Sim2 Lumis 3 Chip DLP little Test - Page 7 - AVS Forum
Forum Jump: 
Reply
 
Thread Tools
post #181 of 371 Old 01-26-2009, 03:44 PM
AVS Club Gold
 
Christy Warren's Avatar
 
Join Date: Oct 2001
Location: Almaden Valley in San Jose, CA
Posts: 89
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 0 Post(s)
Liked: 10
Quote:
Originally Posted by Digital2004 View Post

hello
does the LUMIS fit the SCHNEIDER CINEDIGITAR lens ?
or vignetting ?
the C3X 1080 with ISCO II is slightly vignetting.

ps: stratospheric on off is an academic discussion imho once you're in scope config with black room. + ANSI gives so much more pop in most scenes, the 3D depth. way above CRT which never had good ANSI CR.
if one is in a non scope configuration, on off at high levels is required.

my very very humble opinion.
... and FTL rules !! how could people live with 6-8FTL on so many machines in the past and still now !

I own a Schneider and with some luck I'll be able to give an answer next week if I'm lucky and can get access to a Lumis.
Christy Warren is offline  
Sponsored Links
Advertisement
 
post #182 of 371 Old 01-26-2009, 05:46 PM
AVS Club Gold
 
GetGray's Avatar
 
Join Date: Aug 2004
Location: Mid-South USA
Posts: 5,428
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 20 Post(s)
Liked: 46
Christy: Didn't you report that a CineDigitar worked with a C3X? I felt the CineDigitar is simply too small to use with these projectors. The prime is just too recessed and the beam size too large at the lens. It may fit but if it does, it will just barely and you may have edge effects from the borderline vingetting. Perhaps if one "stuck it in the hole", but then you would be very limited on tilt, yaw, etc. tunability. Or if you were at the long end of a T3, maybe. It would still be too close for comfort I believe.
GetGray is online now  
post #183 of 371 Old 01-26-2009, 05:46 PM
AVS Special Member
 
odyssey's Avatar
 
Join Date: Jan 2001
Posts: 1,088
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 0 Post(s)
Liked: 10
Art,

Where did I say that you should toss sequential and ANSI CR measurements? I said that I thought that sequential, ANSI, and MTF; all three of which measure contrast, are equally important. I should correct that to sequential and MTF only, since ANSI CR performance (at least relative performance) can be derived from a square wave MTF at 2 cycles frequency.

The overall comparison of the Lumis and a 5000:1 sequential CR Barco DP-1500 is of great interest to me and I will be doing it very carefully. I purchased the Lumis, so I obviously think that it's a good design. I don't think that my conclusions will be very different from Wolgang's. It would be better if there was no difference in the sequential CR of the two Barco units, but we will deal with that somehow.

It will be tomorrow before I can post the references I mentioned.
odyssey is online now  
post #184 of 371 Old 01-26-2009, 05:54 PM
Advanced Member
 
mlang46's Avatar
 
Join Date: Aug 2006
Location: Boulder, Colorado
Posts: 905
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 0 Post(s)
Liked: 11
Quote:
Originally Posted by Alan Gouger View Post

Its nice to acknowledge the relevance and importance of MTF which no one is doubting is it TI or the manufacture that keeps this information from being published.





.

TI has published theoretical papers on contrast ratio of each pixel but has not to my knowledge measured the Mtf directly. The new chips which have 10.3 micron centers and swing +-12 degrees are limited in contrast ratio by the diffraction pattern of the individual chips, the dimple holding the hinge and the black coating on the backing plate. Edge scatter determined by scatter programs is not a problem and does not significantly degrade the pixel to pixel contrast. The best theoretical results give a pixel to pixel contrast of between 900 to 1000/1

The geometric Nyquist cutoff limit on the chip across 1980 pixels is 41 line pairs/mm. At the center of the field a good projection lens is down to 40 percent MTF and 10 percent at the edge , so its the optics not the chip that is limiting performance

Fortunately at theedge of the field your eye's acuity is down by a factor 3 so the field performance is not as critical. unfortunatley in a 3 chip reflective system misalignment will kill your MTF depending on the colors your looking at because you will see them at the center of your field. say your red is seperated from the blue by one pixel because of misconvergence. when an object is shown with the color green this means the projectors resolution is only equivalent to a 480p projector that is why a .75 pixel misconvergence is really really bad. Mis convergence is much worse than lateral color becuse it is at the center of your vision.

TI is very open about there specifcations but the projector manufacturers for the most part tell you nothing. ANSI contrast is a scatter spec and tells you very little about the performance of the projector and has very little to do with contrast at high spatial frequencies , the frequencies that determine sharpness. I would like to see MTF specifications on all projectors sold for over 10,000 dollars and I would like to see MTF curves sent out for each individual projector at 3 field points. This would give the customer a quantifiable meausure of performans which would guarentee quality. Right now when you buy a 3 chip reflective projector whether its LCOS or Dlp you are essentially rolling the dice. What is the misconvergent spec on a 3 chip Sim Lumis or a JVC RS20. Do they have one?
mlang46 is offline  
post #185 of 371 Old 01-26-2009, 06:04 PM
AVS Club Gold
 
Mark Petersen's Avatar
 
Join Date: Dec 2000
Posts: 4,671
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 0 Post(s)
Liked: 11
Wow, I don't check this forum for 3 days and next thing I know there is a 10 page thread that I need to catch up on

Wolfgang, many thanks for taking the time to view, measure, write up and discuss the Lumis. As was discussed in Alan's thread, mating DB with top-flight ANSI has created something we haven't seen before. I also have to agree with some of the comments that others have picked up on as far as the RS20 vs Lumis pioneer logo comparison. I think the on/off on the Lumis is higher than 20k:1, that's the only answer that really makes sense to me. It also sounds like the RS20 that you have is perhaps down on on/off as I think most people seem to be reporting measurements of over 30k:1 with the iris wide open, but it's what the RS20 does on a smaller screen with the iris aperture reduced and the contrast up around 40k:1+ that is what makes it special. Bottom line though, I think the Lumis is in another league and obviously so from a price range standpoint. With a really large screen that most people will be using who spend the money for the Lumis, the RS20 isn't even an option.
Mark Petersen is offline  
post #186 of 371 Old 01-26-2009, 06:20 PM
Advanced Member
 
mlang46's Avatar
 
Join Date: Aug 2006
Location: Boulder, Colorado
Posts: 905
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 0 Post(s)
Liked: 11
[quote=Art Sonneborn;15657535]The high spacial frequency of CRTs (some of the best there are) are a murkey mess.The gausian beam spot is the culprit in most instances. Of course you are making my point. There are other PQ parameters that can and do make emormous difference in image quality of a projected image. The fact that CRTs have terrible MTF compared to a DLP and horrendous ANSI contrast they still produce stunning images.

You are also making my point regarding the MTF data being unavailable so saying that x unit is better than Y due it's superior MTF is tough to swallow. Who has the data list you gave above ?

Personally ,I believe this will also be shown to be the case with the higher MTF DCI units and the consumer units we are now seeing like the Lumis.




your missing my point . The low MTF of the CRT is offset by its large focal plain but the MTF it transmits to the screen is very High which is why it looks sharp to you because all front projectors are limited in MTF by the lens because of the high pixel count and small focal plane of digital projectors. a nine inch CRT with the same resolution as a .95 inch chip has a Nyquist cut off frequency of only 8 lp/mm vs 41line pairs/mm of a DLP

A projector with a Low MTF measured at the screen will always produce a poor image always

You want to understand how MTF is used in projector design read pages 314-323 Projection displays by Brenholtz and Stupp second edition and you will understand where the list came from.

who has the data list the projector manufacturers of course.
mlang46 is offline  
post #187 of 371 Old 01-26-2009, 06:20 PM
AVS Club Gold
 
Mark Petersen's Avatar
 
Join Date: Dec 2000
Posts: 4,671
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 0 Post(s)
Liked: 11
Quote:
Originally Posted by Michael W. View Post

Somebody did a graph on an RS1 and it showed the 15,000:1 on/off was only good up to a 2% APL after that it got progressively worse than a DLP projector it was compared against. This is clearly visible in the photos from Wofgang that shows the much brighter black bars on the JVC vs the Sim2, and thus are hardly a surprise to me. The deep black floor of the RS line quickly goes away as the APL rises.

Yes that was the graph that William Phelps and I did comparing full white intra-image contrast using an arbitrary test pattern that varies by APL. When I get some free time I plan on measuring the RS20 to see what it does and it would be interesting to see what the Lumis does as well and where (and if) there is a crossover.

Incidentally, there is also a companion graph that measures intra-image contrast using a fixed, low APL pattern where white is varied from 0-100IRE which is very useful in seeing what happens with dynamic technologies like dynamic gamma and dynamic lamps. The results can be directly compared with a native (non-dynamic) projector. It would be really interesting to see what the Lumis with DB does on this sort of a test pattern suite.

Quote:


Such a contrast graph would also serve to dispel any myths of outrageous contrast claims of future LED based projectors that could turn off the LEDs on a 0 IRE image. I think contrast graphs will give people a lot better idea of what the actual contrast capabilities of a projector are rather than just a quick glimpse at certain points that manufacturers can tweak to pacify the crowds.

Yes exactly! on/off contrast and intra-image contrast at low APL's are not the same.
Mark Petersen is offline  
post #188 of 371 Old 01-26-2009, 07:33 PM
Senior Member
 
Michael W.'s Avatar
 
Join Date: Nov 2005
Posts: 210
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 0 Post(s)
Liked: 10
Quote:
Originally Posted by Mark Petersen View Post

Yes that was the graph that William Phelps and I did comparing full white intra-image contrast using an arbitrary test pattern that varies by APL. When I get some free time I plan on measuring the RS20 to see what it does and it would be interesting to see what the Lumis does as well and where (and if) there is a crossover.

Incidentally, there is also a companion graph that measures intra-image contrast using a fixed, low APL pattern where white is varied from 0-100IRE which is very useful in seeing what happens with dynamic technologies like dynamic gamma and dynamic lamps. The results can be directly compared with a native (non-dynamic) projector. It would be really interesting to see what the Lumis with DB does on this sort of a test pattern suite.



Yes exactly! on/off contrast and intra-image contrast at low APL's are not the same.

MLang seems to be stuck on the idea that the black floor of a projector is the end all be all to determine how great an image can look. But, this metric by itself really means very little. Take LED DLP where theoretically the LEDs can be turned off on a 0 IRE image, the black floor is true black in this case, but when an actual image is displayed the black floor will be quite different.

It is easy for manufacturers to doctor up numbers for 0 IRE and 100 IRE images where the specs they state mean absolutely nothing. Take the new LG plasma with 1 billion:1 contrast ratio, is one to assume from MLang's comments that this LG plasma beats out everything on earth to date?

I think if the black floor at 0 IRE can be maintained throughout the APL ranges than what MLang is stating could be closer to the truth, but this is never the case and often quite far from the case. What is relevant is how the black floor is handled throughout the APL range, and I think this is why DCI projectors with measly 2,500:1 contrast rations are able to look so good compared to machines with 40,000:1 contrast ratios. Because, through most of the APL range the DCI projector actually has a better contrast ratio, and therefore better black floor in each particular scene, than the 40,000:1 projector.

Overall I think on/off contrast ratio is a metric that really tells very little of the story. After all, who watches a 0 IRE screen?!? What we should have is a simultaneous contrast ratio measurement graph with a test pattern that can approximate different APL scenes, i.e white box in center with black in the rest of the image. A 1% APL pattern should get pretty close to a native on/off contrast ratio measurement. The graph will show how much fluff that on/off ratio is, i.e. how quickly does that contrast ratio break down. Would you rather have 40,000:1 on 1%-2% APL scenes and 500:1 on 3%+ scenes, or 9,000:1 on 1% - 20% scenes, etc... (Not saying this is what it is, just an example to make a point.)
Michael W. is offline  
post #189 of 371 Old 01-26-2009, 08:02 PM
AVS Club Gold
 
Mark Petersen's Avatar
 
Join Date: Dec 2000
Posts: 4,671
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 0 Post(s)
Liked: 11
Yes I agree that on/off has little meaning with dynamic technologies and this is something that I've been preaching for quite awhile. As an example, someone could easily build a projector with 1000:1 native on/off and use an iris to fully close the light path and achieve infinite on/off. But no real image improvement is obtained other than full black outs. Seeing what happens from low APL through ANSI is very useful in my opinion as is seeing what happens at lower IRE's when dynamic technologies are employed.
Mark Petersen is offline  
post #190 of 371 Old 01-26-2009, 11:29 PM
Senior Member
 
Michael W.'s Avatar
 
Join Date: Nov 2005
Posts: 210
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 0 Post(s)
Liked: 10
Quote:
Originally Posted by Mark Petersen View Post

Yes I agree that on/off has little meaning with dynamic technologies and this is something that I've been preaching for quite awhile. As an example, someone could easily build a projector with 1000:1 native on/off and use an iris to fully close the light path and achieve infinite on/off. But no real image improvement is obtained other than full black outs. Seeing what happens from low APL through ANSI is very useful in my opinion as is seeing what happens at lower IRE's when dynamic technologies are employed.

I think it is useful even without dynamic technologies. The RS1 doesn't use dynamic technologies but the contrast ratio takes a nose dive above 2% APL scenes. I think the graph of contrast ratios from low APL through ANSI is useful for every projector and display technology made as it gives you an overall picture of what the display is capable of.

Take the lumis with DB turned off, it starts at an on/off of 5K:1 and ends up at 1k:1 at ANSI. Take the RS20, it starts with an on/off up to 42k:1 and ends up at 250:1 at ANSI. What the graph will tell us is where the crossover point is. I suspect on DCI machines the crossover from 2,500:1 to its ANSI is very low in the APL range compared to the RS20, and that is why the DCI machines can hold their own (contrast-wise) even with such a low on/off contrast ratio compared to the RS20.

It would be a worthwhile exercise to take an RS20, Lumis, and DCI machine and do such a graph (Lumis DB on and off) and see what it looks like. I think it would be very enlightening and could explain a lot of what we are seeing in the visual results.

Of course there is far more to a projector than just contrast, but this would at least provide some insight to those that preach on/off contrast as being the end all of display technology.
Michael W. is offline  
post #191 of 371 Old 01-27-2009, 01:11 AM
AVS Special Member
 
mhafner's Avatar
 
Join Date: Nov 2000
Location: Switzerland
Posts: 4,592
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 1 Post(s)
Liked: 24
Quote:
Originally Posted by W.Mayer View Post

sorry but i will not do such discussion again and again with you.

You hit the nail on the head Wolfgang. You tried to convince me years ago that projectors with 2000:1 or 3000:1 On-Off and no dynamic iris or dynamic black had no (obvious) haze problems, were good enough etc.. When I knew it was simply not true, based on watching such projectors in my room. Now we have just another round of tech improvements and the claims are the same as ever. Deja vu all over.
mhafner is offline  
post #192 of 371 Old 01-27-2009, 02:00 AM
AVS Special Member
 
mhafner's Avatar
 
Join Date: Nov 2000
Location: Switzerland
Posts: 4,592
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 1 Post(s)
Liked: 24
The specific Lumis had either a higher On-Off than the specific RS20 in that room or pure black is brighter on the Lumis. That assumes the room is not playing tricks on these black fields with relevant illumination contributions from ambient light and reflections caused by the two projectors on each others projections. If the room plays tricks the comparison is invalid by definition. If the actual On-Off on the Lumis is higher it contradicts Wolfgang's own measurements but not SIM's 30000:1 which can be higher than the RS20 depending on how it's set up and calibrated.
If the native contrast of 5000:1 is correct the max factor used by the Lumis with dimming and iris is 4-6. That assumes for black fields the maximal factor is used, which makes sense. With the maximal factor the Lumis has darker whites than the RS20. For actual intra image contrast with all kinds of material the contrast results depend largely on how badly the RS20 pollutes his blacks as soon as any > 0 IRE content enters the picture. ANSI contrast gives an indication for a special tough case. A case not representative of say, 0 IRE black space plus a couple of dozen 100 IRE stars. If the Lumis beats a good RS20 in that case too, CONGRATULATIONS! I'M IMPRESSED!
With top whites equalised, to get the black level down to RS20 levels requires a factor of 6-10 for lamp dimming, iris closing or a combination thereof. Unless you claim some stars suddenly double or triple the black level of the RS20. Then a factor of 2-3 would be enough. It still makes the whites darker on the Lumis than on the RS20. And it introduces white crush. With the same black level now the RS20 will win. It has brighter stars. No white crush. Despite a supposedly very unstable black level.
So how does the Lumis beat the RS20 in this case???
Anyway, if the Lumis comes close it's good news. The JVC can add lamp dimming and dynamic irises as well. And improve ANSI with better optics. What the SIM can do the JVC can do too in a future model. Starting with a panel that has 6-10 times more native On-Off contrast. Cool, no?
mhafner is offline  
post #193 of 371 Old 01-27-2009, 02:43 AM - Thread Starter
AVS Special Member
 
W.Mayer's Avatar
 
Join Date: Mar 2002
Location: Germany
Posts: 1,781
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 0 Post(s)
Liked: 12
i think if its possible(shift range from the barco) to
do one more screen shot.

barco cinema pr. with 2400:1 on off and arround 800:1 ansi cr. dim down to
the same light out the rs20 have by a nd8 filter side by side
with the rs20 and this black picture the pioneer offers at pause.
the same picture i had post withe the lumis and the rs20.

lets see if the shift is as big that i need from the barco cinema unit.
the pr itself shoud be not a problem but the opening arround the
lens will be the bottleneck.
W.Mayer is offline  
post #194 of 371 Old 01-27-2009, 02:52 AM
AVS Special Member
 
coldmachine's Avatar
 
Join Date: Nov 2006
Location: Waiting in the weeds
Posts: 5,807
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 0 Post(s)
Liked: 11
Quote:
Originally Posted by mhafner View Post

And improve ANSI with better optics.

The low ANSI wont be cured with better optics.

The JVC 4k units and the Meridian 810 have vastly superior optics but still only manage the same 250 ANSI. Its an inherent limitation of the technology, as is the soft image.

Meridian tried, unsuccessfully, to improve both. They were successful in improving the JVC noise exacerbation issue but that has come at a huge price, as anyone who has properly viewed this unit with anything other than the specially selected demo material, will tell you. The damage to the image is considerable and the 810 is way behind the best 3 DLP units.

--------------------------------------------
"Wow, do you think you are Adonis"...... "Baby, I'm not A-donis, I'm THE-donis"
coldmachine is offline  
post #195 of 371 Old 01-27-2009, 04:11 AM
AVS Special Member
 
Digital2004's Avatar
 
Join Date: Dec 2003
Location: Gotham City
Posts: 2,779
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 0 Post(s)
Liked: 11
Quote:
Originally Posted by GetGray View Post

Christy: Didn't you report that a CineDigitar worked with a C3X? I felt the CineDigitar is simply too small to use with these projectors. The prime is just too recessed and the beam size too large at the lens. It may fit but if it does, it will just barely and you may have edge effects from the borderline vingetting. Perhaps if one "stuck it in the hole", but then you would be very limited on tilt, yaw, etc. tunability. Or if you were at the long end of a T3, maybe. It would still be too close for comfort I believe.

that's my fear too
i have an ISCO II here with a C3X 1080 for test and it vignets slightly (but it's place at short throw, 5.6meters from a 4meter screen scope.
is the schneider like the ISCOII or even smaller "eye" ?
Digital2004 is offline  
post #196 of 371 Old 01-27-2009, 04:25 AM
AVS Special Member
 
Digital2004's Avatar
 
Join Date: Dec 2003
Location: Gotham City
Posts: 2,779
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 0 Post(s)
Liked: 11
Quote:
Originally Posted by Mark Petersen View Post
Wow, I don't check this forum for 3 days and next thing I know there is a 10 page thread that I need to catch up on

Wolfgang, many thanks for taking the time to view, measure, write up and discuss the Lumis. As was discussed in Alan's thread, mating DB with top-flight ANSI has created something we haven't seen before. I also have to agree with some of the comments that others have picked up on as far as the RS20 vs Lumis pioneer logo comparison. I think the on/off on the Lumis is higher than 20k:1, that's the only answer that really makes sense to me. It also sounds like the RS20 that you have is perhaps down on on/off as I think most people seem to be reporting measurements of over 30k:1 with the iris wide open, but it's what the RS20 does on a smaller screen with the iris aperture reduced and the contrast up around 40k:1+ that is what makes it special. Bottom line though, I think the Lumis is in another league and obviously so from a price range standpoint. With a really large screen that most people will be using who spend the money for the Lumis, the RS20 isn't even an option.
exactly
different machines different use different "clients" different home theaters:
big giant screens with punchy images imply tons of brigthness. the domain of the LUMIS or much more expensive machines.
see attached file.
recent lcd barely bright enough for a 200-220cm screen

 

screen size lumens.pdf 8.4072265625k . file
Digital2004 is offline  
post #197 of 371 Old 01-27-2009, 04:32 AM
AVS Special Member
 
Digital2004's Avatar
 
Join Date: Dec 2003
Location: Gotham City
Posts: 2,779
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 0 Post(s)
Liked: 11
Quote:
Originally Posted by mhafner View Post

The specific Lumis had either a higher On-Off than the specific RS20 in that room or pure black is brighter on the Lumis. That assumes the room is not playing tricks on these black fields with relevant illumination contributions from ambient light and reflections caused by the two projectors on each others projections. If the room plays tricks the comparison is invalid by definition. If the actual On-Off on the Lumis is higher it contradicts Wolfgang's own measurements but not SIM's 30000:1 which can be higher than the RS20 depending on how it's set up and calibrated.
If the native contrast of 5000:1 is correct the max factor used by the Lumis with dimming and iris is 4-6. That assumes for black fields the maximal factor is used, which makes sense. With the maximal factor the Lumis has darker whites than the RS20. For actual intra image contrast with all kinds of material the contrast results depend largely on how badly the RS20 pollutes his blacks as soon as any > 0 IRE content enters the picture. ANSI contrast gives an indication for a special tough case. A case not representative of say, 0 IRE black space plus a couple of dozen 100 IRE stars. If the Lumis beats a good RS20 in that case too, CONGRATULATIONS! I'M IMPRESSED!
With top whites equalised, to get the black level down to RS20 levels requires a factor of 6-10 for lamp dimming, iris closing or a combination thereof. Unless you claim some stars suddenly double or triple the black level of the RS20. Then a factor of 2-3 would be enough. It still makes the whites darker on the Lumis than on the RS20. And it introduces white crush. With the same black level now the RS20 will win. It has brighter stars. No white crush. Despite a supposedly very unstable black level.
So how does the Lumis beat the RS20 in this case???
Anyway, if the Lumis comes close it's good news. The JVC can add lamp dimming and dynamic irises as well. And improve ANSI with better optics. What the SIM can do the JVC can do too in a future model. Starting with a panel that has 6-10 times more native On-Off contrast. Cool, no?

why didnt JVC manage to improve ansi contrast ? lcd and even the sony HW10 did.
"panel 6-10times more on off" is "info" ?
i remember when i tested my SIM2 D80E or even a BENQ W5000 with blu ray of GHOST RIDER.
previously i had an HD1 for 6months.
the DLP made the images so deeper, 3D ! with only 3000-5000:1 on off
so more realistic. ( remark: in a scope format so no black bars). So again
it has a lot to do with the presence of black bars or not (scope screen/LENS)
Digital2004 is offline  
post #198 of 371 Old 01-27-2009, 04:49 AM
AVS Special Member
 
Lawguy's Avatar
 
Join Date: Sep 2004
Posts: 5,705
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 0 Post(s)
Liked: 11
I wish that we could start a seperate thread discussing these MTF/ANSI/on-off issues because we are getting far off topic.

EDIT: I started an MTF thread HERE, so perhaps we should move the discussion there to avoid getting this thread further off topic.

I will try to summarize this MTF discussion in simple English that I can understand.

If I understand Mlang correctly, MTF is the true metric by which a projector's performance should be judged. This is so because the MTF measurement tells us the level of accuracy by which a source can be transmitted to the screen.

MTF can vary by APL level. Thus, to fully describe a projector's performance, we need to measure its MTF at different APL levels. If we were to measure a projector's MTF at the screen at different APL levels, we would have a precise metric of how a particular projector performs, with bright material, dark material and everything in between. If these measurements were taken, we could empirically compare the accuracy of two different projectors in transferring a source to the screen. A particular projector might have astronomical MTF with bright material but average or bad MTF with darker material or vice versa.

Furthermore, if I understand Mlang, existing CR measurements really only accidentally describe aspects of a projector's performance. When we say that a projector has high ansi or great on/off we are really describing their MTF at different APL levels.

While this makes perfect sense to me, because these measurements really only give us very limited information, it appears to have the effect of rendering the language we currently use to describe a particular projector's performance kind of obsolete and inaccurate. To make matters worse, we don't really have any viable way of measuring MTF (at least that I know of). So, we are kind of back to where we started.

One other thing that I did not appreciate (and would really like tested in some way because it seems extrememe) is the extreme importance of good convergence in a 3 chip device. Misconvergence of .75 of a pixel in the center of a screen has the effect of turning a 1080p projector into a 480p projector.

Affable Nitwit
Lawguy is offline  
post #199 of 371 Old 01-27-2009, 05:02 AM
AVS Club Gold
 
Art Sonneborn's Avatar
 
Join Date: Jun 2000
Location: Battle Creek,MI USA
Posts: 22,293
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 6 Post(s)
Liked: 44
Quote:
Originally Posted by Mark Petersen View Post

Yes I agree that on/off has little meaning with dynamic technologies and this is something that I've been preaching for quite awhile. As an example, someone could easily build a projector with 1000:1 native on/off and use an iris to fully close the light path and achieve infinite on/off. But no real image improvement is obtained other than full black outs. Seeing what happens from low APL through ANSI is very useful in my opinion as is seeing what happens at lower IRE's when dynamic technologies are employed.

I know in your contrast thread it appeared that there was an issue comparing dynamic contrast with native that translated well into on screen performance ?

Art

My HT


iRule rules my theater
 

"If she's amazing she won't be easy,if she's easy she won't be amazing"

 

Bob Marley

Art Sonneborn is offline  
post #200 of 371 Old 01-27-2009, 05:07 AM
AVS Club Gold
 
Art Sonneborn's Avatar
 
Join Date: Jun 2000
Location: Battle Creek,MI USA
Posts: 22,293
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 6 Post(s)
Liked: 44
Quote:
Originally Posted by Lawguy View Post

Furthermore, if I understand Mlang, existing CR measurements really only accidentally describe aspects of a projector's performance. When we say that a projector has high ansi or great on/off we are really describing their MTF at different APL levels.

While this makes perfect sense to me, because these measurements really only give us very limited information, it appears to have the effect of rendering the language we currently use to describe a particular projector's performance kind of obsolete and inaccurate. To make matters worse, we don't really have any viable way of measuring MTF (at least that I know of). So, we are kind of back to where we started.

Yes, and therefore having no MTF data makes it all but usless in projection discussions here.

I can tell you that the MTF of the DCI units will be touted to be higher than the consumer units but ,as I said previously ,I bet head to head when we are looking at nearly double the ANSI contrast and nearly an order of magnitude increase in sequential contrast that most will find devices like the Lumis to produce a better image.

Art

My HT


iRule rules my theater
 

"If she's amazing she won't be easy,if she's easy she won't be amazing"

 

Bob Marley

Art Sonneborn is offline  
post #201 of 371 Old 01-27-2009, 05:17 AM
AVS Special Member
 
Lawguy's Avatar
 
Join Date: Sep 2004
Posts: 5,705
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 0 Post(s)
Liked: 11
Quote:
Originally Posted by Art Sonneborn View Post

I can tell you that the MTF of the DCI units will be touted to be higher . . .

But DCI units will have lower MTF at lower APLs. This just seems to make sense, doesn't it?

As far as not having any way to measure MTF, Mlang referred to some $800 device in another thread.

Affable Nitwit
Lawguy is offline  
post #202 of 371 Old 01-27-2009, 05:33 AM
AVS Special Member
 
odyssey's Avatar
 
Join Date: Jan 2001
Posts: 1,088
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 0 Post(s)
Liked: 10
MTF has nothing to do with APL and sequential CR is still needed and is important. MTF has to do with spatial frequency, which is nothing more than detail, starting with the coarsest and going to the finest. It shows the quality of the resolution at all levels of detail by showing the contrast of that detail.

Again, there is no unexplained mystery about the seeming discrepancy of the Pioneer logo images and the sequential CR measurements of the Lumis and the RS-20. It's obvious that you can't use just the center measurement for the RS-20, because it's not representative of the image. If you changed the measurement method for both projectors and measured the entire image, the resulting contrast ratios would correlate well with the visual images.
odyssey is online now  
post #203 of 371 Old 01-27-2009, 05:36 AM - Thread Starter
AVS Special Member
 
W.Mayer's Avatar
 
Join Date: Mar 2002
Location: Germany
Posts: 1,781
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 0 Post(s)
Liked: 12
Quote:
Originally Posted by mhafner View Post

You hit the nail on the head Wolfgang. You tried to convince me years ago that projectors with 2000:1 or 3000:1 On-Off and no dynamic iris or dynamic black had no (obvious) haze problems, were good enough etc.. When I knew it was simply not true, based on watching such projectors in my room. Now we have just another round of tech improvements and the claims are the same as ever. Deja vu all over.

i long long time ago give up to convince you as i guess no one can do it.
so the opposite is true.

you are not open for news that for some reason you not like or not understand or not belief or whatever.


you are keeping a crt pr. longer than anybody else and till you got your
jvc you always sad the all digital pr. are not good at all.

now as you have a jvc lcos you reaktion is now that you are against
all dlp pr.or at least to this new lumis 3chip dlp.

but at least with one i can agree with you
"""deja vu all over""".
W.Mayer is offline  
post #204 of 371 Old 01-27-2009, 05:38 AM
AVS Special Member
 
odyssey's Avatar
 
Join Date: Jan 2001
Posts: 1,088
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 0 Post(s)
Liked: 10
Quote:
Originally Posted by Art Sonneborn View Post

Yes, and therefore having no MTF data makes it all but usless in projection discussions here.

I can tell you that the MTF of the DCI units will be touted to be higher than the consumer units but ,as I said previously ,I bet head to head when we are looking at nearly double the ANSI contrast and nearly an order of magnitude increase in sequential contrast that most will find devices like the Lumis to produce a better image.

Art

You seem to have a continuing problem with simple math. Going from 5000:1 to 18,000:1 is nearly an order of magnitude? Going from about 850:1 to 1000:1 is nearly double? You also seem to have already decided on the outcome that best suits your needs.
odyssey is online now  
post #205 of 371 Old 01-27-2009, 05:41 AM
AVS Special Member
 
coldmachine's Avatar
 
Join Date: Nov 2006
Location: Waiting in the weeds
Posts: 5,807
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 0 Post(s)
Liked: 11
Quote:
Originally Posted by Lawguy View Post

If I understand Mlang correctly, MTF is the true metric by which a projector's performance should be judged.

You seem to be putting a lot of faith in a source that I feel, based on past performance, doesn't warrant it.

--------------------------------------------
"Wow, do you think you are Adonis"...... "Baby, I'm not A-donis, I'm THE-donis"
coldmachine is offline  
post #206 of 371 Old 01-27-2009, 05:48 AM
AVS Special Member
 
coldmachine's Avatar
 
Join Date: Nov 2006
Location: Waiting in the weeds
Posts: 5,807
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 0 Post(s)
Liked: 11
Quote:
Originally Posted by Lawguy View Post

One other thing that I did not appreciate (and would really like tested in some way because it seems extrememe) is the extreme importance of good convergence in a 3 chip device. Misconvergence of .75 of a pixel in the center of a screen has the effect of turning a 1080p projector into a 480p projector.

could you provide a reference for that, or a source.

I simply cant accept that, especially after testing the MC adjustment on the RS20 and the Lumis to induce errors, and looking at the results. Neither unit suffered that sort of catastrophic performance reduction.

--------------------------------------------
"Wow, do you think you are Adonis"...... "Baby, I'm not A-donis, I'm THE-donis"
coldmachine is offline  
post #207 of 371 Old 01-27-2009, 05:53 AM
AVS Club Gold
 
Art Sonneborn's Avatar
 
Join Date: Jun 2000
Location: Battle Creek,MI USA
Posts: 22,293
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 6 Post(s)
Liked: 44
Quote:
Originally Posted by odyssey View Post

You seem to have a continuing problem with simple math. Going from 5000:1 to 18,000:1 is nearly an order of magnitude? Going from about 850:1 to 1000:1 is nearly double? You also seem to have already decided on the outcome that best suits your needs.


No, I don't,you continue to have a problem saying a regular Barco has 5000:1 which it does not. Of course Peters modded Barco has something around that the last I heard (or was it 3000:1?) but his Barco is sacrificing ANSI to get that the Lumis does not.

I'm looking at this thread at it's getting repeated that the Lumis likely has higher than 20,000:1 .

What is the sequential contrast of your DCI unit ?

Art

My HT


iRule rules my theater
 

"If she's amazing she won't be easy,if she's easy she won't be amazing"

 

Bob Marley

Art Sonneborn is offline  
post #208 of 371 Old 01-27-2009, 06:01 AM
AVS Special Member
 
coldmachine's Avatar
 
Join Date: Nov 2006
Location: Waiting in the weeds
Posts: 5,807
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 0 Post(s)
Liked: 11
Quote:
Originally Posted by Art Sonneborn View Post

I'm looking at this thread at it's getting repeated that the Lumis likely has higher than 20,000:1 .

It does indeed. There are 2 units measured at 25k and 27k.

Wolfgangas iris is inoperative, and its designed to fail open, so the 18k would likely be 20k. His unit, as explained by the engineer, is 25% lower on CR as its not been optimized like the serial units. That would take him to 25k too.


As an aside, I directly compared the Lumis to a Barco DP-1500. The same one Ive quoted here a number of times. The Lumis is simply superior in term of overall PQ. The Barco is brighter and has better optics, but that's it. Translated to real world viewing, the difference is very obvious.

I have found DCI units, other than huge light, have now fallen behind the very best domestic units. They are still clearly superior to budget units, but that will change too as the cheaper DLPs improve.

--------------------------------------------
"Wow, do you think you are Adonis"...... "Baby, I'm not A-donis, I'm THE-donis"
coldmachine is offline  
post #209 of 371 Old 01-27-2009, 06:07 AM
AVS Special Member
 
Lawguy's Avatar
 
Join Date: Sep 2004
Posts: 5,705
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 0 Post(s)
Liked: 11
Quote:
Originally Posted by coldmachine View Post

could you provide a reference for that, or a source.

I simply cant accept that, especially after testing the MC adjustment on the RS20 and the Lumis to induce errors. and looking at the results.

Here.

I also find this kind of hard to believe.

Affable Nitwit
Lawguy is offline  
post #210 of 371 Old 01-27-2009, 06:25 AM
AVS Special Member
 
coldmachine's Avatar
 
Join Date: Nov 2006
Location: Waiting in the weeds
Posts: 5,807
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 0 Post(s)
Liked: 11
Quote:
Originally Posted by Lawguy View Post

Here.

I also find this kind of hard to believe.

Ah. I doesn't actually say it will be a 480 machine. Hes saying in an area where the diverged colors would be needed to combine, there will be a reduction. Thats a different thing. Thats already well understood.

This is the very reason the PJ manufacturers attempt align for max performance at screen center. At fairly close viewing distances a good bit of MC is undetectable yet single pixel lines are still resolved.

Try it on your RS20. You'll find it doesn't have as dramatic effect as you would think. It certainly didn't on the RS20 i tried.

Obviously any MC is detrimental, but it does get overplayed sometimes. No different to focusing on any single performance parameter as being of disproportionate importance, like the current BS we see regarding CR. Fortunately real world viewing proves much of these hypotheses to be bogus.

--------------------------------------------
"Wow, do you think you are Adonis"...... "Baby, I'm not A-donis, I'm THE-donis"
coldmachine is offline  
Reply Ultra Hi-End HT Gear ($20,000+)

User Tag List

Thread Tools
Show Printable Version Show Printable Version
Email this Page Email this Page


Forum Jump: 

Posting Rules  
You may not post new threads
You may not post replies
You may not post attachments
You may not edit your posts

BB code is On
Smilies are On
[IMG] code is On
HTML code is Off
Trackbacks are Off
Pingbacks are Off
Refbacks are Off