The average real IPD of humans is 64 millimiters, with a range of 54 and 72 milimiters.

On the other hand, we have the virtual IPD: the distance between the cameras. Developers play with this parameter to alter the sense of scale of virtual objects in relation to the viewer, increasing or decreasing the size of the virtual environment as desired. Look this happy user testing the difference of real and virtual IPDs.

static1.squarespace.com/

Lastly, to obtain the stereoscopic effect, each eye must see an image a little bit different from the other one, as occurs in actual reality. In more technical words, there is a difference in the apparent object’s position viewed along two different lines of sight.

This is called parallax, and together with the virtual IPD, is the most relevant parameter to obtain a sense of real presence within the virtual world. Here is a very representative example.

video-stitch.com/

Hopefully, you will make the next choice of VR headset with more conscience of what’s behind!

Did you know that most 360 videos need to be “stitched”?    

The consume of 360º videos and virtual reality experiences has increased dramatically in the last 2 years. Hundreds of new viewers are engaged daily, but only a few of them wonder what’s behind this incredible technology.

Maximum FOV (Field of View) of current lenses is around 180º horizontally*, while the human eye covers only 114º.

Thus, evidently this makes impossible to take 360º pictures with only one lens. So as a minimum we require two of them, and then bring them together in a proper way.

This process is known as “stitching”. As my grandma stitches my winter jumper, VR professionals stitch videos. And it is almost as difficult as stitching jumpers!

Stitching turns the individual videos into a single seemless panoramic video. The resulting video typically covers 360×180° FOV in a “world map” style of geometric projection. Stitching can be done automatically by specific software, but best results are obtained when edited “manually” by a professional to not miss any detail.

There are recently launched services that make the stitching process more smoothly: VideoStitch and Kolor are the most well-known software.

Besides being properly stitched, our videos need to be also perfectly synchronized. A gap of a few milliseconds between both videos would led on a total mesh for the viewer.

All in all, post processing a 360º video is not an easy task, but with knowledge or powerful programs we can achieve good results!

To get a taste of how stitching takes place in action, pay attention to the following video! The good stuff begins at 2:08.

*We are not considering omnidirectional cameras due to lack of commercialization.

How fast should be virtual reality to look like actual reality?

In low range VR headsets as Google Cardboard, there is a perceptible delay between the movement of your head and the image shown on the screen.

It is called “lag”, that directly impacts the quality of experience of the VR user.  It is the result of the existing latency that our headset or mobile device have.

lag and latency

Current non-VR videogames normally provide around 50 milliseconds of latency between the controller input and the image generated on the screen, which is more than enough for having a smooth experience on stationary displays.

But in order to achieve a feeling of “actual reality” within “virtual reality”, much more precision is required: ideally a VR device should provide a delay in the range of 7-14 milliseconds to look like reality, but this has not been achieved yet. John Carmack, CTO of Oculus, says that 20 milliseconds of latency provided by their Rift is deemed acceptable by most of the users.

There is also a physical limitation with actual LCD screens. No matter how fast is your video processor, these kind of displays show traces of old pixels many tens of milliseconds after the image was changed. That’s’ why all VR devices as HTC Vive, PS VR or this Oculus Rift use OLED screens.

Let’s see an interesting experiment applying lag in real life!

Stay tuned for the next VR Short Lesson!

Source:

http://oculusrift-blog.com/john-carmacks-message-of-latency/682/

http://image.slidesharecdn.com/presentation-virtualreality-160401180906/95/making-immersive-virtual-reality-possible-in-mobile-36-638.jpg?cb=1469737483