CurtPalme.com Home Theater Forum

[mp20748]

[WTS]

Hi Mike,

Good to hear you picked up some tips from Scott to tweak your boards alittle more.

Yeah the G90 is another PJ with BW bottlenecks which make me wonder how they can make certain specs claims.

[mp20748]

[Nashou66]

[Nashou66]

[WTS]

Hi Athanasios,

Yes the spec sheet says you can go above the 500 mark but only in one configuration and that is not the configuration being used here. The data sheet is very lax in performance specs period. Not only that but AD doesn't even have spice files for these devices, I called them once about this and they wouldn't go into why they wouldn't supply them or why they didn't have one for them, hmmmm.

[WTS]

Hi Mike,

Sounds like an idea. I'm not really into messing around with these boards very much anymore, but nevr the less I'd like to hear what you've been up to lately and what you've picked up from Scott.

[wallace123456]

[Nashou66]

[Nashou66]

[VideoGrabber]

You're hitting this pretty hard, Dave. I understand where you're coming from, especially with bloated bandwidth claims, and I agree with parts of it, but you seem to be going a bit overboard. I was glad to see you admit...

> I'm not saying it can't look decent with typical HT content, it just will not be razor sharp.. <

since, in fact, it can not only look "decent" with HD content, but excellent. In spite of the fact that a test pattern is not "razor sharp". In fact, digitally rendered HD content, with hyper-sharp edges, is actually a distortion. Bjoern Roy explained this often overlooked point very well, many years ago. While this digital "sharpness" may look appealing to you, it isn't accurate, as can be seen if you look at a spectral plot. It will expose all kinds of 'spectral hash' that wasn't present in the original image, that was created by the hard edges.

> Search for Chiem's bandwidth write up on AVS--he details this pretty well. <

I think you may need to go back and re-read that (and my followup comments) yourself, since you seem to be suffering from selective recall.

> the real tell-tale sign is the thick white box around the test patters. Look at the outer left edge of that--way soft. Imagine that is a planet surrounded by space--it is going to look more like a cotton ball than a planet! This part of the picture proves... <

Actually, what it proves is that in your rush to debunk, you're not being as observant as you usually are. Who claimed that the square box on the periphery had hard edges? How do you know that border isn't intentionally rounded edged, specifically to avoid driving the amps hard before being presented with the actual test signals? If your claim about cotton balls was correct, then it would be just as soft and blurry on the right-hand edge of the test patterns! And it's not. Not even close.

There's something strange about that test signal anyway... in the 2x2 vertical pattern, the max peak amplitude on the green was 244, and the minimum valley was 137. Yet immediately after that pattern terminates, there's a vertical black band with a minimum amplitude of only 47 (mostly in the 50-55 range). There's something fishy going on there, if we're being asked to believe that the test signal for each of the vertical black bands is supposed to be going to full black. Yet only drops to ~140 during the pattern, then drops to ~50 afterwards.

That's all I have time for tonight.

[Nashou66]

Tim, I agree , I stated this on another thread here about the sharpnes of digital, to me it looks fake.

[Person99]

In the end though, I'm really questioning the need for 1080p/72 output in devices (VPs). If I made a VP, I would not try to support that as the consumer base for it is very small. But, even in that consumer base, they don't need it. Few CRTs will show the whole 1920. Kal saw some of the same content on a JVC product and realized he was not seeing the whole 1920 on his Cine 8 Onyx. Now, given that vertical resolution is more important than horizontal and downscaling is easier than upscaling, you could downscale to something like 1700x1080@72, get no less detail, but a sharper picture than 1920x1080/72 on ANY 8" and most 9". So, why not do that?

In the end, we went into the woods with what MPs specially selected and meticulously set up machines can do (which from every screen shot I've seen look soft so I don't know what everyone else is looking at--my Cine 8 Onyx was much sharper at 1080i with Tim's spot shape trick than any shot I've seen MP post to "prove" how sharp he can get an 8" doing 1080p) or what tse's special new CRT not available to any of us (at a reasonable price can do). Those are red herrings and don't matter. The discussion was really my questioning the "need" for 1080p/72 output.

[mp20748]

[WTS]

Hi MP,

Just curious to know what improvements you saw when you removed the 835 from the G90, was it in the direct video signal path? I assume there was one per channel (RGB). Where in the signal path were these located?

[VideoGrabber]

Dave,

thanks a lot for all the clarifications. You are correct that some of your comments could have been misleading, and not really what you were intending to say. As you know, I am in complete agreement with you that 1080p/72 is neither necessary nor optimal on 8" CRT PJs. You, myself, Kal, Mark_A_W and a handful of others are all in that camp. (Mark has even gone so far as to set up his PJ at 1080p, just to see if he could eliminate the slightly perceptible line-flicker that he was attributing to running interlaced... only to find the same flicker running progressive.)

There are some who feel that 1080p/72 is the holy grail, and if only they had that on their PJ it would be perfection. Of course, it might require wholesale mods to their entire video chain, and likely lense and tube changes, but it would "obviously" be well worth it. And of course must be superior to 1080i. They then point to the exceptional results that MP has achieved as their proof.

And yet, as Mike just pointed out, "No one is or was saying an 8" looks better at 1080P/72hz". That's pretty significant! That has apparently never been his claim. Though I suspect some have taken that as implied. Thanks for clearing that up, Mike. I'm sure you recall, Dave, that some folks had previously proclaimed that an 8" was incapable of handling it, and Mike and Scott (tse) both managed to prove them wrong, by example. That doesn't mean that what is possible on one highly-optimized machine is the best choice for other PJs. Though I think that earlier comments from Mike (e.g. Sep'07) "A truly resolved 1080P native image is much more punchier, with more depth and 3-dimensional than 1080I any day." may have contributed to that perception.

The one disappointing thing for me is that both Dave and I have requested, and you indicated you would generate, pics of the same impressive patterns at 72p, only at 72i. That way, we could actually see the differences in MTF, resolving power, and contrast ratios. Unfortunately, you never found the time to do those tests (or, at least, post the results). You did say in a discussion on 1080i vs 1080p back in Oct'07 that, "Good points. I'll be mindful to try that at 1080I tonight.". And later confimed in Dec'07 (on the A Forum), that "I gotta admit, much to my surprise 1080I is not bad on my lowly 8500." But we never got to see the proof of the pudding, or your evaluation of where you thought the best tradeoffs were for stock 8" PJs.