|
As this forum is rarely used anymore, we've locked it. Feel free to browse and read. Questions? Please reach out to us directly. Cheers! |
|
 |
|
|
| Author |
Message |
macgyver655
Joined: 22 Aug 2007
Posts: 8508
|
| Posted: Fri Jan 08, 2010 12:34 am Post subject: |
|
|
| garyfritz wrote: | | macgyver655 wrote: | | Ya know CJ. I think if they can get some more light out of the LEDs and put them in a 3 chip DLP it would be a killer projector. |
But if you go to 3-chip you end up with converence issues. Single-chip DLPs have a huge advantage there. Get rid of the rainbow disadvantage and I think 1-chip DLP is the sweet spot, at least for home HT.
| ecrabb wrote: | | Gary, that 1,024 level brightness limit is probably a time-imposed limitation, with human eye sensitivity also a factor. In essence, that may be for a single DMD running monochrome with no color wheel. Perhaps throwing a 6-segment color wheel in the light path and forcing the DMD to represent 6 different luma values in the same time it would usually represent one, may cut the color to something like 7-bits/per pixel. |
?? I thought it meant the DMD was capable of flapping its mirrors to represent 1024 levels of grayscale. That's 10 bits of grayscale **per color**. You can represent 1024 levels when the color wheel is showing red, 1024 when it's showing blue, etc. So that should give you the equivalent of 10 bits per color, or 30 bits of color resolution. |
First, my thoughts on the convergence, there are already 3 chip DLPs and 3 panel LCDs and they dont have convergence issues so why would a LED dlp?
Second, Its my understanding that the affect of rainbow is caused by the milisec times that the single chip is only displaying one color. The affect is your seeing the flash of the chip when only reflecting one of the 3 colors. If you have 3 chips on at the same time then there is no rainbow.
Somewhere in my crap of documents I have the color gamut of single chip and 3 chip DLP. I'll see if I can find it.
|
|
| Back to top |
|
 |
ecrabb
Forum Moderator
Joined: 13 Mar 2006
Posts: 15909
Location: Utah
TV/Projector: JVC RS40, Epson 5010
|
| Posted: Fri Jan 08, 2010 12:46 am Post subject: |
|
|
I'm not finding much. I'm guessing it's not exactly something TI or any of the projector manufacturers want to advertise. However, many 1080p 3-chip DLP systems are now being advertised an sold as 30-bit-per-pixel displays. If a DarkChip3 3-chip projector is a 30BPP display, then a single-chip machine based on the same DMD is a 10BPP display... meaning 1024 colors. 10-bit can look pretty good if it's done well, but 30-bit it is not. That's probably the biggest reason I dislike single-chip DLP so much. I'll keep looking later and see what I can find.
SC
|
|
| Back to top |
|
|
cmjohnson
Joined: 03 Apr 2006
Posts: 5180
Location: Buried under G90s
|
| Posted: Fri Jan 08, 2010 12:51 am Post subject: |
|
|
I don't want to talk out of class, but Scott didn't say that anything he showed me was confidential, either.
But out of courtesy, I will not say exactly what I COULD say, but I will say this: There is an attractively packaged version of this new
VDC LED DLP projector, in addition to the regular industrial packaged version. It is obviously intended for the home theater market.
It has an onboard control pad and is a sharp looking unit. It is a version being made for one of the other "named companies" that has
previously sold repackaged Marquees. You can ask but I won't reveal which one. I actually didn't see a badge on it anyway. I also
can't say (because I don't know) if there are any internal differences. My guess is that the differences are mostly cosmetic...except for
the onboard control panel.
As for 3 chip DLP units, their panels have to be aligned during the assembly of the optical engine but this is a one time
alignment process. In normal usage it will never change, BUT some DLP projectors have limited convergence
capability in their menu systems. However, it's only shifting the image a pixel or so. And it can't shift the image
LESS than one pixel.
TI's convergence spec, if my info is right, is that the total misconvergence allowed is no more than one half pixel width
between any two panels. In practice, this is LESS than the apparent shift in pixel positions from red to green due
to the imperfectly achromatic lenses used.
CJ
|
|
| Back to top |
|
|
macgyver655
Joined: 22 Aug 2007
Posts: 8508
|
| Posted: Fri Jan 08, 2010 1:09 am Post subject: |
|
|
| cmjohnson wrote: | I don't want to talk out of class, but Scott didn't say that anything he showed me was confidential, either.
But out of courtesy, I will not say exactly what I COULD say, but I will say this: There is an attractively packaged version of this new
VDC LED DLP projector, in addition to the regular industrial packaged version. It is obviously intended for the home theater market.
It has an onboard control pad and is a sharp looking unit. It is a version being made for one of the other "named companies" that has
previously sold repackaged Marquees. You can ask but I won't reveal which one. I actually didn't see a badge on it anyway. I also
can't say (because I don't know) if there are any internal differences. My guess is that the differences are mostly cosmetic...except for
the onboard control panel.
As for 3 chip DLP units, their panels have to be aligned during the assembly of the optical engine but this is a one time
alignment process. In normal usage it will never change, BUT some DLP projectors have limited convergence
capability in their menu systems. However, it's only shifting the image a pixel or so. And it can't shift the image
LESS than one pixel.
TI's convergence spec, if my info is right, is that the total misconvergence allowed is no more than one half pixel width
between any two panels. In practice, this is LESS than the apparent shift in pixel positions from red to green due
to the imperfectly achromatic lenses used.
CJ |
I think if they really wanted to develop the 3 chip version they could produce a chip that had a few extra rows of mirrors surrounding the regular resolution mirrors to allow for shifting the image for convergence. There is no reason to need geometry adjustment in this type display so it wouldn't need to be complex.
|
|
| Back to top |
|
|
cmjohnson
Joined: 03 Apr 2006
Posts: 5180
Location: Buried under G90s
|
| Posted: Fri Jan 08, 2010 2:33 am Post subject: |
|
|
3 chip DLP units have been around for a long time. In fact, they were the FIRST DLP units to be made. They started with the First Generation TI 3 chip engines at 800x600 resolution. I've had a couple of those units, in 800x600 and then later in 1024x758.
Right now I have two Barco SLM R6 3 chip DLP projectors in my garage. They're 1280x1024 and are considerably more advanced than the earlier engines. Their panel misalignment is TINY. The chromatic aberration in the lens is greater. There's no need for any extra mirrors
to handle convergence, and it wouldn't work well anyway as the typical errors are a fraction of the size of a single mirror.
It would weird to make a DLP chip with a row of insanely tiny mirrors surrounding each "normal" mirror, just to fine-tune the convergence. It would make the design logarithmically more complex.
CJ
|
|
| Back to top |
|
|
garyfritz
Joined: 08 Apr 2006
Posts: 12088
Location: Fort Collins, CO
|
| Posted: Fri Jan 08, 2010 2:51 am Post subject: |
|
|
| ecrabb wrote: | | I'm not finding much. I'm guessing it's not exactly something TI or any of the projector manufacturers want to advertise. However, many 1080p 3-chip DLP systems are now being advertised an sold as 30-bit-per-pixel displays. If a DarkChip3 3-chip projector is a 30BPP display, then a single-chip machine based on the same DMD is a 10BPP display... meaning 1024 colors. |
No no no. You can't have a projector with 1024 colors unless it's monochrome. You wouldn't know ahead of time what color palette to select. A 1024-color RGB projector would look horrific -- literally like a cheap cartoon.
*IF* the DMD could provide 1024 levels for each color segment, a 1-chip DLP should support 30 bits of color resolution. So you're saying you don't think the mirrors can do that?
Hm. Maybe. I thought the mirrors oscillated at MHz speeds -- 16MHz sticks in my mind -- but I can't find a hard reference. dlp.com just talks about "up to thousands of times per second." With a 4x 7-segment wheel running at 60Hz, that's 4*7*60 = 1680 segments per second. I would think the mirrors would have to oscillate at *least* 1024x faster, 1.6Mhz, to provide full 1024-level grayscale, and probably at least 2x or 4x faster. 16MHz should be enough to support 1024 levels per segment, but I don't know if the mirrors actually flop that fast.
|
|
| Back to top |
|
|
macgyver655
Joined: 22 Aug 2007
Posts: 8508
|
| Posted: Mon Mar 29, 2010 4:44 pm Post subject: |
|
|
| cmjohnson wrote: | There's no need for any extra mirrors
to handle convergence, and it wouldn't work well anyway as the typical errors are a fraction of the size of a single mirror.
It would weird to make a DLP chip with a row of insanely tiny mirrors surrounding each "normal" mirror, just to fine-tune the convergence. It would make the design logarithmically more complex.
CJ |
Dont know how I missed this unless it was when I went offline for a while.
I didn't mean extra mirrors surrounding each current mirror. That would be ridiculous.
I meant an extra row or 2 of mirrors around the perimeter of the DMD mirrors to allow for some image movement to correct minor misalignment and still retain resolution. The software addition would be almost nothing.
|
|
| Back to top |
|
|
macgyver655
Joined: 22 Aug 2007
Posts: 8508
|
| Posted: Thu Jun 24, 2010 2:07 am Post subject: |
|
|
I'm resurrecting this thread because I found the documents that pertain to the DMD function and to the comments that I made.
I was questioned on 2 comments I made in this thread. One was more then 1 mirror per pixel and the final outcome was there is 1 mirror per screen pixel but can be more then 1 mirror per input pixel on any resolution below the native resolution of the DMD. And that's a fact.
The second comment was that group mirrors are used to control contrast. I knew my memory was telling me this was correct but since I couldn't find the document, I had no proof. So I edited my posts as such and left it at that. Well after rereading my documents my comment was correct. Group mirrors are used to control contrast. Brightness is controlled by mirror on/off speed.
Questions are welcome.
|
|
| Back to top |
|
|
Spanky Ham
Joined: 22 Mar 2006
Posts: 5643
Location: Comedy Central
|
| Posted: Thu Jul 29, 2010 7:32 pm Post subject: |
|
|
I didn't really want to start a new thread on this, so I will post here.
I stopped by and saw Zeus, I mean Scott on Tuesday. He showed me around and I got to see some brand new vertical split pack 9500s. Since we were the last couple of people to leave, I should have slipped one or three out to my van. These may be in Nash's future in about ten years.
I got a couple of demos of the LED pjs. They had just done a presentation for the Navy with four blended pjs that looked really nice. The blend zone was darn near invisible. We also looked at the Asian test blu ray in his lab. The motion blur system works really well. He displayed the 1080p moving image. Without motion blur, I think it could only do around 500p. Engage motion blur and you could easily see the 1080p test pattern. VDC is the only ones that have this motion blur system at the moment.
On 3 chip LED dlp, Scott said he researched it. Unfortunately, you would have to actively cool the LED chips to 15C. On a 1 chip dlp, the LEDs are only on 1/3 of the time and can therefore be cooled by a heat sink and fan. If anyone has any novel ideas that Scott may not have heard about to cool the chips, then I am sure he is all ears.
|
|
| Back to top |
|
|
Nashou66
Joined: 12 Jan 2007
Posts: 16171
Location: West Seneca NY
|
|
| Back to top |
|
|
AnalogRocks
Forum Moderator
Joined: 08 Mar 2006
Posts: 26706
Location: Toronto, Ontario, Canada
TV/Projector: Sony 1252Q, AMPRO 4000G
|
| Posted: Thu Jul 29, 2010 11:10 pm Post subject: |
|
|
| Nashou66 wrote: | what about those piothermicsomethingoranother chips? the ones that get real cold on one side and hot on the other? could a water cooled heat sink system be used in conjunction?
Athanasios |
That's Greek to me, do you mean a Peltier cooler 
_________________
Tech support for nothing
CRT.
HD done right!
|
|
| Back to top |
|
|
Nashou66
Joined: 12 Jan 2007
Posts: 16171
Location: West Seneca NY
|
|
| Back to top |
|
|
stefuel
Joined: 07 Mar 2006
Posts: 3353
Location: Green Harbor MA USA
|
| Posted: Fri Jul 30, 2010 12:07 am Post subject: |
|
|
That was GEEK you fool
_________________
Chip
A Barco is only a AmPro with training wheels
Card carrying member of the AVS chain gang.
|
|
| Back to top |
|
|
AnalogRocks
Forum Moderator
Joined: 08 Mar 2006
Posts: 26706
Location: Toronto, Ontario, Canada
TV/Projector: Sony 1252Q, AMPRO 4000G
|
| Posted: Fri Jul 30, 2010 12:10 am Post subject: |
|
|
Damnd typo's!
_________________
Tech support for nothing
CRT.
HD done right!
|
|
| Back to top |
|
|
Spanky Ham
Joined: 22 Mar 2006
Posts: 5643
Location: Comedy Central
|
| Posted: Sun Aug 01, 2010 11:07 pm Post subject: |
|
|
|
I should have also mentioned that CRT is going to be around for awhile for those that want to stick with it. Scott says they are still getting orders for CRTs. He said that these multi million dollar sims are built for a crt pj, so they can't just stick another pj in its place. If one is ordered today, then it will still be projecting probably ten years in the future.
|
|
| Back to top |
|
|
|
|
|
|
|
You cannot post new topics in this forum
You cannot reply to topics in this forum
You cannot edit your posts in this forum
You cannot delete your posts in this forum
You cannot vote in polls in this forum
You cannot attach files in this forum
You can download files in this forum
|
Forum powered by phpBB © phpBB Group
|
|
