CRT monitor vs. CRT TV (also SNES region modding)

Discussion of hardware and software development for Super NES and Super Famicom. See the SNESdev wiki for more information.

Moderator: Moderators

Forum rules
  • For making cartridges of your Super NES games, see Reproduction.
User avatar
rainwarrior
Posts: 8732
Joined: Sun Jan 22, 2012 12:03 pm
Location: Canada
Contact:

Re: Two Ship mode 5 hires + interlaced comparison

Post by rainwarrior »

I'm curious why S-RGB pin 19 doesn't need to be part of the mod described in Oziphantom's link... I found some pinouts on a wiki here, but the descriptions are light.

So, the PPU has a 50/60hz select. The S-RGB / S-ENC / BA6529F all have pin 19 PAL/NTSC pin. Does this select an internal oscillator for colour encoding? I thought TmEE was suggesting they needed an external oscillator. The japanese datasheet for BA6529F seems to have a diagram at the end with pins 13-15 attached to an external one, but the pinout description on that wiki says they're ground, not connected, and video out?

(I'm kinda wondering what it would take for a full PAL/NTSC switch, rather than just the PAL60/NTSC50 possibility.)
lidnariq
Posts: 11432
Joined: Sun Apr 13, 2008 11:12 am

Re: Two Ship mode 5 hires + interlaced comparison

Post by lidnariq »

In NTSC SNESes, PPU2 divides the reference clock by 6 to generate the chroma clock, but in PAL-B that's the wrong frequency. (Should work in PAL-M or PAL-N)

As seen in arzi84's schematic, on 3-chip PAL SNESes there's a separate chip (S-CLK) that takes a 4xPAL crystal and multiplies that by 6÷5 to generate the master clock, and separately also makes the PAL colorburst frequency.

But I also recall someone having a board that used a PAL crystal using the BA6592's on-board crystal driver. (Was that rework? I have no memory)

The only thing that's clearly true is that the chip itself needs two different phases on the "VB" and "VC" inputs, and both the BA6592 datasheet and jwdonal's redrawn SNES schematic show this. Getting the wrong amount of phase delay will somehow make the color wrong.

(Also: S-RGB and S-ENC have different pin orders)
rainwarrior wrote: Sat Jan 21, 2023 5:23 pm Does this select an internal oscillator for colour encoding?
Not the oscillator - only whether color correctly inverts the encoding of the V component from one scanline to the next.
User avatar
rainwarrior
Posts: 8732
Joined: Sun Jan 22, 2012 12:03 pm
Location: Canada
Contact:

Re: Two Ship mode 5 hires + interlaced comparison

Post by rainwarrior »

Ahh, so the PPUs control the synch rate, and the S-RGB/etc. are only responsible for colour encoding?

So to go all the way, you don't just need to switch that pin on the S-RGB, but also they have a colour clock input which is derived from the main onboard oscillator. So, it's not just swapping a crystal, but also whatever multiplier/divider stuff is needed to translate that clock for both the CPU and colour clock...

I can see how this quickly becomes a lot more of a problem than only changing the refresh rate. :S
lidnariq
Posts: 11432
Joined: Sun Apr 13, 2008 11:12 am

Re: Two Ship mode 5 hires + interlaced comparison

Post by lidnariq »

Hence why PAL60/50 mods were common... and otherwise people seemed to just use RGB.

It does leave a fair amount of interesting options for subtle tweaking the exact timing on the SNES that isn't possible on the NES, because the colorburst frequency isn't strictly tied to the pixel clock. Normally the SNES has the exact same 227⅓ or 284⅙ chroma periods per scanline that the NES does, but one could instead feed a master clock of exactly 1364·15625Hz for "perfect" PAL-B/-N scanlines or 279MHz÷13 for "perfect" NTSC/PAL-M scanlines without the diagonal dot crawl artifacts.

Was there ever a SNES or superfamiclone that emitted PAL-N or PAL-M?
User avatar
tokumaru
Posts: 12427
Joined: Sat Feb 12, 2005 9:43 pm
Location: Rio de Janeiro - Brazil

Re: Two Ship mode 5 hires + interlaced comparison

Post by tokumaru »

lidnariq wrote: Sat Jan 21, 2023 11:26 pmWas there ever a SNES or superfamiclone that emitted PAL-N or PAL-M?
All SNES consoles officially sold in Brazil were PAL-M. It looks like the only difference from NTSC is the crystal and the SCLP/SCLN jumper:

https://youtu.be/UopyK0BUoLs
lidnariq
Posts: 11432
Joined: Sun Apr 13, 2008 11:12 am

Re: Two Ship mode 5 hires + interlaced comparison

Post by lidnariq »

I'm really confused by the chroma artifacts that he shows in the footage of PS1 Alundra 2. Why would there be red and green fringes on everything on the PAL-M PS1's evidently-incompetent chroma encoding?

Anyway, it's gratifying to confirm that the difference between PAL-M and NTSC-M is exactly as simple as I assumed it would be here - tell the encoder to use PAL vs NTSC color encoding, and use a crystal that differs by 910/909.
User avatar
rainwarrior
Posts: 8732
Joined: Sun Jan 22, 2012 12:03 pm
Location: Canada
Contact:

Re: Two Ship mode 5 hires + interlaced comparison

Post by rainwarrior »

Obviously an interesting topic, but maybe we should split this off into a new thread with a more descriptive name at this point?

Suggest splitting here: viewtopic.php?p=285325#p285325

Suggest name: "NTSC vs PAL CRT and SNES region modding"
Pokun
Posts: 2681
Joined: Tue May 28, 2013 5:49 am
Location: Hokkaido, Japan

Re: CRT monitor vs. CRT TV (also SNES region modding)

Post by Pokun »

Nikku4211 wrote: Wed Jan 18, 2023 7:31 pm
regiscaelus wrote: Wed Jan 18, 2023 4:19 pm Thank you @linariq. It looks like I need to find a CRT monitor, then.
Note that CRT monitors and CRT TVs are completely different, so ideally, you'd want a CRT TV (if you already have a real console) to test real 15 kHz interlacing.
rainwarrior wrote: Wed Jan 18, 2023 9:10 pm A TV is a monitor that has a channel tuner and loudspeaker. That capability is probably not important for this purpose?
...
TmEE wrote: Thu Jan 19, 2023 12:41 am I imagine a PC monitor was thought of and they have one big difference, which is much lower persistence phosphors since they are not meant to display interlaced image. They can still show it but it will be even more flickery than a TV is.
From what I've heard, a professional 15 kHz monitor such as a Sony PVM may have more flickery interlace than a consumer TV due to the higher TVL they tend to have, which refers to the TV's general bandwidth and horizontal resolution. A high TVL means scanlines becomes more distinctive and doesn't meld together as much but also more flicker in interlaced images. A consumer TV typically has less than 200 TVL while PVM may have 200 to 1000 TVL or more, on around 1000 TVL you can see the black bands between scanlines clearly.

Mine is 600 TVL and SNES interlaced image is bearable, but noticeably more flickery than on a consumer CRT TV.
User avatar
rainwarrior
Posts: 8732
Joined: Sun Jan 22, 2012 12:03 pm
Location: Canada
Contact:

Re: CRT monitor vs. CRT TV (also SNES region modding)

Post by rainwarrior »

Yes, I'd expect a sharper picture to have a stronger perception of flicker. Blending vertically is the best way to mitigate it, and a vertically fuzzy picture output can accomplish this for the same reasons that a vertical blur of the source material does.

I think TVL is technically a horizontal measure of precision? I'm sure that in practice it is very much correlated with focus/blur along the vertical axis too.

I think brightness is a similarly related factor, i.e. an increase in brightness usually reduces sharpness?
Pokun
Posts: 2681
Joined: Tue May 28, 2013 5:49 am
Location: Hokkaido, Japan

Re: CRT monitor vs. CRT TV (also SNES region modding)

Post by Pokun »

Yes, as I understands it TV-lines (TVL) is the TV's ability to display distinct alternating black and white vertical lines within a given width. As Wikipedia says it's defined to be the number of vertical lines that can be resolved on a width that is the same as the full visible height of the tube. So for example a TV that has a 20 cm tall screen and is rated at a 400 TVL should be able display up to 200 white and 200 black distinct alternating lines on a 20 cm wide area of the screen.

And yeah even though it's a measurement of a form of horizontal resolution, in practice a high TVL supposedly gives sharper scanlines that are less likely to bleed together, thus it also increases the size of the black scanlines between the video scanlines in 240p signals (which are hardly visible on consumer TV-sets) besides affecting interlace flicker visibility. I guess both qualities of the phosphor coating and of the electron gun mechanism are responsible for the TVL score.

Arcade monitors typically has pretty high TVL which is why they have so distinct scanlines, and so has Sony BVM and such professional monitors used by TV broadcast companies or others that works a lot with video and needs this level of precision.

I'm not sure how brightness affects sharpness.
User avatar
rainwarrior
Posts: 8732
Joined: Sun Jan 22, 2012 12:03 pm
Location: Canada
Contact:

Re: CRT monitor vs. CRT TV (also SNES region modding)

Post by rainwarrior »

I mean that an increase in brightness, with all other factors the same, is accompanied by a spreading of the beam. At least this seems to be true for CRT's I've used.

One CRT can definitely be both brighter and sharper than another, but as a single adjustment parameter on a CRT the brightness setting appears to have a significant effect on sharpness.

Though, maybe a brighter picture would also make interlace flicker more apparent in a way that counteracts the mitigating effect of blur...
Post Reply