hegykc

I just had to deal with screen resolution for a personal project. After a ton of research I can see how people misunderstand certain aspects. When comparing a 24'' screen to another 24'' screen, sure you only want the resolution numbers. BUT, when comparing image quality of a 24'' to that of a 7'' screen, one needs to look at PPI ( pixels per inch) or the size of the dot pitch. And just for an example, a 24'' 1080p monitor has 92 pixels per inch or pixel size of 0.28 mm. While an Oculus rift 7 inch 1280x800 screen has 216 pixels per inch or 0.12 mm dot pitch. So if you want to look at picture quality, Oculus rift has 2.4 times better picture quality than a 24'' full hd monitor. And small 1080p screens are coming. But how it looks in real life, remains to be seen. I am sure it will be of equal or several times better than our huge monitors where every spec of side distraction or light ruins the immersion. Think of it like this: Two cars, one has 1080 horse power, the other 800. Can you say which one is better by only those numbers. But the first one is 2.400 kilograms and the other one 700 kilograms. What about now?

I wouldn't be so sure about it. Convergence at 300, and shooting at 150 is not as big of a problem as having convergence at 150, and shooting at 300. If your convergence is at 150 yards and you shoot at 300 yards, your left/right bullet groupings are 5 yards/meters apart. You're not shooting anything down with that. If you put the sight on the center of the target, you're missing by 2 yards/meters on both sides. And if you count in dispersion, your hits might just be harmless. On the other hand, having the convergence at 300 and shooting at 150, your hits have tighter groups, just that the left/right bullet groupings grow apart. It's really not that simple as saying I want 150 yard convergence. You can forget about Il2 here. I mean I have the bullet paths, groupings and harmonization all modeled in colour 3D and I still wouldn't go that far as guaranteeing something being better.

It all depends on what kind of a pilot you are, and how you like to fly. If you want action and dogfights for fun, you are probably better off with the current setup. Because you can engage anyone, anytime and anywhere. Me personally, I have no problem flying for an hour with a wingman to get to an altitude advantage, and with the sun on my back pick a target below, unaware of my presence and try a sneak attack. If he spots me, brake off, head home and fight another day. Also I don't mind practicing for days and weeks with my 150 yard setup before going online. But I don't know how fun that would be for others that don't have the time or interest and just want to shoot things down. I also wouldn't call this close range convergence theory rock solid. It just might turn out too damn hard to get into those kind of close ranges and keep at it. And not have the ability to shot at 200, 300 yards+.

This is absolutely correct. Once you get to these very close convergences you loose the buffer zone. You can no longer engage the enemy +/- 150 yards from the setup. The spread patterns make it a waste of ammo. You engage at 100 yards -/+ 20%, or not at all. I don't know how many sim pilots have the discipline to do that. It is definitely a setup for a pilot who knows what he's doing. No room for error.

Yes I would say that would account for most of the problems, in the simulation. It was hard in real life to keep your finger off the trigger and get closer, when the enemy is clearly in your sights. And if you missed or he spotted you before you got close enough, Germans had far more experience and it was likely you would get killed. That's why most of the pilots shot as early as possible and they were trained to do so. It is also why the harmonization patterns are not a single dot. And it was like that for a reason. In a simulation there is no fear or real risk so you can play arround to get as close as you want, far too close for the given convergence. But that's not a problem of DCS. It's a combat simulator and if real combat was done at 1000 feet then that's the way it should be. BUT, as I've heard from many people reading the historical accounts of dogfights and pilot aces, every one of them had their convergence set up at much closer ranges. 100-150 yards. Because once they got experienced they could do much more crucial and concentrated damage with closer convergences. The thing is, in a simulation you can get ALL the practice you want, and basically everyone of us will become an 'ace' eventually. No matter how many times you get shot down. So we'll all outgrow the safe 1000 feet convergence eventually. So a much closer convergence would make sense in a sim world. Now I'm no mathematician, but I can get the numbers needed to copy the same logical pattern seen at 300 yards, to let's say 200,150 and 100 yards. But I'm not gonna put out something that might turn out to be crap. So when I do this (really busy now, but I'll definitely have it before the FW190 comes into play) I'll send the data and the pics/graphs to Yo-Yo and he can decide if its 'DCS worthy' :book: Until then, don't play aces:pilotfly:

Oculus rift + this: and you're all set!

Yep, it actually feels like you're really flying

DCS: Crack cocaine It will ruin everything else for you, but it feels goooood :)

Then we will see who knows how to shoot things down. Taking off, getting in a dogfight or two and returning to base and landing safely. All the while watching for CEM, structural failures and all the things that a study sim brings into the mix.

What are you, dream police ? :D Let us dream :)

Goodbye social life.

Things are coming together! We have a candidate programmer for an interface program with a teensy board. The plan is to cut the costs of production down to a limit where it requires compromise in quality and/or realistic looks. Than I stop. That limit is somewhere around 30$ per single pointer gauge. Less when I leave out some of the details like illumination, for those with a really tight budget. Not every gauge needs to have its own interface, a limitless stepper motor or be injection molded or CNC'ed out of metal. Its all overkill in my mind, and keeps it out of reach for the most of us, in the price department. Of course, high end stuff 100% mil spec replicas can be designed too, if there is enough interest. Right now I want to make something for everyone to be able to enjoy. And this is not limited to DCS or combat sims only. ADI has 11 motors driving 11 different moving parts, so its like 11 gauges in one. Similar for HSI. The most important thing, it can be done. My rough estimate would be 80-120$ per a complex instrument like that. I have also teamed up with a member who has a 3D printer, and we are working hard and plenty on making all (and I mean all) the sticks, throttles and collectives (including force trim and ffb) in the current and coming DCS lineup. The idea being that for the price of the warthog, one could buy a base and 5-6 different sets of sticks and throttles. The machine should also help a lot with instrument building. Due to these recent developments I might have to hold off making this panel, for a few weeks at least. All in the goal of making it even simpler, cheaper, better. Plenty of big surprises coming in the next weeks and months!

I knew something was up when I didn't recognize the camo pattern :D

No flickr account, and I can no longer see the pics. So I guess the problem is on their side...

I see them fine. Nice centerpiece with the camo keyboard :)

@danilop - yes, that's the main motivator behind doing Russian gauges, being there are no available ones out there. The price, in my mind, 50-60$ per gauge is acceptable, 30-40$ would be reasonable, and sub 30$ gauges are my goal, even less with DIY kits. It is within reach now. Mig-21bis is the aircraft of my country's air force .I do believe there are still a couple left in flying condition after 2 collided and crashed last year (Yep, you can't touch us). So I'll definitely include gauges for it. @agrasyuk - yes, many challenges. But being that I'm out of work, for at leas another 10 months, I have 24h a day to tackle them. Which is basically what I'm doing now :) I'm on this probably 16 hour a day on average, plus 1 or two times a week I'll go for 25-30 hours straight when I'm onto something and can't stop till I finish.

I'm no electrician but, I could just take the servos power wires, connect them in parallel to a 5V 6A power adapter? 1 servo is 5V and peaks at 500mA. @Gadroc - I tried finding free quality templates for CAD gears but to no avail. Even standard solidworks gears do not fit nicely. These template programs are quite pricey, 1,200$. So I found a local company that cuts gears and can laser cut them for me, I just give them the ratio and number of teeth. @Veljko . yes, I have designs for dual needles on one axis, and 3 needles on one axis, like the accelerometer. And much more, like the mechanical counter wheels on the altimeter, HSI an alike...

@Gadroc I plan on using the cheapest analog rc servos with 180° of travel, and opencockpit cards control them with 10 bit accuracy, so 1024 positions over 180° travel. Geared up 2:1 it comes down to 0.7°, or 0.25 mm or 0.0098 inches in needle movement which I think is acceptable. Accuracy would be a problem if they were under great load. I think they'll be just fine for this application. But if tests prove me wrong, I'll replace them with better ones. All the gears will be custom cut with laser. Steppers are better, but if these servos will do, then they're a bit overkill, needing more complex and expensive drivers. Opencockpit cards can take 5V i think, I'll have to re-check this. But it would be no problem at all to serial connect the servos to a 5V adapter. I've looked at the pololu boards, they're an option too. Although teensy seems to be the cheapest option since it can control up to 24 servos (and more as I'm told), or 4 steppers at 24$ per board. I had 2 members contact me about programming between DCS/LUA and teensy/arduino boards so here's hoping they come up with something.

Looking into Ka-50 gauges, ADI and HSI especially. Shouldn't be a problem. Are the military specifications for Russian instruments available?

You magnificent bastards! You plan on adding motion platform under that room while you're at it?

It's about 2$ per holder, 3d printed in black. So it might be worth it.

The angles you see on "fake" light holders are achieved by sanding down the acrylic. I'll make both fake and functional versions of light holders. 3d printing is an option if laser cutting won't do.

Well thanks for bursting my bubble :) Poor research on my part I guess. As I read on their webpage, you can add up to 24 servos to it. Is that 24 servos that can be controlled individually? Yep, agrasyuk that's right. Middleware that will send DCS export data to the teensy board. Servos and steppers mainly. Next are LEDs if possible. Because if it can control up to 24 individual servo motors, that's 90% cheaper per servo than opencockpits or other servo controllers.

Isn't teensy an HID/yostick off the shelf? Like the LeoBodnar boards? just connect your buttons/potentiometers/hall sensors and your ready to go?

I guess that's what I mean by "no direct way". Opencockpit cards have SIOC as a way to talk with the sim. Nothing yet for the teensy boards.