Retro Consoles - 0 or 7.5 IRE for Black?

[Speedy]

Do games on NTSC retro consoles like SNES and Genesis output using the SMTPE 7.5 IRE setup/pedestal for black.

It seems to me like it would be logical for them to output true black at 7.5 IRE since that was the US NTSC standard when these were current, but I'm honestly confused after reading the descriptions on things like the 240p Test Suite.

It's been difficult for me to figure out which they're using because it's also not easy to tell in most retro games whether shadow or highlight details are being crushed due to them being so low-res that there naturally aren't very many details.

I have all my more modern consoles hooked up as well which default to RGB Limited and the Wii U is RGB Limited only.

My preference is to calibrate my CRT for one setting and not have to adjust when switching between consoles, but I'm wondering if that is even possible.

[fernan1234]

This doesn't matter if you're using RGB, rather only if you're feeding your monitor an NTSC signal (e.g. composite, S-video).

When I do use a composite or s-video source it's usually Japanese system/content, and therefore I set the monitor to 0. The difference ends up being hardly noticeable on a CRT though.

[Speedy]

The difference seems substantial to me.
If I calibrate my CRT to 0 IRE then anything intended to use a 7.5 IRE pedestal is too bright.
If I calibrate my CRT to 7.5 IRE then anything that is expecting true Black to be at 0 IRE is too dark.

I've determined that a US PlayStation 2 outputs true black at 7.5 IRE (53.57 mV) if I can trust the THX test pattern on my SSX3 tricky disc.
If it's true that SNES outputs true black at 0 IRE (0mV) then this seems like something that should be considered when calibrating a CRT.

[fernan1234]

Maybe the difference can be substantial depending on the sources, but if you're using RGB from your PS2 then the issue would have to lie elsewhere, as again NTSC (or NTSC-J) pedestal settings/calibration only pertain to NTSC (J) signals.

[Speedy]

I'm thinking that the Wii U might be the only device I have that is RGB Limited only the more I think about this.

I'm going to try running it through an HDFury X3 to see if its "Colour Depth Upscaling" feature can properly convert the Wii U's signal from RGB Limited to RGB Full (I'm currently using an HDFury Nano GX which doesn't support this feature).

If this ends up working out then I think I'll be able to have all my consoles/devices operating with 0 IRE as true black and I'll be able to stop constantly adjusting the brightness on my display.

[fernan1234]

You may be mixing up a few things. RGB limited vs. full pertains to digital signals displayed on flat panels, and is a completely separate matter from analogue NTSC 0 vs 7.5 IRE. Neither one corresponds to the other. If you actually send limited or full RGB to a DAC and then to a CRT monitor you may not even be able to tell which one is which.

Regarding the HDFury's color upscaling, I used to think it could do this, but I've also heard that ironically (and stupidly) that function only works for YPbPr signals and not RGB ones.

[Speedy]

Agreed that RGB Limited vs. Full is a concept for digital signals, but I also have digital consoles like Wii U connected to my CRT in addition to fully analog older consoles.

Are you sure that the digital RGB Limited vs Full isn't related in to the 7.5 IRE vs 0 IRE being analog true black respectively though?

My understanding is that...

• RGB Limited's 16-235 translates a digital value of 16 to 53.57 mV which represents the analog 7.5 IRE being true black.
• RGB Full's 0-255 translates a digital value of 0 to 0 mV which represents an analog 0 IRE being true black.

That's a bummer to hear that the HDFury might not be able to translate 16-235 to 0-255 unless the source is YPbPr (you're right that this would make no sense).

[maxtherabbit]

my Portta HDMI to VGA DAC does not handle limited range color - it just treats all inputs as full range and outputs them as such

if you feed it limited range you get truncated whites and gray blacks, I suspect most of the basic HDMI DACs do this

[Speedy]

I had that same experience with the Portta HDMI to VGA converter truncating RGB Limited. Ran into this as well with a few other adapters like the Tendak and more before settling on my current HDFury Nano GX which passes through the digital RGB Full & Limited as-is to Analog RGB, but I'm realizing now that this is causing me issues solely with Wii U.

Really keeping my fingers crossed that stepping up one more level in HDFury's range to the X3 will solve this issue by converting the digital RGB Limited's 16-235 to 0-255 before outputting in analog.

[fernan1234]

For what it's worth, I use an X3 and leave the "color upscaling" on just in case, and with WiiU (which as you know only outputs limited RGB via HDMI) outputting 1080i to a BVM the black levels look correct, neither crushed nor gray blacks. I have no experience with other DACs + WiiU to compare though.

[Speedy]

The HDFury X3 arrived yesterday and I just finished testing it out.

Some things I've noticed:

• It doesn't appear to convert 16-235 to 0-255 as I hoped. I've set the "Colour Depth Upscaling" DIP switch to the ON position, but I've tried both settings and neither modifies RGB Limited or RGB Full. I'll have to continue toggling a darker BRIGHTNESS preset on the BVM whenever play Wii U, but clearly not the end of the world.
• I'm happy that it's able to pass 1080i & 480i without any issue though. I wasn't able to do this previously because the HDFury Nano GX output RGBHV which I ran through an Extron RGB device (sync combining) which struggled with interlaced content. I'm now letting the HDFury X3 output YPbPr through a RetroTINK COMP2RGB and I'm having no issues at any resolution (240p, 480i, 480p, 720p, and 1080i).

With that said, if anyone knows a solution to convert HDMI RGB Limited to HDMI RGB Full then I'd love to hear about it.

I'm testing brightness calibration from two different devices:

• Oppo 203: I'm using a PLUGE test pattern from a Digital Video Essentials Blu-ray and flipping between "RGB PC Level" (RGB Full) and "RGB Video Level" (RGB Limited).
• Xbox One S: I'm using the built-in brightness calibration test pattern and flipping between "PC RGB" (RGB Full) and "Standard (Recommended)" (RGB Limited).

These are the brightness settings I end up at on my D32:

• RGB Limited calibrates to a BRIGHTNESS setting of ~880
• RGB Full calibrates to a BRIGHTNESS setting of ~960 (used for everything but Wii U)

[Fudoh]

RGB limited remains a crappy format, that's not supposed to exist. Blame Nintendo for using it in the first place. It would have been so much easier if limited range was reserved for YCbCr color transmission, while RGB remained at DVI specs with full range.

As far as I recall (it's been a long time though since I used a HDF X3), you get an output without a black level offset if you're using HDMI YCbCr on the input side. If you're using RGB on the input instead the X3 probably just assumes full range by default, so it's doesn't remove the black level pedestal.

The "color depth upscaling" option refers to the treatment of 4:2:2 input signals.

In this posting* the OP claims that the Tendak converters recognizes the limited range input and properly converts to the full analogue output range.

* https://gbatemp.net/threads/getting-ful ... ry.539154/

[fernan1234]

If you really want to be sure, and maybe even get HDFury to release a firmware that converts limited to full RGB, you can talk to them through their discord channel. That's also where you can ask them for a firmware that will work with sources like PS3 (basically stripping HDCP, just don't ask for that since it's "illegal"). Here's an invite to their channel: https://discord.gg/ekqxup

I'm still not convinced RGB limited matters that much when converted to analogue though. If you have a Switch, you can change the theme to dark so it's easier to see, and change the output in the menu back and forth between full RGB and limited RGB. The difference is very small, one just looks a bit darker than the other, but you don't see the huge obvious difference of an overly dark picture or a gray washed up picture as it occurs when there is a limited/full RGB mismatch between source and flat panel. As I mentioned earlier, when I use a Wii U the black levels look fine for me (with the brightness set to 955 on my particular BVM, D24, which looks good for almost everything).

[Speedy]

RGB limited remains a crappy format, that's not supposed to exist. Blame Nintendo for using it in the first place. It would have been so much easier if limited range was reserved for YCbCr color transmission, while RGB remained at DVI specs with full range.

We can agreed to disagree that RGB Limited is a crappy format
I'm pretty sure that RGB Limited existed originally as an analog broadcast format (PVMs and BVMs in Studios using the format) and then later a digital broadcast format. I believe it was largely used so that BtB (Blacker-than-Black) signals could be calibrated on CRTs to align with the 7.5 IRE YUV NTSC broadcast standard.
I also personally appreciate having a signal that is BtB available for dialing in a PLUGE calibration.

As far as I recall (it's been a long time though since I used a HDF X3), you get an output without a black level offset if you're using HDMI YCbCr on the input side. If you're using RGB on the input instead the X3 probably just assumes full range by default, so it's doesn't remove the black level pedestal.

The "color depth upscaling" option refers to the treatment of 4:2:2 input signals.

Ahhh, thanks for sharing. That lines up with what fernan1234 had stated as well. Kind of seems like false advertising on HDFury's part though.

In this posting* the OP claims that the Tendak converters recognizes the limited range input and properly converts to the full analogue output range.

* https://gbatemp.net/threads/getting-ful ... ry.539154/

That's awesome. I actually remember (and forgot) I had previously realized this myself and originally thought it was a defect. I think you and I actually discussed this in another thread on Shmups. I think I have a Tendak lying around somewhere and will give this a try specifically for Wii U.

If you really want to be sure, and maybe even get HDFury to release a firmware that converts limited to full RGB, you can talk to them through their discord channel. That's also where you can ask them for a firmware that will work with sources like PS3 (basically stripping HDCP, just don't ask for that since it's "illegal"). Here's an invite to their channel: https://discord.gg/ekqxup

Thanks! I forgot that they had a Discord channel. I sent them an email yesterday, but I'll also reach out on Discord and see what I can accomplish there.

I'm still not convinced RGB limited matters that much when converted to analogue though. If you have a Switch, you can change the theme to dark so it's easier to see, and change the output in the menu back and forth between full RGB and limited RGB. The difference is very small, one just looks a bit darker than the other, but you don't see the huge obvious difference of an overly dark picture or a gray washed up picture as it occurs when there is a limited/full RGB mismatch between source and flat panel. As I mentioned earlier, when I use a Wii U the black levels look fine for me (with the brightness set to 955 on my particular BVM, D24, which looks good for almost everything).

I personally think the difference is substantial, but you can easily switch back/forth yourself and test it if you have an Xbox One by doing the following:

• Keep your D24 at 955 BRIGHTNESS as you have already calibrated and flip between "PC RGB" and "RGB (Recommended)" and look at the built-in open-eye/closed-eye calibration patterns
• You should see on "PC RGB" that everything looks as it's supposed to (no closed eye), but that on "RGB (Recommended)" (RGB Limited which the Wii U used) that the Closed Eye will be way too bright.
• In a very dark room, just manually using the knob set the brightness to ~880 (you can check the number by toggling out of the permanent setting screen - just don't press enter) as I have and I think you'll see that RGB Limited looks correct now.
• Then just toggle back and forth between your 880 & 955 brightness settings and look at how far apart they are - You likely won't be able to unsee it.

[fernan1234]

I personally think the difference is substantial, but you can easily switch back/forth yourself and test it if you have an Xbox One by doing the following:
Keep your D24 at 955 BRIGHTNESS as you have already calibrated and flip between "PC RGB" and "RGB (Recommended)" and look at the built-in open-eye/closed-eye calibration patterns
You should see on "PC RGB" that everything looks as it's supposed to (no closed eye), but that on "RGB (Recommended)" (RGB Limited which the Wii U used) that the Closed Eye will be way too bright.
In a very dark room, just manually using the knob set the brightness to ~880 (you can check the number by toggling out of the permanent setting screen - just don't press enter) as I have and I think you'll see that RGB Limited looks correct now.
Then just toggle back and forth between your 880 & 955 brightness settings and look at how far apart they are - You likely won't be able to unsee it.

I don't have an XBOne to do this test, and will admit I have not done any pattern-based testing for this, but to me the difference cannot be called substantial on actual content. If you think about it, on our BVMs the difference between brightness at 950 and 880 is actually not much (the range is from 0 to 2000, right?), and so if that alone is all it takes to bring down to correct black level a limited RGB digital source converted to analogue, then the difference can't have been that big in the first place, certainly not even close to when there is a range mismatch on a flat panel where it is painfully obvious.

Either way, if a small brightness adjustment is a complete solution, you just need to copy your picture settings from channel 1 or whatever you use for RGB to another channel, and then change the brightness setting on that channel, which you can devote to limited RGB only sources like the Wii U. It's just a single button press on your BVM's controller unit. I do this also for analogue sources that are a bit too hot, like SNES and N64. This is also what I do for sources that have different centering, since you can set H position per channel as well.

[Fudoh]

I'm pretty sure that RGB Limited existed originally as an analog broadcast format (PVMs and BVMs in Studios using the format) and then later a digital broadcast format.

The 7.5 IRE offset on analogue tapes was a thing indeed (though I don't appreciate it), but I can't think of any storage format or transmission standard what would use a digital limited RGB range. SDI in general uses a YCbCr format and all digital media formats do so as well. Digital Betacam uses YCbCr 4:2:2 and so do all other tape formats. Higher compressed formats use YCbCr 4:2:0.

If you can think of any broadcast device or storage that actually utilizes limited range RGB, let me know.

[kitty666cats]

Also there's the X-Vue Box1020 transcoder



I @'ed you in the CRT Discord, Speedy, but I figured I'd post these in here as well. Very helpful with combating the gray-ness. If you're using the Gamma Boost dongle it absolutely has to be right before any YPbPr transcoder you're using, and if you're using a HDMI to VGA I've found it is ideal to connect it directly to the output. If you're using the Portta & the Portta is plugged in, the dongle doesn't need external power. I can't vouch for the transcoder as I haven't tried it, but considering it's made by the same people I'd say it is probably quite good.

https://www.ebay.com/i/193321326624

Here is a listing I found for the transcoder if anyone wants to give it a whirl

[Speedy]

I don't have an XBOne to do this test, and will admit I have not done any pattern-based testing for this, but to me the difference cannot be called substantial on actual content. If you think about it, on our BVMs the difference between brightness at 950 and 880 is actually not much (the range is from 0 to 2000, right?), and so if that alone is all it takes to bring down to correct black level a limited RGB digital source converted to analogue, then the difference can't have been that big in the first place, certainly not even close to when there is a range mismatch on a flat panel where it is painfully obvious.

I locked my camera's sensor and took some photos to illustrate why the difference seems substantial to me. If you don't adjust the brightness then the image is either significantly too bright or you're crushing shadow detail.

Check out these photos below and let me know if you still think you can't see a significant difference:

Spoiler

Too Bright (RGB Limited - Brightness 960):


Too Dark (RGB Full - Brightness 880):


Correct (RGB Full - Brightness 960):


Correct (RGB Limited - Brightness 880):

----

If you can think of any broadcast device or storage that actually utilizes limited range RGB, let me know.

I don't know of any, but how about uncompressed video coming off an ultra high-end digital film camera like an Arri Alexa SXT M or Arri Alexa LF.
Do you think those output YCbCr or RGB?
If RGB, then 0-255 or 16-235?
If those don't output RGB though then you're right and I'd imagine that basically everything in Broadcast is YCbCr.
----

[img]Very helpful with combating the gray-ness. If you're using the Gamma Boost dongle it absolutely has to be right before any YPbPr transcoder you're using, and if you're using a HDMI to VGA I've found it is ideal to connect it directly to the output.

Thanks! I'll look into these.
Don't these adjust Gamma though instead of Brightness?

[Fudoh]

how about uncompressed video coming off an ultra high-end digital film camera like an Arri Alexa SXT M or Arri Alexa LF.
Do you think those output YCbCr or RGB?

I think YCbCr, but it's a little far fetched to reference those in order to call limited range RGB over HDMI a meaning full signal.

Would be interesting to learn where limited RGB in HDMI showed up first. I suspect that with the move from DVI to HDMI, TV manufacturers added YCbCr support, but weren't careful enough to always use different calibration slots for RGB/DVI vs. YCbCr/HDMI. This caused the usual "black crush/washed out dark scenes" problems and since TV manufactures give a shit, manufacturers of the sources (either Blu-Ray players of gaming consoles), eventually started to offer limited RGB as an option to counter TVs not capable of showing either the correct way.

[fernan1234]

Yeah looking at those pictures RGB limited at the brightness level where RGB full looks correct does indeed look very bright there. And of course you shouldn't be looking at RGB full with brightness lowered that much, so that one is irrelevant. I wonder if it's something about the XBone's output though. Would you say you see a similar difference in the case of the Wii U as you do with the Xbox One? That is, does the Wii U look that bright when you have brightness set at 960?

[kitty666cats]

I don't have an XBOne to do this test, and will admit I have not done any pattern-based testing for this, but to me the difference cannot be called substantial on actual content. If you think about it, on our BVMs the difference between brightness at 950 and 880 is actually not much (the range is from 0 to 2000, right?), and so if that alone is all it takes to bring down to correct black level a limited RGB digital source converted to analogue, then the difference can't have been that big in the first place, certainly not even close to when there is a range mismatch on a flat panel where it is painfully obvious.

I locked my camera's sensor and took some photos to illustrate why the difference seems substantial to me. If you don't adjust the brightness then the image is either significantly too bright or you're crushing shadow detail.

Check out these photos below and let me know if you still think you can't see a significant difference:

Spoiler

Too Bright (RGB Limited - Brightness 960):


Too Dark (RGB Full - Brightness 880):


Correct (RGB Full - Brightness 960):


Correct (RGB Limited - Brightness 880):

----

If you can think of any broadcast device or storage that actually utilizes limited range RGB, let me know.

I don't know of any, but how about uncompressed video coming off an ultra high-end digital film camera like an Arri Alexa SXT M or Arri Alexa LF.
Do you think those output YCbCr or RGB?
If RGB, then 0-255 or 16-235?
If those don't output RGB though then you're right and I'd imagine that basically everything in Broadcast is YCbCr.
----

[img]Very helpful with combating the gray-ness. If you're using the Gamma Boost dongle it absolutely has to be right before any YPbPr transcoder you're using, and if you're using a HDMI to VGA I've found it is ideal to connect it directly to the output.

Thanks! I'll look into these.
Don't these adjust Gamma though instead of Brightness?

I would suppose so, there's some high res pics of the packaging out there that can explain the device better than me. I've just found that it's been excellent for black crush / brightening the OSSC's image when scanlines are enabled (this is while using my NEC PC CRT). I used it briefly with Nintendo Switch on said monitor and it definitely helped with turning washed-out grays black again

[Speedy]

how about uncompressed video coming off an ultra high-end digital film camera like an Arri Alexa SXT M or Arri Alexa LF.
Do you think those output YCbCr or RGB?

I think YCbCr, but it's a little far fetched to reference those in order to call limited range RGB over HDMI a meaning full signal.

Completely agree that it's a ridiculous comparison That's why I brought it up though just to mention the furthest possible scenario from what we're dealing with.

Would be interesting to learn where limited RGB in HDMI showed up first. I suspect that with the move from DVI to HDMI, TV manufacturers added YCbCr support, but weren't careful enough to always use different calibration slots for RGB/DVI vs. YCbCr/HDMI. This caused the usual "black crush/washed out dark scenes" problems and since TV manufactures give a shit, manufacturers of the sources (either Blu-Ray players of gaming consoles), eventually started to offer limited RGB as an option to counter TVs not capable of showing either the correct way.

Agreed. I'd be interested in learning more about that as well.

----

I would suppose so, there's some high res pics of the packaging out there that can explain the device better than me. I've just found that it's been excellent for black crush / brightening the OSSC's image when scanlines are enabled (this is while using my NEC PC CRT). I used it briefly with Nintendo Switch on said monitor and it definitely helped with turning washed-out grays black again

Just be careful though with using a Gamma adjustment to correct a Brightness issue. You always want to set brightness correctly first and then adjust Gamma only after brightness is correct.

----

Yeah looking at those pictures RGB limited at the brightness level where RGB full looks correct does indeed look very bright there. And of course you shouldn't be looking at RGB full with brightness lowered that much, so that one is irrelevant.

I don't know that I'd call them irrelevant... just wrong... right?

That's my point that RGB Limited and RGB Full require different brightness settings and when using their opposite calibrated brightness level the requirement for calibrated brightness settings becomes that much more obvious.

As an aside, there are a lot of people in the Retro Gaming community who seem to believe that Brightness calibration of the black pedestal/setup level isn't as important with CRTs as it is for digital displays, but I'm confident that isn't accurate. The black pedestal/setup level is critically important on all displays to getting a good image... our eyes are way more sensitive to brightness than to color after all.

I wonder if it's something about the XBone's output though. Would you say you see a similar difference in the case of the Wii U as you do with the Xbox One? That is, does the Wii U look that bright when you have brightness set at 960?

Yes, I see an identical situation with the Wii U, Switch, Xbox 360, PS3, PS4, and Oppo BDP-93 as well.
Correct, the Wii U looks that bright at 960. I'm confident you'll find the same on you D24 if you explore this some more.

[Ryeno]

I use the dreamcast and calibrate calibrate for 0 IRE. Always calibrate for 0 and if a game happens to display 7.5 than accept that black will be a little washed out. Games are developed on PC aka 0 IRE so on average you're better off assuming the dev didn't bother to readjust to 7.5 IRE.

[masek]

Sorry to resurrect this oldie; but, I was looking for the same answer and may have it.

The Japanese consoles you mentioned were in fact NTSC... But originated from NTSC-J, not NTSC-M.

The reason why this might matter for you and me, is because NTSC-J and NTSC-M were identical EXCEPT in black levels. NTSC-J used a 0 IRE black, NOT 7.5 like NTSC-M.

Assuming this is the reason, I imagine some pixel art content never made a specific conversion to NTSC-M black levels. The question may now be: did any? Maybe NTSC-M firsts or exclusives ? (were there any?) Probably not worth trying to find out, and assume 0 IRE for consoles that originated in NTSC-J.

Probably not going get a response on an old thread, but wondering if anyone else has anymore details or confirmation. Specifically, more details around PS2 RGB output and video levels. Yes, it's RGB and you would assume full range, but that doesn't matter unless the PS2 has an LUT expand 16-235 to 0-255 if the content was mastered for NTSC video levels.

Edit: If I calibrate my CRT for 7.5IRE video black, Final Fantasy VI on SNES is way to dark and contrasty (beautiful color pop, though, but it will give you a major headache). So, further confirmation from me the content was considered only for NTSC-J. Unless someone was programming and testing the pixel art with sunglasses on.

[fernan1234]

Probably not going get a response on an old thread, but wondering if anyone else has anymore details or confirmation. Specifically, more details around PS2 RGB output and video levels. Yes, it's RGB and you would assume full range, but that doesn't matter unless the PS2 has an LUT expand 16-235 to 0-255 if the content was mastered for NTSC video levels.

It's easy to get all of these things mixed up, but they're all separate.

For RGB, IRE settings won't have an effect. If you are using RGB with analogue equipment such as a CRT, you will have correct black levels regardless (provided that brightness/black level is set up correctly on the display).

Full range vs. limited range will not apply to analogue equipment, it will only be relevant if digitizing the signal (which in turn can be in RGB, or YCbCr color spaces with various chroma sampling rates) for digital displays, in which case you only need to match the ranges between the digital output and the display.

And this is not to be confused with IRE 0 vs 7.5, which applies to NTSC analogue signals, namely in the CVBS (composite) or YC (S-video) type of signals, not to be confused with RGB signals regardless of the region (NTSC-J, NTSC-M, or PAL) of the source content. Also on the analogue side, YPbPr can too have different settings for Betacam levels, but those will not be relevant for gaming sources.

[masek]

Hi fernan,

Thank you for that. I think I understand the difference, maybe. But in the case of PS2 supporting YPbPr, Y/C, and RGB, would it not matter what the content was intended/mastered for? If you use Y/C, then we would assume 7.5 IRE if in NTSC-M. Then if you switch to RGB, would we assume non video(NTSC) levels? In this case, calibrate our CRT with a 2.2 or sRGB gamma curve in mind? (Lowest black to 80-100 nits peak)

Edit: I'm assuming the answer is yes. Since the general consensus for PS2 era was content created with PC RGB monitors. And the lucky ones to have a 15khz monitor can enjoy the full dynamic range of PS2 RGB 240p content. Where as the users with consumer CRTs just got what they got. At least, this would be the easiest solution on the mind. But, did they instead use NTSC master monitors or TVs to finalize the end result? If mastered on NTSC-J monitor it would not matter for RGB monitor users around the world, but for NTSC-M, everything would be darker. Does it make sense where I'm going with this? (As off base as it may be)

[fernan1234]

Thank you for that. I think I understand the difference, maybe. But in the case of PS2 supporting YPbPr, Y/C, and RGB, would it not matter what the content was intended/mastered for? If you use Y/C, then we would assume 7.5 IRE if in NTSC-M. Then if you switch to RGB, would we assume non video(NTSC) levels? In this case, calibrate our CRT with a 2.2 or sRGB gamma curve in mind? (Lowest black to 80-100 nits peak)

Edit: I'm assuming the answer is yes. Since the general consensus for PS2 era was content created with PC RGB monitors. And the lucky ones to have a 15khz monitor can enjoy the full dynamic range of PS2 RGB 240p content. Where as the users with consumer CRTs just got what they got. At least, this would be the easiest solution on the mind. But, did they instead use NTSC master monitors or TVs to finalize the end result? If mastered on NTSC-J monitor it would not matter for RGB monitor users around the world, but for NTSC-M, everything would be darker. Does it make sense where I'm going with this? (As off base as it may be)

To keep it simple, if you're using either RGB or YPbPr you don't need to worry at all how the game was mastered, or what region the game and/or console is. If using Y/C just set it to what looks correct.

Regarding gamma, CRTs will not have 2.2 gamma. CRTs have their own gamma, and 2.2 was an early attempt to approximate it on digital displays. A more recent attempt is the BT. 1886 gamma curve which is a better approximation in general, again for digital displays only. sRGB is a color space that is also only applicable to digital displays. CRTs again were their own thing when it comes to color space, which depended on the type of phosphors that it used (most commonly P-22, but also later on professional monitors EBU and SMPTE-C phosphors). This is also something you don't need to worry about when it comes to games--though most likely the majority of games where developed on PC CRTs using P-22 phosphors.

[masek]

I guess I'm just a stickler for this kind of thing. I really want to know if I'm viewing it with the creator's intention.

I get sRGB is a color space standard, but I thought it was generally thrown around with and assumed approximately 2.2 gamma since most CRTs assumed to behave similarly(with correct brightness/contrast assumed). The point I guess I'm trying to make, is if the content creators used an RGB monitor (regardless if CRT or a digital display calibrated to CRT reference), the final result on NTSC TVs would look off - most noticeably, clipped blacks. Less noticeably (most cases), color accuracy.

Regardless if I should worry about it or not(I probably should not), would that be correct?

Edit: sorry, meant clipped blacks, not raised on final NTSC result.

[fernan1234]

I guess I'm just a stickler for this kind of thing. I really want to know if I'm viewing it with the creator's intention.

I get sRGB is a color space standard, but I thought it was generally thrown around with and assumed approximately 2.2 gamma since most CRTs assumed to behave similarly(with correct brightness/contrast assumed). The point I guess I'm trying to make, is if the content creators used an RGB monitor (regardless if CRT or a digital display calibrated to CRT reference), the final result on NTSC TVs would look off - most noticeably, clipped blacks. Less noticeably (most cases), color accuracy.

Regardless if I should worry about it or not(I probably should not), would that be correct?

Edit: sorry, meant clipped blacks, not raised on final NTSC result.

Viewing these things with the creator's intent will be tricky to say the least, but the #1 recommendation would be, for the majority of game systems and games produced in Japan, to set the monitor's white point to D93 aka 9300K (cooler white than the warmer D65), the standard in Japanese game and video content throughout the standard definition era (and even up to the present?). No one can say for sure, but I do find that colors feel better with this white balance. We can be sure that consumers in Japan were (are) using TV sets with approximately this white point (though virtually no one would have calibrated it exactly) even while using RF or composite as the vast majority did.

I'd say anything else hardly matters and is not something you can really know if it matches what the creators' were expecting when referencing their test displays, or imagining what most people at home would see. Everyone was using NTSC signals, again mainly RF or composite, and they all had different chroma and phase (or saturation and hue) settings on their TV sets. When you see photos or footage of old developer stations, you often see them having an old Japanese PC CRT (including the types that could sync to 15khz and 24khz along with 31khz) next to small consumer TV sets, sometimes a color video monitor, where they were most likely seeing the result via RF or composite, which means they had the chance to check that the work they did on their RGB PC CRTs (which, just to clarify again, had a CRT gamma and CRT phosphor color space rather than 2.2 gamma or sRGB color space used on digital displays) looked as expected on the TV/monitors they referenced with NTSC signals, almost surely set to IRE 0. This means that you will not get issues with blacks, either clipped or elevated, when using RGB/YPbPr, or when using NTSC with IRE 0 (aka no "pedestal"), for games developed in Japan.

Oh, and yesterday I said that the phosphors on the PC CRTs devs used were likely P-22, but that only would apply to approximately the mid 90--early 2000s. No idea what they may have been using before that, but all phosphor color spaces are pretty similar, and especially for those older systems which used so few colors it doesn't really matter at all and that's just an added factor that makes it clear there's no way to know what was "intended" regarding colors. But for blacks there's no issues.

[masek]

Thank you fernan,

I'm following you. Your response gets to the details I was wondering. And I guess it further confirms to me we should assume 0 IRE these consoles to be safe. However, I'm not sure I can stand 9300K after using 6500K for so long. I will look into that.

Thank you!