AVS Forum banner
1 - 20 of 25 Posts

·
Registered
Joined
·
33 Posts
Discussion Starter · #1 ·
I noticed that 120 hz refresh rates were a big thing at CES this year with many of the new televisions on display.


Of course this would make a big difference with motion blur and other artifacts but it ocurred to me that another issue would be potentially solved with this change.


It would be possible to feed both 1080p24 signals and 1080p60 signals to the projector and both could easily be scaled to 120 hz without any need for further processing. Just double the frame rate for 1080p60 video signals and quintuple the frame rate for 1080p24 film-based material.


Does this make sense or am I missing something here?


It just seems to me that the faster frame rate would solve many of the evils we see with combo film/video material (like extras on HD-DVD's that are from video and not 35mm film).


Does anyone know of any front projectors out now or in the near future that will have 120 hz refresh?


Thanks



Michael
 

·
Registered
Joined
·
352 Posts
We used to talk about this as the holy grail 5 years ago on here for connecting our projectors to HTPCs. I'm not sure if they have projectors that will sync to that yet. I'm still trying to figure out which projectors will do 72hz like my current Davis DLS8 does.
 

·
Registered
Joined
·
9,223 Posts
You can't add pictures that were not there to begin with. With film for example, you have 24 unique pictures per second. Now you can flash those same 24 pictures twice, three times, even 4 times per 1/24th second, i.e. at 48hz, 72hz, 96hz but it's still only 24 unique pictures. In otherwords, showing the same frames multiple times does not smooth motion artifacts. That damage is done during capture and nothing will restore the exacy motion between the missing frames.


Now with film projectors and CRT based video displays you must run at least 48 fps to make flicker tolerable. 72fps is even better. But this is a different problem. Film projectors chop the light on and off between frames. This causes flicker. The faster the chop rate, the less the flicker. CRT's make an image by scanning an electron beam across phospor. The phospor starts to decay in light output if the beam doesn't return fast enough. This is why 24hz is not viewable on a CRT monitor.


However DLP, LCD, LYCOS, SXRD do not have this problem. The pixels stay lit until a frame comes up thet requires them to be off. Because of this there is no flicker. On a full white screen, the pixels never turn off. So it doesn't matter if you send 24, 48, 72, or higher frames per second. The image stays the same.
 

·
Registered
Joined
·
8,441 Posts

Quote:
Originally Posted by Dave McKean /forum/post/0


Glimmie,


We know that. It's that 120Hz is a multiple of both 24 and 30 so both video and film based sources will be displayed with smooth motion.

By smooth motion you mean no pulldown judder while using the same scan rate as a video source. No need to switch from 48Hz to 60Hz for example.
 

·
Registered
Joined
·
8,441 Posts

Quote:
Originally Posted by tahustvedt /forum/post/0


It would require twice the bandwidth of 60 Hz, so most cables would probably introduce artifacts or not work at 1920x1080 120Hz.

It would be internal up-sampling. You would send 1080 24fps, 24Hz, 48Hz, 30Hz, or 60Hz and all would be up-sampled to 120Hz. The only question is what to do with PAL 50Hz?
 

·
Registered
Joined
·
6,092 Posts

Quote:
Originally Posted by Glimmie /forum/post/0


You can't add pictures that were not there to begin with. ...

Actually, that's exactly what some of the demos did - created frames that weren't there to begin with.


Most of the demos showed 60 fps to 120 fps conversion using simple moving images and the results were very impressive. With these bright plasma displays, you can clearly see the benefits of 120 fps over 60 fps. However, most of demos were simple; the image was just a scrolling test image were any algoritm could easilly determine what should be in the imbetween frame. (The original 60fps video scrolled left to right about 10 pixels every 1/60th of second; so the processor just created 120fps video that moved 5 pixels every 1/120th of second.)


However, there was an expection. Micronas put on a pretty impressive demo where they took 30p movies and made them 60p. (see attached image)


In their demo, you had much more complicated images to work with. In the attached image you will see a lady walking as things pass her by. With the Micronas processing, the 60p created image looked significantly better than the original. They also had a scene from Lord of the Rings that looked just as good which not only had frame correction but 3:2 judder ellimination as well.


Anyway, these are just demos but the technology has some potential. I would love for Micronas to use the SMPTE Italian wedding footage for a true torture test.


One place were this technology could work best is not necessarilly for external video processors but MPEG4/H.264/VC-1 decoders. H.264 and VC-1 both have block motion vectors that state how the image is moving from one frame to the next; typically these vectors are only used for creating the 24p image but they could be used to create intermediate frames in a way similar to what Micronas is doing.


All of this sounds like science fiction but it did work at CES to some extent.


-Mr. Wigggles
 

·
Registered
Joined
·
1,622 Posts

Quote:
Originally Posted by William /forum/post/0


It would be internal up-sampling. You would send 1080 24fps, 24Hz, 48Hz, 30Hz, or 60Hz and all would be up-sampled to 120Hz. The only question is what to do with PAL 50Hz?

Ah, ok. PAL could possibly be slowed down in many cases.
 

·
Registered
Joined
·
634 Posts

Quote:
Originally Posted by tahustvedt /forum/post/0


Ah, ok. PAL could possibly be slowed down in many cases.

After speeding up in the first place?
No, the only solution is 600Hz displays!
 

·
Registered
Joined
·
4,644 Posts

Quote:
Originally Posted by nietzscheman /forum/post/0


Does anyone know of any front projectors out now or in the near future that will have 120 hz refresh?

The InFocus DepthQ accepts and displays 120Hz signals. It was introduced in April 2005.
Quote:
Originally Posted by MrWigggles /forum/post/0


Micronas put on a pretty impressive demo where they took 30p movies and made them 60p. (see attached image)

Is it better than Digital Natural Motion (a.k.a. Trimension DNM ) which Philips introduced several years ago?
 

·
Registered
Joined
·
6,092 Posts

Quote:
Originally Posted by Erik Garci /forum/post/0


Is it better than Digital Natural Motion (a.k.a. Trimension DNM ) which Philips introduced several years ago?

I think my Phillips HD CRT has that processing (called "Pixel Magic" or something like that) for its SD inputs. It looks o.k. but I hardly ever use my TV's SD inputs.


The Micronas demo looked great with real world images BUT that doesn't mean it will work well with all images. Without seeing how well it works with brutal test footage I can't really tell you how well it works for everything.


The SMPTE Italian wedding footage is the most brutal I can think of and at the same time it could really use intermediate frames. It is shot 24p with a pretty small shutter angle (I would guess less than 120 degrees) and there some very difficult scenes that really show the limitations of 24 frames per second shooting as much as anything. I've never seen the SMPTE footage in a darkened theater only 30ft-L or higher displays and I've yet to see any of the displays handle it well. (Here's a sample image from the movie; various companies use it to show off their image quality: http://www.smpte.org/smpte_store/tes...tEM_Flyer3.pdf ) I would love to see how well the Micronas could work with that image.


With these type of technologies, the batting average has to be high or the viewer will hate the artifacts of the additional frames more than the "judder" of the original source. So only after I've seen a technology like this shown with a multitude of sources can I say its good or bad. Right now its simply "promissing".


I want to reiterate that the best place to perform this frame-creation technology would likely be at the MPEG et al encoded source. The encoding process (which is rarely realtime and much longer than the decoding process) puts a lot of work into figuring out how much simularity there is from one frame to the next. That information could be very valuable to a frame creation algoritm. So in princple I really think the best location for these creation algoritms is in the source NOT so much the displays.


-Mr. Wigggles
 

·
Registered
Joined
·
4,644 Posts

Quote:
Originally Posted by MrWigggles /forum/post/0


I think my Phillips HD CRT has that processing (called "Pixel Magic" or something like that) for its SD inputs. It looks o.k. but I hardly ever use my TV's SD inputs.

It's probably Pixel Plus, which is a term that Philips uses to refer to their video processing technologies, such as scaling and sharpening. DNM is one of those technologies, but it is not always included on devices with Pixel Plus. In other words, some devices have DNM, but some don't. Also, even if the device has DNM, it is not always enabled, since the user can choose to disable it in the menu.
 

·
Registered
Joined
·
6,092 Posts
Erik,


My TV does have the DNM technology. At times it is very interesting while at other times it doesn't seem to do much. However, my CRT is now 4 years old so I doubt I have the latest version.


It seems that most big name companies at CES had some sort of frame insertion technology. I don't remember if Philips did or not. The demos were typically so similar that I don't even remember whose was whose at this point.


-Mr. Wigggles
 

·
Registered
Joined
·
3,106 Posts

Quote:
Originally Posted by MrWigggles /forum/post/0



The Micronas demo looked great with real world images BUT that doesn't mean it will work well with all images. Without seeing how well it works with brutal test footage I can't really tell you how well it works for everything.


I heard Micron TruHD chip will be in the Sharp D92 line which support 120Hz. If so, we should know the result very soon.
 

·
Registered
Joined
·
9,223 Posts

Quote:
Originally Posted by MrWigggles /forum/post/0


Actually, that's exactly what some of the demos did - created frames that weren't there to begin with.


-Mr. Wigggles

Well of course you can always interpolate and with technology advancements there is no reason you couldn't have this capability in a low cost display. Probably by next year!


I guess I should say you can't add accurate frames where they don't exist
 

·
Registered
Joined
·
33 Posts
Discussion Starter · #19 ·
I am still confused. Isn't the idea to just take the existing frames and flashing them faster on the screen. You don't have to add any new information or frames to do that.


If you have film-based 24 frame per second material, you just increase it to 120 frames per second and the same for 60 frame per second video.


Am I missing something again? Is this more difficult technically than it seems?
 

·
Registered
Joined
·
4,644 Posts

Quote:
Originally Posted by nietzscheman /forum/post/0


I am still confused. Isn't the idea to just take the existing frames and flashing them faster on the screen. You don't have to add any new information or frames to do that.

That's the easier way, to simply replicate the existing frames.


There's also the harder way, to fabricate intermediate frames by tracking motion between the existing frames.
 
1 - 20 of 25 Posts
This is an older thread, you may not receive a response, and could be reviving an old thread. Please consider creating a new thread.
Top