Is it just me, or I see going from V2 to V3 was like going from 8K X to 8K? Maybe Pimax forgot to put “X” in prototype naming, or wanted to mislead backers? You choose option. V2 was public demo, V3 private(closed) demo. I just relay info, I am only reader/messenger. Don’t kill the messenger xD
My motivation for using brain for this compilation, interest in advances in VR(potential gen 2 HMDs).
My sources: data collection post(reason for unofficial in title).
Where did you get those specs for PV3 from?
.
I read this as 2160x3840(DSC 2:1). Because we know about BW limits of 2xHDMI 1.4 . Can send this data with 4:2:0 color compression to HMD. Original solution was 2xHDMI btw, so this is from that point in time. Later they write 2x1440p, when they found out DSC impossible on input on anx7530. At least 2:1, 3:1 is possible with downscaled in to anx7530 then encoded by engine on board to feed the panel IC “DSC” signal, half res input to DSC algorithm before encoding(V2 vs V3)… Big deal for sure.
Very nice compilation.
Will be very entertaining to see how much Pimax will be able to deliver of those specs.
I have also bad news for 4K owners, real spec of input(per HMD), not marketing number.
Maybe let them guess from this sheet, I am a bit tired 
.
I will give them hints DSC 2:1(1xMIPI) , DSC 3:1(2xMIPI), let’s see if anyone can give me answer. What is sent from GPU through HDMI 1.4 cable(real res, and color data). BTW data about DSC was marked as *Confidential*, why? It’s just a tech Pimax, why hide, we know “in principle” how it works, you don’t wanna say if it works? Like with Brainwarp? Let’s use logic then, is DSC used on P4K_HMD and Brainwarp on P8K_V1-3? Challenging? (I like puzzles . )
suggested title change
Compiled specs sheet [“8K” V2 vs “8K” V3 vs “Vive” vs “8K” X CV vs “8K” CV] PULLED OUT OF MY ASS
@destraudo
I accept there are people who would rather trust company with name P-IMAX, with proven record of marketing HMD as 4K, when input was 1080p(vertical), than person who gathered all available data, solved some puzzles, and presents the answers (real tech specs, not specs as marketing). And has HW/SW background, and has shown good analytical skills. You pick between my specs, and marketing specs, but don’t force other how to pick. Since Pimax will not give real specs, as didn’t with 4K HMD (to this day, how long from launch?) .
They could hide from some people, for some time, not from all people, the whole time.
Since I am mildly annoyed, by fanboy-ism, I will write analogy for non VR tech hard core people to understand. Analogy, from the console world:
Pimax 4K v1 (1080v res), that is like using XBOX 360 on QHD TV.
Pimax 4K v1.2 (1440v res), that is like using PS3 on QHD TV.
Pimax 8K v1-2 (1080v(DSC)) is like using PS4 Pro on 4K TV. (upscaling most games, ~30fps)
Pimax 8K V3(1080v downsampled) is like using XBox One S on 4K TV. (900p to 4k upscaled)
Pimax 8K CV, is like PS4 pro but with checkerboard rendering(4K CKRB), good enough to wait for PS5.
Pimax 8K X CV, is like XBox One X on 4K TV, some games run native, some upscaled.
And for fans of gen 1 tech, here it goes:
Rift CV1 is like XBox One - with Full HD TV.
Vive CV is like PS4 on 21:9 Full HD monitor, a bit wider FOV(to Rift).
Some compare Pimax 8K to gen 2, gen 2 will be, PS5, XBone 2X 
4K gaming on 4K TV, 60fps expected and HDR.
You can also compare HDR to SDR LCD(pimax) vs OLED(Rift/Vive/PS VR).
How can I dare to make analogy, who am I, to do that? I can do what I want, you can read what you want. Comment how you want. Some might find it interesting(confirmed by some backers). So I continue… I am on backers side, not company side. I will use logic+available data+assumptions, you can use personal attacks as your weapon of choice to defend company.
Edit: I won’t waste another post on explaining how you attack me instead of present new source of better data. That is not marketing(check marketed in my sheet).
I can do thought experiment, which I did, and from that I debunked whole 1080p to 1440p update via firmware. And scaler is part of DSC(decoder) engine, no separate HW scaler IC on P4K P_any_K …
But it is not about trusting pimax. It is about not trusting someone who is talking out of their ass. are presenting a data chart, compiled as unofficial, from data that you posed in another thread as being complete speculation.
speculation about how brainwarp works
'Brainwarp theory:
Input(60FPS/Hz), simulated 90Hz/eye, “180Hz Brainwarp” ’
speculation that pimax never intended to release a 90 hz headset
‘‘I am having HMD with 80Hz refresh rate. I don’t want to market it as 80Hz, because standard is 90Hz for 6DOF VR in 2017. How I hide that I can’t run HMD at real 90Hz that gives better MTP, I put this idea about using Brainwarp, and numbers like 75Hz(this per HMD possible), then put 90Hz and after that “per eye”, then explain what “per eye” means, by referring to Brainwarp and put even higher numbers like 150/180’’
You literally admit that your entire spiel is based on speculation on top of speculation .
‘‘I have really great conspiracy theory, this will be fun, maybe not true, so I say now “This is only for entertainment purpose conspiracy theory lovers - enjoy”
I spend some time bashing my head, why use 2xHDMI on demo, when plan is 1xDP. So here is fun theory:
Laptop output DP 1.4 with DSC, DP to 2xHDMI adapter used, HMD accept 1440p@90Hz(DSC), per eye signal, give it to Driver IC, and here we go, we have answers about 90Hz issues with 1xDP input.’’
And then you present your speculations in a chart and talk about how you have good analytical skills while listing your sources for a data chart as your brain and the pimax 4k.
And then you have the comedy genius to ask people to present a ‘more reliable source’ for information that is not available, and which you have speculated wildly on?
I didnt care when you were posting this stuff as speculations and theories, but presenting it as anything more than that is ridiculous.
That’s a pretty long formula for it to equal ‘conjecture’.
I appreciate you diving into this as much as the next guy and it most definitely makes for an interesting read, but please make sure you clearly label those data points you have no reliable source for as just that. If that means you have no source at all, then either use your time more prudently or mark the entire post as ‘fiction/conspiracy’.
bingo. Unofficial makes it sound like hard data coming from a source outside pimax. Really its a pile of assumptions and speculation layered onto 20 percent actual numbers.
So why are you here again?
To prevent again debacle from 4K with 1080p in upscaled launch. They can at least delay 8K until solve DSC on input (DP 1.4), as was marketed from start of Kickstarter. I know they do false market a lot, but I hope if someone present this to them, and alternative solution. they may accept it. This is not their money, this is people’s money 4M+. Better to give them products that is compelling with better use case than P_4K(NSFW). Solution is into making, that is why 8K X is delayed. Give options to backers, wait or get this “Pimax 4K UltraWide” marketed as Pimax 8K product. I would wait for that, you? Analogix is making new bridge chip, that will allow DSC (DP 1.4) on DP input(current is DP 1.2), I say wait for that to give to backers what is shown in V2. Understand now my motivation to fight every personal insult people may throw at me for not having official specs, that is not available? Is it my fault because Pimax only give to public marketing(“specs”)… That is their business practice, so far. I fight them with same weapon. I give unofficial data, but that is closer(much closer) to real stuff. As is ~30ms MTP to 15ms they are marketing.
Why would we use your speculations to guess on the 4k as it has been reviewed by Proffessionals?
Any more than we would use your speculations as real on unknown hardware componets? Only real details we have is the bridge chip being used on 5k & 8k. No info in the un demoed 8k-X.
Why not tell us the lastest problems with the Ps6? 
Wish we could stop feeding the trolls…
It was interesting at first but now he’s just finding new ways of repeating the same nonsense he’s said before.
Wish we could have some moderator action over this. It’s time to stop this user in this forum.
Ok, I see I can’t gain traction, no people with enough tech knowledge to back me up. And marketing team and fanboys are majority here, or backers just want to blindly trust on claims vs what is compiled as solutions in prototypes/4K, actual implementation assumptions. If you wanted to use Kickstarter knowing that many projects fail to deliver promise, so can you. And will sacrifice Pimax brand, to be honest is not doing very well now, <1% market share, P-imax, 4K debacle with 1080p in signal etc. I know you have backup brands, such as “Aukey Cortex 4K”, and maybe don’t care about users care only about business deals you make from this “hype train”, based on idea how you can deliver “8K” VR first to market, when we know tech giants with thousands of engineers/researchers can’t, cause they wouldn’t risk hiatus if input is 2560x1440/eye(then done upscaling on horizontal axis, downscaling on vertical) - on PC, and marketing it as “8K VR” or cutting edge, or w/e. This is my last attempt to wake Pimax up. First time was to send risk analysis about 80Hz issue to all backers(via KS Update), not put on forum/hide on bottom of page that don’t trigger email notification. Now I read marketed is changed to 75 to 90 but per eye. Even fake marketing specs is getting brainwarp treatment just to keep real info hidden from public. GOOD JOB. I found DSC 1.1 used in P4_K, with chroma subsampling 4:2:0 done on GPU, + downscaling to feed the Panel IC “FAKE DSC”, so it can decode it(upscale if you want) to 4K res. Same thing with V2, but use GPU DSC encoder, so this is “REAL DSC 2:1” almost losseless algorithm, comparing to “junk” downscaling with 4K HMD and V3. Since anx7530 don’t support fake nor real DSC, cause no 4:2:0 color format supported. This thread is unofficial cause Pimax is using “confidential” mark on some of this data, and claim I made it up, when it has ground on tech basis. Their specs has ground on what? Wishes, desires to warp brain that 8K is real 8K, that there is separete upscaler chip on HMD, that MTP is 15ms with BW when Vive with OLED(and real 90Hz) has 21ms. Ok. Just leave this for people who would want another source of “speculation”. Before we see reviews of final product. I rest my case.
@jorgenRe http://community.openmr.ai/t/compiled-specs-sheet-8k-v2-vs-8k-v3-vs-vive-vs-8k-x-cv-vs-8k-cv-unofficial/4641/3?u=specsreader .
You need to keep your sources available on all threads you start. Otherwise they are just as they state: “Speculation” seeing as lots of what you say has not been confirmed. Atleast kudos for saying they are unoffical…
EDIT: By sources i mean actual links to where you got the info not where you had the connected dot soloution eureka.
Yes, everyone who disagrees is a fanboy or on the marketing team. I recall you accusing someone else of being sponsored by pimax in another thread.
‘You are sponsored by Pimax, 100%’
Maybe instead of being pimax agents and fanboys, maybe some people just don’t find your case that compelling. And that with every post, your ability to ‘connect the dots’ on other things, like people being sponsored by pimax makes everything you say less compelling.
When you have a case built on a solid foundation i will be happy to listen. But you do not have that now. You have, assuming x works like this, and assuming they are using same panel , etc etc and on and on. Everything you say may end up being true but at this point its too much ‘connecting the dots’ for me personally to give it any weight.
I as tech guy, like to know how existing tech works. Pimax 8K V1-V3 is existing tech, so not pure speculation(marketing). So you know difference between my specs and Pimax specs(envisioned product specs vs specs derived from existing product in prototype stage). Here is my tech overview of tech used in existing products at V2 and V3 stage as demo footage proves.
Agree or disagree, not for lay men, requires understanding of DSC, panel/driver IC/HW understanding (educational background preferred). Here it comes:
Pimax “8K” prototype stage tech analysis:
Single " 4K panel specs* " : We have CLPL holder that is containing 4 slices of screens, 4K panel is made by taking 4 1920x1080p panels tilted 90degrees, and cut in half on (920h x 1080v x 4) is actually “4K panel”. And each slice, has its own backlight, that through firmware update can be made to produce strobbing effect, that will mask ~4ms GTG pixel response time of single slice(Low persistence effect). This is SW upgrade that Valve is referring to when claims LCD is “good enough” for high end VR now. Now, lets inspect the connections per CLPL holder that contains driver IC.
"Driver IC specs* ": Pimax SW engineer, pointed out some limitation of Driver IC regarding to input on MIPI-DSI ports. If we feed native signal to 4x1K panel(4K from now on), we are limited to 3840x2160p@30 interleaved to 60. So actual refresh rate, when transferring native input res is 30Hz/slice limit, 60Hz/2 slices(if input is 30hz interleaved, not optimal for VR use), or 120Hz/4 slices. We use technique called “interlacing” to achieve double perceived refresh rate. About interlacing check John Carmack remark on interlaced displays. I derive from this that Pimax wants to push this idea to reality. That is how you get brainwarp numbers like 150-180, that is refresh rate not of (whole)display, but rate at which 2 slices(2x1K) b-light is on, so per 2K slice we get 150(75+75)-180(90+90) sync backlight on/off rate, we have 1 MIPI per 2K slice, and that is how Pimax gets confidence to put that number on Brainwarp specs in refresh rate specs. This is confusing to lay people, since they don’t know as much as engineer who wrote 150-180 on some spec sheet(without * next to it). Marketing team is using this number not in user friendly way, I criticize Pimax for that(I am pro-truth not pro-_any_company). To get real benefits from this approach we need to use at least 1MIPI port per 2K slice, at native res(4k) input. For interlacing effect not interleaving effect. And to go from 30(original num) to 60Hz(i*) i* - interlaced pixels mode. This is why Pimax advertise with 4K HMD, real deal 60Hz(up to 90 async), to achieve it they use shutter to sync(to eyes) the 90async/2K slice refresh rate. You get downside, loss of brightness(as reported), comparing to native 60Hz(with 60Hz/whole panel sync b-light). I don’t know specs of the each 1k x 1k slice(panel) but CLPL holder I think uses same Driver IC as with 4K HMD in 8K with updated firmware to get real CLPL(not simulated with shutters).
That is why we get /eye in current marketed tech specs of 8K CV.
About actual input per 2K slice, this is also problematic. Driver IC has specs* to support DSC 1.1 version/FBC - Qualcomm similar tech to (de)compress the signal from MIPI to actual individual 1K panels/slices. To feed in more than 30Hz input(what is limit per 4K panel), by using same number of MIPI per 2K slice, Pimax uses two techniques. In 4K HMD, uses actual Driver IC as upscaler, this is famous “scaler chip” in fact is just IC put in front of the CLPL holder. So they use same tech 2x. BTW designing custom IC(ASIC) is cost effective in 1M+ units, so I don’t criticize Pimax for this move. What I criticize is that they are hiding behind 1.4b bandwidth limits, to use as excuse why Pimax 4K HMD can’t have 90Hz input like Rift/Vive. Real reason is, Driver IC needs DSC signal(to feed the panel higher than 60Hz input). They take HMD signal with 4:2:0(pixel compression), feed it into “DSC (en/de)coder IC/marketed as scaler”. So throught cable they send 2160v x3840h(4:2:0)@60 signal for both eyes to Driver IC used as scaler, because it doesn’t get native 4K, but compressed(on PC side, (up)down-scaled, (up)down-sampled w/e) 2x2560h x 1440v formated for “DSC 3:1 input per 2K slice” (360x3840) (after real DSC treatment) and 1080v x 3840h per eye, after 2:1 encoding. So input per “scaler” is 2x1080v x 3840h with 16b/pixel color(4:2:0). Data flow: 2x2560hx1440v(GPU render target, no SS) - 2x1080v x 2160h(down/upscaling on Pimax post-render pipeline) - 2160v x 1080h RGB6@60 (data through 1xHDMI 1.4 to HMD) - 1080h x 3840v DSC 2:1 (DSC encoder/scaler in Pimax words) - fed to 360x3840(DSC 3:1) input to DSC decoder on CLPL holder/4K panel holder/driver.
How many steps, 1,2,3,4,5 complicated for people who haven’t used algorithms before . I get that, but marketing render target as input res is misleading at best.
With 8K V2, actually used is different (smarter approach), not perform (down)upscaling on PC, but use GPU DSC engine to feed the CLPL holder with DSC 2:1 signal, if use 1 MIPI ports 2K slice, or DSC 3:1 if use DSC 2:1 with 1 MIPI per 2x 2K slices.
How they used actually GPU DSC engine, on V1-V2, vs using “fake DSC process” in P_4K and P_V3.
Here are components of 8K V1-V2:*
1x DP 1.4 to 2xHDMI splitter, unknown maker(performs almost no logic).
2x HDMI to DisplayPort Bridge Controller ANX7678, in specs noted it is designed for use with ANX7530. Specs.
2x ANX7530 . Used to accept out from anx7678(2160v x3840h (DSC 2:1)) and sent to MIPI, actual DSC(real deal, used Algorithm not just math to reduce BW) to each MIPI 1080v x 1920h, 2MIPI per eye, so we get not uspcaled, but decompressed 4K per eye output res. Not native, but decoded/decompressed/restored res. Much better than dumb upscaling.
Why this method won’t work with 1xDP and direct DP to anx7530, because anx7530 is not designed for native DSC 1.4, but to be fed via anx7678, then the claim DP 1.4 compatible have sense. This guys design HW they are correct always or HW is junk.
That is why I write we got migration from "8K V1-V2 X P"prototypes, that are shown on Public demo, to V3 that is “8K V1 P” that is shown to have 90Hz input probs, uses upscaling vs DSC etc etc. From 4K upscale to 8K, input become in marketing data, 2x2560x1440. Same BS as with 4K HMD.
Debunked or not, make your call.
*- unofficial specs, reverse engineering used. (common practice), not pure speculation as some pro-Pimax people claim. On KS projects you can do pure speculation on specs, aka marketing specs, as I note in my specs sheet(top post).
Goal: Get Pimax to understand backers will not be happy if get same experience as with 4K, only better tracking in double FOV with same input res as half FOV res. Or give backers option, 2xHDMI with short cables (for use like PSVR but with good head/motion tracking). But no roomscale. Or give option to wait for DSC 2:1(real DP 1.4 input), and room scale, expected at date as “8K X” ship date.
If solution DP1.4-2xHDMI to 2x anx7530 is too expensive. Give option to trade all rewards for actual DP1.4 to b/o box, to 2x anx7530. I think most backers will decide, if want to wait for rewards, or wait for product as was shown to public and they think they backed.
Give backers options, if can’t solve DSC on DP input without using HDMI and B/O box.
Option 1: Trade all rewards for “8K V2 CV” HMD, to get same experience as “8K V2 P”, 1920x2160/eye(after DSC 3:1 compression) restored to to 4K resolution, not simply upscaled, that is non tech correct for DSC algo to call it. This is same as 8K X, but in upscale mode, but on horizontal res, like Checkerboard rendering, so not so simple as divide/multiply Res. Proper Mid gen VR product. With maybe 2nd gen FOV, this is speculation/assumption ofc, I can’t know specs of future products that are not marketed!
If possible with GPU out DSC encoded 4K to HMD directly(SW encoder), if not, use B/O box(DP 1.4 to 1xHDMI 2.0), trade some KS reward for it.
Option 2: Get “8K V3 CV” HMD with 2x2560x1440 target render res, ASAP. Wait for rewards. Expect nVidia/Pimax to provide SW DSC encoding solution at later date, to give promised product specs. Probably no ETA from Pimax possible, heard there were plans nVidia wants to retrofit DP 1.2 cards with DSC capability… Rumors to be clear.
I am pro-backers, for some reason I prefer to defend people from companies, than companies from people xD.
TL:DR
Sorry can’t perform Fake DSC, it would lose valuable data(display bad picture about my reverse engineering skills) .
This is in depth explanation of my logic, without all sources, cause it would be huge time consumer to explain every tech detail in my logic. This is digest version, imagine amount of data I have in head if this is DSC 10:1 to it . If you want lay-man version, console tech analogy, read this.
