4K Blues

[Xer Xian]

The RadiancePro line debuted in 2015 I believe, and it supported 4K60 input and output. The Radiance 21XX line supports 4K60 output (but no input). I think neither of them has 4:4:4 internal processing, but it makes sense given that UHD content is distributed with 4:2:0 encoding. These devices generally aren't aimed at videogames enthusiasts. Also they cost $$$. But I don't know what orange808 is referring to.

That said, I have a feeling that 4K upscalers aren't strictly necessary for retrogamers. It seems that many UHD TV sets play along with the OSSC just fine, and with no added lag.

[orange808]

The RadiancePro line debuted in 2015 I believe, and it supported 4K60 input and output. The Radiance 21XX line supports 4K60 output (but no input). I think neither of them has 4:4:4 internal processing, but it makes sense given that UHD content is distributed with 4:2:0 encoding. These devices generally aren't aimed at videogames enthusiasts. Also they cost $$$. But I don't know what orange808 is referring to.

That said, I have a feeling that 4K upscalers aren't strictly necessary for retrogamers. It seems that many UHD TV sets play along with the OSSC just fine, and with no added lag.

I haven't seen the new Lumagens. Looks okay.

Honestly, I'm not shopping most 4k machines because they don't do anything interesting. Many machines are using full frame buffers without any good reason. A full frame buffer for progressive inputs with no geometry features isn't worth my time.

I was hoping that the move from prefab commercial scaling chips to custom FPGA designs would deliver faster processing, but that hasn't been the case. New machines are still using frame buffers and taking their sweet time.

There's also a tendency for manufacturers to just blatantly lie about their products.

I'm watching the video wall machines now. Really good video wall and geometry (rotation) features are the only missing link in my setup.

[copy]

I made the leap and got the 49" Sony X900E a couple of weeks ago. I haven't spent a great deal of time with it yet, but so far I'm very pleased. I just did some quick SNES testing through the XRGB-Mini, and everything seemed to play nice together. It looks good in both 720p and 1080p modes.

Now that I have a TV that supports Deep Color, I'm really loving how that setting in the XRGB-Mini eliminates the color noise problem. It makes a huge positive difference.

Interestingly, the X900E accepts the Mini's 50 Hz output modes. Not that I'll need them for gaming, but it's pretty cool knowing it's a universal TV.

(I have an OSSC order on the way, so I'll be able to try that soon too.)

It's true that the Graphics picture mode does a straight 2x scale of 1080p to 2160p. However, I think I prefer the normal Game mode interpolated scaling, which is quite decent.

I think I'm close to dialing in all the settings how I like. One thing I've found is that I prefer the Auto Local Dimming feature set to "Low." The default of Medium seems to change the brightness too much and too often. I noticed this most clearly during the intro for Mortal Kombat II on the SNES. The clouds being lit up by lightning caused very noticeable and distracting dimming changes on Medium, whereas the Low setting kept everything more consistent while still having an acceptable black level.

Standard Blu-ray looks amazing as well.

[FinalBaton]

Nice! thanks for the writeup copy. can't wait to read your thoughts on the OSSC handling by the set, once you receive said contraption

With game graphics havin large areas of uniform solid colours, it makes sense to lower the local dimming I think.

I wonder if there's a 720p test pattern picture someone could make for you, so that you can test it in game/graphics mode to test how sharp 720p upscaling is in that mode

[Xyga]

No need for a 720p specific, you can draw one yourself in paint with a bit of patience. make lines of 1, 2 , 3, 4 pixels etc, horizontal and vertical, you an add circles, cencentric, triangles, text etc. then it can be used at various resolutions.

You can make very elaborate ones with plain colors also, but we essentially need to see if the scaling will mess up vertical and horizontal integers to the point visible artifacts would be visible to the naked eye, or not.
With Sony sets it seems there's always one or two settings that are 'clean' enough even in effective scaling modes, for every resolution.

(sorry I had a couple .png ready somewhere but I cannot find them anymore, too many hard drives duh)

[FinalBaton]

Xyga, I was referring to the Rtings pattern in their x900e review. They used a 1920x1080 image with a diagonal lines made up of 1 pixel than 2 then 1 then 2 etc. This is to test if the upscaling from 1080 to 4K is non-interpolated(sharp pixels) or interpolated(if it was then the line would have been smoothed out in their test result. but it was not). I want something similar but for testing handling of 720p on a 4K set

I guess I can make one myself in paint, yeah. In fact I think I'll do just that

[Xyga]

Where? I don't see that test in their review. Actually I've been wanting them to do something like that for a long time because the little squares bottom right of their useless boat picture definitely are not enough.

Anyway you just need one image file with a quantity of lines and shapes, then display it over different signals, this is what matters.
Prad.de for instance have been doing just that for ages.

If you google you'll find some but not very complete, it's better to make your own, one a la prad is good.

[FinalBaton]

Where? I don't see that test in their review. Actually I've been wanting them to do something like that for a long time because the little squares bottom right of their useless boat picture definitely are not enough.

Anyway you just need one image file with a quantity of lines and shapes, then display it over different signals, this is what matters.
Prad.de for instance have been doing just that for ages.

If you google you'll find some but not very complete, it's better to make your own, one a la prad is good.

down in the comments of the review page

[Xyga]

Ah ok but - gah - that doesn't help much since we already knew about the first part, those guys don't really know what to look for. *sigh* one day they'll learn and actually post scaling test we can use.

EDIT: damn, they don't have the set anymore so it's no use asking them to do it.

[FinalBaton]

we need to draw one for 720p

[Xyga]

I think you have yet to realize there's only one to make to test all resolutions...

[FinalBaton]

hmm... if you want to see how the set handles 720p signals, then you gotta make a 1280 pixel x 720 pixel image.

You need one for every resolution, otherwise how are you gonna test how it handles upscaling?

(speaking about pixel to pixel mapping here, not general sharpness/definition test. What Rtings did here is a 1-pixel wide line in a 1920x1080 resolution picture. This is the only way to precisely determine if the upscaling algorithm applies interpolation or not. Had the picture size not been in the res we want to test upscaling for(in this case 1920x1080), then they wouldn't have been able to test the upscaling of 1080p material. Same will fly for 720p. If you draw a "720p" upscaling pattern in a 1920x1080 picture, then you're not testing 720p upscaling : you're testing 1080p upscaling).

Unless you mean having them all in the same image, but exporting just the portion you need for each size, which... well is the same damn thing as "having one for each res", like I was mentionning. You're still gonna load different file on your set. Who cares if they're all crammed into 1 huge file, it's no less work to draw and you're gonna have to export it into different file sizes anyway. But from what I understand, you're speaking of using the same file for all res upscaling tests. Which is just wrong. Maybe I'm not understanding you correctly

btw drop the attitude. "I think you have yet to realize" is condescending af. I can do without that

[Xyga]

I won't even bother replying with an explanation because 1. I've actually already explained and gave a trusted example but that escaped you, I can't do anything about that, 2. you're being a real dick here, maybe you'll realize later but knowing how proud you are and how easily you take offense every time someone tells you you got something wrong, I doubt we'll ever know that.
I'm the one tired of the attitude of people on this fucking forum, you try to say something useful and you get that shit.

[Fudoh]

If you draw a "720p" upscaling pattern in a 1920x1080 picture, then you're not testing 720p upscaling : you're testing 1080p upscaling).

the other way around: your test image can be bigger than your screen resolution as long as you set the viewer to display it 1:1. You'll only see a portion of your test image, but that's no problem, is it ?

[FinalBaton]

the other way around: your test image can be bigger than your screen resolution as long as you set the viewer to display it 1:1. You'll only see a portion of your test image, but that's no problem, is it ?

had completely forgotten about 1:1 display modes, since my TV doesn't have one. But does the x900e has it? (from what I remember seeing as far as it's zoom/display modes, it didn't have one. but maybe I'm wrong)

EDIT : it has Wide/Normal/Full/Zoom modes. Does one of them display sources 1:1? In usual Sony nomenclature, none of them should. but maybe there's a separate option for that somewhere else in the menu?

[Fudoh]

You need a 1:1 display mode in your viewer, not on your TV.

[FinalBaton]

You need a 1:1 display mode in your viewer, not on your TV.

Also, if we have a 1920x1080 picture with a pattern to check all res(or a 4K one with similar patterns), doesn't showing it in 1:1 mode make the 1080p and 720p patterns occupy just a part of the screen real-estate, and not all of it? I thought we wanted to blow up a 1280x720 pixels picture on the whole screen to test 720p

[orange808]

You need a 1:1 display mode in your viewer, not on your TV.

Also, if we have a 1920x1080 picture with a pattern to check all res(or a 4K one with similar patterns), doesn't showing it in 1:1 mode make the 1080p and 720p patterns occupy just a part of the screen real-estate, and not all of it? I thought we wanted to blow up a 1280x720 pixels picture on the whole screen to test 720p

FWIW, you made valid points and you didn't deserve that. Xyga was right, but you are too. An explanation would have been nice, but you just got a fucking rant. There are two regulars here that do this shit to people. You didn't deserve that.

Maybe I'm making it three, but fuck-a-doodle do! You didn't deserve that.

As if partially quoting you and putting the snarky end on your sentence wasn't picking a fight.

You picked a fight, Xyga. Fuck.

[Fudoh]

I think I'm losing you - or you're just thinking around two corners.

Your point was to have a 720p source with a matching pattern to see that a 4K screen does with 720p content, right ?
The argument above was you just need one test pattern. If you want it to full fill your screen (which isn't even neccessary) then a 1080p pattern is enough to test all output resolutions up to that. If you set your source to 720p and make sure the 1080p pattern is displayed 1:1, then you're just seeing a portion of your pattern, but what you see is mapped 1:1 to the output signal (not to the display), so it's what you want.

If you don't want to rely on software rendering, then you can use a test pattern generator. Every generator is able to produce a 1:1 black/white line pattern, which is perfect to check the scaling quality. All DVDOs have this built in.

[FinalBaton]

If you set your source to 720p and make sure the 1080p pattern is displayed 1:1, then you're just seeing a portion of your pattern, but what you see is mapped 1:1 to the output signal (not to the display), so it's what you want.

okayyyyyyy that's the concept that I couldn't grasp before. I didn't think about doing all that stuff, display-wise, on the source's side. Like, it didn't even cross my mind.
I was solely considering feeding the TV a picture displayed in it's entirety and having to do the adjustments on the TV's side only. Looks like I was omitting a factor in the equation.

That makes sense to me now, thanks for breaking that down

[Xyga]

You picked a fight, Xyga. Fuck.

Since you're meddling I have a right to answer: No, fuck you. There was 00,000% intent on being condescending or whatever in anything I've posted until that sensitive idiot with too much imagination got offended.
Who came and picked a fight while we were discussing normally? It's in plain view you biased ass; that idiot out of nowhere asked me to check my 'attitude' calling me 'condescending'. I haven't done a fucking thing.
And this isn't the first fucking time he's messed with me like that so yeah that fucking pissed me off.

But you wouldn't be objective orange since I've had a feeling for a while that like him you've been offended by something I must have said somewhere, since our few exchanges have been giving an uneasy vibe. The reason must be just as wtf idiotic like I've disagreed on something technical whatever. So you waited for an opportunity to team up w/ someone because you couldn't man up and argue with me? Too bad if you had maybe there would have been some respect, instead you had to pop up here and be a mean dick.

Want an attack, really intended this time? here's one: you're two are annoying pussies offended by ridiculous things, you're behaving like SJW would, like you're entitled to tell people how they should behave and speak to you so they don't scratch your fragile egos. Pretentious patronizers.
Don't like me or whatever I have to say about displays? Fine. it's a bit hallucinatory that it would come from such insignificant things, I don't want to believe it's another case of members who can't stand me for political/ideological reasons, circling around and sending me negative vibes instead of saying fuck you to my face. The ever-growing shmupsreich sub-community would love to kick me out of here but I don't think they're the majority of this community yet, nor that you two are with them. Anyway whichever the reason you have only one choice: ignore me, you'll get the same silence in return. I think that's a good deal. If you don't want that, well I'm definitely set on proving members who dislike/hate me they're right to do so.

[FinalBaton]

until that sensitive idiot with too much imagination got offended.

Go fuck youself.

The fact that you didn't realize you were acting condescending, changes nothing to the fact that you absolutely did come across that way(it's evident to everyone but you). So my 'attitude' comment was 100% justified

[Xyga]

You're really too easily offended, if you want to see this as offensive maybe you could read back and realize you have ignored what I said about the unique file twice and then stubbornly repeated 'we need one for 720p' without giving a fuck about my posts, coldly ignoring people like that could pass as quite annoying. offensive would be much, except for pussies.

Seriously, it's the second time you've been against me saying I'm offensive towards you when I absolutely didn't mean to, someone disagreeing or insisting isn't an offender. So you're just as I say: pretentiously schooling the person you're having a conversation with because you don't like what they say, you make a tantrum about - oh my god - the tone of their sentences? Who the fuck do you think you are, again; check your own way to deal with people.

Exactly as I described; the mentality of a SJW.

[FinalBaton]

I absolutely didn't mean to

then why didn't you just said so?
Instead you just exploded in anger

seems to me like you're the one who gets offended...

[Xyga]

You're really dishonest. You were the abusive one right from the beginning plain and simple; look at the fucking chronology, you're the one who began by ignoring my posts only to finally care to reply starting with 'xyga,...' like I was some clueless idiot, I could have taken offense but I don't get really mad for so little, I first assume it's a bit of miscommunication and resume the conversation. But then you came and criticized my attitude for a slight increase in tone from my part, which I insist was simply insistance because you were purposedly ignoring me, and yet you had the nerve to teach me how I should address your Majesty, calling me condescending.
And again that's the second time you annoy me like that for such a fucking stupid little thing, last time I was saddened by how negatively you saw my point but I just gave up thinking fighting with you wasn't worth because I thought you were an interesting dude to exchange with. Seems I was wrong, you see I don't explode in anger randomly for no reason, I do when someone is being a annoying prick with me. Who's got an attitude problem? Now I offered you a deal that we each go our own way to end this. Do you want to go on? I don't mind.

[FinalBaton]

here's the thing : my attitude comment is a little thing. It's a warning. It's not a nuke by any stretch of the imagination

whatever, let's each walk our own way now

[Xyga]

You're a specialist at twisting reality at your advantage, too bad it doens't actually work here.
You've been having a problem with me for a while anywyay, I now have one with you, so let's indeed ignore each other from now on.

Fuck, people should just either consistently ignore the posts of the members they dislike, or be honest and say out loud to the other member "I don't like you so I don't want to talk with you" instead of pulling that sort of shit.

[copy]

Back to the X900E. I got my OSSC this week, and have finally had a chance to try it briefly today. Some very quick initial impressions with the SNES:

Passthru and 2x modes don't seem to work at all. The TV reports no signal.

3x works, but with frequent dropouts. Interestingly, the TV gives me a choice of two "Wide mode" (image size) options, "Normal" and "Full 1." Normal windowboxes the image at about 67% screen height (doubling 720 to 1440?), while Full 1 displays it at full height.

4x works perfectly with no dropouts. Now the TV offers sizes of Normal (~88% screen height -- doubling 960 to 1920?), Full 1 (full height), and Full 2 (full height, but stretched to 16:9).

5x also works with no dropouts. The only size available is Normal.

I'm surprised that 2x and 3x have problems. Maybe this is due to the SNES's quirky video? I'll have to try other systems, and I need to learn more about the OSSC's settings to see if they would have any effect.

In any case, so far the 4x mode seems rock solid and looks fantastic. The image is sharp but scrolls smoothly, and the colors are great. I could happily play 240p games in 4x mode all the time on this TV. It's very nice indeed!

Oh yeah, and I think I can feel the lack of lag versus the XRGB-Mini.

I wonder if there's a 720p test pattern picture someone could make for you, so that you can test it in game/graphics mode to test how sharp 720p upscaling is in that mode

I do have a few calibration discs with 720p native patterns, like Digital Video Essentials, so I'll take a look at that sometime soon.

[copy]

Hm, so it turns out I am getting occasional dropouts in 4x mode, although they're much less frequent than in 3x. It's a bit worrying, though.

Also, I forgot to mention that the X900E does not like the HD Retrovision SNES cable at all. I can only get a second or two of picture at a time, and then long stretches of nothing. Bummer.

[Fudoh]

It's unlikely that you get dropouts on the TV because of the resolution you're sending to it. There are options on the OSSC (like sync lpf) which you can try to adjust and which are likely to help. Check the wiki for all sync options.