Skip to content

Posts tagged ‘projector’

25
Mar

Multi-Touch Table – The Projector

Finally it’s time to talk about the projector.  Why devote a whole post to this one component of the table?  Because picking your projector is the single most maddening, difficult, and long lasting choice you will ever make about any part of your table build.  The projector is forever, Bitch!

Why is the projector so damn hard to choose?  Because many of the choices here will dictate entire sections of the rest of your build – overall table size, overall table height, bounce mirror(s), and camera placement to name a few.  Why does the projector choice affect all of these items?  Two words – Throw Ratio.

Throw Ratio

Simply put, the “throw” of a projector is a measure of the distance from the projector lens to the projection surface.  It’s how far the image is being “thrown” onto the surface.  The throw ratio of a projector is determined by taking the throw distance and dividing it by the  projected image width.  Projectors with large throw ratios will need to be further away from the screen to get a certain projected image size, projectors with small throw ratios (particularly those less than 1.0) are referred to as “short throw” or even “ultra short throw” and can often produce a 40-60″ diagonal image while being only 12-18″ away from the projection surface.

Why does it matter?

The simplest way to get a projected image onto a tabletop surface is to simply aim the projector straight up at the bottom of the surface.  There’s a catch with that though, what if you want the table surface to sit 25″ from the ground?  Well, let’s say you have a projector that’s 10″ long from front to rear.  If you take that projector and point it straight up at the bottom of the table the lens is probably going to be sitting about 12-13″ from the bottom of the surface (remember you need to leave room for the cables to go into the back so that the actual front to rear length of the projector is going to be closer to 12-13″ rather than the 10″ length of the projector itself) leaving you with 12-13″ of “throw” to the bottom of the tabletop.   Let’s say you want a 45″ 16:10 diagonal image – to get that in this setup you’d need a projector with a throw ratio around 0.31 – 0.32.  This is because a 45″ 16:10 image is about 38″ wide.  12″/38″ (Throw/Width) = 0.3157.  This is a pretty short throw projector!  Well that’s great you say, you know the throw ratio you need, now just go find a projector with that ratio and you’re done!  Well, no, not exactly.  You also need to worry about the offset of the projector.

Offset

The offset of a projector is a measure of the distance between a line shot straight out of the center of the lens and the actual bottom (or top if you’re talking ceiling mounted) of the projected image.  This is really hard to show without an image so thankfully the nice people over at InFocus have gone and made one that I will now share with you here.  That doc does a far better job explaining offset in a few pictures than I ever could in words.  Here’s the important takeaway – a projector with a zero offset means the image is evenly split in height around the center of the lens.  This would be great for just pointing it straight up – center it in the table and go.  Only slightly less good than an offset of zero is any offset less than or up to 100.  With an offset of 100 you just center the projector width-wise and place the center of the lens aligned to the bottom (or top) of the projection area and you’re good.  This works because most tables have a lip of some kind typically a few inches wide and most projectors are only 3-4″ tall, so you can easily “hide” the 2″ or so the projector sticks out past the projection inside the table.  Problems arise as the offset increases and in almost every case it will be greater than 100.  This means if you just point the projector straight up the distance it has to be offset from the actual projection area gets larger and larger and soon the entire projector will need to be placed “outside” your table enclosure to get any kind of image within your desired projection area.  Now we have two problems to deal with – throw, and offset – so how do you deal with them?

Mirrors

I’ve already spoken about mirrors in a previous post and I’m just too lazy to link to it here but you can find it if you poke around.  Mirrors can do two things for you – increase your throw distance and/or correct for an offset angle.  So great, just tear the bathroom mirror off the wall and you’re set, right?  No, not at all really.  Most of the mirrors we deal with in daily life are what’s called second surface mirrors.  This means the actual reflective surface is placed behind some kind of clear protective layer – typically glass, but often times acrylic.  The reflective layer is typically very thin and fragile so putting it behind something makes a lot of sense – it allows you to clean the mirror without rubbing off the very thing that makes it a mirror!

A problem arises though if you try to use a second surface mirror as a way to collapse a projector’s throw distance.  You get ghosting.  This is maybe best explained by thinking about the old rear view mirrors in cars (not the fancy electrochromic auto-dimming ones we have today) that had that little lever on the bottom for “night driving mode”.  These mirrors worked by exploiting the fact that the front surface of the transparent layer of the mirror ALSO acts like a mirror it’s just way less good at reflecting (meaning it’s less bright) so you typically can’t see it.  The fact that it’s less bright is a good thing for a rear view mirror but this effect – when sitting a few inches away from a 2500 lumen projector is pretty noticeable.  So what you get is a reflected image from the rear of the mirror, and a second “ghost” reflection from the front of the mirror.  There’s an easy (but not cheap) way to solve this and that’s to use a… wait for it… First surface mirror.

A first surface mirror, as the name would imply, moves the reflective layer up to the front of the mirror substrate.  This way there can be only one (!!!) reflection since the image never passes through to the back of the substrate and is only reflected off the “first” surface.  The problem with first surface mirrors is that they are really expensive and also really fragile in the sense that it’s easy to damage the reflective layer, not that they are more prone to breaking – thought hey do tend to be pretty thin.  First surface mirrors are used a lot for guiding lasers and hence tend to require really strict optical properties which the dumb-ass like me who just wants to bounce a projected image around has to pay for.  There are some cheaper variants around but quality really can vary so be sure of the product before you buy.

All that being said – for just futzing around and general “aiming” of your image a second surface mirror can and will work.  I plan to do some rough planning and design with a cheap IKEA acrylic mirror before dropping the cash on a good first surface mirror.  Yes there will be ghosting, but you can scratch and scrape that cheap mirror without much worry during the early stages of a build and not cry about wrecking it.

As you can guess – if the offset of my projector is too much (and I am confidant it will be) I plan to use a single mirror to adjust and bounce the image properly to my screen.  There may be some challenges to this but hopefully less than the original two-mirror bounce I was planning on the previous table build.  The more mirrors you add the harder it becomes to solve for the right angles at all points.

Are We Ready to Buy a Projector Yet?

Well, no.  So we’re already worried about throw ratios, lens offset, mirrors and angles – what else can there be?  In addition to those items you also need to decide on the projector brightness (which is further impacted by your choice of projection material, the thickness of your acrylic “sandwich”, the expected brightness of the area the table will be used in, etc.) which is measured in lumens, the resolution and aspect ratio of the projector, the actual projection technology (LCD, DLP), the cooling and airflow, the available inputs and outputs, and even the overall size and weight.  And let’s not forget flat out COST too.  Oh and then there’s optical zoom, lens shift, keystone correction, networked/serial/remote control options, and probably about 20 other potential pluses and minuses to consider.  Now you begin to understand why this one piece of technology can be a f$%king nightmare to decide on.

Ok, Are We Ready to Buy One Now??

Yeah, sure.  I’m sure you’re thinking, armed with all this knowledge, that I spent days upon days planning, searching, and optimizing every aspect of my projector choice… nope.  Here’s the reality – you do need to be aware of all the factors involved in your choice but when it comes down to it, you’ll have a budget, and you’ll go to some of the big projector reseller sites on the net and you’ll filter on price, then you’ll filter a bit more, maybe for projectors with a short throw, and then projectors with the resolution you want, and then you’ll have like 5-6 choices.  NONE of those choices will be the projector you want.  The projector you WANT is going to start somewhere around 5x your budget and it will be 3x larger in size than your table design can accommodate.  But you’ll have your 5-6 choices and now you need to start doing your homework.  Pick the “best” of those choices.  Hooray, you’re a winner!

The good news is that you are not the first person to try and build one of these tables and there are places you can go to see what projectors other people have used.  The NUI Group forums are a great place to do this.  However, if you are a dumb-ass like me you’ll avoid that completely and try a brand new, just released, totally untested projector.  Which brings me to…

The Dell S300 Projector

FINALLY!  This is what I bought.

Meet the S300

The Dell S300 is pretty damn new.  It wasn’t even shipping from Dell when I ordered it but the specs all seem pretty good and the price was what made me willing to take the risk of buying it.  The S300 is a short-throw, 1280×800 (16:10) DLP projector.  It supports both digital and analog inputs as well as network and/or serial control.  The light output can reach 2200 lumens and it even has support for 3D.  Typically a short-throw projector with this resolution and feature set would cost over $1000, probably closer to $1500 (and many still do), but the S300 can be had for a mere $899!  It’s almost enough to make you think that it’s LESS than $900!!  Seriously though, $899 is not a bad price so I went for it.

The projector is both light and small which should help with placing it.  The throw ratio is right around 0.51 meaning I can get a 45″ diagonal image (my desired size) with the projector only about 19-20″ from the screen.  The catch here is that the smallest image the projector does is 44.6″ diagonal.  Why is that bad?  Because it’s really damn close to 45″ and if I’m off just a bit the projector won’t be able to focus the image.  The throw distance between a 44.4″ image and a 45″ image is going to be a fraction of an inch but the complete and utter madness of wondering why the image is just never quite focused properly and tweaking it (accidentally moving the projector 1/8″)  and then trying to re-focus and seeing the image too big… it’s going to suck.

How can you not be drawn in by the sumptuous curves of that lens?

The projector sports a pretty significant array of connections which I hope to use to control it from the PC.  This way when you power on the PC it will them power on the projector, and when you shut it down, it’ll power it off.  It’s a small thing but it’s a nice feature to have rather than trying to rig up some physical or purely electrical way to do the same thing.

In my rigorous and extensive testing (playing XBOX360 games through the HDMI port) the S300 looks and works quite well.  I am able to get  something around an 8′ – 10′ diagonal image with the projector around 3-4′ from the screen (wall).  There is no tearing or ghosting that I have noticed (in both Left4Dead and Left4Dead 2) but that might be related to the fact that I’m just thrilled as hell to have an 8′ gaming screen.

This thing has enough inputs and outputs to make a porn starlet blush!

I’m not going to say that my “testing” hasn’t had any effect of delaying the build of the table.  I’m just going to let you figure that out by yourself.  Go ahead, figure it out, I’ll wait.

So why is this a really dumb thing to do?  Because I still don’t know what the real offset of the projector is.  It’s not really addressed in the manual and I have not yet tried to measure it out myself.  Either way, while the throw really is short, it’s not so short that I can just stand it on end and have the image size I want within the table height that I want.  So I will almost certainly have a mirror in play.  But depending on the offset adding that mirror might be really easy, or it might be incredibly frustrating – I simply don’t know yet.

Poorly photographed and smudged with fingerprints

I might be the first person to try a table build with this projector so there’s no one else’s experience to draw from.  There is no reference design for these tables.  That’s part of the fun, but it’s also part of the risk.  I may have dropped $900 on a projector that won’t actually do what I want/need it to do.  I might get it aimed and bounced perfectly but find it’s not bright enough, or too bright.  But this is what hacking things like this table is about – picking the best parts from what you can afford and then finding a solution to the engineering challenges that arise from those choices.  I’m sure I can make this thing work, I just can’t tell you exactly how yet.

Next time I’ll cover some of the other parts I already have that are key to the working of the table.  I’ll also try to make a post that’s less than 2500 words.  Even I’m getting sick of reading this post at this point.  Until next time…