OPTOMA HD91 FULL LED DLP full hd 2D 3D Ready september 2013 - Page 2

I'm talking about native contrast with an iris fully open. The Planar PD8150 with it's dynamic iris in use is higher than the Marantz or the Sharp. If you were to have the iris fully open it should be more in line with the white paper. The native on/off with the iris fully open should be similar to the Planar PD8150 but it's dynamic iris can dig deeper than either manual iris. You get the best of both worlds (high brightness and low black level) while you need to choose one or the other with the Marantz or the Sharp. Currently I'm using a Marantz VP-11S1 with a High Power 2.8 Da-Lite screen so I can clamp down the iris and still get decent peak white levels. So far it seems to be a great combination. Sometimes I'll put the Planar PD8150 on low lamp mode with this screen but it's still a lot brighter than the Marantz (almost too bright).
Edited by Seegs108 - 2/1/13 at 12:18am

Well you have to compare apples to apples. Native contrast to native contrast, both the Sharp and the Marantz have significantly higher native contrast than the Planar (3000:1? with iris clamped down). If you would implement a dynamic iris in them they would have significantly higher dynamic contrasts as well.

But they don't have one and that's my point. With those two models you have to pick your poison. High peak whites with mediocre black levels or decent black levels with mediocre peak white. With the dynamic iris you're getting the best of both worlds...always.

They have basically the same native contrast when there is no iris on but the Planar still has better black levels when any of the machines use an iris. If you check out the review posted above, it measure just a hair under 12000:1. This is still quite a step up over either the Marantz (~8000:1) or the Sharp by atleast (9000:1).

These numbers obviously aren't groundbreaking compared to some of the DLP machines out there today. The one thing the Planar has over those current DLP models with better on/off numbers is an undetectable dynamic iris. I've talked about this before in another thread, but that stealth lends itself to how Planar implements it. As you can see it only boosts the contrast by a factor of 4. Some of the newer model DLP with on/off numbers measured at 30000:1 only have native contrast ratios of ~1500:1 (maybe a tad higher depending on the model) which means they have to be MUCH more aggressive and thus are normally unusable in any practical situation like during a movie. They are simply too distracting to use. I've witnessed this on a couple of the newer DLP models. Mitusbishi's HC7800D was very bad. Slow to kick on and too aggressive intra-scene (tons of pumping). The only current DLP model under $5000 I'd purchase is the Sharp XV-Z30000 because of the same reason I like the Planar. It implements it's DI in a very smart way. You only get ~9000:1 on/off but that is great for a machine running a .65" DMD.
Edited by Seegs108 - 2/1/13 at 12:46am

And what about Mitsubishi HC8000?

I haven't seen it in person, but from what I've read it's very similar to the Mitsubishi HC7800D. I wasn't all that impressed. I'm a huge DLP fan, but I'd rather own a low end JVC (X30 or X35) over Mitsubishi's current DLP offerings. For these DLPs to compete you need to be able to use it's DI and unfortunately they are just implemented awfully. I've read the DI in the 9000D (Mitsubishi's LCOS projector) has an awfully implemented dynamic iris as well. Luckily those units have high native contrast (~10000:1) and I think you can use the iris in a manual fashion with plenty of different stops so you do have an option to get better contrast if the DI is too noticeable for you. Unfortunately Mitsubishi's DLP models don't allow you to use it manually.

Segs 108

You're missing my point. I'm talking from a designs perspective with regards to how offset affects contrast. How good the planars di is doesn't come into the equation.

I'm confused by your point. What is the point you're trying to make? The Planar is just like the Marantz and the Sharp. It doesn't have any lens offset and I've already pointed out that the native contrast off all three is 3000:1 which is pretty decent even though they have a telecentric architecture which the white paper says could be better if there was offset used. Though all of these units use great lenses and have quality light paths and relay optics which is a major contributing factor to that high native contrast.

The only reason I brought up the DI is because you were claiming the other two had better on/off contrast and a lower black floor. I continued that discussion to answer someone's question.
Edited by Seegs108 - 2/1/13 at 1:47am

Seegs,

I think I got confused when the Planar was introduced into the discussion.

My point it this:

As far as I understood, according to the white paper you have to have a significant offset to get the best contrast out of a single-chip DLP projector. Having it without offset ruins the contrast. However, the projectors that have the highest contrast off all single chip DLPs don't have any offset. How does this compute? To me it seems that offset doesn't have the impact on contrast as is claimed otherwise the Marantz shouldn't be able to smack the offset DLPs over the head with much higher contrast.

I didn't read the white paper in depth but I think that you're under the impression that the difference in contrast will be HUGE when in reality there isn't much of a difference between center set lenses versus offset ones. The InFocus IN83 is a DC4 .95" DLP projector with a pretty hefty offset. 36 percent and the native contrast is still only ~3500:1. The drawback to using offsets like this is that there's usually no allowance for lens shift which makes placement a pain sometimes. A projector is much more marketable if it has a decent amount of lens shift as it can be used in a lot more spaces a projector like the InFocus IN83 can't be. Plus with a centric lens setup you get much better light uniformity. With a properly introduced iris the lower contrast can be alleviated while still keeping the benefits of a lens centric design.

If you check out the IN83/IN82 dedicated threads on this forum you'll see TONS of people complaining about the fixed offset and that was the main reason people didn't buy it. It just simply won't work for many peoples' setups.

Seegs, in you opinion how much on/off contrast will give you satisfying blacks?

I don't know, that's just the impression I got from Mark.

Read the damn paper and think. Basically the paper gives a designer the way to maximize contrast and contrast is not just on off contrast. And what do you guys know about contrast once we leave the realm of on off. A little but you can't generallly can't spout comparative numbers. this trounces that.


There are two things a designer must be concerned with. The lens, its quality, and its cost. Namely its cost. And two flexibility of placement or you won't sell a lot of machines. Obviously there are other things. But if you cab spend a lot of money on the lens, you can negate some of the effects of other contrast lowering compromises. The marantz used a very expensive lens. The planar and Samsung ditto. The Sim2s etc use expensive lenses. A way to make up for choices that lower contrast is using a DI Using a DC4 chip over a DC3, an expensive thing to do for a designer.


Don't worry. Whatever DLP you have is great. And you will find it satisfying but you will consistently be reminded by fan boys and othesr, it won't be the JVC blacks and you know deep down in your heart that you have failed miserasbly because for 2D its all about the blacks.

Read the paper. I don't think a projector has been built using a DLP to maximize the contrast over all other considerations.


Think about the projectors that require large offsets. There are cheap and use cheap lenses.

Using a good lens allows one to make contrast compromising choices.

remember the white paper is not a tell you how much. If you do this it will cost you X contrast.

Its for designers. Its not for you guys who want in their hearts to be designer and thing that they are brighter than the designers and engineers. Those that want to employ their superior intelligence to solving the puzzle, to understand. Its harsh.

You know what's harsh. A low income Marywhowho salesman from CA called me last night seeking advice on what projector to buy. Namely he could buy one of the Optomas on Ebay which had a $2500 or so MSRP for $1100 and use it on his ebay $40 screen. Big screen too with the specs giving a range of gains hovering around 1.0. The guy had no clue about offsets, He lucked pout there being about 4 ft above screen top available. Guy didn't know exact screen size. And I am suppose to know about cheap projectors.

what I am an expert on are CHEAP forum members. Not you Moo Moo. Just a statement to err increase the contrast.

I think Mark is one of the most shameless posters here. He uses sarcasm to illustrate his point so well. Mark, you almost seem miserable talking about this stuff hahaha.

Yes, DLP projectors are the worst because they can't compete with JVC in on/off contrast Didn't you know that black levels are the only thing that matter? (See I can do it too )

I don't think the above post was aimed towards me, but if it was, I agree with with you said.

You're really asking two separate questions here. On/Off doesn't determine how good black levels will be. An on/off contrast measurement is just comparing peak white 100IRE versus how little light is hitting the screen at 0IIRE. You can have a 1000000:1 on/off and still have a crappy black level.

But to answer the question you're looking for, most machines with good black levels typically have contrast ratios above a measured 8000:1 on/off. What I usually look for in a review is it's black level before looking at on/off numbers. What I want to see is a number below .003 ftl for a black level. I think anything below that is good enough and satisfying for me. I just hope that peak white is at least 16 ftl when a projector is able to get a black level that low. With DLP machines that's usually hard to come by and there aren't a whole lot under $10000 that can do this.

That is what I look for. That doesn't mean something better or worse won't satisfy you. There are too many people who read into numbers and argue about trivial differences. I'm sure I've been guilty of that at least a few times.

The fact of the matter is that most of the products talked about on the forums these days will put out a very satisfying image. Unfortunately, the technology has matured in a way that to get the most satisfying product we must have a never ending debate about the trivial advantage on projector A and how its better than than trivial advantage on projector B. Black levels mean the world to some people. I'm perfectly happy with what a good DLP machine can offer even though it's far behind many LCOS and a few LCD models.
Edited by Seegs108 - 2/1/13 at 7:56am

Amen,

a short question:rolleyes:,

can we talk about the optoma, may someday again !

Greetings

][_,{[]}][_, Awesome post! Yeah, that's what I was asking. For 1000000:1 display to have not satisfying blacks it's peak luminance level should be 3000 ftL or higher BTW, Panasonic's 65VT50 has 0.002 ftL MLL: http://www.*********************/images/pdf/Shootout%202012%20contrast%20ratio.pdf But I'm not sure if it was measured on 0 IRE or 1 IRE.

Determining how pleasing intrascene contrast is in dark scenes is very much part of Native On/Off. BTW, the problem with those newer .65" sharps has been focus uniformity so I have heard, this has been an ongoing issue. I appreciate Seeg's input on the matter, but I politely disagree that on/off ratios don't determine black levels, of course they do with all else equalized (calibration, fL, clipping, brightness, black level controls set correctly, etc...).

When we discuss projectors in here for black levels, it is automatically assumed you do what is necessary to get the peak fL balanced and that you set the black level and white peak correctly in a completely dark room, so there is no clipping. It's not that you have to, it's just that FOR comparison sakes we don't look at black floor, we look at the ratio (because the ratio is what is most meaningful, otherwise you are just looking at numbers that are variable in every setup). That's what ratios are for, for EQUALIZING the comparison numbers. You can compare the native on/off of a PJ without the IRIS to the dynamic on/off of an IRIS, but this is not a 1:1 comparison because of INTRASCENE contrast.

I adamantly will say that native on/off is still important EVEN with an IRIS, even though this is not because native on/off determines how dark the black levels go (but actually it does this as well if there is no IRIS, though some dimming without an IRIS may occur delayed on some PJ's). That said, the best indicator of how intrascene contrast comes out in dark scenes is NATIVE on/off not DYNAMIC. EVEN though the on/off ratio is just a full white field measured, and then the full black field measured, a side effect of the measurement is that it is the BEST indicator of how intrascene contrast will look in dark scenes (even though this wasn't the original intention of this ratio). Projectors with LOWER native on/off even with IRIS's cannot produce the same INTRASCENE white to black contrast in dark scenes even in the upper-peaks even with only a 4x multiplier. The reason people get confused on thinking an IRIS that can do 25,000:1 is as good as a projector with Native On/off that does 25,000:1 is because they are thinking in terms that the intrascene contrast would be the same. It won't though, the JVC will have brighter whites within that starfield because it has higher intrascene contrast in dark scenes because of higher native on/off, and it doesn't matter what the IRIS + Dynamic Gamma of a PJ does to counteract this, it still cannot fully compensate. This is because DI's don't change intrascene contrast much in most cases, though there is some affect depending on the room (because intrascene contrast goes up as reflections off the walls decrease, hence as the PJ goes dimmer). So technically even though a DI doesn't change contrast, it sort of does in a way DEPENDING on the room.

The point of measuring without the IRIS isn't to see how DARK the projector can go on a plain black screen, but coincidentally due to the way our machines work, the NATIVE on/off is also a very decent indicator of how that intrascene magic is going to look overall in the darker scenes.

IMO, 1/2 of native on/off is as valuable as double the dynamic on/off even versus the most PERFECT of IRIS's like the Sony, and with all things equalized and speaking of 4x multipliers. In reality, you can set whatever multiplier you want, because some of these IRIS's are too visible even with a 2.5x multiplier that you won't be much further distracted. It is more so the real smooth IRIS's that try hit those perfect multipliers, because a poorly coded IRIS even at the perfect multiplier is still a POORLY coded IRIS.
Edited by coderguy - 2/4/13 at 7:52am

A good DLP system, even one that only reaches 3000-4000:1 native on/off will still reach an ANSI contrast ratio (intrascene contrast) typically twice as high as the best JVC. So even though a lot of what you've just said should make sense, the numbers don't agree with you.

I also think you're missing the point of what I was trying to say about black levels. Considering that the black level reading is apart of achieving an on/off number, that doesn't HAVE to mean that if the number is high, black levels will always be low. That was my point. In real life however this typically ISN'T the case. I was just stating a fact. There can a light cannon with a massive amount of light coming out of it but it has a poor black level. The on/off number could be very high considering the brightness which would give a misleading impression of what the black level is to the average person who thinks because the on/off number is high that HAS to mean it's black levels are great. Does this happen often? No, but it can. That's all I was saying.
Edited by Seegs108 - 2/4/13 at 8:29am

ANSI contrast isn't the same as intrascene contrast.

Intrascene contrast is the difference between the darkest and brightest pixel in an ACTUAL video scene (or brightest to darkest cluster).
Some have incorrectly over the years (even some experts) stated that ANSI = intrascene, but they are incorrect and it's very easy to prove it.

ANSI contrast is only the intrascene contrast on that specific test pattern and is not a good indicator of the INTRASCENE contrast ramp in dark scenes. We discussed the scientific nature of this in another thread, we went into far detail and argued it all out. The intrascene contrast ramp is a set of derivative measurments on the brightest and darkest pixels in a scene relating how other colors and dynamic gamma responses are indicated when the projector is displaying ABSOLUTE black compared to how much % of the screen is brighter and how bright that "brighter is" (or even just the brightest and darkest pixels in a given scene).

It just so happens that in actual dark scenes (not the checkerboard pattern), the NATIVE ON/OFF measurement for front projectors is actually a BIGGER indicator of instrascene contrast in dark scenes than ANSI contrast. You are mistaken here my friend, but a lot of people make this mistake.

The JVC will blow away the DLP's intrsacene contrast on average starfields, fact not fiction. Unless you put something bright on the screen like the sun or a bright planet or large cluster of stars, the Native On/Off has a more pleasing look. This is one area the JVC cannot be matched, and the IRIS does not help with this, the IRIS only helps darkening the overall image to try to compensate for the LACK of Native on/off.

And why do you think some people test for ANSI contrast? This figure is a great indicator of what intrascene contrast will be like. That's why they do it. The star field scenes are not something that show off good intrascene contrast. They show off good black levels considering that 98% of the screen is pure black.

I can't do this argument again, we did it in the other thread (search JVC vs Sony ANSI in the forum), you will find the thread and read all our answers.

It is much easier to see the difference for intrascene contrast in darker scenes than brighter scenes. If I play the game X3: Albion Prelude (Which is all space scenes), I can see a huge immediate difference in contrast. If I play any variety of DLP's that have different ANSI, I will see very little to ZERO benefit in this dark scene, and probably also cannot tell in the brighter scenes.

Why do people test for ANSI, because they don't know better
Normal people can't measure ANSI correctly between two projectors, it's almost impossible, it requires carbon-like black-coated cone devices over the meter and $10,000 equipment.
The ANSI numbers you see online are hilarious jokes.

The ANSI pattern is not a good indicator of intrascene contrast in average scenes, it is only an extreme indicator. The reason it works only in extremes is because the pattern itself is an extreme (Checkerboard of abs black and abs white).
The reason native on/off is a good indicator even though it is also an extreme pattern, is not an intuitive reason, but it's because the 2-frame black/white response actually closer mimiicks intrascene contrast in dark scenes, than ANSI mimmicks intrascene contrast in bright scenes.

You don't have to argue because you aren't going to change my mind on the subject. So let's just agree to disagree.

OK, but it's been proven the readings are false, including in our math from that thread mentioned.

There has to be something really bright (near ABS white) on the screen and that takes up a significant portion (usually greater than 10% to 25% of the screen) for ANSI to even matter in any scene that is not really bright. If you speckle stars on the screen or have too much darkish objects, what matters is native on/off to get an idea of how the end-result intrascene contrast will come out.

The checkerboard is 50% white, 50% black as a proportion of total screen area, there are VERY few scenes in real life in movies that match this. Even moving the checkerboards to a different pattern starts to give LCOS projectors higher contrast readings, ANSI is a faulty pattern that matters much more in very very bright scenes.
Edited by coderguy - 2/4/13 at 9:17am

Intrascene contrast isn't static and it's going to look different depending on the scene. I'm not saying that high ANSI contrast means that black levels will be amazing. In certain scenes it can give the impression that black levels are very good. Like you said it doesn't matter on darker scenes but it does matter on some scenes where there is a great amount contrast. While ANSI truly shines on that test pattern there is plenty of real world content that can show off great ANSI contrast.

I'm not an expert on the subject and don't claim to be. What I usually say in posts reflects real world viewing from what I've personally seen on the projectors I've owned. Two of the units I own have particularly high ANSI contrast and from what I've seen blow away all three of the JVC units I've had on all but the darkest of scenes. Maybe I just see things in an odd way? I don't know, but from my eyes a high ANSI contrast seems to make a world of difference and in all but very dark scenes the intrascene contrast seems to be higher than the JVCs. Maybe I'm wrong but not to my eyes.

http://home.roadrunner.com/~res18h39/intrascene.htm

Another issue is that the projector's with higher ANSI are generally the ones with better LENS's, which means you will see that as well. DLP's also have higher MTF and a wider pixel gap fill, so all that plays into what we see. ANSI is part of it I think, but it's hard to say how much, it's not as much on some scenes as people would like to believe. I am not ALL-OUT JVC FAN, but I do know there are some scenes the JVC just eats up other PJ's on. Too bad it has limitations.

The calculator above is not accurate (because SCREEN ANSI will be much less than PJ ANSI), but it's interesting none-the-less.
Edited by coderguy - 2/4/13 at 9:19am

FWIW, this is the best info/thread I've ever seen, and the most empirical data I've seen on the subject of contrast with front projectors (Credit to Mark Peterson for the awesome work):
http://www.avsforum.com/t/852467/avs-contrast-thread-now-with-dynamic-contrast-results

Now that said, there must be something with the measurements that's not captured. In the $20k+ forum, Wolfgang posted pictures of a Sim2 Lumis next to a JVC RS20, side by side in the same room with the peak white levels equalized. The results there are rather interesting given he measured the RS20 at 26000:1 on/off contrast and the Lumis at only 18000:1.
http://www.avsforum.com/t/1112115/sim2-lumis-3-chip-dlp-little-test/90#post_15650586

It seems there's more left to understand WRT contrast.

Yes, the lens and on my Marantz (which gets close to 1000:1 ANSI) has amazing lens coatings on top of the quality lens to yield great results. Here is a typical JVC (X3 or X30 for example) versus real numbers measured for my LED projector:



As you can see everything but the star field is significantly higher. Just like I was saying before.

The real experts on contrast are actually in the field of radiology where scientists are trying to further the techniques, that is where you will find actual hard-scientific fact that has no subjectivity relating to contrast, however the papers are so boring to read they will put you to sleep (and I've tried to read a couple, but had to pass). It's crazy how far these radiology papers go into the subject. It goes far beyond what any pure video person has ever written on it.

Again, that calculator is not accurate because you are entering the projector's ANSI from the lens measurement instead of the screen measurement. When you take an ANSI reading from the lens, it's corrupted because the light bleeds across the pattern giving logarithmically higher readings than actually how it plays out to the screen itself.

ANSI contrast has more intrascene ramp affect on enclosed sets than it is on front-projectors, though the room is the most important with FP;s.

If you had a white unity gain screen wouldn't the bad effects translate equally to both projectors? Meaning, wouldn't they still appear to have the same amount of difference when viewing even if the numbers didn't stay that high?

Well:

The room itself even coated in the most pitch black materials is still reflecting light, hence our screens do not have light isolation like back-lit TV's and some other devices do. Back-lit TV's can use light rejection coatings rejecting the light pollution outward away from your eyes, and creating a sort of "black hole" effect. That is one reason on back-lit devices the contrast is so much higher looking in ambient light. Also, our screens cannot use the same type of reflective + anti-reflective techniques that TV's do because the light is bouncing off the screen directly in front of your eyes, instead of from the back. The light is sprayed onto the screen without isolation and that is where a large amount of ANSI is lost, as the room reflects and mixes that light back onto the screen. The white checkerboard pattern bleeds over the blacks more so on the screen than taken from the "air" perse or off the LENS. It's a doubling effect, we get a loss of ANSI from the light path initially, but the ANSI is still much higher before it hits the screen because the screen pollutes itself (even with no reflection on the wall). When showing a checkerboard pattern, the white shadows over the black on the screen raising the black levels more than a back-lit screen.

Some TV's or LCD monitors with relatively low on/off BUT very HIGH ANSI will be very competitive to the black levels of front projectors in starfields, because the intrascene ramp of a back-lit device is more extreme (it does not take nearly as much bright white to start evening out the contrast game, this is because ANSI is better preserved off the screen on back-lit devices).

IMHO, the biggest confusion of ANSI contrast in front-projectors is because people are equating it to how ANSI works on TV's and back-lit devices, and it's two very different things. The end result would be the same, except that with projectors the ANSI is toasted at the screen, no matter how high ANSI you start with at the lens.

That simulation I posted would be much more accurate if using it for TV's, than it would for projectors.
Edited by coderguy - 2/4/13 at 10:00am