Sony VPL-VW1000ES vs. JVC DLA-X90R shoot out - Page 2

Owen, in your opinion, and from what you seen/read, what has the best MTF? DLP, DILA, SXRD, or LCD?

The Samsung a900b is the sharpest DLP ever tested I believe. MTF is not really used to test projectors or to rate them, it was used in a few studies. It is true that is a common thing to be used in imaging and for cameras, but AFIK MFR's don't use it for projectors, no idea if they use it internally in engineering testing, possibly, but they may use other tests.

There is little data to go on but I very much doubt that the underlying technology of the projector makes much difference, implementation and lens quality will likely dominate the outcome.

OK, I've got to ask some questions, for my own education.


Let me ask you this, MTF is an additive/cumulative function, I'll probably be wrong on the math here, but if you had say a camera with 50% MTF at Nyquist, scanner with 50% MTF at Nyquist, and then displayed by a display with 50% at Nyquist, you end up with a system with about 12.5% MTF at Nyquist (drastically oversimplified).

Why is it not appropriate to isolate just the display/projector with an artificially generated test pattern with perfect 100% MTF all the way up to Nyquist? Why should we evaluate any individual component's MTF performance with test content already hindered by another device's imperfect MTF?

I'm not saying I disagree with the idea that it's the real world content that matters.


I'm curious, have you seen Mark P's measurements with a line scan camera?
http://www.videovantage.com/?p=819

I've been looking for that article. I read it a while ago and lost the bookmark. thanks for posting!

IIRC some measurements done around the release of the first 1080p LCOS chips so 5 or 6 years ago had CRTs at 30%, DILA/SXRD at 85% and DLP at 90% of the Nyquist limits. Do a search of the forum.

Anyone wanting to see a great visual comparison of the representation of different MTFs on an image see slide 26 of this excellent slide pack by Matt Cowan which I know have posted many times.

http://etconsult.com/papers/Black%20...tt%20Cowan.pdf

It's all too outdated IMHO.

I worked in rendering a bit. IT is actually very difficult for people to determine sharpness by looking at a video, the video has to have specific traits. Also a moving image does not test just sharpness, it is testing other things such as motion resolution or motion noise sensitivity levels, so you cannot do it like that. I mean you can, but the first thing to do is to use a still image, any pattern that attempts to test sharpness on a moving image would only be an addition and another attribute.

The ISO Image I posted is actually part of a standard sharpness testing process used in professional applications. Some people don't use it, some do, it is even used in labs sometimes. So I wouldn't totally discount everything I'm saying here just because I tend to downplay the importance of MTF. I am very much a real-world application guy, if I can't use something I'm not going to obsess about it, and I have no use for MTF at all because I do not have the equipment required, and neither do any of us.

The circuit board I use is very good for this because it has contrasty white lines in the image, multi-dimensional pop-out over the lines and text at the same time.

I also disagree about that article being discounted entirely, the guy made some typos but who doesn't. He is a German photographer since 1977, a well known one at that, and the Germans are really good at these sort of things. I've never met a dumb German engineer, or even someone that isn't an engineer that was German

Now that is funny!

To that point, it would be nice if there were higher quality lens options available (for the JVCs); either interchangable lenses or as top model in the line. I'd much rather be paying for a superior lens/optical system than 3D tools and extra warranty as is done with the RS65.

I have no idea what point you are trying to make here.

Sorry if you've already posted this, but have you checked actual content on DVDs or BDs to see if the encodings of the pixels levels themselves actually don't go above 10% MTF in small areas (or do you have a link to an article with values from actual encodings of video/film content on DVD of BD)?

I haven't read all the posts here, but I brought up this issue with Joe Kane when he was presenting his A900B projector and talking about how sharp it is at the pixel level. IIRC, I mentioned that video shouldn't have black pixels right next to white pixels and his response was basically that there isn't an enforcement of the levels only changing so quickly between adjacent pixels in the encodings.

He was using a scene with a tweed jacket from Seven Years in Tibet to demonstrate how good his A900B is and talking about how much better it is than the JVCs, which couldn't do that scene right and shouldn't have been called 1080p projectors according to Joe. I later tried an A900B I owned against a friend's RS20 with splitscreen and honestly couldn't see much difference at all from reasonable viewing distance with that detail n the jacket, but didn't have any trouble at all finding scenes where the on/off CR advantage of the JVC seemed pretty obvious. So I quickly sold the A900B.

To be clear, I'm not saying Joe was right, but I am curious about whether these limitations really end up being there in the final encodings in all cases. I think it would be interesting to see the levels for that tweed jacket. One of these days I should run it through my Panasonic AE1000 and use the waveform monitor to see if I can figure out approximately how fast the levels change in the encoded values from pixel to pixel there.

It seems like when I have looked at the actual stars in starfields they rarely or never are encoded to completely drop-off from near-white or white to black pixel-to-pixel, which would support limited encoded MTF at full resolution even in those scenes.

Thanks,
Darin

Sorry, I didn't mean to say what you guys were talking about has no value. The earlier references was to testing sharpness by users or reviewers, that's how this started. I am just saying that to test sharpness with MTF is beyond the capabilities of most, so it has no real benefit other than in a lab scenario.

Yah, that's another problem is the camera gradiates the items, unless we are looking at animated content or fake looking renders, it's probably not going to drop right off. Even in animated content (unless it's old Bugs Bunny cartoons or something), they are going to apply a light bloom around the star to make it look realistic, otherwise it would look like a piece of white lint on a black piece of cardboard.

Sounds correct. I believe tse told me those numbers are in the ballpark. I remember measuring a CRT (with Darin), but I can't remember the number at 1080p.

I recall looking at the scope and I believe the white was around 2x or 3x the black. Actually, I should say green instead of white since we were only measuring the green gun and also at these levels of MTF the black looks pretty gray from up close. The number 3 sticks in my head, but I'm not sure if it was 30% MTF or 3x bright lines to dark lines. I'm thinking the 2x is right with the picture I have in my head of the scope at best focus and that comes out to about 30% with (2 - 1) / (2 + 1) by my calculation.

Also, I think that was a pretty modded 9" tube and would expect many CRTs to give much lower. After calibration blue could have lower MTF from defocusing it while increase light output from that tube, but I don't think MTF with blue matters nearly as much as it does with green.

--Darin

Owen James Cameron is moving to New Zealand. Perhaps you could drop by and see him to put things into perspective. Ask him if he is going back to NTSC or PAL?

Seriously people look at computer data screens just as much as we look at video. Did you forget this aspect?

The competition (flat panels) have no lens to soften the image quality or three panels with all sorts of added (and easily observable) alignment distortions. Projection system should be transparent as possible or else consumers will naturally switch over. This means the optical path should be of camera quality.

Any softening of the image should be left to the engineers. The transmission, storage or reproducing display system should in in no way alter the image.

HD camera and mastering techniques are constantly evolving. Most of us can see the result of improved MTF/contrast in every stage of the capturing and mastering chain. Thank you James Cameron for raising the standards.

Audio went through a similar evolution. Digital filters improved greatly over time and are able today to capture high frequencies close to the Nyquist limit. Again, its the mastering engineer who selects to add low level dither or increase bit depth.
Consumer audio systems should not be adding noise, distortion or resample. Realizing just how bad mp3 degraded a generation of listener enjoyment, Steve Jobs was planning to go to 24 bits. These are the leaders who are making our lives more enjoyable.

This is what high fidelity or high technology is all about. People will pay for the best quality experience. Hopefully the very expensive Sony 4K display will resolve 4K resolution. (the less expensive Sony projector have issues resolve 1080. see Kris Deering's review of the 95ES). If not 80-90" flat panels will - and for much less.

I was hoping you would remember it better than me. Honestly, I thought it was less than that, but it has been five years. I seem to remember MTF being pretty low until the resolution was dropped to like 1024x768. The Marquee was modded, but I don't believe it was performance modded. Usually Scott was doing customer specific mods. Also, I am not sure which lens was used. The best HD10 gives 12 lines per mm instead of 10 for the regular one.

Here is another comment from the AVForum about the Sony 4K. Sounds like this could also be the Solo that is being compared.

"Our Sony is roughly 8m back from our screen and our seating is from 3m-5m (large L-Shaped Sofa)
Our screen size is 175.2" Horizontal.

The picture quality is equally good standing 6' away or standing under the PJ at around 24'.

We also have another projector to compare, that is over twice the price is 1080P, looks great, but nowhere near the quality that the Sony 4K is producing.........these are just my findings and I have been sitting I front of it all day watching a variety of film clips.

LOTR trilogy looks incredible. "

Ive mentioned this on a different forum... But it kind of gets my goat when people mention the eshift feature and 4k together.

Jvc have been clever in marketing their pjs last year and mentioned the eshift and 4k slogan left right and centre. The jvcs are ultimately 1080 displays. Yes the image is manipulated but it is still a 1080 display prsented in a neater manipulated manner. It has diddly squat to do with 4k. It doesnt have the independant pixels to be labelled a 4k machine and it wouldnt know what to do with 4k signal!

Hence any material related to the jvcs either instruction manuals or the pjs box wont mention 2k4k as its misleading.

Granted its a stepping stone towards the higher res age-but its still a 1080 machine

These comments arnt in the same context as the person who had made the earler comments on the sim2 3ds.

The projector mentioned here at twice the cost could be any number of machines

The JVC takes a 2k signal upscales it to 4k and displays the single frame in 2 flashes. Give or take a few pixels.

The Sony takes a 2k signal upscales it to 4k and displays the single frame in 1 flash.

Consider the JVC is displaying the frame in a sort of interlaced fashion....the end result is still 4k.

Of course the Sony can take a 4k signal and display that too, where as the JVC can't , perhaps next year.


This beast also uses E-Shift to produce an 7680 x 4320 end result:

http://www.3dfocus.co.uk/3d-news-2/s...-ibc-2011/5433

If JVC is not 4K, the resolution looks higher than 2K

i'm not really concerned about what the technology is called, as long as it makes a noticeable difference on a large screen - which it does. I was hesitant to give this technology any credit until I saw it for myself in my own home theater. In my particular setup, it would be easy to pick out for most folks in these threads vs. the other projectors in this price category.

The e-shift's benefit of an increase in perceived contrast gets lost in still photos that are heavily cropped.

I'll be seeing the VW1000 soon next to the RS55 to get an idea of the real world difference between these 2 projectors. There's only so much we can draw from tech specs, forum conversations and translated reviews.

I'll be keen to read about your observations. I'm guessing you will do you test with and w/out Sony's DRC enabled?

definitely. I am excited to see the Sony. I am glad they did this since they are one of the few companies who has the driving force to push 4K content. Hopefully in the next 2 years or so.

Watching 'The Art of Flight' shot on the Red 4K with E-shift @ 3 is amazing to see on my 142" @ 1.25 SW and makes me wonder how great it could look with native 4k material on the Sony.

I'll accept the X70/X90 as 4K projectors the moment they can be fed native 4K signal which is displayed with full 4K resolution intact. They can not display 4K images, hence they are not 4K projectors.


Where did you get that from?

Click for PDF: http://www.google.com.au/url?sa=t&rc...XSSFmbSfDDSTRA

and more detailed PDF: http://www.google.com.au/url?sa=t&rc...JrFQR91EapClpQ

Interesting; thank you.

And this is how you record 8K. Going to be in action for the London Olympics.

I'am actually owning the VPL-VW1000ES since last friday.
And I only can say, that is as the most impressiv black level I have ever seen so far, the brightness is awsome and the 3D performance sets a new standard for shutter technology. There is no, and I mean really no, visible pixel structure on my screen (185") - projection distance 9m. Also its absolutly quite, you have to stand right beside it to hear it wisper .

Awesome. I'am really impressed.