Previous Thread
Next Thread
Print Thread
Rate Thread
Page 3 of 4 1 2 3 4
Re: 1080 i vs 720P
michael_d #215985 07/23/08 01:25 AM
Joined: Dec 2007
Posts: 7,786
axiomite
Offline
axiomite
Joined: Dec 2007
Posts: 7,786
 Quote:
1080i is the same resolution with the same pixel content, but missing 50% native data

Unless I have forgotten something, this is not quite correct unless you add "in any given frame" to the end of that sentance.

Any interlaced signal should contain all the data, just spread over two frames. Deinterlacing simply takes the two frames and merges them back into one frame with all the data.

So a progressive signal gives you all the data 60 times per second, an interlaced signal gives you all the data 30 times per second, but split over alternating frames.


Fred

-------
Blujays1: Spending Fred's money one bottle at a time, no two... Oh crap!
Re: 1080 i vs 720P
fredk #216038 07/23/08 03:24 PM
Joined: Jul 2004
Posts: 4,444
Likes: 16
M
connoisseur
Offline
connoisseur
M
Joined: Jul 2004
Posts: 4,444
Likes: 16
Good point, my statement was misleading. But even to say it with your preposition isn’t correct either as each video frame contains two fields. I was trying too hard to keep it simple and generic and avoid discussing interlacing, 2:2, 2:3 telecine pull down, reverse pull down, line doubling, 24 fps doubling….etc.

Re: 1080 i vs 720P
michael_d #216074 07/23/08 08:12 PM
Joined: Oct 2006
Posts: 375
R
Riker Offline OP
devotee
OP Offline
devotee
R
Joined: Oct 2006
Posts: 375
Mdrew, so much for being bored with the subject hey ? Great post, thanks so much for the explanation. This stuff can get pretty hard to understand.

So, since the broadcast signal is 1080i only, there should not be much difference between 1080i and 720p right ? because the VP is converting 1080interlaced to 720progressive. Now if the signal was 1080Progressive to start with, then 720progressive would not look as good..

That's if I understand what you are saying.

Bottom line, broacast HD signal kind of sucks either way. My older CRT display probably hides some the SD flaws so my "overall" impression of picture quality is quite good, the HD signals I have seen on my father in law new Panasonic are a bit sharper, but with a lot of noise, artifacts etc.. So, that's the main reason I'm still on standard digital signals and not hd digital signals yet.

Now, I'm sure a native Blue Ray at 1080P would be nice on his 1080P Panasonic. It might even look good on my 1080i CRT !!

Thanks again, Most informative.


Acoustic Zen Adagio, Veritas center, Axiom EP500, QS8s, Anthem AVM20, MC20,Adcom GFA7400
Re: 1080 i vs 720P
Riker #216075 07/23/08 08:34 PM
Joined: Apr 2008
Posts: 274
local
Offline
local
Joined: Apr 2008
Posts: 274
Broadcast is mostly 720, no? With content coming from varying sources and being sliced, diced, and otherwise processed before reaching the viewer via any medium, I'm kinda skeptical that any de-interlacing can do much good in the general case. I also wonder how many HD-DVD and BD titles are actual HD transfers, and not rescaled SD DVD transfers. There are certainly lots of crummy DVD's out there (eg. Aliens, The Abyss) that are just VHS/LD 4:3 transfers so I have to wonder if the same laziness applies to HD media.

<--- grumbles about how his Samsung was advertised as 1080p yet can't actually take 1080p input in any useful way.

Re: 1080 i vs 720P
Riker #216105 07/24/08 01:44 AM
Joined: May 2002
Posts: 10,654
shareholder in the making
Offline
shareholder in the making
Joined: May 2002
Posts: 10,654
Stephane, I see that this one has new life, and yes, there's no reason to become bored with the subject as long as the technology involved isn't grasped firmly, and unfortunately some of the comments here have been as incorrect as the contention of the salesman that started this. There's no "native data" problem with respect to 1080i transmissions; all 2 million plus pixels are the original real ones. None of this is "discarded" when the two fields of the 1080i transmissions are combined into one full frame.

A CRT HD set is inherently interlaced in its operation and the persistance in the screen phosphors results in the two fields being automatically combined for viewing; no deinterlacing processing is necessary.

HD sets other than CRTs are inherently progressive in their scanning(i.e., the full frame is displayed in one swipe, not two), so when presented with an interlaced input it has to be first deinterlaced electronically to combine the two fields into one full frame. When properly deinterlaced the full 2 million plus original pixels are reproduced without loss and the 1080i material has greater resolution than 720p material, which has only 900K plus pixels. As Nick pointed out(probably he had Gary Merson's tests for HomeTheater Magazine in mind)not all sets properly deinterlace, and in that case a better picture results from having a player, receiver or separate video processor do the deinterlacing instead, as has been frequently pointed out here.

So, the botton line again is that 1080(i or p makes no difference in regard to resolution)resolution is superior in resolution to 720(rapid motion is a separate subject), as the several charts and graphs of resolution/screen size/viewing distance combinations illustrate.


-----------------------------------

Enjoy the music, not the equipment.


Re: 1080 i vs 720P
JohnK #216106 07/24/08 02:04 AM
Joined: Apr 2005
Posts: 1,155
connoisseur
Offline
connoisseur
Joined: Apr 2005
Posts: 1,155
Hey John, its good to be back. I've been very busy for a number of reasons.

I have not seen that specific chart before and I don't really know what factors they computed to determine those viewing distances.

I may be able to find the reference that gave the relative bandwidths of 1080I and 720P as nearly the same. I still maintain that for a fully utilized bandwidth, the data content should be the same. However, the viewing formats may be considerably different leading to different optical "visions."

Anyhow, it appears that the salesman did not correctly state what was going on.


The Rat. M80s, VP-150, QS8s, SVS PC 20-39+, OPPO, Onkyo 703s, Harmony 880 Sony 60" SXRD HDTV
Re: 1080 i vs 720P
ratpack #216115 07/24/08 02:37 AM
Joined: May 2002
Posts: 10,654
shareholder in the making
Offline
shareholder in the making
Joined: May 2002
Posts: 10,654
Bernard, hope that you'll be able to participate more often now. There's no question that, as you say, the required bandwidth for transmission of 720p and 1080i are similar. Although the full 1080 frame contains over twice the pixels than does the 720 frame, 1080i is transmitted in two separate fields in sequence, so its transmission doesn't per unit of time require over twice the bandwidth, as would be required if all the pixels were transmitted simultaneously.


-----------------------------------

Enjoy the music, not the equipment.


Re: 1080 i vs 720P
JohnK #216233 07/24/08 07:40 PM
Joined: Jul 2004
Posts: 4,444
Likes: 16
M
connoisseur
Offline
connoisseur
M
Joined: Jul 2004
Posts: 4,444
Likes: 16
I used pore words and verbiage to keep an incredibly complex subject simple, which was obviously a mistake. It takes a great deal of time and effort to explain it in a way that others can understand, which is why I did not want to get into this. I do not claim to be an expert.

Seeing how I’ve dug a hole for myself with poor / incorrect verbiage, I’ll attempt to dig the hole even deeper.

In short, using the word eliminate/delete was a very poor choice of words, and as John pointed out, just wrong. Native information was also a poor descriptor. In my mind, ‘native’ means that a frame is in the original format as shot without any rearranging of pixel content in any way, shape or form. 1080P/24 is the only ‘native’ film format currently available for home entertainment that I am aware of. 1080P/30 and 1080P/25 are two ‘native’ video formats that I am familiar with. Why I chose the word eliminate is harder to describe and I’ll probability just create more confusion trying to explain my rational. It has always been a mind set of mine, and one I need to reconsider.

Per my understand………(maybe John can educate me differently)

All data that was converted from 1080p/24 is there in an interlace 1080i video stream. The data is there, but it is rearranged and also duplicated for further rearrangement during the 3:2 telecine pull down process to convert 24 fps to 25 Hz and 30 Hz video, which is further processed to get 60 Hz for NTSC and 50 Hz PAL displays (and also 72 and 120 Hz displays). Frames in an interlaced video stream are composed of two separate fields each, one field representing the odd number of horizontal lines of information, the other field representing the even number of horizontal rows of information. Some frames contain the odd and even rows from the same frame, and some frames contain fields from two separate frames. For every four frames of film, telecine conversion from film to video creates 5 frames, consisting of 10 fields for 30 Hz video. Instead of simply doubling two fields and tripling the other two fields to get the ten video fields for five frames [(A-A-A, B-B, C-C-C, D-D), which would make the video stream appear uneven (pause and hold affect)], the fields within each video frame are staggered in a sequentially reoccurring order for smoother playback (A1-A2, A1-B2, B1-C2, C1-C2, D1-C2). Legacy CRT displays would light up the fields in sequence, but not the whole frame. New digital displays light up the entire frame. When an interlaced stream is converted to a progressive stream, these fields are re-woven in an attempt to arrange them back to the original film based frame sequence. The process of reverse telecine to put all the fields back together in the proper sequence still baffles me, so I will not even try to describe it.

Digging my hole deeper, there are industry standards for the different 1080P formats which also specify different timing for each frame rate. Timing relates to the number of pixel clocks in the horizontal lines of information. [Vertical timings are standard to all 1080P formats.] In each 1080 horizontal line there are 1920 pixel clocks and also a blanking period that is composed of a front porch, a sync pulse and a back porch. Blanking is there for CRT displays to make sure each horizontal line (or field) is placed properly as it retraces across the screen from right to left. (the beam is turned off during the retracing period, which is where the term blanking came from). Pixel clocks for the sync pulse and back porch remain the same for all formats, but the front porch clocks vary from format to format. The front porch clock variation difference is for 48, 50 and 60 Hz to ensure proper sequential line placement. The sync pulse is what triggers the retrace and is towards the middle of the blanking period and the back porch is at the end of the blanking period. The front porch is at the front of the blanking period and varies in length. The back porch keeps the electron beam turned off long enough for the entire passage of the beam from right to left (maintaining proper sequential line placement timing). This blanking period no longer applies to modern digital displays, but the legacy signal / timings remain to standardize and maintain proper pixel clock timings of all the different types of display formats (1080P48/50/60/72/120 Hz).

While you’re probably thinking, “so what, how does this apply to the 1080i / 720 – 1080P comparison?”, it has to do with what you see. Timing to maintain proper horizontal interlaced line placement frequently gets out of sync. It also applies to the process converting an interlaced video stream to a progressive video stream. This is why I maintain the position that, in MY words, NATIVE 1080p/24 or any down conversion of that format will always yield a better picture than any other format, especially any interlaced video stream. Only the very best video processors can take a, in my words ‘non native’ video stream and convert it back to an original 24 fps or doubled/tripled 24 fps format, or apply progressive scanning for 60 Hz output.

DVD, HD/DVD, BR, broadcast HD, broadcast SD, video, gaming and film all vary to some degree in how these processing parameters interact and how well or poor the video processor will interpret each feed and lock onto it for proper conversion and output.

I'm putting my shovel away and warming up the back hoe......

Re: 1080 i vs 720P
michael_d #216252 07/24/08 10:24 PM
Joined: Jun 2003
Posts: 8,488
T
axiomite
Offline
axiomite
T
Joined: Jun 2003
Posts: 8,488
Here ya go, Mike.



I'm a helper!


bibere usque ad hilaritatem
Re: 1080 i vs 720P
michael_d #216258 07/24/08 11:14 PM
Joined: Dec 2007
Posts: 7,786
axiomite
Offline
axiomite
Joined: Dec 2007
Posts: 7,786
AHHHHH! duck and cover, duck and cover, the porches are coming!!

Mike, you write rather well with a shovel, I can hardly wait for the backhoe version.

I had never thought about the implications of converting a signal optimized for the analog world to the digital world. Now that we are in a world of arrays instead of beams blasting at phosphor dots you would think that getting all the data on screen properly would be simpler. Looks like we are not quite there yet.

John. You have a real tallent there: minimum words, simple explanations, maximum understanding. Not many people are able to write that effectively.


Fred

-------
Blujays1: Spending Fred's money one bottle at a time, no two... Oh crap!
Page 3 of 4 1 2 3 4

Moderated by  alan, Amie, Andrew, axiomadmin, Brent, Debbie, Ian, Jc 

Link Copied to Clipboard

Need Help Graphic

Forum Statistics
Forums16
Topics24,945
Posts442,486
Members15,617
Most Online2,082
Jan 22nd, 2020
Top Posters
Ken.C 18,044
pmbuko 16,441
SirQuack 13,840
CV 12,077
MarkSJohnson 11,458
Who's Online Now
0 members (), 990 guests, and 3 robots.
Key: Admin, Global Mod, Mod
Newsletter Signup
Powered by UBB.threads™ PHP Forum Software 7.7.4