Samsung F4500 plasma bargain for 480p sources

[Josh128]

The photo post on here sold me on this set, but of course I missed most of the blowout pricing a few months back. I was even looking for a decent 720p set specifically for older games too.

Dang.

If you are in the US Im pretty sure you can still get it online from Best Buy for $399. I saw some advertisements on Black Friday. If not, hit up Craigslist man-- The F4500s were first produced in 2013 (AFXZA model) and into 2014 (BFXZA model). I have both and they are both phenomenal.

[Xan]

I compared the manuals of the two and the H4500 does actually have the noise reduction option... it's strange they'd just remove it from their advertising.

Did any owners here try updating the firmware? A really long shot but maybe they added 288p support.

[Josh128]

Totally missed that the pillar bars are gray on those NES and SNES pictures, this further reinforces my opinion that a transcoder on a VGA CRT would be the better choice for 4:3 480p stuff, if one can deal with setting it blindly.

See above-- you can make them pitch black if you so choose.

As far as "Digital Clean View", you dont want that anyway as it causes some strangeness with dark scenes in 480i content.

I have yet to see an H4500 in person, so Im not 100% sure but if I had to put money on it I would take that bet. The F4500s base is indeed black.

About 288p mode-- what games/systems would require that? So far, I have used NES, Genesis, SNES, N64 (the set REALLY does wonders with the 64), GC, Wii, Dreamcast, and X360 and have not found a game that either its 240p, 480i, 480p, or higher modes would not handle.

[Xan]

I don't care about that function of course, I just tried to check whether the H4500 is the same thing. Their naming and marketing is really confusing, but I guess that's not unique in the TV market.

288p is just the PAL equivalent of 240p.

[Josh128]

Do you - by any chance - know for sure what's the difference between the F4500 and the H4500 models ?

It physically looks identical to my F4500, but of course it has DVB-T and C tuners vs. the ATSC tuners of the F4500. Power consumption is identical to the F4500. It also has "Football Mode", which I dont have.

I cant say for certain, but I'd be highly surprised if it uses a different scaler/scaling algorithms than the F4500, BUT I dont know jack about the PAL system that you guys have over there (is it still 50Hz?) I dont know if it would have 288/50p but being a UK television I'd assume it should.


http://www.cclonline.com/product/149355 ... n/PLA0723/

[Josh128]

288p is just the PAL equivalent of 240p.

Yeah, I just remembered you guys have 288 and 576p over there!

[bobrocks95]

The photo post on here sold me on this set, but of course I missed most of the blowout pricing a few months back. I was even looking for a decent 720p set specifically for older games too.

Dang.

I forget if you're in America Einzelherz, but if you are, every Best Buy store near me has one in stock, and you can always get it shipped to a store near you for free. They have the 51" for $400 with free home delivery. Might've been cheaper during Black Friday or something, but I'd still say it's a good price (if it's as good as Josh claims).

[Josh128]

Some photos-- please understand Im using a 10 year old 3.2M pixel camera, and it is a pain in the ass getting the photos to look decent without blurring- these photos do not do the set justice. Some of the dimmer shots I used flash, and you can better see the detail in those shots. All photos are on games running on original hardware except for the GC games, they are running on the Wii. Also some of the double shots are with flash on and flash off. I wish I had an iphone 6, as they can take stunning pictures of plasma sets.

[22point8]

Wii Component














Wii RGB


PS2 (Snakebyte SCART) @ 480p > CSY-2100 > Profigold Component cable (http://www.profigoldcables.co.uk/catalo ... -p-50.html)(only paid £2 from cpc farnell)

[Josh128]

Nice. Are you the same 22point8 from AVS Forums??

[22point8]

Indeed I am, in fact I'll post my guide here for anyone to read so they can learn more: https://drive.google.com/file/d/0B_p3a0 ... djRE0/view

The camera on my phone is rubbish, my phone is feeling old now so I'll look for one with a better camera and 1080p video recording.

[Fudoh]

OK, I ordered a 43" one to give it a little spin over the holidays.

[BuckoA51]

he reason the Wiis output gets such a bad rap is that everyone always tests it on 1080p tvs, ALL OF WHICH HAVE HORRENDOUS 480p UPSCALING.

Also want to add, that's not true, the 480p scaling on the newest Panasonic and Sony sets I've seen is really good. I used to consider the ABT/DVDO Edge 480p scaling to be better than what most TVs can do but it's quite mediocre compared to newer sets.

[wildchild22]

Does Killer Instinct Gold for the n64 blank out when it switches between 240p and 480i?
(When choosing your fighter (480i) then fighting (240p))

[Josh128]

he reason the Wiis output gets such a bad rap is that everyone always tests it on 1080p tvs, ALL OF WHICH HAVE HORRENDOUS 480p UPSCALING.

Also want to add, that's not true, the 480p scaling on the newest Panasonic and Sony sets I've seen is really good. I used to consider the ABT/DVDO Edge 480p scaling to be better than what most TVs can do but it's quite mediocre compared to newer sets.

Depends on what you consider "good" I guess. My 2014 Samsung 1080p plasma doesnt look very good with Wii or GC 480p, nor do any other (Ive tried a few) LCDs Ive seen. Some of that may be to personal taste though. 480p upscaled to 1080p just looks very unnatural to my eyes.

[Josh128]

Does Killer Instinct Gold for the n64 blank out when it switches between 240p and 480i?
(When choosing your fighter (480i) then fighting (240p))

It sure does, because the game switches from 480i in teh select screen to 240p for the fights. Its annoying, but thats about the only game that does it so often.

[Josh128]

OK, I ordered a 43" one to give it a little spin over the holidays.

Cool, Ive never tried the 43" version, but I would have to assume it should be essentially identical. Remember to turn off the ECO mode and tweek the cell light and sharpness.

[Fudoh]

I was torn between the 43" and the 51" version. EUR 299 vs. 399, but since I don't intend on keeping the display, the 43" is easier to handle. I assume their processing is identical, maybe with a little less of a screendoor effect on the smaller one.

[Ed Oscuro]

If you can compare with the recent Panny, with its better lag time, that'd be great too!

[RGB0b]

I'm also curious about lag. Can anyone post the results from the manual lag test in the 240p suite?

[bobrocks95]

Well, a more scientific method is preferred, though I know plasmas muddle the readings a little bit because of how they work.

[Josh128]

he reason the Wiis output gets such a bad rap is that everyone always tests it on 1080p tvs, ALL OF WHICH HAVE HORRENDOUS 480p UPSCALING.

Also want to add, that's not true, the 480p scaling on the newest Panasonic and Sony sets I've seen is really good. I used to consider the ABT/DVDO Edge 480p scaling to be better than what most TVs can do but it's quite mediocre compared to newer sets.

I just re-visited your post and I dont think we are on the same page-- Im talking about how 480p video games look upscaled on 1080p sets. This is very different from 480p/480i video upscaling, such as from a DVD or Blue Ray player. My new 1080p plasma actually does a very good job with 480i and 480p video/DVDs, excellent, in fact.

The issue is solely with 480p to 1080p video game upscaling. My 2014 PN51F5300 looks way better than the 1080p LCDs Ive tested with the Wii, and I'd be willing to bet it probably looks as good as or better than the Panasonic and Sony LCDs you speak of with 480p game sources-- problem is, I just dont like the look-- I think it makes the games look "fake" or "cartoonish".

The PN51F4500 on the other hand, looks much more natural with 480p and 480i games. By natural, I mean closer to "native", or as they look on a 31kHz 480p CRT. To my eyes theres simply no comparison. I would venture to guess that the only way you could get a 1080p set to look as good as the F4500 would be to use some sort up upscaling device that would fit the 480p into a 960p frame, and add dark gray or black scanlines in between each row of pixels. That would probably look quite good.

I heard the Framemeister is capable of such upscaling, but when I googled some images of it, it looked extremely dim/not so good. Perhaps Fudoh can elaborate on that. I would imagine though, if it would look as good as I believe it should, that people would be raving about it for the Wii and GC quite a bit more than is happening now.

[Josh128]

I'm also curious about lag. Can anyone post the results from the manual lag test in the 240p suite?

One of the first photos I posted show a single component feed split using Y splitter cables, feeding the F4500, F5300 and a 480i CRT. On screen, if you look, Im running the FZero SNES game. The games timer has a precision of 1/100th of a second (10 milliseconds). If you look carefully at the snapshot, you will see the plasmas lag the presumably "lag free" CRT by 3 hundredths of a second, or 30 milliseconds.

Now, I realize the SNES screen only updates every 16ms, so its possible to have 16ms of error in the test. I ran multiple tests, and sometimes the plasmas lagged 4 hundredths of a second. Because we measure our lag in gaming in the US in 1/60ths of a second-- I have concluded that the sets can vary between 32 and 48 ms of lag time from a component 240p source. Note I ran the tests both with and without game mode, and there was no difference whatsoever with either. It is said that if you rename an input to "PC", lag will decrease even more than game mode, but this is not possible with the component input, only HDMI.

What all the lag tests I have seen dont take into account is that there is very likely a difference between HDMI input lag and component input lag. I do more retro gaming on the set than anything else, this is why I did the FZero test. I would be curious to see what the Wii Homebrew lag test would show, run at least 10 times (because the lag does seem to vary a bit), with and without game mode. My lag is vs. a CRT, which I am just assuming is lag-free. (It certainly should be).

[Fudoh]

What all the lag tests I have seen dont take into account is that there is very likely a difference between HDMI input lag and component input lag.

that's unlikely. What should make a difference is a 15khz versus a 31khz signal. The 31khz (480p) signal might lag a little less than the 15khz one - despite the screen understanding what a 240p signal is.

I heard the Framemeister is capable of such upscaling,

no. But I think you could achieve the Samsung's look pretty much exactly as you like it on a 1440p screen with MAME and HLSL or other new shader techniques.

[Josh128]

What all the lag tests I have seen dont take into account is that there is very likely a difference between HDMI input lag and component input lag.

that's unlikely. What should make a difference is a 15khz versus a 31khz signal. The 31khz (480p) signal might lag a little less than the 15khz one - despite the screen understanding what a 240p signal is.

I heard the Framemeister is capable of such upscaling,

no. But I think you could achieve the Samsung's look pretty much exactly as you like it on a 1440p screen with MAME and HLSL or other new shader techniques.

I'd be quite surprised if an analog video feed and a digital video feed would have identical lag, as they use different circuits to decode / encode the signals. Not saying its not possible, but every analog to digital device I have used has introduced some noticable amount of lag, (I havent used an XRGB though) while the analog SCART to component converter I use appears to add zero lag.

You may be completely right though, I dont have a way to test HDMI lag.

[Xyga]

@Josh128
I think you just got a set with a well-engineered scaler, and on a plasma it is naturally even more enjoyable.
The plasma having lower pixel pitch but larger 'rounder', brighter and more contrasted dots, also better perceived motion thanks to the way plasmas are flashing the image, this all makes the experience look better.

The recent good quality 1080p sets like the Sony W series do scale 480p well, very, very well actually (and not just 'games' 480p, they're good at everything).
But the higher resolution makes it look sharper and blockier, with individually less powerful/colorful and contrasted pixels, and inferior perceivable motion.

I believe you have experienced what for instance maniac audiophiles call 'synergy'.
In your case:
The right display in terms of size/pitch for a signal and resolution its built-in scaler happens to handle very well.

It's awesome when that happens, but one cannot generalize on the topic of one native resolution being inherently better or worse for one particular job.
You are definitely not the first person I read saying how better video games look on his old 'hd-ready' set, just, every time the people reporting a similar experience with their particular set, seem to be exceptions.
It's a whole chain of things to do right, some engineers, sometimes, just do it...and it happens.
Unfortunately it is not written in big red letters on the products we buy, and we don't have enough reviewers in the whole world to try every set with every device to find out and inform us.

Regarding lag: I definitely did read about different inputs giving different lag test figures, but that was a long time ago (was it on SRK or avsforum ?) and those were only camera/stopwatch tests.
What later lag tests seemed to show is that different testing methods indeed give different results, but the reasons given dive deep into theory, with stuff about 600Hz temporal dithering and how soon our eyes are able to detect movement on a plasma etc.
Whatever...

[bobrocks95]

I just re-visited your post and I dont think we are on the same page-- Im talking about how 480p video games look upscaled on 1080p sets. This is very different from 480p/480i video upscaling, such as from a DVD or Blue Ray player. My new 1080p plasma actually does a very good job with 480i and 480p video/DVDs, excellent, in fact.

The issue is solely with 480p to 1080p video game upscaling. My 2014 PN51F5300 looks way better than the 1080p LCDs Ive tested with the Wii, and I'd be willing to bet it probably looks as good as or better than the Panasonic and Sony LCDs you speak of with 480p game sources-- problem is, I just dont like the look-- I think it makes the games look "fake" or "cartoonish".

The PN51F4500 on the other hand, looks much more natural with 480p and 480i games. By natural, I mean closer to "native", or as they look on a 31kHz 480p CRT. To my eyes theres simply no comparison. I would venture to guess that the only way you could get a 1080p set to look as good as the F4500 would be to use some sort up upscaling device that would fit the 480p into a 960p frame, and add dark gray or black scanlines in between each row of pixels. That would probably look quite good.

BuckoA51 runs a site called Video Game Perfection. We're on a gaming-centric forum. He's almost certainly talking about how well 1080p sets upscale 480p games.

I'm also curious about lag. Can anyone post the results from the manual lag test in the 240p suite?

One of the first photos I posted show a single component feed split using Y splitter cables, feeding the F4500, F5300 and a 480i CRT. On screen, if you look, Im running the FZero SNES game. The games timer has a precision of 1/100th of a second (10 milliseconds). If you look carefully at the snapshot, you will see the plasmas lag the presumably "lag free" CRT by 3 hundredths of a second, or 30 milliseconds.

Now, I realize the SNES screen only updates every 16ms, so its possible to have 16ms of error in the test. I ran multiple tests, and sometimes the plasmas lagged 4 hundredths of a second. Because we measure our lag in gaming in the US in 1/60ths of a second-- I have concluded that the sets can vary between 32 and 48 ms of lag time from a component 240p source. Note I ran the tests both with and without game mode, and there was no difference whatsoever with either. It is said that if you rename an input to "PC", lag will decrease even more than game mode, but this is not possible with the component input, only HDMI.

What all the lag tests I have seen dont take into account is that there is very likely a difference between HDMI input lag and component input lag. I do more retro gaming on the set than anything else, this is why I did the FZero test. I would be curious to see what the Wii Homebrew lag test would show, run at least 10 times (because the lag does seem to vary a bit), with and without game mode. My lag is vs. a CRT, which I am just assuming is lag-free. (It certainly should be).

Are you using a 15KHz CRT for that comparison, or one of the 480p/1080i sets you've been talking about a lot? I believe the HD consumer sets had a digital chassis, and thus had input lag (please someone correct me if I'm wrong on this one, it's confused me for a while).

I'd be quite surprised if an analog video feed and a digital video feed would have identical lag, as they use different circuits to decode / encode the signals. Not saying its not possible, but every analog to digital device I have used has introduced some noticable amount of lag, (I havent used an XRGB though) while the analog SCART to component converter I use appears to add zero lag.

You may be completely right though, I dont have a way to test HDMI lag.

From what I can tell it depends on the type of ADC (Analog-to-Digital Converter) used in the television. Wikipedia says Flash ADCs will convert everything in one parallel process, which would be much faster than serial-based ones. Supposedly an 8-bit Flash ADC could convert in tens of nanoseconds.

Of course, I have no clue how/if any of this is applied or what types of ADCs are in televisions, but it at least appears to be possible for a very low lag analog input conversion.

[Ed Oscuro]

Any difference between the component and HDMI inputs will be down to the set using different processing on different inputs - like the magic "use input #5" or "use VGA" suggestions.

The games timer has a precision of 1/100th of a second (10 milliseconds).

That's what it suggests, but we can't know that for certain. We don't know how it works - when it updates, how precise or accurate it is, or even whether it accurately tracks seconds! (Probably not an issue for really short runs like those here, but I'm not sure we can use this timer without better information on how it works. Time for some frame stepping tests...)

Now, I realize the SNES screen only updates every 16ms, so its possible to have 16ms of error in the test.

Well, that's an error in only one direction. You should be able to easily compute a partial "worst case" lag figure by dividing the timer result by 16ms and rounding up the remainder to the next full 16ms. Best case would be attempting to interpret the remainder as the actual amount after the last whole 16ms (assuming that is actually a full frame - that the timer is reliable). However, plasma is going to make the timer look faster than it really is, at least for that part of the screen. The CRT is updating those scanlines at the top of the screen close to 0ms after the source signal is received, while the plasma is drawing the screen top no faster than nearly 16ms after (the full push-to-screen refresh of a plasma described implies this). At the bottom of the screen, however, that will be displayed closer to 0ms after the source signal is received, same as the LCD, discounting other sources of processing lag for each device.

In any case, this site pegs the input lag at 50ms. DisplayLag has it at 46ms.

[bobrocks95]

Question: due to the way plasma screens update, does a plasma "feel" like it's less laggy than it really is? Like given an LCD and a plasma that had identical lag, would the plasma feel like it had less? I know my current one has a good 3 or 4 frames of lag, but it doesn't feel nearly as horrendous as it should, and things I've read seem to suggest the eyes are almost sort of tricked into seeing frames connected more smoothly due to the subfield drive and partial renderings and whatnot.

[Xyga]

@Ed; Both websites use only the Leo Bodnar tester though (displaylag average of three bars, and rtings only middle bar), which gives significantly higher figures on plasmas.
Hdtvtest during the year or two they still did review plasmas and used both the camera/timer and LB tester, reported differences of over 16ms (18ms precisely) between the two methods, stating the camera/timer was more accurate.
If hdtvtest are right then this set's lag is somewhere around 30ms.
Even so that doesn't take into account the theory saying the eye sees the picture start moving even before the plasma is done flashing all the sub-fields and peak brightness is even reached. Which of the camera or the eye sees it first then, and what's the error margin ?
It's probably only about 1/2 a frame, maybe less, idk, bet we will never know with enough precision.

@bobrocks95; well, there's that theory ^