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

Great article on contrast, but as far as the Wolfgang thing though:

Well, the main problem with his test is that 18,000:1 and 26,000:1, only about a 45% increase in native contrast isn't very much to the eyes when we are already above 15,000:1. So if you have a high enough spread of ANSI and native on/off that close, then yah we can see some better blacks when there is that much light in the room (because the two projector's are polluting each other's light source in his PJ A on / PJ B on testing method. With light pollution from both projectors the on/off advantage will be lessened, so that test isn't really the best one. The projectors need to be compared by taking a spot measurment with an accurate light meter in A then B, not A+B at the same time.

All this said, I am not saying we cannot see ANSI contrast between a DLP and a JVC, I am adamant that we can, but I'm just saying that after that, like on a Sony vs. most DLP's, you're not going to see it much according to my tests. The JVC is really the only projector to worry about lower ANSI on, and even then it still doesn't affect all that much scenes in movies, more so in documentaries and day-light stuff.
Edited by coderguy - 2/4/13 at 10:30am

More interesting is Lumis vs the RS20 comparison. Crazy to see what high on/off AND high ANSI contrast is like. While that may be true of the Lumis there is a very small amount of projectors that can perform close to that performance of high on/off and high ansi contrast. If I were to change the LED modulation to the next step I think my NuVision would perform very similar. It should achieve close to 22000:1 on/off and keep the high ansi at the same time. But alas, I'm still waiting for Vivitek to send it back to me...

I don't think the Lumis vs. RS 20 comparison is valid at all, the on/off ratios with two projectors in a room are affected by not only the gamma response and level of the proejctor opposing (which then causes an uneven light distribution for on/off pollution). He has introduced a lot of variables that shouldn't be there.

On/Off ratio remains the same in a darkened room with no other polluting light sources, in this case the two projectors are polluting each other. As noted, it should have been spot measured individually, instead of compared next to each other. If one projector has a gamma of 1.8 and the one to the right has a gamma of 1.9, and we are displaying a very very dark scene, then suddenly one projector is receiving far more light pollution than the other.

Now it wouldn't matter if comparing a JVC to a 3000:1 on/off DLP, but 18,000:1 and 26,000:1 are too close that the pollution corrupts / ruins the test.

I didn't read the whole article but I would image both were calibrated. Wasn't it noted was that they were lumen matched and the black levels (with DB on) and ansi were higher on the lumis which means that the JVC would be polluting the Lumis more.

The projector with the higher native on/off would receive more pollution,not less, especially as it appeared to the eye. This is not debatable. The Lumis is close enough in on/off to the JVC that it could beat the JVC with the ANSI boost, but it wouldn't be as profound as shown by that test.

This is very easy for me to test, if I turn my Benq and JVC on at the same time in split-screen on my screen, even without overlapping the images, the BENQ starts quickly catching up to the JVC in blacks.

You can measure the peak white fL with 100 IRE on both projectors, but the Benq will still be brighter and outputting more light, and therefore when you increase additive light over the Benq, you are making the whites brighter while raising the black floor without even re-compensating the black floor calibration of the projectors in doing so (test doesn't make sense).
Edited by coderguy - 2/4/13 at 11:07am

For instance, in one shot even on two matched calibrated projectors:

Intrascene White Peak = 5.6 fL of JVC, while black level = 0.01 fL
Intrascene White Peak of Benq = 7.0 fL while black level = 0.2 fL .... This is just an example, this can happen even when both projectors are trying to show darker scenes.

Now what happens if we set them side-by-side, well you raise the black floor very near the BENQ's black floor (because they are side-by-side after all), but the Benq retains the higher white peak, and ends up with the better on/off contrast. This is just an example and doesn't really apply to a Benq vs. JVC because they are too far apart in contrast, but it surely would when it's that close as the Lumis and JVC are:

I'm confused. Wolfgang points out that they are lumen matched very closely. Wouldn't that mean 100IRE on both is the same fL reading and then all subsequent lower IRE except very close to 0 IRE or does that change? It looked like the black level was deeper on the lumis which would mean the JVC would have a more detrimental effect on the Lumis? Was the on/off for the lumis with or without DB on? I think that's the key thing I'm missing out on.

100 IRE does not guarantee a match of the IRE levels in the intrascene contrast when you add either abs white or abs black (doesn't even have to be both). You have also polluted the black floor of the JVC by the higher black floor of the lumis, and given the lumis a higher white peak in intrascene contrast because the lumis had the higher black floor which would generally equate to higher white peak in intrascene contrast if both are at the same peak fL at 100 IRE.

Realistically the JVC may look far better in intrascene contrast compared to one projector in a dark scene even with darker whites (hence the JVC may appear to have brighter whites), when really the JVC just had darker blacks + darker whites, it's all relative and depends on the intra-scene. Taint the black floor enough and the PJ with brighter intrascene white will win, even though the other PJ would win by itself.

Like I said if the spread were great and 3,000:1 vs 25,000:1, the test would work, but not with a 45% spread.
Edited by coderguy - 2/4/13 at 11:26am

Only in scenes with some [edit] bright [edit] objects; otherwise the intrascene CR will depend on the on/off CR.
Edited by noah katz - 2/5/13 at 11:02am

Coderguy, you are saying that Wolfgang's test isn't valid because the Lumis is polluting the high on/off of the JVC. But surely, if we are talking in practical terms, where most peoples theaters do not have absolutely perfect light control - wouldn't the real world results be more improtant than numbers? If we trust Wolfgang's camera and his personal impressions, in an 'imperfect' enviornment the Lumis produces better perceived black levels. Isn't that what realy counts in the end?

Wolfgang's comparison test was valid. In fact, I believe Coldmachine also had a JVC and a Lumis Host, and came to the same conclusion.

If actually comparing two projectors side by side, lumen matched, on a huge screen ( isn't Wolfgang's screen 20' wide ) with the same signal feed, isn't valid, I don't know what would be.

I'm not saying the JVC would look better than the LUMIS, I'm just saying the test shown is a bit ridiculous (especially the fact they took pictures of it to show the comparison and people are deriving results from looking at that screenshot), if they say the LUMIS looks better I'll give them the benefit of the doubt. There is no way the camera could have captured the visible difference, we would need reference level monitors and a perfect photo.

The on/off is too close 18,000:1 vs. 26,000:1 is barely a visible difference even in an A/B, when you put them side-by-side you corrupted the test.
Again, on/off is not polluted from reflections of the wall but from external light sources. Having both PJ's on at the same time is absolutely major pollution.

Essentially what he is saying is
1) With blown ANSI, the Lumis looks bettter.
2) With raised on/off, the Lumis looks better.

The ANSI contrast is majorly blown out with two PJ's projecting side-by-side, and the black floor is raised. This test works great for sharpness and color and various things, just not for contrast. Like I said, it works when projectors are farther apart in contrast because the difference is so big that the lost contrast does not favor one side enough, but not when they are so close.
Edited by coderguy - 2/5/13 at 5:21am

It's completely valid, if you understand what it's trying to compare and what it isn't.

Starting with what's not valid, the measured contrast ratios of either machine in this comparison are not valid for anything other than comparison in this test. As noted each machine will pollute the other, thus the black levels for both will be inaccurate (higher) than either machine alone. Now that said...

What is valid is the relative comparison of their performance to each other. As you noted later, the projector with the higher on/off would be polluted more. This would cause the two machines to appear closer in performance than either on their own. Now what this would not do, nor explain, is why the black floor of the JVC, with it's higher on/off (and matched peak white levels) would appear higher than that of the Lumis. No matter how much more light pollution a given projector A may be producing, it can not make the black level of a projector B higher than itself, a given projector will always pollute itself more than it's neighbor since pollution drops as you move away from the image.


But that doesn't explain how the Lumis would have lower black level in that very low APL scene. Unless ANSI is more important than some want to believe.

But I'm sure the JVC is always better. The BenQ, no matter how much worse than the JVC simply can not increase the JVC's black level more than it's own.

I don't understand this comment at all. If you put a 100IRE screen on both projectors and calibrate them both (via IRIS/lamp/ND filters/etc) to be the same ftL, they are the same ftL and are putting the same maximum amount of light on the screen*.

Well, one small exception, JVCs reflect light back into their lamp assembly so a small 100 IRE patch is actually brighter on a JVC than a full-screen 100 IRE field (the "recycled" light from the black area boosts the white a bit).

No you don't, not for this comparison. This is via a single image with both projectors captured at the same instant. It's completely valid to use that to compare specific attributes (like black level) between the two projectors as they were photographed. You don't need reference monitors or perfect photos for this. Since both projectors are on the same screen at the same time, and one photograph was taken, all you need to do is expose it "correctly" to capture the difference in black level between the two and it's a completely valid comparison.

Now obviously you wouldn't want to evaluate color performance or anything like that, but for the limited comparison of black level between the two projectors, it's just fine.

It's pollution, but again, there's no way for the the projector with "worse" contrast (Lumis) to raise the black level of the projector with "better" contrast (the RS20) above it's own black floor.

But again, there's no way for the Lumis to raise the JVC's black floor above it's own. If the Lumis were putting out so much pollution as to raise the JVCs black floor as high as it is in the photos, the Lumis's black level would have to be even higher still. And beyond that, it couldn't give the RS20's blacks the blueish tinge they seem to have that the Lumis itself doesn't have.

It seems like some just don't like what the photos show pretty clearly IMO.

There seems to be a lot to contrast we don't understand or know how to measure or something.

Can someone point me to wolfgangs test, I'm interested in this?

^^^ post 56 of this thread has a link.

So this will be released in Sept or when?

I'm afraid that when we see what the true calibrated lumens are on this projector, it too will be not bright enough. I sure wish that better brighter LED's were available, That would have been / will be more revolutionary than either 3D or 4K IMO. Let lamps fade away - forever!

Quite agree.

This is very exciting, for me.

All previous LED projectors have been either (a) over 10k, or (b) not true home cinema jobs.

The model number (HD91) suggests a dedicated home cinema model, and the manufacturer (Optoma) suggests sub-£3k.

I joked earlier about the HD82. Actually, this does looks like the H82, but in turn that looks like the HD87. From what I've seen of the HD87, I think it's as good a projector as I'd want, other than very occasional RBE problems the lamp dimming over time. I think I see RBE little enough that any reduction at al will make it about nil.

I've seen better projectors, but nothing better enough for me to warrant spending the extra. I can dream of a Sim2, but it's never going to happen this side of a lottery win. From what I've seen of it, I could live with the HD87. Forever!

I've been holding off upgrading my 720p HD73 until a proper, affordable LED model turned up. I nearly went for an HD33 as a stop gap. But now I think I'll wait to see what the autumn brings.

In that spirit, if anyone gets any more news on this, please let us know.

Steve W

For the record, the 83 and 87 had claimed lumens of 1600 and 1700 respectively, and were generally considered to be extremely bright. I tend to find that manufacturers exagerate equally across all models.

And, of course, that would have dropped below 1300 (even an equivalent 'claimed 1300') pretty quickly; the 91, being an LED, should remain fairly solid.

It would appear that most lamps lose about 20% of their brightness fairly quickly, then slow down. Indeed, I've heard of some people who won't have their projector calibrated until it's reached that stage. 20% off 1600 & 1700 lumens leaves 1280 & 1360 lumens respectively. In other words, if Optoma are exagerating the light output by a similar amount, this model will be about the same brightness as their previous bright models were after a few hundred hours. That's not to mention the consistent view that LED projectors look brighter than they measure.

If this is the case, that'd do me.

Steve W
Edited by Pecker - 2/5/13 at 9:43am

Maybe. I just found my receipt for my Optoma H79 from Feb. 2005. It was $ 6,200.00.


It will be interesting to see this projector.

edited to add a crucial missing word

Please note I was quoting £ not $.

I'm not from the colonies.

I don’t think any of Optoma’s ‘HD’/Themescene projectors have had a rrp of over £2,999 for a long time, but I might be wrong.

If I remember rightly, at launch the HD82 was £2,999 , the HD83 was £2,400 and the HD87 was £2,700. I think maybe the HD86 was £3,400 way back in 2009.

A few times I’ve heard people saying the LED light source costs a few hundred dollars/pounds more than a standard lamp, so conceivably this could push £3,500, but that’d still be a different league to the £10,000+ LED home cinema projectors we’ve seen so far.

I’m going to have a guess this will launch at £2,999, possibly just over £3,000 with more exotic lens options.

Steve W

Does Optoma list this projector LED light source lifetime (half lumens lifetime)? Anyone knows the number?

Still six months until launch.

Does anyone have any news, or even rumours?

If you don't, just feel free to make stuff up to keep me going.

Steve W

Okay.

Man, this is going to be the best projector ever. Keep going. =)

I've heard it is made of 95% post-consumer content, is solar powered and fully compostable.

The compostable part is critical - it will help dispose of the unit when the next gazingus pin comes to market.

Will HD91 have a demux function like the HD33, where you can select one to display the left channel and a second unit displaying the right for a stereoscopic dual-passive setup?