**1.09 BETA IS OUT**

Post by **b4thman** » Sun Sep 07, 2008 7:01 am

- Asus 8800 ULtra + 177.83 forceware drivers
- Windows XP + SP3
- Gigabyte GA-M56S-S3 + 2Gb ram + AMD athlon 64 X2 5600+
- X-Fi Xtreme music
- 2x Viewsonic VA-1716s (lcd 17" monitors tested with planar) OR Zalman trimon 22" 3d monitor (tested with interleaved horizontal).

Good luck.

Post by **BlackShark** » Sun Sep 07, 2008 7:11 am

Hi,
I've tried to 3D mark the driver in anaglyph mode to see the performance improvements, unfortunately 3Dmark06 crahes during the Deep freeze scene so i couldn't get any results results. I also noticed that the "Return to Proxycon" demo is much too bright when the driver is enabled (even in 2D mode), this happens for both 1.08 and 1.09beta.

The only other game i have with a benchmark is Devil May Cry 4, i think there is also a benchmark feature in UT3 but i don't know where.
Unfortunately i was about to publish the scores when i realized that my 1.08 results were made with the demo instead of the full game so i have to reinstall 1.08 to test.
Between 1.08 DMC4-demo and 1.09b DMC4-retail, i noticed a 3-5 fps increase on an average of 30 to 50 fps in 3D mode (about 10% increase) but i don't know if it's all the work of iz3d or if some part of it is because of last minute optimisations for the game and / or drivers...

will publish the whole updated bench this evening.

I will also make a detailled report on how great DMC4 looks and the few little bugs I've found.

I also have a little report on FRAPS !
It still doesn't work for me, but since 1.09b i can actually see fraps' FPS meter appearing about half a second when a game launches, and then Iz3d kicks in (iz3d's FPS meter), and fraps disappears.
Annoying but not an emergency, fix shutter support and you'll see about fraps later...

Post by **BlackQ** » Sun Sep 07, 2008 4:20 pm

Hi, guys!

First of all thank you for congratulations!! It is big pleasure for me

I also want to thank you for so active help of debugging - I'm running between forum collecting bugs now

this weekend will help us a lot.

We moved a lot of profile edition operations to control center to make users' life easier (I can not edit XML

) and it created more link inside the code - it is easy to correct them - and this is a good news.

Additionally I just want to confirm that shutter trial will be longest (60 days instead of common 30 days) and I hope we all together can make it good

OK - let's me run not to make a final list of bug for tomorrow - oops! for today already

Post by **winch8** » Sun Sep 07, 2008 4:30 pm

black q there is not a maximum separation option with this realease?

Post by **quadrophoeniX** » Sun Sep 07, 2008 4:50 pm

VadersApp wrote:
ignatius wrote:Hey VadersApp, these are the shutter glasses I have, http://www.ray3d.com/graphics/arcade_glasses.jpg
These glasses look like the IR revelator. What i don´t understand is, are revelators working in interlaced mode? I

from the size of the active LCD screen they rather lok like the H3D version (same as Metabyte Wiccked 3d's).The LCDs almost look like rotated by 90° compared to the revelators - the adapter included with the H3ds or Metabytes could do page flip as well as interlaced (color coded). If Ignatius adaptor is rather BIG that's very likely...
just pulled out my H3Ds .-.. it's them - if the picture is right, that is. Ignatius may confirm:The icon above the right eye shows a large H with "3D" in a globe...

Post by **The_Doctor** » Sun Sep 07, 2008 5:37 pm

BlackQ wrote: Additionally I just want to confirm that shutter trial will be longest (60 days instead of common 30 days) and I hope we all together can make it good

It would be good to keep it free until it works good so that people can keep using it and provide feedback. From all the outputs this will be the hardest to figure out.

Post by **chrisjarram** » Mon Sep 08, 2008 3:58 am

BlackQ wrote:Hi, guys!
Additionally I just want to confirm that shutter trial will be longest (60 days instead of common 30 days) and I hope we all together can make it good

Gulp, I really, really hope that is not implying we could be waiting another 2 months for a shutter solution after the amount of time we've all been chomping at the bit!

Post by **chrisjarram** » Mon Sep 08, 2008 4:00 am

The_Doctor wrote:Race07 & GTR Evolution - no anaglyph or interlaced. Nothing happens, works in anaglyph with 1.08

GTR Evolution works fine for me with 1.09 in anaglyph. Triple screen, 1280*1024 per head.

Post by **suchy** » Mon Sep 08, 2008 4:04 am

It didn't work for me at all in iZ3D mode. Only one panel was utilized, I was seeing the game screen only with right eye. Desktop on the other panel. When I turned on 3D, double image was rendered on a single panel - still desktop on the other (FEAR) or crashed (Armed Assault).
Did anyone else had such problem? So far I got back to 1.08.
My specs are:
WinXP SP2, DirectX 9.0c, 3GB RAM
AMD 64 X2 6000+
NVIDIA 8800GTX 768MB
iZ3D Monitor 22'

Anaglyph seemed all right.

Post by **chrisjarram** » Mon Sep 08, 2008 4:06 am

BlackShark wrote:Hi,
I've tried to 3D mark the driver in anaglyph mode to see the performance improvements, unfortunately 3Dmark06 crahes during the Deep freeze scene so i couldn't get any results results. I also noticed that the "Return to Proxycon" demo is much too bright when the driver is enabled (even in 2D mode), this happens for both 1.08 and 1.09beta.

The only other game i have with a benchmark is Devil May Cry 4, i think there is also a benchmark feature in UT3 but i don't know where.
Unfortunately i was about to publish the scores when i realized that my 1.08 results were made with the demo instead of the full game so i have to reinstall 1.08 to test.
Between 1.08 DMC4-demo and 1.09b DMC4-retail, i noticed a 3-5 fps increase on an average of 30 to 50 fps in 3D mode (about 10% increase) but i don't know if it's all the work of iz3d or if some part of it is because of last minute optimisations for the game and / or drivers...

will publish the whole updated bench this evening.

I will also make a detailled report on how great DMC4 looks and the few little bugs I've found.

I also have a little report on FRAPS !
It still doesn't work for me, but since 1.09b i can actually see fraps' FPS meter appearing about half a second when a game launches, and then Iz3d kicks in (iz3d's FPS meter), and fraps disappears.
Annoying but not an emergency, fix shutter support and you'll see about fraps later...

3dMark06 in anaglyph for me turns into a slideshow... From 90fps+ to more like 15-20. As does COD4, from 200fps to _7_ fps.. this can't be right.
On the flipside, rFactor, GTR Evolution, Bioshock and Timeshift all seem to work v well in anaglyph performance wise (though this is a bit bloody useless for me until the damn shutter mode is working the way it is supposed to!

)

Post by **winch8** » Mon Sep 08, 2008 5:33 am

about the maximum separation issue i will let you see how i am obliged to play far cry to get a good 3d stereo effect...
http://www.mtbs3d.com/gallery/displayim ... ?pos=-1691

Post by **winch8** » Mon Sep 08, 2008 5:35 am

with a maximum separation option like nvidia drivers' one, the stereo images of the rifle would'nt e bo separated, but much more near!

Post by **The_Doctor** » Mon Sep 08, 2008 6:53 am

It seems that some people can get something to work and some can't, not only gtr evolution. The results are a little inconsistent. Can you start rfactor at all (if you have it)?

chrisjarram wrote:
The_Doctor wrote:Race07 & GTR Evolution - no anaglyph or interlaced. Nothing happens, works in anaglyph with 1.08
GTR Evolution works fine for me with 1.09 in anaglyph. Triple screen, 1280*1024 per head.

Post by **ignatius** » Mon Sep 08, 2008 8:14 am

Hey quardropheoniX, I can confirm that the glasses I have are most definitely the ELSA 3D Revelator shutter glasses. Here's a good link about them: http://www.stereo3d.com/revelator.htm It also shows the IR emitter adapter here, http://www.stereo3d.com/revelator_cables.jpg that i have. It's a very tiny little adapter about 1inch in length that i place on top of my tv. Oddly enough, my glasses don't say ELSA above the right-eye like it does on that website. It's simply blank. But the glasses and the adapter are identical to the ELSA shutter glasses.

So far, I've only been using the 3D DLP format with the iz3D drivers and it seems to be working great for most of my games. But i will try out the other formats like interlaced/page flip and see what I get.

By the way, I've been doing a bit of gaming as much as I can to test out these drivers. And for the most part I haven't had much trouble except for a few things. When playing Half-Life 2, it works brilliantly for the most part but, every time i get near water the game becomes instantly unstable and lags like crazy. The only way I could partially solve this issue was by turning down the "Texture" Detail to Low or Medium and that seemed to smooth things out a bit but, there was still a bit of noticeable lag. I should note that I was able to turn every other video setting to MAX (including AA + Shadows) without any problems in 3D. The only time i had a problem was when I got near water in the game and I'd have to simply turn down the "texture detail" or turn off the 3D stereo completely in order to play that part of the game properly, which kind of sucks.

I was also wondering if there were any graphic settings I should turn off for Gears of War in 3D? For some odd reason, after a bit of gameplay the game would randomly crash. I'm guessing there's rather a problem with the iz3D drivers, a setting I'm not supposed to enable or the game itself needs to be re-installed.

Anyway, hope to hear back from you guys, I'm off for now and if anyone has any questions about my setup, let me know. I'd always try to help fellow 3D enthusiasts as much as I can

Cheers,
Varinder

Post by **Neil** » Mon Sep 08, 2008 9:08 am

For Gears of War, make sure you download the latest version of Windows Live client (I think that's what it is called). It will solve these problems. It's mentioned in the Gears of War support section.

Regards,
Neil

Post by **ignatius** » Mon Sep 08, 2008 9:26 am

Cool, thanks for the help Neil! I'll get that right now.

edit - I would just like to confirm that what Neil suggested did fix the Gears of War trouble i was having. I couldn't find the support thread you mentioned for Gears of War but I did find another very helpful thread which is located here: http://www.gamespot.com/pages/forums/sh ... d=26422869 The last post is the most helpful and explains everything you need to fix this Windows Live problem and has the appropriate links to get the right files.

By the way, Neil or anyone else who can help. I'm still having trouble getting Crysis to work in 3D. I've opened up Crysis in DX9 but enabling 3D stereo doesn't do anything. Do you have any suggestions?

-Varinder

Post by **chrisjarram** » Mon Sep 08, 2008 11:08 am

BlackQ wrote:Hi, guys!

First of all thank you for congratulations!! It is big pleasure for me

I also want to thank you for so active help of debugging - I'm running between forum collecting bugs now this weekend will help us a lot.

We moved a lot of profile edition operations to control center to make users' life easier (I can not edit XML ) and it created more link inside the code - it is easy to correct them - and this is a good news.

Additionally I just want to confirm that shutter trial will be longest (60 days instead of common 30 days) and I hope we all together can make it good

OK - let's me run not to make a final list of bug for tomorrow - oops! for today already

Hey BlackQ,

Have you been following the 1.09 shutter thread here out of interest? I noticed you hadnt commented on this yet but its certainly the hottest topic at over 1500 views in the last 3 days! I think between us here we'e ascertained the marker format is not going to work (either WLC, BLC or your own) as this could frequently result in flicker unless every users machine is capable of running every game at a fast rate.

The only solution therefore is to implement the low level calls nVidia are using to be able to pageflip 2 separate back-buffers, and then populate them separeately with the rendering train. This should be pretty easy, if not a case of getting this info out of nVidia there are tools for monitoring device driver execution and low level API calls / interrupt routines which should help you do this. I think it was assumed this had been done by iz3d already as it seems a little pointless having a simple shutter mode without it(?), but I guess this was overlooked.

Could you at all give an idea of when you going to be looking at fixing this? It would put a lot of peoples minds at rest!

Thanks for all the great work, hopefully it won't take long to get all this sorted...

Post by **BlackQ** » Mon Sep 08, 2008 12:47 pm

Hi, guys!

I'll report about debugging and progress in 1.09 thread of iZ3D forum in few minutes - let me just combine everything into one

Post by **ceashure** » Tue Sep 09, 2008 9:22 pm

The beta driver doesnt appear to work at all for the iZ3D Monitor. I can see the desktop image through the game and non of the toggles work. I also tried with default set to enabled, with no luck.

I've got Vista 64-bit and a Nvidia 7950GT. Was trying a whole buntch of HL2 based games, and I did try a restart, clicked the "setup screens" button several times, and made sure dualview was on in Nvidia control panel.

I have noticed that the Contrast settings are MUCH better though...

Post by **The_Doctor** » Tue Sep 09, 2008 10:00 pm

Sometimes when changing output type (for example Anaglyph to Interlaced) it doesn't work at all after the change. After switiching some things on and off, clicking apply several times it works again.

Some keys don't work for key bindings (for example CTRL + ~).

Post by **BlackQ** » Wed Sep 10, 2008 3:40 pm

iZ3D 1.09 beta 2 will be released in few minutes - 99% bugs are corrected (we hope) - sorry guys for short comment - two sleepless night and 5C no sun in Moscow ...

good time for release

CU all! Have a good test and last one for 1.09 I hope

Post by **yuriythebest** » Wed Sep 10, 2008 3:52 pm

cool waiting

wow izCold in russia atm- here in ukraine atm it's a nice 13 deg

Post by **chrisjarram** » Thu Sep 11, 2008 4:59 pm

BlackQ wrote:iZ3D 1.09 beta 2 will be released in few minutes - 99% bugs are corrected (we hope) - sorry guys for short comment - two sleepless night and 5C no sun in Moscow ... good time for release CU all! Have a good test and last one for 1.09 I hope

99% of bugs? That include the shutter mode then?? Its definately the biggest bug there after all!!

I know you said you had someone at iz3d with the experience to investigate that at the end of this week - have you made any headway now? Really hope you guys have managed to fix this or if not will do in the next few days - I reckon by the end of the weekend the mtbs thread regarding shutter support will have reached over 4000 hits!!

Post by **BlackQ** » Sat Sep 13, 2008 7:29 am

Hi, chrisjarram!

We met the guy (he is helping us with consoles) and he confirmed that low level of GPU is not accessible for us to send any sync signal...

we continue to think what else we can use as solution of this problem. We knew it can be like that, but we decide to release shutter mode also, even originally we stated that problem may come - this is not a bug, this is something we need to improve to reach compatibility - We'd happy to solve this problem alone, but we may need to have help from shutter glasses manufacturers.

Post by **yuriythebest** » Sat Sep 13, 2008 7:39 am

BlackQ wrote:Hi, chrisjarram!

We met the guy (he is helping us with consoles) and he confirmed that low level of GPU is not accessible for us to send any sync signal... we continue to think what else we can use as solution of this problem. We knew it can be like that, but we decide to release shutter mode also, even originally we stated that problem may come - this is not a bug, this is something we need to improve to reach compatibility - We'd happy to solve this problem alone, but we may need to have help from shutter glasses manufacturers.

have you tried contacting eDimensional? their (old) driver does just that so perhaps you guys could work out a deal or something like "19% off for (new) buyers of eDim shutterglasses" in return for them providing their driver source code to you, or could be the other way round- eDim could bundle your driver with their glasses (not as free software but with protection like a serial number that will only work with a single pc ID), and in return you get a cut from their sales. This is of course wild brainstorming, just my 2 cents.

Post by **chrisjarram** » Mon Sep 15, 2008 3:05 am

BlackQ wrote:Hi, chrisjarram!

We met the guy (he is helping us with consoles) and he confirmed that low level of GPU is not accessible for us to send any sync signal... we continue to think what else we can use as solution of this problem. We knew it can be like that, but we decide to release shutter mode also, even originally we stated that problem may come - this is not a bug, this is something we need to improve to reach compatibility - We'd happy to solve this problem alone, but we may need to have help from shutter glasses manufacturers.

Why do I get the feeling we're going backwards here?

It is not the sync signal that is the issue, it is access to the page-flipping buffers. Sync can be generated via a simple square wave converted from the V-sync output of any card - there are extrenal modules out there which convert a vertical sync output to the standard 3-pin DDC signal required by a lot of shutter glasses.

There is no help any shutter glasses manufacturer can give you to solve this problem, you need to access the page-flipping capability of the graphics card to guarantee a LRLRLR output - thats your side of things. This is certainly a bug as it renders the feature useless without it.

Post by **chrisjarram** » Mon Sep 15, 2008 3:10 am

yuriythebest1 wrote:
BlackQ wrote:Hi, chrisjarram!

We met the guy (he is helping us with consoles) and he confirmed that low level of GPU is not accessible for us to send any sync signal... we continue to think what else we can use as solution of this problem. We knew it can be like that, but we decide to release shutter mode also, even originally we stated that problem may come - this is not a bug, this is something we need to improve to reach compatibility - We'd happy to solve this problem alone, but we may need to have help from shutter glasses manufacturers.
have you tried contacting eDimensional? their (old) driver does just that so perhaps you guys could work out a deal or something like "19% off for (new) buyers of eDim shutterglasses" in return for them providing their driver source code to you, or could be the other way round- eDim could bundle your driver with their glasses (not as free software but with protection like a serial number that will only work with a single pc ID), and in return you get a cut from their sales. This is of course wild brainstorming, just my 2 cents.

You know what, this is actually a good idea. Personally I've had an 8800 working on a CRT with 8800 Ultra running PRAY, with no problems whatsoever, so I know for a fact eDimensional have page-flipping working with their own driver. Certaintly if they can do it, iz3d can.

Post by **BlackQ** » Mon Sep 15, 2008 7:58 am

Hi, chrisjarram!

it sounds good to have an "access to the page-flipping buffers" - but no one knows how to interrupt single thread process... we did not find such expertise worldwide yet.

Post by **Hugo** » Mon Sep 15, 2008 10:08 pm

Perhaps a stupid idea, but what´s about a ISR for the GPU IRQ?

Post by **BlackQ** » Tue Sep 16, 2008 9:16 am

nop - only GPU manufacturers can try this way ...

Post by **TheLostBrain** » Tue Sep 16, 2008 3:54 pm

Matching to Vsync... just some hypothetical 1000ft level thoughts...

Use a combination of the Directx function :: WaitForVerticalBlank and also a simple timer for tracking to provide a concerted effort of keeping things in sync..

So... hypothetically...

You'd start with an initialization process where you'd match a timer (running on a separate thread) to the interval returned by a couple polls of the WaitForVerticalBlanks.

You'd then spit out the L or R on every tick of that timer. (L for even ticks and R for odd ticks for instance)

Every once and awhile you'd re-poll the WaitForVerticalBlanks and rematch the timer interval as needed.

This would also match Left eye renderings to even ticks and Right eye rendering to odd ticks, etc to resolve reversal issues you wind up when it comes time to recover from frame rate drops, etc.

At the very least... just the ability to recover shortly after a frame rate drop would be awesome... and I don't know but maybe you can make something work with that even/odd tick check for LRLRLR synchronization.. Just some random thoughts.

Here's a code example of someone doing something similar to what I'm describing...although much less low level, etc.
http://www.codeproject.com/KB/GDI/teari ... ect=752625

Post by **BlackQ** » Wed Sep 17, 2008 11:57 am

Hi, TheLostBrain!

Thank you for your recommendation - we tried similar way today with no success - render was blocked by second thread... but let us try your way too

Post by **KindDragon** » Wed Sep 17, 2008 12:06 pm

TheLostBrain wrote:Matching to Vsync... just some hypothetical 1000ft level thoughts...

Use a combination of the Directx function :: WaitForVerticalBlank and also a simple timer for tracking to provide a concerted effort of keeping things in sync..

So... hypothetically...

You'd start with an initialization process where you'd match a timer (running on a separate thread) to the interval returned by a couple polls of the WaitForVerticalBlanks.

You'd then spit out the L or R on every tick of that timer. (L for even ticks and R for odd ticks for instance)

Every once and awhile you'd re-poll the WaitForVerticalBlanks and rematch the timer interval as needed.

This would also match Left eye renderings to even ticks and Right eye rendering to odd ticks, etc to resolve reversal issues you wind up when it comes time to recover from frame rate drops, etc.

At the very least... just the ability to recover shortly after a frame rate drop would be awesome... and I don't know but maybe you can make something work with that even/odd tick check for LRLRLR synchronization.. Just some random thoughts.

Here's a code example of someone doing something similar to what I'm describing...although much less low level, etc.
http://www.codeproject.com/KB/GDI/teari ... ect=752625

DirectDraw method WaitForVerticalBlank() not available in D3D9.

1.09 BETA IS OUT

Matching Vsynch and recovering from frame rate drops..ideas.

Re: Matching Vsynch and recovering from frame rate drops..id