[chrisreeves]

Andrew Robinson claims single chip DLP is more "film like" than 3 chip - huh?

This is from his review of the Sim2 Nero 3D-2. I do not understand what the words coming out of his mouth mean:

"Because of its single-chip nature, the way in which the Nero portrays an image is more film-like than what is provided by most multi-chip and/or panel-based front projectors. Via the Nero, things like film grain actually appear to look like film grain rather than subtle pixilation [sic], not that you're seeing the pixels, per se. The simplicity and focus inherent with single-chip designs is something that, once experienced firsthand, is hard to live without."

Say what? Does he just mean it's sharper because there are no convergence issues?

http://hometheaterreview.com/sim2-nero-3d-2-single-chip-dlp-projector-reviewed/?page=2

 

[Seegs108]

I agree with him. I think a well built DLP projector with a good lens and light engine can produce the most "natural" film-like image out there so far from current digital front projection technologies. And I think you're on to something about what he's trying to say. I believe he's trying to say that the inherent potential sharpness a single chip DLP projector can have (and does have in the case of the Nero 3D-2) can portray material, and more specifically film grain, in a more "true" way. In a very real sense, a single chip DLP has some of the least processing and modulation if you will, from inputting a signal and getting an image. Other than it's current short-comings with on/off contrast I believe DLP is the best digital front projection technology out there currently, but only when it's done well. I hope TI has some tricks up their sleeve to help gain some on/off contrast in the future with 4K DMDs. From what I hear, they do, but who knows how much they'll help?

 

[mark haflich]

However, single chip designs tend to be more fatiguing and have other limitations with trade offs occurring based on the color wheel selection to favor certain parameters. Many commercial theaters use 3 chip DLP engines and a home 3 chip DLP machine can emulate the performance characteristics of them but the end product won't look the same having higher on/off (better contrast) but not the benefits of DCI server source (rec 709 vs P3) and other limitations.

 

[chrisreeves]

Ahhh. I thought the consensus was that high end 3 chip DLP's (i.e. $30K Sim2 models) have an more natural image with more saturated and subtle colors than high end 1 chip DLP's (i.e. SP-A900B, PD-8150, Nero etc). I wouldn't know personally because I've owned several single chip DLP's, but never viewed an at-home 3 chip DLP with the limitations of rec 709 content. The only 3 chips I've seen are in commercial cinemas.

Something about the color wheel degrading the image, making for less subtle gradations of tone and shade, introducing noise, etc.

So that's why AR's article threw me - when he seemed to say single chip DLP is more natural looking than 3 chip. If that's the case I'll stop lusting after the 3 chip Sim2's, to which people are continually having orgasmic reactions, and be satisfied with my JK .95 DLP. P

 

[mark haflich]

I don't get, and never will, get the bit about more saturated colors. If your source is rec 709, you calibrate to rec 709, if a color is reproduced outside the triangle color space, it is indeed more saturated but it is wrong. If a display can display more saturated colors than in the rec 709 space, depending, it maybe able to reproduce say P3 but for accurate colors you would need to calibrate for that space and then have a source coded for that space.

 

[Seegs108]

Ahhh. I thought the consensus was that high end 3 chip DLP's (i.e. $30K Sim2 models) have an more natural image with more saturated and subtle colors than high end 1 chip DLP's (i.e. SP-A900B, PD-8150, Nero etc). I wouldn't know personally because I've owned several single chip DLP's, but never viewed an at-home 3 chip DLP with the limitations of rec 709 content. The only 3 chips I've seen are in commercial cinemas.

Something about the color wheel degrading the image, making for less subtle gradations of tone and shade, introducing noise, etc.

So that's why AR's article threw me - when he seemed to say single chip DLP is more natural looking than 3 chip. If that's the case I'll stop lusting after the 3 chip Sim2's, to which people are continually having orgasmic reactions, and be satisfied with my JK .95 DLP. P

I think he was referring to "3 chip" as 3LCD and 3 chip LCoS. Most people say that 3-chip DLP looks like it has more saturated colors. This is sort of true and it has to do with the typical high brightness these 3-chip DLPs usually have. Higher brightness=more saturated looking colors to the eye. So even if both are calibrated to the same standard of color, typically speaking, the one with higher brightness will look more saturated to the eye. That makes sense to me. The more light our eye receives about the color the stronger our eye will perceive it as I guess? 3-Chip DLPs can have more gradations/shades of color, but so can 1-chip DLPs. What you're referring to deals with the content, not the projector. Though both projectors would need to have appropriate processing and other hardware to properly display such color. With 8-bit blu-ray you're limited to 16.7 million colors. This is the same no matter which projector you view it on. Now one of them may have better dithering techniques to "upsample" the color to 10 or 12 bit, but on a technical level the 3-chip DLP does not have an inherent color advantage. Like I said, this has to do with the content being fed to the projector. At the cinema you get 12bit 4:4:4 luminance/chroma information and a wider color gamut (P3), so yes, typically at the cinema color will look subjectively better, but does not mean a single chip DLP can't give you the same experience. We just happen to be stuck with 8 bit 4:2:0 on blu-ray.

Personally speaking, I find nothing wrong with his write up. I just think he should have been a bit clearer as to what he meant. I think you're getting caught up in the semantics of how he writes.

 

[zombie10k]

he got a whole bunch of folks here in an uproar over this article

http://hometheaterreview.com/sony-vpl-vw1000es-4k-sxrd-front-projector-reviewed/

 

[Seegs108]

Yeah, I think without a JVC with a DI, the Sony most likely would have looked far better. Or at least that's what I expect. The JVCs at the time (RS50) were around 25000:1 native at max brightness. The Sony has about 40000:1 usable dynamic contrast with real content (aka not an entirely black image) according to cine4home. Combine the dynamic contrast with far superior ANSI contrast, lens quality, RC/upscaling, and better motion I just don't see how the JVC could have looked "that good" against the 1000ES.

 

[chrisreeves]

Now one of them may have better dithering techniques to "upsample" the color to 10 or 12 bit, but on a technical level the 3-chip DLP does not have an inherent color advantage.

Seegs you're just a better writer than he is - you explain with precision. You ought to write yourself actually.

As for your quoted section, thank you - that is one of the things I've always wondered. Do some projectors and/or external processors "upsample" using some kind of AI algorithms that examine frames and say "hmm, that gradually increasing sweep of blue looks like a sky, and that blockiness/banding in that section of blue in the frame looks like compression artifact, not signal - therefore I, projector machine, will replace the banding with several more shades of blue to make it a more gradual transition on screen."

 

[chrisreeves]

I don't get, and never will, get the bit about more saturated colors. If your source is rec 709, you calibrate to rec 709

Neither did I but I assumed I was ignorant and/or stupid.

 

[zombie10k]

it was an RS20 used for the comparison.

A quick check-up on the JVC's light output revealed that it was projecting a mere eleven foot-lamberts at the screen. To ensure a level playing field, we dimmed the Sony's light output by closing the VW1000ES' iris until we matched the two projectors' light output at roughly 11 foot-lamberts.

heads were exploding.

 

[Seegs108]

Oh wow, for some reason I thought it was an RS50. Yeah, the Sony should have demolished it.

 

[mark haflich]

Seegs you're just a better writer than he is - you explain with precision. You ought to write yourself actually.

As for your quoted section, thank you - that is one of the things I've always wondered. Do some projectors and/or external processors "upsample" using some kind of AI algorithms that examine frames and say "hmm, that gradually increasing sweep of blue looks like a sky, and that blockiness/banding in that section of blue in the frame looks like compression artifact, not signal - therefore I, projector machine, will replace the banding with several more shades of blue to make it a more gradual transition on screen."

No.

 

[Seegs108]

Seegs you're just a better writer than he is - you explain with precision. You ought to write yourself actually.

As for your quoted section, thank you - that is one of the things I've always wondered. Do some projectors and/or external processors "upsample" using some kind of AI algorithms that examine frames and say "hmm, that gradually increasing sweep of blue looks like a sky, and that blockiness/banding in that section of blue in the frame looks like compression artifact, not signal - therefore I, projector machine, will replace the banding with several more shades of blue to make it a more gradual transition on screen."

Honestly I'm not sure, but if I were to guess I'd say no, they're not that complex. I would venture a guess that something like this would be pretty heavy duty processing-wise and might not be do-able in real time with out current video processors. But again, I'm really just guessing here.

 

[mark haflich]

Neither did I but I assumed I was ignorant and/or stupid.

I am too tired to argue with you.


As to Seegs, brighter is not more saturated. For any given color the saturation does not change with its brightness. To the eye, increasing the light say shined on a color will make the color appear lighter (less dark or pure) and to the eye it might appear less saturated but it ain't.

 

[Seegs108]

I am too tired to argue with you.


As to Seegs, brighter is not more saturated. For any given color the saturation does not change with its brightness. To the eye, increasing the light say shined on a color will make the color appear lighter (less dark or pure) and to the eye it might appear less saturated but it ain't.

Try and tell this to the guys in the $20K+ forum hahaha

 

[Andreas21]

it was an RS20 used for the comparison.

A quick check-up on the JVC's light output revealed that it was projecting a mere eleven foot-lamberts at the screen. To ensure a level playing field, we dimmed the Sony's light output by closing the VW1000ES' iris until we matched the two projectors' light output at roughly 11 foot-lamberts.

heads were exploding.

With this simply stupid review Andrew Robinson lost all credability. When he also said the Sim2 M.150 is better than the VW1000 it is easy to see he is bought and payed by Sim2. As I have seen the M.150 side by side with the VW1000 I know it is very far from being true.

I also saw the M.150S a month ago side by side with my X500 and the X500 demolished it. I find the Sim2 M series to be very overrated.

 

[Andreas21]

And Seegs108 when you say brighter projectors look more saturated I wonder if you have ever side by side tested, like Mark says the brighter actually look less saturated even if it is not. I have seen this every time I side by side test and have not yet brightness matched.

 

[Elix]

With more brightness, given everything else is the same, you get more... brightness. In terms of color it's luminance. Not saturation. To give a live example, here's the same picture with the exception of a full stop of exposure has been added (meaning more brightness in photography):

Here's the same original image with added +100 luminance to all color channels:

Looks similar to the previous adjustment, right?

And here's the same original image with an added +100 saturation to all color channels:

So Mark is correct, saturation doesn't change. It's for you to decide whether increased luminance creates a perception of more/less saturated colors. I think it's not so simple (when is it simple when things comes to human perception?). Take a look at this +1 exposure adjustment in another image:

Can you honestly say that all colors look more saturated in a brighter image? I cannot. For example, while tomatoes might look more saturated to me, lemons definitely look lees saturated and more white. This is not a clean experiment as I don't know what Lightroom algorithms might be taking place. Ideally, an exposure adjustment should be linear in that it should change all channels equally.

The problem with saying that two projectors calibrated to REC.709 standard should look the same color-wise, despite the technology, is that nobody actually went to such lengths as to compare each of the 16.7 million colors that this standard defines. It's usually a 100% color sweep, in best cases it's 25%/50%/75%/100% sweep. And rarely the reviewer compares all 3 aspects of color: hue, saturation and luminance. The latter is disregarded more often than not. So who knows how better a 3-chip DLP actually is when it comes to hitting that REC.709 target with each and every color on all three dimensions: H, S and L. But given the nature of a DLP projector I'm more inclined to believe it will be easier for a properly engineered DLP PJ to be more accurate at hitting each of those 16.7 million color targets than other current technologies.

 

[12B4A]

it was an RS20 used for the comparison.

A quick check-up on the JVC's light output revealed that it was projecting a mere eleven foot-lamberts at the screen. To ensure a level playing field, we dimmed the Sony's light output by closing the VW1000ES' iris until we matched the two projectors' light output at roughly 11 foot-lamberts.

heads were exploding.

Was it because it was a false comparison like: Car A goes 100mph and Car B goes 120mph. In order to fairly evaluate driver experience we decided to drive car B at 100mph.

Or was it that the JVC was actually projecting ~eleven foot-candles at the screen?