Calling all CRT owners: photos please!

[MajorPainTheCactus]

@UnixRoot

That’s why I took my photos with 6500k. It’s the white point almost every console and PC is designed for. I don’t know why you decided to use 5000k as a standard. 6500k also is the standard for sRGB on the PC, where we’re comparing these pictures.

Yes you’re absolutely right 6500K is the standard for CRT displays and more generally. However 5000K is seen as neutral white (although opinions vary) and this is for the camera i.e the camera is neither adding nor removing warmth.

We could set the camera warmth to 6500K but then you’d be effectively adding 1500K (i.e 6500K - 5000K if you accept 5000K as the neutral white point) onto the TVs temperature in the photos.

However really all we care about is that we know what settings a photo is taken with so we can adjust if needed locally.

[UnixRoot]

D65 corresponds roughly to the average midday light in Western Europe / Northern Europe and is used in all colorimetric calculations requiring representative daylight. I really wouldn’t call 5000k neutral, especially not indoors.

But enough of this. I may have a calibration OCD.

[ToD]

One of the issues with providing CRT color temperature information is that CRT displays are able to display a wide range of color temperatures, just like any modern display. So (for many CRT displays) it doesn’t make sense to have some simulated value. In fact, BVMs are typically factory calibrated to D65 and D93. Per the manual:

They don’t run warm or cool, they are calibrated to multiple standards. This applies to many CRT PC monitors as well and most of the professional 15KHz screens. The BVM-20F1U has three color temperature presets, which are all user adjustable.

I realize most consumer CRT TVs did not have color temp controls (at least, not without going into a service menu). But even for those, I imagine there’s a lot of variance there even between any two of the same models.

I think some misconception regarding CRT display white balance has arisen due to people taking pictures using auto white balance settings, and cameras often guessing a WB which is way too cool.

[MajorPainTheCactus]

Yes maybe Im getting the wrong end of the stick as to what you’re both getting at but from my perspective what the CRT white temp is has little to do with the white temp we’re asking that the camera is set to.

To a certain degree they’re kind of independent - one changes the WB of the console output signal and the other changes the WB of the image from the cameras sensor.

You could I suppose make up for WB differences introduced by the tube and room lighting (in between) by adjusting the camera WB but we can rule out room lighting by having none and then we kind of want to capture what the tube is doing to the signal so we kind of want that information ie that the tube makes it warm/cool.

Maybe there’s something I’m not getting here/missing?

[MajorPainTheCactus]

I’d say midday light is warm as in its the peak of the day it’s not neutral. I do agree we can probably argue about what is neutral until the cows come home although most references I’ve seen have it at or under 5000K - quite a few light bulb manufacturers have it at 4000K. 🤷

[UnixRoot]

Now I get it. You’re probably mixing up sunlight with daylight or neutral light in print vs digital media.

Yes, the sun color itself is closer to 5000K but D65 is also paying attention to atmospheric scattering and indirect illumination in more overcast conditions. This more or less reproduces real daylight conditions and not only the color of the sun.

D50 normlight is used in printing while D65 is the correct white point to use for digital media.

I work in a graphics agency. We use D50 lamps in the proof printing area and on the presses to compare the prints.

In our digital department (where I work) there is D65 standard lighting in all rooms and all monitors there are calibrated to 6500K.

So you’re not wrong with your 5000K if you want to compare prints. Now the cows are back home, I think.

[MajorPainTheCactus]

Just to be clear I’m thinking of the scenario of taking a picture of an emissive display in a pitch black room. I’m not sure how much of the above applies to that scenario?

I get the end viewing conditions of the photos are also important but isn’t that arguably beyond the scope of this?

If I’ve understood correctly I think the above instructions in the initial post are still sound and we’re not messing anything up with them. I hope that’s right at least!

[Duimon]

@UnixRoot

I work in a graphics agency. We use D50 lamps in the proof printing area and on the presses to compare the prints.

And if someone wanted to take an accurately colored photo of your team comparing the prints, the camera would need to be white balanced.

The old analog method would be to balance the camera on a large sheet of 50% gray stock that was temporarily introduced to the scene.

The 5000K setting is taking the place of the gray stock. I has nothing at all to do with the ambient lighting in the room, any daylight standard, or any display calibration.

It is a control for a standard output image.

If you worked in film instead of digital graphics your team would do the same. (Or white balance in post, which we can’t do here.)

---

Just my two cents. I have been a graphics designer for 27 years. (I started with Macromedia 1.0 and 3D Studio in DOS.)

[ToD]

One of the pleasant things about shooting in RAW is that white balance has not yet been applied to the image. The camera will store what the white balance was set to at the time the picture was taken to the EXIF data. RAW editors will default to that value, but it can be changed to any value desired. If a neutral gray or white area is present in the source of the image, that can be used as the white point.

Finding a neutral gray is a little trickier with close up and in focus images taken from a CRT, since there’s really no white. Only the individual phosphors are visible. One way to go about it is to take a photo from a known gray or white screen, like test patterns from the 240p test suite, and ensure the camera is out of focus so no phosphors, moire or other lines are visible. Use that in the RAW editor as your white point for other pictures taken with the same camera and display settings.

I don’t mind setting my camera to 5000K if that is more useful for your purposes, however.

[MajorPainTheCactus]

Just to note EXIF data is stripped out of images on this forum and a lot of upload sites (I’m pretty sure that happened last time I checked at least). That’s actually the main reason I say to note settings alongside the photo.

As for whether you should change settings I’m not sure - ideally it would be 5000K I suppose but because we can post edit the images to some common WB value for comparison/reference purposes as long as we know the original settings it’s not too much of a problem.

[RealNC]

If anymore with a CRT NTSC TV is reading this, I’d love some feedback on whether Super Castlevania IV looks like this:

Title screen

Health bar

Or like this:

What I’m trying to figure out is whether the NTSC Adaptive shader (as used by crt-guest-advanced in this case) should have its phase parameter set to 2 or 3.

[Cyber]

Based on my reading, I think SNES uses a 3 phase output.

The Guest-Advanced NTSC thread

There is a quick fix for this by using ordinary (positive parameter value) for 3-phase filtering. I mentioned earlier that 3-phase filtering can’t resolve dithering patterns in the first place, which produces a situation with maybe more viable options. More reading on this topic: The 256px presets use 3-phase NTSC output, which is what the NES, SNES, and N64 output, while the 320px presets reflect the more common NTSC output from the Mega Drive, PC-Engine, PlayStation, and most consumer video…

[Cyber]

@ToD

Looks like a small convergence adjustment is needed here. Not noticeable from normal viewing distance.

Just imagine you want to fix deconvergence while we actually start with perfect convergence then deliberately add deconvergence in our CRT Shader presets in order for them to look more authentic! Lol

[RealNC]

Found some really good, high res photos of a VGA monitor:

Credit: @PhilsComputerLab (I believe these are yours?)

[MajorPainTheCactus]

They’re fantastic pics of a dot mask - great colour - thanks!

[Tathan]

After months (I’m sorry), I’ve finally been able to go home and use my MX4200! Running Sega Megadrive and Master System games on an RGB Modified Wondermega RG-M1.

ISO 80 1/60 7550K

Once again, sorry for the delay!

[MajorPainTheCactus]

No, no, no! These are well worth the wait! Excellent close up shots! Really great reference material. Can you let me know what settings were used for your camera for this - indeed what camera you used?

EDIT: sorry missed the setting part on my first read. Brilliant stuff thanks for these!!!

[Tathan]

You’re welcome ahah, I also find them pretty cool! I would really have loved to give you pictures with a perfect screen calibration and also a Sony PVM but, the PVM I ordered on internet arrived completely smashed (400€ lost…) and the Screen Settings on my MX4200 seem to have been broken, half of the options just don’t do a thing and it’s quite difficult to have a perfect geometry.

(I may have lied a bit on the settings, those are the settings that were stored in my Phone, but maybe the camera settings had changed between the day I made the picture and the day I posted them)

NB: Oh, and, well, don’t know if it’s the right place to post this, but if you know someone near Belgium that might be able to repair a smashed 20" Sony PVM, let me know.

[PhilsComputerLab]

Yea With PC monitors, they developed quite a bit with the smaller dot pitch and higher refresh rates. This monitor is on the newer sign with OSD menu and digital controls.

[MajorPainTheCactus]

Oh sorry to hear your PVM arrived smashed. When you say smashed do you mean the screen or the casing?