When is 720p Not 720p?

Henning Hoffmann writes "HDBlog has an interesting entry about many home theater displays.
Home theater displays around the resolution of 720p (most DLP, LCD, and LCOS displays) must convert 1080i material to their native resolution for display. No surprise there. But many displays do this by discarding half of the 1080i HD signal, effectively giving 720p viewers an SD signal - not watching HD at all! "

Comments

Which Models?

[goldspider]

Is there any way of telling which sets do this? This is certainly something I'd like to know before I dropped that kind of cash.

Re:Which Models?

[David Leppik]

Is there any way of telling which sets do this? This is certainly something I'd like to know before I dropped that kind of cash.

Yes. Go to a showroom and look at the displays. If you see some that have greater vertical resolution than the non-HD models, there you go. If you can't see a difference, then it doesn't make a difference.

If there is a difference you can't see but could learn to see, don't learn; it will not bring you joy, it will only make you miserable or annoying. Long ago I learned to see the FFT distortion in JPEG and MPEG images. Has it made me happy? No. I end up making the JPEGs on my website bigger than everyone else's so I won't see wrinkles on people's faces that are apparently invisible to everyone else. And I can't stand to watch satellite television on a big screen TV because of the annoying compression artifacts.

ED displays

[justforaday]

I haven't had a chance to read through the full article/blog/whatever yet (I'll do that at lunch), but this sounds like something I noticed over the weekend while browsing the Best Buy site. Companies are now producing ED-compatible TVs. They list all sorts of compatible display modes (1080i, 720p, 480p, etc), but then mention that they downscale them for display on the TV. Is this just some way of offering half-assed support to unsuspecting consumers?

only HD is HD

[Anonymous Coward]

Manufacturers better refrain from selling not HD capable displays as HD displays. This is clearly false advertising and there have been several succesful lawsuits lately where people who have been stupid enough and bought into this got their money back.

Re:Resampling

[gyg]

The only effective limitation comes from linear algebra - there are only as many degrees of freedom as there are pixels, so if you downsample, you *always* lose data, like it or not.

However, even this is not a problem in practice since in real-world pictures nearby pixels are not independent. By using an appropriate encoding dictionary such as wavelets, which zoom in on sharp edges and economize on flat surfaces, you can shrink a typical picture by something like 90% without visible quality loss.

Now since wavelets are actually continuous functions, you could then convert from wavelet representation to upsampled signal, with no information loss. I imagine this could give sharper edges as wavelets are better at egdes than Fourier.

The problem of upsampling well is very similar to making a blurry image crisp - called deconvolution. The problem in doing that is that any noise in the blurry image gets amplified, must be filtered out etc...

Moral is that clever algorithims have been around for ages - the effective limit is whether they can be done in realtime.

How to prove/disprove it easily.

[Stavr0]

Simply provide a 1080i test pattern made of alternating horizontal lines where the left half is black lines on even scanlines, white lines on odd scanlines and the right half is reversed (white-even,black-odd)

The results may be one of the following:

You will get a screen full of tiny, shimmering horizontal lines that shift in the center of your screen Congratulations! Your HT gear is showing a true 1080i picture You will get a full screen of gray, possibly with a line in the center Not bad, your gear is properly downscaling the signal Half your screen is black, the other is white Uh oh. Your gear is taking the easy way out and dropping half the scanlines to downconvert (Bele and Lokai) I call that the Cheron Test.

CD Audio all over again.

[karlandtanya]

Remember when CDs first came out?

Oh, this was going to be great. Fidelity like you never had it before. No scratches. No groove wear. Dynamic range you won't believe. Crystal clear highs. Thunderous lows, with no rumbling feedback even if sat your player on your speaker.

Remember the little logos? AAD? ADD? DDD was the best you could have (digital recording, digital mastering, and (obviously) digital media in your hand). And a lot of hard work on the part of the engineers operating the mixing boards. It's that last part that costs time and money. Now, all the equipment is digital. So, it's all great, right? Sorry--the technology is not the limiting factor in sound quality anymore.

The limiting factor is apathy. Most people can not really hear the difference. And fewer people care.

Exactly the same thing is now happening in video.
Since we can't improve the functionality (well, we could, but you'd never notice). It's pure hype from here on out.

Now, where'd I leave that case of speaker spikes and green markers? Gotta get 'em up on ebay; David Hannum was right.

Re:misty

[Phreakiture]

Mfgrs usually tout their amps with having "200 watts of pulsing music power" which usually means 100 watts per channel peak.

I call this the Radio Shack method of describing the system, because it was originally used only by Radio Shack, but now is used by everyone -- the system power is described as the sum of all channels, not the value of one channel.

As such, I have a 420W sound system in my living room that, 15 years ago, would have been described as 70W. More likely, it would have been described as 70/70/70--70W main, 70W surround, 70W subwoofer (yes, it is a small system, but fully appropriate to the room).

On the peak/RMS issue, good amps give RMS and have peak listed on the specs page. Sometimes the peak exceeds the typical sine-wave ratio of RMS=.707*peak because not all sounds are sine waves. My previous amp was a good example. It was listed as a 60W (per channel) amp at 8 ohms (15W surround, no sub, no center) but could, according to the specs page, deliver 120W on a peak to the main channels. This system was replaced last Decmber after 15 years of service. Coupled with a set of Harmon-Kardon speakers, it rocked pretty hard.

Decide for yourself

[DumbSwede]

I have a NEC1351 projector and can display in 1080i 1080p and 720p. I have looked in the show room and noticed a huge range of qualities in HDTV quality. Upconverting standard definition is by far the largest area of quality difference between various models, but that is a bit of an aside to this discussion. I have always been surprised at the number of show room models that wimp out at 720p and not even true 720p but some weird non 1280 horizontal number that has to be extrapolated no matter what the source. Three beam rear projection tube models generally support all resolutions, but have far less actual resolution than the raw numbers they sight on their spec sheets. My 1351 looks stellar compared to anything you will find at BestBuy, but I have done a lot of tweaking, and it is still short of absolute perfect 1080 performance, having to do with a range of factors I won't go into here. The point is even a high-end machine like mine is currently not maxing out the quality of a true HDTV signal yet. The HDTV signals themselves are of so huge a variance of quality you have to wonder what kind of Rube Goldberg solutions they are using behind the scenes in some cases.

I have seen pros and cons on how these sets do their sampling. Here is my advice -- go look at the picture on a set with a good HDTV source. Use the specs as a guide but don't trust them. Get what looks good to you. My father would never have been able to see anything more than the quality of a good DVD. He couldn't see the difference between crappy digital cable and DVD. Some people like me are so manic about visual quality we will devote huge amounts of time tweaking our systems. While my system is probably limited to about 1500 lines of resolutions due to the lenses, I find its image much warmer, uniform, and pleasing to the eye than the pixilated look of some very high-end flat screen solutions that go for 10-20k.

About the only thing that really shows how good HDTV can be is material that is shot originally with HDTV video cameras. Upconverting film inevitably introduces a softness that is exaggerated by systems like mine. For now you can only see a few things on the Discovery Channel and a few musical events in true HD (meaning not upcoverted from film). I mention this because while I advise you to go see for yourself (if possible) most stores don't really offer a good enough HD signal to display the difference. If you can hold out a little longer I would wait until either HD-DVD or Blu-Ray players hit the shelves and then demand a demo with these HD sources to make a decision.

One final note, I haven't noticed that 1080i hasn't had as much comb-artifact during motion as I would have expected, but there still is a noticeable blurring during camera pans (maybe this is just combing in disguise). I'm sure I will get a little boast in quality when I can play off of a true 1080p source. If I were to design the next generation of video recorders I would introduce variable framing rates in playback. The picture being refreshed as high as120fps, but the actual picture updates depending on need for frames to eliminate motion blur. About motion blur, storing 120fps would be inefficient and overkill. The system I propose would make true frames at like 30-60fps, but as the camera moves, the edges would be scrolled in as needed to keep a smooth fluid motion. A really intelligent system might be able to also track one or two moving objects across this field and give them higher frame rates as well. At 2 mega pixels, I think we need to retrench and try to slay motion blur before going onto higher pixel counts.

Shameless Atari / Amiga Plug

[tjstork]

Back in the day, on the Atari and Amiga, you could actually do stuff in that interrupt time. The most common thing was to swap display buffers for double buffering. This made for rock steady hardware scrolling, an effect that still lacks somewhat in today's PC's, believe it or not, as there was absolutely no tearing of the display whatsover. Just a beautiful effect.

Re:Shameless Atari / Amiga Plug

[Shalda]

That's a a bit of troll, wouldn't you say? In the EGA/VGA (I think CGA as well) spec, you can check to see if the card is in a vertical retrace and swap buffers then. It's rock steady as well. Compare this with the approach taken in the Commander Keen games where they achieved very smooth side scrolling by adjusting the offset of the video buffer on the card. One only had to draw a few percent of the screen and move some sprites.

Re:Shameless Atari / Amiga Plug

[awfar]

And, I would add if IIRC, the Amiga had an intelligent chipset and a local bus (chip vs. fast ram) for video and sound, and the autonomous chipset would DMA from/to using an abstracted display (command list) list, fully using interrupts to inform of events like retrace, both horizontal and vertical, I think, full hardware sprites and collision support; the audio could dma and play audio autonomously; set registers and a dma pointer and go. The CPU was doing other very important things like OS work and calculations, etc., which with the OS made it the first true multitasking home PC OS I had ever used, and it was smooth; not perfect, but it was doing everything, mouse, gui, video on many screens ,stereo sound on a 10MHZ 68000!

Re:When is 720p Not 720p?

[Ibiwan]

the DCT is lossless; it just serves to separate the block information into lower-freqency components at the top left of the block, gradient to high-freq components at the bottom right of the block. The artifacts come from a lossy process where they divide the matrix, element-wise, by a matrix with ones in most places but higher and higher numbers as you go toward the bottom right. The end result is, you throw away partial information on your higher frequencies, in a pattern that the JPEG committee's research decided people can't see. Of course, you can further raise the factors from there, which most people do, which is why too many images show the blocking.

Incidentally, the image is further compressed by

• converting the block to a linear array of data by reading in a zig-zag pattern from the top left
  
• using (lossless) run-length encoding on that bit stream (or on the data as bytes or words, not too sure on this point)
  
• compressing THAT output with a (lossless) entropy coding; I believe it's the huffman family of algorithms.

Re:For the inevitable /.ing

[Infinite Entropy]

Just wait for the thin CRT technology to come out. It uses millions of tiny electron guns instead of one, and if it fulfills its promise it will destroy plasma and LCD.