Sorry if this is a repeat (I searched and didn't see anything) and this may belong in the >$20k forum, but I'm wondering if anybody has any information about this DLP:

http://www.zeiss.de/c12567b00038cd75/Contents-Frame/916a1bd1307bf2acc1257480005213b4
powerdome®VELVET from Carl Zeiss is the world's first and only video projection system with an absolutely black backdrop based on the DLP® technology of Texas Instruments. A Carl Zeiss in-house design, VELVET fascinates not only by its measured contrast ratio of 2,500,000 : 1 – a tremendous leap compared to other projectors offering up to 30,000 :1 –, but also by its expanded color spectrum thanks to the latest BrilliantColor™ technology of Texas Instruments. In the projected image, very bright and extremely dark areas are excellently resolved at the same time.
...
The leap in contrast to more than a million to one has been made possible by an innovative optical design that excludes stray light and, thus, a gray background from the start.
...
The new projectors will become available at the beginning of 2009.I'm wondering whether they are doing something that allows the black to be extremely dark even with some pinpoints for stars being very bright, or whether they are doing something that dims the whole screen and so lowers the maximum white point as it lowers the absolute black level. I think a local dimming system could work well for their application if it had enough zones (wouldn't need 1:1 pixel mapping), but I have no idea what they are doing. Given that they talk about edge blending it would seem that they need to do something beyond just pulling the native CR range for the whole image down to really do this right.

I also wonder if we'll see whatever they are doing in home theater models.

--Darin

Thanks for posting this,

I have no idea how Zeiss made it, but i do believe in a native cr without any local dimming features. The reason for this is becuase,in late 2007, a well known projector company demonstrated the effect of using a enhanced DLP design using two TIR prism "infront" of the dmd. By doing that, the DMD flat state (or what in the article is refeering to scattered light could be pretty much eliminated to enter the lens).

Yes, this works for both singel and 3chip DLP (and a added bonus for singel chip would be to integrate a second light path (carry the rejected light at the CW) connecting to the two prisms and thus avoid throwing away 2/3 of the light when projecting white and also increase the secondary color energy)

However, this is only true for very small/narrow light paths ..basically adding more bulbs into the DLP archticture (we have heard about duARCH and similar concept) you are also increasing the entendue and hence needs to deal with the increased scattered light that follows in the DMD flat state.

I do fully believe that Zeiss could have designed the light source in such a way that they avoid any entendue increase(one could assume the ANSI lumens is somewhat high in that projector) and then carefully applied the two TIR PRISM with a special designed lens.

who knows, i dont think Zeiss will tell...hehe...

and there is no mention of cost......

not cheap I would suspect.....

and there is no mention of cost......

not cheap I would suspect.....


Definitely belongs in the 20K plus forum;). Probably in the in the 200K plus forum, perhaps even in the 2000K+ forum, as these machines will most likely not become available as a single unit, but only as a complete set-up, and Carl Zeiss is one of the highest end planetarium builders around.

Compare it to the pricing indications given at Infocomm by E&S, for its first and upcomming second generation laser based planetarium projectors.

Zeiss intricate laser starprojectors have in the past cost one or even several million a piece.

Hadn't heard of these machines before, last time I checked the Carl Zeiss website was a few months ago. It all depends on the scale of the machines, one or two per dome or many, and the light/optics involved.

Edit: just read the stuff at the linked site, and it mentions that other applications are also possible, and it relies on a multichannel (i.e. multiple projectors) set-up to light up the dome, so it will be available seperately, and not be enormous (and enormously expensive), still expect it to be very likely these units would fit in the 200K+ list forum.

The English language release doesn't provide much more info: http://www.zeiss.com/C1256A770030BCE0/WebViewTopNewsAllE/51B77724D5160A94C125748300388B3A?OpenDocument.

The English language release doesn't provide much more info: http://www.zeiss.com/C1256A770030BCE0/WebViewTopNewsAllE/51B77724D5160A94C125748300388B3A?OpenDocument.From that link it looks like it goes way beyond what a dynamic iris for the whole screen could do and is more like I was hoping:
In Chicago, planetarium directors from all over the world expressed their enthusiasm about the image quality: the blackness of the background makes white text appear like fluorescent writing, and objects seem to float in the room. Furthermore, the BrilliantColor™ technology from Texas Instruments, which mixes six colors instead of only three, ensures a high degree of colorfulness.

Unlike the total blackness achieved by VELVET, practically all standard projectors provide a gray background. Therefore, an image frame always remains visible. For dome projection, several partial images are put together to form one large image, with the image edges having to be blended into each other. With VELVET, these edges are completely invisible for the first time.The 2nd paragraph describes the problem I think they would have if they were doing full screen dimming to get the 2.5 million:1. They might be using it partially to get that high, but from what I've seen of an RS2 that was probably at least 13k:1, it didn't have nearly enough to get black backgrounds with at least small or medium sized white text or to have the edge blending problem from raised absolute black levels go away. Might be enough with a full screen of white text on black if the ANSI CR was high enough, but if this link is right then it seems like a big step forward. Even if it won't be coming to home theater in the short run I still like to see manufacturers achieving things like this.

--Darin

1 In the past Zeiss used LDT from Jenoptik. That is laser display technology. That includes a very expensive laser generator that also modulutes the laser acoustically and then blends the r+g+b channels into a fibre. This light is then 2D scanned by a polygonal scanner and a galvanic mirror. This technology has high bit depth and contrast but is limited in resolution. It tops out at HD for a single channel.
Resolution limited

2 E&S GLV ESLP2 projector has enormous resolution and good bit depth. Sony has taken GLV to 30000:1 but I doubt E&S is there with ESLP2. Sony has shown very decent contrast is possible.
Low contrast for planertarium applications

3 This new DLP thing by Zeiss will be somewhat resolution limited and that is why several channels are necessary.

4-panel lcos does 1M:1 so I am pretty sure Carl Zeiss is doing something like
4DLP. Scattered light go everywhere that is why I think it has to be actively blocked by one extra panel and sent to lighttrap number two.

However it's done, it is exciting. Perhaps we will see trickle down.

Art

Art S.
Could it be tthat Wolf cinema might be using this kind of technology?

Ohlson, I am surpirsed to see no-one picked up on this, but the key might be in the unusual configuration of the lamphousing being seperated from the projector. This is not the standard two part config of lampbalast underneath or at the back of the lamphousing/optic part.

Unless the PRs have put a misrepresentating spin on this.

"The technology of VELVET projectors has been particularly designed for use in dome theaters. The lamp housing, which is independent of the projector, permits the installation of various lamp types with various performance features. Inclined installation positions of the projector do not influence the service life of the projection lamp. For brightness adjustment of the projectors in a multichannel system, the lamp is controlled instead of the video signal, once again leading to improved image quality of the multichannel projection."

The LASER system, Adlib: http://www.zeiss.de/c12567b00038cd75/Contents-Frame/916a1bd1307bf2acc1257480005213b4

The SRX system (Skyskan offers similar two projector full dome set-up, perhaps collaboration like with the DefinityTWIN?) http://www.zeiss.de/c12567b00038cd75/Contents-Frame/916a1bd1307bf2acc1257480005213b4

The lamp housing, which is independent of the projector, permits the installation of various lamp types with various performance features.

Id love to see this become standard ( sadly it never will ). Projectors with no light source and a open aperture in back for a lamp house. Just like with 35mm film projection.
We would choose a lamphouse to fit our needs ether xenon or UHP from a LH manufacture. Over the years you would keep the same lamp house which would be a universal fit between upgrading projectors.

One question.If,4DLP, what good does a Brilliantcolor algorithm do for you ?
How do they do it in 550 mm x 550 mm x 350 mm

Aero
Mentioning Brilliantcolor is only TI advised spin doctoring in my opinion. Mentioning the stability of dlp which is true is also TI advised spin. I think TI will push on how stable dlp technology is compared to the competition.
I can not be sure it is 4DLP but I can not see how you with good optics can just suck out the bad photons like dust with a vacum cleaner and know which photons are good ones and let them be on their merry way.

Aero
Mentioning Brilliantcolor is only TI advised spin doctoring in my opinion. Mentioning the stability of dlp which is true is also TI advised spin. I think TI will push on how stable dlp technology is compared to the competition.
I can not be sure it is 4DLP but I can not see how you with good optics can just suck out the bad photons like dust with a vacum cleaner and know which photons are good ones and let them be on their merry way.I think the part about BrilliantColor does make it more likely that this is a 2 chip/panel in series solution, not 4. But you could be right.

--Darin

darinp2
I was thinking 3+1 in series but you make a good point about BrilliantColor so 1+1 is probably more likely.
But BrilliantColor could be a buzzword added by the PR people when the engineer just said the projector is capable of an enlarged gamut.

this is top news you cant found anywhere at the moment.

i think a lot if i shoud post it here in avs because there are some
people that upset me a lot with wrong and stupid postings but
here is it.

i know some people at zeiss since long time and i get the permission to post
here at avs some details.

the system is based on at least 4 pr. for a dome better 6 pr.
reason is one pr. have only 900 lumen light out.
the use one 350 watt uhp lamp in each pr.
dmd resolution is 1920x1200.

such a system is about with all the drive and source electronics
1.2 mil. euro.

so i guess that a singel pr. is arround 150000 euro.

its a single chip design but they not tell me how to get this
2500000:1 cr. but i guess the may use a secound dmd for this.

i got a invitation to see some of the first samples next week but as my
healt sitoation is not that good i cant go.
but lets see......

please note that we talk about zeiss numbers not consumer pr.

when zeiss say 900 lumen you will found may 1000 lumen for sure.
when other consumer company say 900 lumen you may found 600 lumen
if you are lucky.

but i will tell them that for high end home cinemas we need 2000 lumen
to feed a bigger lets say 4.5 m screen.

tomorror i will have a phone call with one of the top people there to talk
about the use of this kind of technology in a high end home cinema pr.
as this is the end of talk about cr. for sure.

i can tell that zeiss already think about it!

there is a big chance for zeiss to enter this market as there is demand
for pr. that are in the bigger than 50000$ range IF the pr. is very good.

the best news is that this proves that there is a way to go to infinity cr.
with dlp and it will be just a matter of time when we will see this in
a home cinema pr.that are in the normaly price range.
but i guess it will take some years.

900 lumens is not a bad start. Much brighter then most current Home Cinema projectors but as you mention for those with larger screens more lumens are needed.



the best news is that this proves that there is a way to go to infinity cr.
with dlp and it will be just a matter of time when we will see this in
a home cinema pr.that are in the normaly price range.
but i guess it will take some years.

Im getting old so I hope its soon:) but yes knowing it is possible is the underlying importance of this post.




such a system is about with all the drive and source electronics
1.2 mil. euro.

so i guess that a singel pr. is arround 150000 euro.

I wonder who in the 20k + forum will be the first:)

Good post. Thanks for the info, keep us posted.

W.Mayer
Thank you very much for the effort and the information!
Please try and confirm that they use 1+1 dmd in series configuration. While you are on the phone ask them when they will ditch the UHP and replace it with three lasers and throw away the color wheel.

Uptime is important in applications such as planetarium, simulation and command and control. A lamp and a spinning color wheel is not the optimal solution.

Well I'd call this the UberKontrast. I'd certainly love to see it in action and ,as Wolfgang says ,if it lives up to what they've said no doubt we would have to find something other than contrast to talk about.

Art

If this is realized by 2 DMDs after each other then the native contrast ratio per DMD seems to be 1,581:1, I guess (square root of 2,500,000:1). The native contrast ratio of current DLP single chip projectors seems to be about 3,000:1, so that sounds reasonable to me.

That makes me wish for an LCOS projector with 4 panels... ;)

That makes me wish for an LCOS projector with 4 panels... ;)

Convergence nightmare:eek:

Convergence nightmare:eek:
It might not be that bad, if the fourth panel has a lower resolution anyway (similar to how HDR displays work).

but i will tell them that for high end home cinemas we need 2000 lumen
to feed a bigger lets say 4.5 m screen.I agree for a lot of the high end. They could still probably find a market with even 700 lumens for under $20k if the CR is as impressive as they make it sound though.
the best news is that this proves that there is a way to go to infinity cr.Do we know that they can get huge CR in a case like a few bright stars on a black background and aren't doing something like a dynamic iris to get their 2.5M:1? I know it sounds like they can get huge CR in individual images, but I would be more comfortable if I knew how much CR they can get in an individual image. For instance, if they are doing 2 DMDs in series then I think they should be able to get over 100k:1 intra-image with just a few bright stars on a black background.

--Darin

Convergence nightmare:eek:
I don't really think so. LED LCD-TVs with local dimming are using just a dozen of LED elements and it still kinda works (with some artifacts). The forth panel wouldn't necessarily have to do pixel perfect dimming. It could do smooth dark/light gradiants. If combined with a clever gamma algorithm (similar to those used for adaptive iris) that should work pretty well, I'd guess. And it would totally eliminate any need for a mechanical adaptive iris.

I agree for a lot of the high end. They could still probably find a market with even 700 lumens for under $20k if the CR is as impressive as they make it sound though.
Do we know that they can get huge CR in a case like a few bright stars on a black background and aren't doing something like a dynamic iris to get their 2.5M:1? I know it sounds like they can get huge CR in individual images, but I would be more comfortable if I knew how much CR they can get in an individual image. For instance, if they are doing 2 DMDs in series then I think they should be able to get over 100k:1 intra-image with just a few bright stars on a black background.

--Darin

may a led based dlp pr. is a good alternative.
its for sure much cheaper but it will may be only at 100000:1 cr:)

may i know more tomorror.

The Zeiss Velvet uses 2 single chip DLPs in series. One with 6 color segement wheel, the other monocrome. Simple idea - great implementation. Projector is a great design with flexibility for light sources and lenses.

Unfortunately Zeiss is as big and slow as an elephant. By the time they figure out how to market it and distribute it to the masses - the components may be obsolete. If they insist on only complete systems they will lose a very good opportunity.

The Zeiss Velvet uses 2 single chip DLPs in series. One with 6 color segement wheel, the other monocrome. Simple idea - great implementation. Projector is a great design with flexibility for light sources and lenses.



So in the future we can go dome or flat screen AND we are not stuck with lamps forever. Great news!

This is much bigger than anything similar to DI operation. Dimmable LEDs that act globally are of limited use but still very good.

The Zeiss Velvet uses 2 single chip DLPs in series. One with 6 color segement wheel, the other monocrome. Simple idea - great implementation. Projector is a great design with flexibility for light sources and lenses.

Unfortunately Zeiss is as big and slow as an elephant. By the time they figure out how to market it and distribute it to the masses - the components may be obsolete. If they insist on only complete systems they will lose a very good opportunity.

nice infos seams you know a lot about it:)

yes i know zeiss long long time and you are right about the elephant.

one problem can be that as zeiss are working with some consumer
companys together like sony and other to produce lenses and more.
now if zeiss produce there own consumer pr. this companys will not like it
and that can stop zeiss from doing it.

Often Zeiss planetariums have special effects like lightening etc. I noticed the color wheel. Does this mean that Zeiss will throw in a rainbow effect for no extra charge? :)

i just hang up with one of the vips at zeiss.

they are very open to modify this dome system to a home cinema pr.
but even with a normaly much less expensive lens for home cinema
and bigger quantity such a pr. will be very likely in the 50k range.

i talk about lumens and they told me that they will increase the light out
to 1200-1300 lumen very soon.
again this are true numbers not marketing numbers.

good for screens gain 1.0 till to 4 m wide to get arround 14ftl.

they already will go with this product into the simulations market as there
is for sure a business because many people there still use crt pr. to
get very high cr.

if i cant go to them may they will visit me with a pr. in the next weeks.

so all in all after 1 hour with this guy i looking forward to hear more from them
and i hope that the start soon to modify this system to use in home cinemas.

sad that at the moment there is not enough light to drive my 7m screen
with this new development.

they try also laser and led but this guy told me the time is not ready to do it.

they know all our home cinema models from jvc from sony and all the
dlp pr.because they use it for domes and for some the build optical parts like
the optic and othere things.
so they know what is going on.

This is great news. I wonder if Zeiss will go it alone or partner with an existing projector manufacturer?

All I know is that the Adler in Chicago uses multiple CRTs and it just is so dim that it is totally unengaging.

Art

If they can do a machine for $50K to $70K MSRP that will light up a 4m or slightly larger (14ft wide) screen adequately, they will sell a bundle and capture the market share owned by others at this price point. Of course if there is no patent protection to using two DLP chips in series coupled with great optics or patent protection for their innovative optical design to achieve the world's highest on/off CR, others could do it as well.

All this in a two chip machine. Why no 4 chip machine? I would thing the given the need for pin point stars, three chip convergence no matter how precise just would't do it. The two chipserial desin given what the second chip does and how it is used, removes convergence as an issue. A great lens is still required for planetarium use. A less lens could be used for HT use I would think.

Also. How many planetariums are their in the world. i mean ones that use multimillion dollar projectors/ How often do these institutions shell out millions to replace their projectors? Hey board of trustees of the museum. Zeis has made a black level breakthough. Time to get a new machine. A new machine will shut up all those AVSers attending our 30 minute tourist show with those MY Kuro has much better blacks. seriously, 99.99% of planetarium goers would ever even think the blackness of the sky was not adequate.

Art. They could solve this problem by fitting a da-light high power to their dome or better yet a Viewtech 6 gain screen. Think of it. The projector is in the center of the floor (projecting at screen center). The CRTs one normally seen on the sides of the planetarium -such as in the NYC planetrium and I believe the adler as well, only use a CRT or rather multiple CRTs for projecting special effects, enlarged planets, enlarged moon yada yada. The main starfield is not from these CRTs but is from the main projectors in the floor center. That big baby that goes up and down hydraulically. Of course I could be wrong here, I haven't been to a planetarium in a few years so anybody feel free to set me straight.

W.Mayer
I understand that they start with lamps. Still it is great to hear that the direction of development is towards the use of lasers.

mark haflich
Your reasoning about why they chose 1+1 dmd series over 3+1 dmd in series is sound.

Note that Zeiss is not unique or first. SEOS was first with their rgbk light engine using 4 standard lcos panels to reach 500000:1. Their current model is called Zorro HC2015 (2048x1536) This projectors also seems rebadged by Barco who specs 1000000:1.

http://www.seos.com/Zorro.shtml
http://www.seos.com/ZorroTheater.shtml
http://www.barco.com/corporate/en/products/product.asp?gennr=1908

What do these 500,000 to one machines cost?

Mark I think $150k US

"nice infos seams you know a lot about it"


Yes I am in an interesting position to know about such things. Maybe with time I'll expand on that....lets say that planetariums are a hobby of mine.

The simulation market is the best bet for them. There is a huge installed base of existing CRT projectors with no replacement technology offering similar performance. Here alone they could make their investment back. But Barco is hot on the trail of this market as well with their Sim7. It is LCOS based and 4:3 ratio to match the CRT sizes. They could have been winning the market already but lost over a year of time while they could not get the LCOS chips in quantity. Then the Belgian supplier went bust early this Summer. So they have just switched to Sony chips and the projector is deliverable finally but very late.

(This is the same problem btw as Seos with their 300.000:1 contrast Zorro. Same chip supplier and Seos can not switch to Sony without redesigning the entire projector - oops! Zorro is a 4 panel DLP or LCOS based on the same principal to use panel one to modulate the contrast. But aligning 4 panels is not soooo easy as 2 single chip DLPs)

Now Zeiss has a better contrast projector than Barco (but I bet Barco tries to make a deal here) for the CRT market. Just no Simulation marketing experience. (Can two such large egos work together?)

As for consumer products it would be such a niche product that Sony or others wouldn't care. The cooperation on consumer cameras has nothing to do with this or even their other higher-end products. The consumer market is ruthless though. I am not sure Zeiss has the stomach for that. Maybe with a partner....

We live in interesting times. Maybe we should start a planetarium section here!

for domes they need 15-20 years a perfect working pr.and
this is very sure a problem for lcos as they are not as
stable as dlps.

yes i think the simulation market is bigger than dome but high end home cinema
is bigger than simulation.

you should send me a pm as we are both in germany and i cant say all
in a public forum.

HD_Domes
JVC is active in the simulation market. As research they have designed a 1000000:1 ultra high contrast projector. Do you know if JVC is going to compete in the extreme contrast simulation market?

JVC has
1 Their own lcos panel, d-ila
2 Very high native contrast at 20000-40000:1 depending on resolution
3 A selection of resolution ranging from 1080p to 8k!

,in late 2007, a well known projector company demonstrated the effect of using a enhanced DLP design using two TIR prism "infront" of the dmd. By doing that, the DMD flat state (or what in the article is refeering to scattered light could be pretty much eliminated to enter the lens).

Aero, do you know if this TIR prism dropped ANSI CR? It might be that it improved on/off but sacrified ANSI contrast in the process.

What JVC was doing already in 2007.
http://www.nhk.or.jp/strl/open2007/en/tenji/t06.html

So is the Barco and Seos pj kind of like the Brightside tech?

Yes and no.

1 The prinicple is the same with serial modulation in two steps.
2 It is different in that Brightside was mostly focusing on HDR. This demands very high brightness levels by one or more magnitued than what is normally used for projection. At the same time HDR manages to retain a low black level thanks to the high dynamic range.

I think it is interesting that JVC is in this game at the research level.

Art. They could solve this problem by fitting a da-light high power to their dome or better yet a Viewtech 6 gain screen. Think of it. The projector is in the center of the floor (projecting at screen center). The CRTs one normally seen on the sides of the planetarium -such as in the NYC planetrium and I believe the adler as well, only use a CRT or rather multiple CRTs for projecting special effects, enlarged planets, enlarged moon yada yada. The main starfield is not from these CRTs but is from the main projectors in the floor center. That big baby that goes up and down hydraulically. Of course I could be wrong here, I haven't been to a planetarium in a few years so anybody feel free to set me straight.

No you are right Mark the central projector does the stars but other presentations are also done in the same dome. All could be done now using the same devices. Yea they could use a HP and three of the 100 seats could see a positive gain.

Art

Actually because the seating is in the round, it would be about 12 seats. I was joking about a gain dome. But there is a point that unlike our HT set ups, a planetarium has its main projector at screen center which man high gain screens need for image optimization.

What is the typical dome size in planetariums these days?

No need to go to the Rai this year, as the dome demo will use 'herkömmliche DLP Projektoren' not the Velvet. Fraunhofer will demo their known blending/warping technology, focussing on autocalibration (3 camera system).