This one takes a bit more observation than usual to appreciate, even on my part when looking at it later. Notice that all outer tribars actually share a center corner, that is the cube in the middle. There are also floating corners between the outer tribars that match them in a different axis than the closest join to the center tribar. It feels like I should point and talk at the same time for it to make sense :)
Since 2013 I’ve been sporting a profile picture or avatar that features me wearing an Oculus Rift DK1. It was originally a photograph for my post about my impressions of the headset. It stuck with me for the longest time but now when consumer hardware has actually shipped I felt the need for an update. The full non-cropped version of it is below.
As I recently did a photo project with the consumer Rift, I decided to use the Vive for my profile to kind of balance things out. I had already featured Oculus hardware for several years so it just seems fair, no? I might swap back to Oculus in a few years or perhaps even to a third brand!
For reference you can see my old and new profile picture next to each other to the left.
The concept itself is based on the camera mode that exists in the Vive headset, dubbed Tron-mode because of the traced outlines and blue color. This is basically the Tron-mode but reversed to show what’s inside the headset rather than outside!
I posted the resulting image to Reddit as I thought it was good enough to share and there I was asked to provide details on how I made it. As I have now written it up once already I figured I could just as well post it here too.
I used my Panasonic Lumix G5 with a 20mm f1.7 lens. This was mounted on a rail with a webcam facing the back of it, this put what was on the camera LCD on my screen through using OBS Studio. The rail itself was mounted to a tripod.
On the PC I ran Virtual Desktop to see what happened on the camera screen while in VR. I used a wireless keyboard so I could press F4 to reset where my desktop was positioned in space, this so I could see my old profile picture to try and strike a similar pose.
To trigger exposures I connected my TriggerTrap dongle to the camera and mounted my phone which ran the TriggerTrap app to a video monopod. I set it to do time-lapse with an exposure every 10 seconds, this to give me time between shots to adjust. I also set the camera to three seconds of review to let me see the result.
I had my flash unit mounted on a separate tripod and I bounced the light off of the white ceiling. The flash was triggered via wireless FlashQ trigger units. The flash was set to manual mode and the second highest gain level.
On the camera I manually set the white balance, aperture, shutter speed and ISO but used single shot auto focus. With the focus area around my nose and eyes and at the same time the front of the headset I could get both shots I needed to be sharp. Manual settings were used to get as close to identical exposures as possible as the result is going to be a composite image.
When I thought I had gotten a good shot with the headset on I would try to stay as still as possible while quickly removing it so I had the same face and pose in the next shot. This was a process I had to repeat a number of times until I was happy with what I got, I actually swapped lenses a few times too. It’s easy to get a mismatch in head angle between the two exposures.
I lifted the shadows in the two selected RAW files and then put them in layers with the no-HMD one on top, setting its blending mode to Difference and matching position on my top row of teeth as they are fixed to my skull.
Then I made a smart object of the top layer which again is only my face. I applied Surface Blur, Find Edges and Invert. Then put the layer in a group, set blending mode of the group to Screen and applied a layer mask to remove everything that was outside the headset profile.
Finally put a new layer above the face layer in the same group and fill it with your tron-mode color of choice, set layer blending mode to Color.
The thumb graphic is just painted, blurred and with blending mode set to Overlay.
Virtual photo shoot
I had to get a background as I did the shoot against a boring dark roller blind. My tripod was still in the same position and angle so I mounted the Vive headset onto the rail to get the same angle of the virtual camera as the physical one I had used.
I booted up The Lab and after some experimentation (and a ton of crashes) I ended up taking a screenshot from the Valve demo room #1. I chose this room as it had similar lighting conditions to my actual room. There are yellow tinted reflected light on me as the photo was shot in a room with yellow walls and the demo room naturally had yellow tinted light.
I scaled up the screenshot and put it in the background, applied blur and brightened it up to match my photo. I extracted myself from my original background with a layer mask so the new background became visible then walked over the edges with a soft eraser to clean it up.
Levels adjustment layer just below the tron-mode group, adjusted all three handles to my liking, it’s basically a more versatile contrast adjustment.
Cropped it to use as a profile picture. Export. Done.
To the left is an animated gif showing the major Photoshop steps, a bit simplified but it’s basically what is going on.
Additionally to the right, as a reference if anyone wants to recreate this, are all the Photoshop layers I ended up .
Phew, well, it took way longer than I expected both to do this and to write it up! I did all in all four different versions of this profile picture before reaching something I was enough satisfied with to use :P Not sure how many hours went into it but more than I want to think about ;)
I’ve hinted on making a buyer’s guide a few times now and I used a few evenings to set up a page for it. Originally I was going to make a folding form you could print and fill out, but considering many of the terms used can be unfamiliar to people I opted for an interactive version with tooltips.
I’ve tried to be as objective as possible making this and have avoided adding features that is in any way subjective or based on personal preference. In the end I recommend to actually trying the hardware out if you really care about getting what suits you the best. Try to find demos in stores or see if any of your friends have bought a system. Personally I’ve had random people from the Internet come by my place to try things out, but that might be tricker to arrange :)
If you think there are misrepresentations in the form please provide me with feedback! I’ve had about 20 people try this out previously through Twitter and instant messaging and I’ve done adjustments accordingly. You can contact me through the comments section below, the email form or Twitter.
To clarify one thing. At the moment I only see the Rift and Vive as the currently available consumer hardware on PC, this as the official OSVR headset while available is still labeled as a Hacker Development Kit. The name alone makes me think it’s not directly targeted towards consumers.
But, the HDK 1.4 does feature positional tracking, a low persistence screen and can be bought by anyone so I guess if someone wants to help me expand the guide to include the OSVR HDK (or any other headset I guess) feel free to supply me with which features in the list it supports and suitable extra features that are not included yet including tooltips. I’ve felt hesitant to add the HDK mostly because I’ve assumed it is not the final iteration yet and I also don’t own it myself so I have little knowledge of the actual feature set and lack the ability to validate things in person.
Sometimes I get this creative urge that drives me to just make something, then photography is a nice outlet, I get to conceptualise and plan something, set up and experiment with image technology, act as a talent and finally dabble around with editing. This entry is about a project that stretched between 2013 and 2016, it turned into a series which I now see as complete. I’ll tell you the story.
Back in 2013, the year I got the Oculus Rift DK1 virtual reality headset, I got the idea to somewhat reenact a very known cyberpunk art piece using the kit, the end result is what you see below.
This is where I live now, 2013-11-19
At the time I had just moved in with a few friends the previous month. In the shot I used the bottom mattress I had to sleep on, placed my work Macbook Pro next to it and emptied the rest of the visible floor, which at the time was done very quickly as my room was almost entirely void of possessions.
The experience seen on the screen was an actual live application and indeed what I saw in the headset. I moved around in the game world until I found a vantage point I thought looked like what you can see on the goggles in the reference art.
To capture this self portrait I used my phone and TriggerTrap to continuously take in-camera bracketed shots, reposing myself between every three exposures and trying to imagine what could look remotely similar to the artwork. I did bracketing because I wanted details in the lit lamp, the screen as well as the very dark shadows.
I did about 4-5 passes, posing, checking the camera for results, posing again until I figured I had something to work with. As I currently don’t have access to my large Lightroom Catalog because the app refuses to work with network drives I have no way to check what kind of editing I did, now 2.5 years later it is quite fuzzy in my mind but I’ll post a small strip below of the three source exposures and the final edit. Basically it’s a simple HDR merge and no other editing outside of basic adjustments and some selective desaturation.
The image itself was made solely to be posted on the Oculus subreddit over at Reddit, an open forum and community about virtual reality and the company behind the headset I was wearing. Just as I mentioned in the previous entry it’s easier to find motivation to be creative if you have an audience you are fairly certain will appreciate what you make.
After posting it, it randomly got a lot of attention over at Imgur, the image host I used. Somehow that score seemed to have transferred to the Reddit post itself, it ended up being the top voted post for the entire subreddit for quite some time to probably everyone’s surprise.
In 2014 I got the Oculus Rift DK2, the next generation development kit, and after a couple of months using it I decided to do a follow-up image to the first one. This time around I had no real idea or concept behind it, so I just went full bonkers and used all accessories I had amassed for VR up till then. You can see the final result below.
Home Improvement, 2014-10-05
One motivation behind making this picture was that I less than two weeks earlier had acquired a GTX 970, this is the the card you see shining blue inside my old desktop, a computer that was brought out of storage for me to have something to play games on. The Macbook was not really built for gaming or VR.
The accessories in the picture that I had bought solely for VR are as follows: Razer Hydra, Logitech G27 wheel, Thrustmaster T-flight HOTAS and the Leap Motion (on top of the PC tower) hand tracker. The Xbox gamepad is in there as well because it was used in quite a few of the VR titles I played and I preferred it above keyboard and mouse when in VR. Fitting enough the consumer Rift would end up shipping with just that, a gamepad.
This time around I figured I wanted more stops of light between the darkest and brightest pictures so I went for five exposures per bracket, this as each step in the bracket can at the most be one stop from the previous one. This is a limitation of the camera I own, the Panasonic Lumix G5.
If my memory serves me I had ghosting issues when merging the five shots so I ended up not using all of them, which was fine as long as the entire range of light to shadow is still covered. As reference I’m providing a similar strip of images as before, or a grid I guess, so you can see what I had to work with. The sixth image is the final edit.
It didn’t help with editing that my brightest image still was quite dark, so increasing brightness introduced crazy amounts of noise in the image, this left the final result quite dark. What I really struggled with was getting a look I thought would fit the scene, I was uncertain if I should keep the sad mood of the first picture or make this happier, mostly from color balance and saturation, this confusion came from me not really having a set theme this time around so it became a little bit of everything. In my opinion this is definitely the weakest picture in the series, I even had a hard time coming up with a sensible title for it.
A fun tidbit about this photograph is that the game on the screen is just a static screenshot. I wanted something flashy and I didn’t have access to an early build of Elite Dangerous, it was to be released first two months later. It is still what I was looking at in the headset though as I was running with a single monitor back then, I tried to angle my head so it would match what I saw.
In 2016 my mind was fired up for the actual Oculus Rift release, the CV1, the consumer product. My original idea was to have me standing in a victory pose of sorts, that things went from dystopian to utopian from the first to last picture in the series. This fell short though as the Rift did not ship with Touch, Oculus own tracked hand controllers, because of that my mind went in the opposite direction and it became perhaps an even more dystopian picture than the first one. See the result below.
This is my life now, 2016-05-07
A lot of things are going on here. The base concept is actually an old idea I had before I wanted to do the utopia one, but where I would be in an ambulance or an actual hospital bed while in VR, kind of what would take place after this picture. I am lucky to know a few people in the medical business so I was able to get my hands on an IV bag and a catheter without too much trouble, prop ideas that came from a nurse I know when I mentioned the project. Add to this some bottled coloured water, a bucket with a toilet paper roll on it, a bunch of cornflake boxes and it’s a picture. I was told by the nurse it was the proper color for matching the urine from someone who has been on IV for some time, otherwise that color was the biggest uncertainty with this shot for me.
The hardware used in the picture is my recently built Mini-ITX machine, an old 4:3 monitor and the Rift CV1. Yet again the monitor is displaying a static screenshot of a game, Lucky’s Tale, this because the actual game is only displaying a very small window on the screen during normal gameplay which would have looked quite lame.
Lumix G5 and webcam mounted on a rail and tilted. Smartphone reverse mounted running TriggerTrap for bracketed bulb time-lapse.
This is the monitor I saw in Virtual Desktop, I used OBS to quickly rotate and crop the webcam image.
In the headset itself I was looking at the secondary monitor, which I also used for tome side light from the left. I was using an app called Virtual Desktop which shows me my desktop in VR and was through OBS displaying my webcam which was aimed at the camera LCD. This way I could actually see how I was posing while wearing a VR headset, similar to what I did for a Christmas card in 2014, even if the camera has no live output.
I also used this to reposition some of the props, it was a bit tricky though due to how much the image had degraded before I saw it in the headset. The images above show the technical setup, it was taken with a crappy camera though as my best gear was part of the actual rig.
Below is a short time-lapse of the setup and shooting process, I messed up in the middle with the classic mistake of turning off the recording when I thought I turned it on and vice versa, but it is fairly complete nonetheless.
Yet again I used TriggerTrap to do my chained bracketing. On a whim I tried a different mode and used Long Exposure HDR time-lapse. Basically I put the camera in bulb mode so it would expose for as long as the shutter release was held, then by using the TriggerTrap dongle and this LE HDR mode I could get way more stops of light between my bracketed shots, which I think is pretty evident in the final picture. It did bring with it a few issues though.
When I the day after shooting merged my selected three shots in Lightroom I would get strange halo effects around the brightest spots of the resulting HDR image which made me wonder if I had messed things up, the source images looked fine though. What I could do was use de-ghosting, it effectively removed the halos but also introduced a whole lot of noise in the darker parts of the image, which means most of it. What I ended up doing was to layer one HDR image on top of the other and mask out the halo effect and bring in the brightest parts from the de-ghosted image. This worked out very well, see below for the source images, the two HDR merges and the final result in a single grid.
As you might have noticed in the final version I also added a painting on the wall, this was a texture downloaded from Textures.com which I then applied some selective masking to. I also replaced all the ICA logos on the cornflake boxes with the Oculus logo as well as adding it to the IV bag. And hardly worth mentioning, I digitally opened the IV tube and valve in case anyone would study this in detail, they were closed as I did not actually have a needle in my arm. After that there was the standard adjustments like color balance, levels and fixing a few blemishes before I posted this as well to Reddit.
The last detail that is fun to mention is that I had gone without shaving for almost a month and grown out my beard and hair since winter because I wanted a specific look for this specific photograph. The day after the shoot I trimmed it all down to get back to looking somewhat respectable!
I got my VR hardware fairly quickly after launch, as I mentioned before, only one week late for both the Vive and Rift. This limited access of course had me thinking about what I could do to have some fun online, and Photoshop was the obvious answer. I worked on both the Rift and Vive and I’ll detail what I did in this entry.
For the Rift I went with some old school stuff, it was an idea I had gotten ahead of time and knew I just had to make, see if you can tell where this is from.
If you couldn’t guess it’s the pre-Kickstarter logotype for Oculus, they’ve changed their that twice since but this is what they had on their site from the beginning. The graphic itself was made by Denny Unger of CloudHead Games in the MTBS3D forums.
After finishing the picture I published it on Twitter, it didn’t gain much traction but no fret as it was a quick job with a fairly obscure reference. I did think more people in the VR community would catch on, but no matter. Below is an animation I made to show how it was done, I’ll mention the most interesting parts in text.
The main thing that made this look authentic was the slight bulge distortion, layer blending mode and most of all the lens blur filter. The blur really does a whole lot to make it look like it belongs due to the shallow depth of field in the photograph. I prepared a gradient that matched depth of the front of the headset and used that as a depth map to gradually blur the logotype.
I published it and thought no more of it, until people on Reddit thought it was real, so I zoomed in and was baffled myself how authentic it actually looked. This is one reason why I made the how-animation.
Next up is the Vive, this was also a no-brainer as even Valve themselves are theming their VR stuff after the Portal universe. As before I took a bunch of photos of the hardware and got to work.
Honestly I didn’t think this would be much different when it comes to attention, but it kind of blew up in my face on Twitter, at least when compared to what usually happens with my tweets (0-1 likes). It’s definitely my most liked and retweeted tweet ever, and will most likely stay that way for a very long time!
I’m assuming the main reason why it caught on is because of the familiar branding. Franchises and IP is so important for recognition, most apparently, and Portal and Aperture Laboratories are well liked brands. This is one reason why making fanart is quite rewarding, you automatically get a (mostly) appreciative audience, I know this myself from back when I was a Halo fanatic and everything I produced was connected to that universe.
Back to the Vive and what I had to do to rebrand the product. This Photoshop job required way more effort than the Rift one, I liked the shallow depth of field there so I tried to replicated the effect. It was way trickier when I wanted to change the entire paint job though, I’ll describe what I ended up doing below.
Luckily I realised the difficulty of what I wanted to do even before taking the photos, so I shot it at both with a shallow depth of field and with sharpness throughout. To change the color of the blurred image I painted the sharp image and then laid that on top of the blurred one and blurred the edges manually to match the image below. It was quite laborious but ended up being the method that made it look realistic enough to satisfy my creative needs.
After that it was a matter of placing the logos and logotypes, I might have rushed that a bit as I was just so worn out after the recolouring and wanted to be done with it, but it looks good enough. I tried applying a curved distortion on the headset logotype but after struggling with that I realised the logo would fit around one of the sensor dimples so I just ran with that instead!
As with the Rift job it caused some confusion if it was real or not both on Twitter and Reddit, which when it comes down to it is the the entire point of Photoshopping in the first place, to make something that is not real appear real. For some reason I’ve done stuff like this for what, more than a decade now, meanwhile it’s still just a hobby.
Hopefully you enjoyed this quick journey into the mind of a Photoshopper :)
I might have said this before, but I’ll describe the way I come up with titles, as it confuses even myself sometimes. For every single figure I have a sketch on paper first, this sketch seldom represents what the final figure ends up looking like, much because of the transition to perspective which automatically changes thing around. The name, in any case, comes from the original sketch. I decide what to name all my files from when I make the first digital file. In this case it was most likely the top shape that made me think of the name, while the entire figure feels fairly unrelated.