720P vs 1080P is there really a difference

[gizzin]

Most people say they can tell the difference. I think I can most of the time. Then other people just say its the placebo effect. So whats the real answer?

[deank]

:) The real answer is that there is a difference if you're using FullHD 1920x1080 screen (TV) connected to a true HD video source (like BD player or PS3) via HDMI connection.

You can really see the difference between 720p and 1080p.

[ron spencer]

depends on how you view....pixel peeping there is a difference, but a 50" 720p and 1080p sitting back 10 feet and there is no difference, some say with anime you can see difference more, but sitting back, no way.

[laserfan]

Most people say they can tell the difference. I think I can most of the time. Then other people just say its the placebo effect. So whats the real answer?The real answer is: "It depends!" :)

It depends on the quality of the source, the quality of the playback device, the quality of the display you are using, and the size of it, and ability of the viewer to discern details i.e. there are "trained eyes" and there are "untrained eyes".

So it's impossible to generalize, but when I am asked to do so I will usually say that below 50" most folks won't see a difference. (Heck, most folks don't see a difference between Blu-ray and DVD on their crummy TV sets!). :p

[rotty]

Yes I agree with Deank, certainly on 1080 screens above 40" you can see the difference. But a good quality 720P picture is still good to look at. You can also see the 720 to 1080 upscaling effects on sharp edges if at angles on a 1080 Full HD TV.

Some TV's as you know are only 720 pixels vertically anyway, and ironically the picture should look slightly better than a 1080 signal because of no down scalling in this case being required.

In fact the simple rule of thumb is that if the signal resolution matches the screen resoloution without having to be changed to do so then thats the best image your going to get. With everything else being equal of course that is.

BUT

In theory, if you had a screen exactly double pixel x pixel in size (in hight and width i.e. 4X amount of pixels). then using intelligent interpolation to look at pixels either side and above and below to add the extra pixel in each position, you could make a picture look better in res than it is but this needs a lot of processor power to do in real time.

[Furiousflea]

I disagee stronly with the above.

Too many ignorant people with opinions (imo :p)

It depends on how good your eyes are, how good the processing is on your output display, how far away you sit...

Put it this way, I have double vision, can't read a number plate from 20 metres, and I can easily tell the difference between 1080p\720p without fail for any source.

Posts like the ones in this thread perpetuate the incorrect idea of how all this works...

[rotty]

I did say "With everything else being equal of course that is" in other words same quality etc.

As for the argument about perception due to vision, well thats crazy, you could take it to its ultimate conclusion and say a blind man couldnt tell the difference and you would be right

[SeeMoreDigital]

Most people say they can tell the difference. I think I can most of the time. Then other people just say its the placebo effect. So whats the real answer?Of-course there's a difference...

A 1280x720p resolution image contains 921,600 pixels and a 1920x1080p image contains 2,073,600 pixels - ie: over twice as many!

When viewing both resolutions on a full resolution 1920x1080 screen, you can certainly see a difference :eek:

[rotty]

of course. I certainly can see the difference, im not saying that 720p is bad or anything like that but I can see the difference.

I love it when people tell YOU what YOU can and cant see.

You can tell us what you can see but dont try and tell ME what I can see or cant, ill tell you that.

[idbirch2]

This has what to do with authoring?

[rotty]

Well as you know authoring sw can allow you to convert 1080 > 720 to save space etc

[Sharc]

Some relevant info is provided here:
http://s3.carltonbale.com/resolution_chart.html

[gizzin]

Thanks for all the opinions. I always knew 1080P looked better :)

[eric97]

depends on how you view....pixel peeping there is a difference, but a 50" 720p and 1080p sitting back 10 feet and there is no difference, some say with anime you can see difference more, but sitting back, no way.

yeah, couldn't agree more
generally 720p is enough for your computer and needn't more higher definition video size

[Ghitulescu]

Well, 720p is endorsed by EBU and supported by JVC for professional uses.

There are TV sets with 720 pixels height, however, I do not own one of these and thus I don't know if they do crop or are Dot-per-Dot, since most of them were sold before HDMI and YUV (and variations) are always overscanned. A 720 pixels TV with Dot-per-Dot coupled to a BD or HD player with Source-Direct would not scale the image and therefore would give the best image.

Now, here it comes the fps. It really depends on the TV if it supports 24, 25 or 30, since most LCDs upsamples these into a higher fps (eg 72, 75, 100), but not for marketing, just for technical purposes.

Now comes the interlacing - ALL flat screens need progressive sources, some do a better deinterlacing than others - if the software deinterlacing is better than the hardware one (TV) then the differences are kept to a minimum.

On the other hand, an 1080p plasma would have no problems in displaying 720p other than upsampling it to 1080.

So it depends. However, downsampling an 1080p BD to 720p than playing it on an 1080 display, well I think you can see the differences, especially when the original source is interlaced (1080i). On a big screen for sure.

But quality is a matter of taste. If you like it, perfect, if not, do some tests.

[Ghitulescu]

Some relevant info is provided here:
http://s3.carltonbale.com/resolution_chart.html

I'm not convinced about this chart. You also need to read the article and to understand that we are not talking here about analog photography, we are talking about how good you are in chopping 720 gold coins for 1080 beggars. We are talking integers here, and how good the "rounding" algorithms can be.

The guy had a 720p projector, so he couldn't have seen real 1080p in action. You can easily accomodate 480i into a 720p (1.5x scale) the very same way you can squeeze 1080p into 720p (again 1.5x), but you're limited to 720p (even less, because there are additional factors).

Now, if Mr. Bale would really wanted to write an objective article, he would have performed the test as following:
1-take 3 projectors, one with 1080p, one with 720p and one with 480p
2-take 3 sources, again one 1080p, one 720p, and one 480p
3-connect them accordingly - preferably by HDMI, since analog inputs are ALWAYS cropped (overscan)
4-assess the quality and detail for each screen, at predetermined distances.
5-he shouldn't forget to test 480i and 1080i (720 is always 720p) as well.

Apparently, Mr. Bale simply made up this chart from an XLS sheet using photographic criteria. I'm not even sure he used his projector to assess the images, well, because he knew the results beforehand.

But the article was written when 720p projectors were still the kings because the 1080p ones were difficult to produce (or very expensive), that's 2006, so a little bit of professional curtesy towards the manufacturers won't hurt anyone, well except the customers.

[Visor]

Of-course there's a difference...

A 1280x720p resolution image contains 921,600 pixels and a 1920x1080p image contains 2,073,600 pixels - ie: over twice as many!

When viewing both resolutions on a full resolution 1920x1080 screen, you can certainly see a difference :eek:
This is the same type of argument that game console companies have been trying to sell people on for the past couple of decades. People can talk numbers all they want, but what is the real-life result? With that, I have a couple of experiences to share...

First, to expand on Ron Spencer's comments, I have a 52" Sony 1080p set that I sit ~10 feet away from. I watched the first 10 seconds of the opening IMAX scene of The Dark Knight in both 1080p and 720p using test files burned to AVCHD DVD+RWs. This is a great shot to use, as the smaller buildings on top of the large building that the camera flies over have grills/divets in their structure. As the camera gets closer to the buildings, the detail in the grills/divets goes from soft hints of grills to being very well defined. I was watching for how long the individual grills took to become visible, and if any aliasing was shown. When comparing the same shot in those two resolutions, I was very, very hard pressed to find any difference. There may have been an extremely slight advantage on the 1080p version, but as Gizzin said, that could've just been the placebo affect screwing with my head. :p

My friend has a Mitsubishi 1080p projector, but unfortunately I don't think his 120" screen is very good (it's white as opposed to a more gray screen that people recommend these days). Anyway, even if the room is completely dark, the picture is too washed out to tell much of a difference between DVD & Blu-ray, let alone 720 vs 1080. That said, I'm not sure how much more a gray screen can really improve on this to really show 1080p to its full potential(?)

I've found that the jump from 480i/p to 720p is much more significant than trying to split hairs to when comparing 720p to 1080p. The main difference between ATSC and NTSC are the colours... ATSC's are so strong, vivid, and realistic compared to the archaic NTSC colour system. Of course, shapness & clarity also makes a huge jump when moving up to ATSC.

Am I against 1080p? Of course not... I'm sure in a few years newer display technologies will improve on the ability to tell the difference (hurry up, OLED!). If you're buying Blu-ray, then by all means celebrate the fact that your movies will look even better in the future. However, considering this site is mainly to do with backing up movies, a lot of people are probably asking the question of just how far they need to go when making, say, AVCHD discs. Is it worth spending the extra $$$ to buy dual-layered DVDs in order to burn 1080p AVCHDs, (and only the expensive Verbatims, as everyone will tell you), when you can burn a 720p version on a single-layered DVD with the same quality?

My two cents,
Visor

[Ghitulescu]

First, to expand on Ron Spencer's comments, I have a 52" Sony 1080p set that I sit ~10 feet away from. I watched the first 10 seconds of the opening IMAX scene of The Dark Knight in both 1080p and 720p using test files burned to AVCHD DVD+RWs.
Quality is a matter of personal taste. If you're happy with 720p why do you care about other persons' opinion?

Secondly, it matters if the images were natively filmed in 720p or 1080i/p, because otherwise a scaling (in studio or you did it at home) occured -> altering one image. A comparizon should be done with native images, on native screens, digitally connected to HD-players that do source-direct. So no scaling occurs at any place between the lenses of the camera and the phosphorus of the TV (LCD, projector ;)).

And you're correct in saying that from 480i to 720p was a bigger advance than from 720p to 1080i/p. It's the same thing as going in audio from a S/N ratio of 50 dB (typical for Cassette Tape Recorders with no noise reduction systems) to 100 dB (studio recorders). You feel it immediately that 80dB is better than 50 but you'd be left in doubt if there's any improvement in going from 80 to 100dB.

The next generation (4k) woudn't look so fantastic in comparison to HDTV (2k), unless you have TV screens of wall size. Because the advantages each new technology brought with were strictly correlated to the screen size: VHS, DVD, BD/HD-DVD, all required bigger screens in order to take advantage of their image improvements.

[rotty]

Hi Ghitulescu
Just for interest, increasing the signal to noise ratio from 80db to 100db would reduce the noise power (in watts) for the same wanted signal by over 97 times.
In other words the noise would be 1/97th the power it was when at 80db s/n when increased to 100db s/n.
A massive decrease.

[Ghitulescu]

Hi Ghitulescu
Just for interest, increasing the signal to noise ratio from 80db to 100db would reduce the noise power (in watts) for the same wanted signal by over 97 times.
In other words the noise would be 1/97th the power it was when at 80db s/n when increased to 100db s/n.
A massive decrease.

The electricity and math parts I know. It happens also that I worked in a studio. It happens that I also have good hearing and vision, so I can spot the defects (sometimes I curse this ;)). Your hearing is allegedly logaritmic, and the real power has very less to do with your sound or visual perception.

Have you ever tried this experiment: put one hand in cold water, the other in hot water, then after a while both in tap water. Your ears and eyes would do the same. All your senzorics would react the same.

You won't be able to distinguish 1080i from 720p if you're far enough that the elementary pixel would be less than 1' of your visual field. But you can spot the problematic issues if you're trained to do this. That's why VHS was ok for 20" TVs, DVDs were ok for 28" and so on. if you watch HDTV on a 20" TV you'll find very hard the differences (not a 20" monitor, because you're sitting close to monitor but 10" to TV).

Again, your capacity of discerning pixels is important when they are about the same as your visual accuity (again 1'). More, you'll see as pixelated, less you won't see the details, only and if the defects (artefacts).

The reason I "asked" for a test with native resolution, is to prevent that scaling/deinterlacing artefacts would negatively affect an image (not its own fault).

[deank]

I think this stupid topic with a stupid question had to be closed long ago.

If there was no difference then 720p wouldn't exist.

or

If there was no difference then 1080p wouldn't exist.

Engineers don't do it just for fun for people to wonder "oh is that really different".

The simple and CLEVER answer to it is: YES THERE IS (unless you are blind and can't see squat).

Dean

[smok3]

Engineers do it cos they are geeks, the actual boundaries of the new media is moved by the more artistic community.

[deank]

Ok, I won't argue about that :)

May be he should ask someone at some 'creativity' forum with a lot of young (or not so young) creative artists, who push the boundaries forward.

p.s. And engineers do it because they are asked/pushed, not because they are geeks. Otherwise you would still use snail mail to let me know what you think, instead of posting here. I'm not an engineer myself, so don't get me wrong.

[rotty]

Ok, I won't argue about that :)

May be he should ask someone at some 'creativity' forum with a lot of young (or not so young) creative artists, who push the boundaries forward.

p.s. And engineers do it because they are asked/pushed, not because they are geeks. Otherwise you would still use snail mail to let me know what you think, instead of posting here. I'm not an engineer myself, so don't get me wrong.

Absolutely
And in fact how there can there be an argument if a higher resolution is better than a lower provided everything else is equal, it is crazy, the answer is in the sentence.

And of course, these resolutions are ephemeral and so quickly replaced.

[Ghitulescu]

p.s. And engineers do it because they are asked/pushed, not because they are geeks. Otherwise you would still use snail mail to let me know what you think, instead of posting here. I'm not an engineer myself, so don't get me wrong.

I fully agree with this. No engineer would create a standard that is incomplete or contradictory, because the engineers are trained to reason. Only the marketing and legal issues force the technicians to do such non-senses.

It was a political decision how color NTSC came to this world, it was a marketing decision why 16:9 then 1:2.35 came, why HDReady TVs exist although no standard ever is fully compatible with their 1366x768 pixels and so on. And it was a marketing decision why you are allowed "in theory" to digitally copy your own creations, but in practice all consumer apparatuses will refuse it beyond the first generation.

People that think otherwise are naïve or have not given up the "Edison" syndrom - do you know that Edison, the genius, had filed more than 1000 patents all over the world and pursuit them very strictly, to prevent the competition by any means (and of course to cash the fees)? If Germany would have not blocked the faimous Otto patent (which US did not BTW with Edisons' ones), the automobiles would have been delayed with 20 years (or maybe more, because of WWI and black friday).

PS: I am an engineer.

[neuron2]

Thread will be closed if these OT ruminations take over.

[Visor]

Quality is a matter of personal taste. If you're happy with 720p why do you care about other persons' opinion?
Because this is a thread discussing the perceptive differences between 720p and 1080p. Regarding why I care about another person's opinion, please re-read my last post.

...because otherwise a scaling (in studio or you did it at home) occured -> altering one image.
Normally I would agree... and the scaling thing is one of the reasons I didn't even look at 720p sources for a while. However, for the sake of curiosity I have started looking at them, and as I commented in my last post, I wasn't able to detect any scaling artifacts in my test clip (or in any of the other 4-5 720p movies that I've also watched). I'm actually impressed with the scaling algorithms that are being used.

Thread will be closed if these OT ruminations take over.
Poor gizzin... I hope you were able to find a satisfying answer in all this! :D

Visor

[gizzin]

I think this stupid topic with a stupid question had to be closed long ago.

****************************

[deank]

:thanks: I'll take the clever advice of yours.

[neuron2]

Closing thread as it is no longer productive and is engendering bad feelings.