NES/SNES 240p dejitter mod

The place for all discussion on gaming hardware
User avatar
Arthrimus
Posts: 106
Joined: Mon Apr 02, 2018 5:49 pm
Location: Arkansas

Re: NES/SNES 240p dejitter mod

Post by Arthrimus »

Today I experimented with injecting dejittered csync into the video encoder on a SNES Jr. Based on the how difficult it would be to lift the csync pin on the 1-chip, I decided to lift pin 7 on the S-RGB encoder and solder a wire directly from it to pin 15 of the CPLD on my dejitter board exactly like maxtherabbit's mod. I added a bit of electrical tape under pin 7 to make sure it stays isolated from the stock Csync pad. This is what my install looks like. Ignore the extra two wires on pins 12 and 17, those are for the S-Video mod.
Image

This install was done with my experimental RGB bypass + Dejitter board, but it should be functionally identical to the standard dejitter board.
Image

I tested this setup using an HD Retrovision SNES Component cable on my OSSC at line5x connected to my Epiphan AV.IO 4K HDMI capture device, which can't normally sync to 5x on a stock SNES. The HD Retrovision Cable uses composite video for sync so it's a perfect test cable for this mod. Below are some captured images from my testing.
Spoiler
Image
Spoiler
Image
Here is a capture of my setup with a normal C-Sync cable for comparison.
Spoiler
Image
I'm pretty happy with the results. I don't see any visual artifacts from the mod.
plus ça change,
plus c'est la même chose,
The more that things change,
The more they stay the same.- RUSH- Circumstances

I install and sell mods at arthrimus.com | SNES RGB Bypass+Dejitter available now! | Watch me live stream my work on YouTube
User avatar
LDigital
Posts: 342
Joined: Sun Jan 17, 2016 9:15 am

Re: NES/SNES 240p dejitter mod

Post by LDigital »

I have a problem with my nes install

I had my nes rgb from ebay, pre installed

I bought the dejitter from videogameperfection so it should be pre programmed for nes. I closed the jumpers as per the instructions on retromodwiki - it doesn’t reallly say why but I did what it says

I just get a black screen when powered up even on my crt.

I have rechecked everything and The only step I cannot be sure of is where it says

“You should also make sure that jumper J8 on the NESRGB is open (not pictured).”

It’s a shame it’s not pictured because I have looked at this thing for an hour and can’t find anything labelled j8

I have checked all continuity and there are no shorts

Any ideas what might be wrong?

Image

Thanks
User avatar
Link83
Posts: 342
Joined: Tue May 21, 2013 2:39 am

Re: NES/SNES 240p dejitter mod

Post by Link83 »

LDigital wrote:“You should also make sure that jumper J8 on the NESRGB is open (not pictured).”

It’s a shame it’s not pictured because I have looked at this thing for an hour and can’t find anything labelled j8
J8 is near the Green/Blue output pins, so on your picture you'll need to disconnect some of your Red/Green/Blue pin headers to see it.

All the NESRGB V1.4 jumper positions are shown on the "Pinout and jumper diagram" linked on the main NESRGB page:-
https://etim.net.au/nesrgb/NESRGB-Pinout.pdf
User avatar
LDigital
Posts: 342
Joined: Sun Jan 17, 2016 9:15 am

Re: NES/SNES 240p dejitter mod

Post by LDigital »

You beauty! That was the missing ingredient. Apparently it has to be TTL for the dejitter to work. Thanks so much
User avatar
maxtherabbit
Posts: 1763
Joined: Mon Mar 05, 2018 4:03 pm

Re: NES/SNES 240p dejitter mod

Post by maxtherabbit »

marqs wrote:
sofakng wrote:Would a sync stripper (Sync Strike) cause a picture shift?
I recall it uses a LM1881 so output sync is slightly delayed, causing picture to get shifted left a bit.
maxtherabbit wrote:Upon further examination, I've discovered this console has jailbars. I don't remember seeing them before I did the mod, but then again I also didn't really scrutinize it prior to now. So my question: does anyone think that my placement of the dejitter board:
SHVC-CPU-01 is known for faint jailbar-like artefact so I doubt it's related to the mod. You could try adding extra ~10uF ceramic SMD cap near PPU2's 5V supply pins since apparently that has helped some people.
just wanted to update anyone following along, I followed this thread: viewtopic.php?f=6&t=62631

and stacked 10uf ceramics on top of the existing decoupling caps for PPU1/2, CPU, WRAM, and both VRAM chips - now my noise issues are completely resolved - the upper deck dejitter mod is good to go!
borti4938

Re: NES/SNES 240p dejitter mod

Post by borti4938 »

I took a look at the implementation today. (SNES version) Is there any chance to get de-jitter implemented for PAL modes instead of simple bypassing the whole functionality?
User avatar
Harrumph
Posts: 368
Joined: Tue Jan 19, 2016 10:06 pm
Location: Sweden

Re: NES/SNES 240p dejitter mod

Post by Harrumph »

borti4938 wrote: de-jitter implemented for PAL modes
AFAI understand, de-jitter is not necessary for PAL 288p mode as it is inherently "non-jittery" (the short length scanline does not exist). IIRC however, the 576i output did have a short scanline, but this mode is barely used (I don't know of any game).
If memory serves, this is described in the nocash full snes documentation.
User avatar
marqs
Posts: 1034
Joined: Sat Dec 15, 2012 12:11 pm
Location: Finland

Re: NES/SNES 240p dejitter mod

Post by marqs »

Harrumph wrote:AFAI understand, de-jitter is not necessary for PAL 288p mode as it is inherently "non-jittery" (the short length scanline does not exist). IIRC however, the 576i output did have a short scanline, but this mode is barely used (I don't know of any game).
If memory serves, this is described in the nocash full snes documentation.
Yes, only 576i has jittery sync but it is caused by a long scanline unlike a short one in 240p. So instead of temporarily stopping the clock, you'd need to speed it up to remove the jitter, but that would likely lead to stability issues. At best one could distribute the jitter over several lines assuming they're nowhere near visible area.
borti4938

Re: NES/SNES 240p dejitter mod

Post by borti4938 »

Thank you very much for clarification :)
User avatar
BuffaloWing
Posts: 50
Joined: Tue Feb 02, 2016 1:31 am
Contact:

Re: NES/SNES 240p dejitter mod

Post by BuffaloWing »

Is there suppose to be a nes dejitter specific firmware in the github? I only see the snes version. I did check the "nes-fix" branch's output_files directory but every file name contain the snes prefix. Am I looking at the wrong place?
nmalinoski
Posts: 1974
Joined: Wed Jul 19, 2017 1:52 pm

Re: NES/SNES 240p dejitter mod

Post by nmalinoski »

BuffaloWing wrote:Is there suppose to be a nes dejitter specific firmware in the github? I only see the snes version. I did check the "nes-fix" branch's output_files directory but every file name contain the snes prefix. Am I looking at the wrong place?
If you've got the ability to rewrite the firmware anyway, why not just try the nes-fix firmware to see if it works?

It's probably a minor oversight that the output filenames weren't changed; I'm sure the focus of that branch was NES compatibility.
User avatar
BuffaloWing
Posts: 50
Joined: Tue Feb 02, 2016 1:31 am
Contact:

Re: NES/SNES 240p dejitter mod

Post by BuffaloWing »

nmalinoski wrote:
BuffaloWing wrote:Is there suppose to be a nes dejitter specific firmware in the github? I only see the snes version. I did check the "nes-fix" branch's output_files directory but every file name contain the snes prefix. Am I looking at the wrong place?
If you've got the ability to rewrite the firmware anyway, why not just try the nes-fix firmware to see if it works?

It's probably a minor oversight that the output filenames weren't changed; I'm sure the focus of that branch was NES compatibility.
I haven't build my board yet, but I will in the near future. I'm just trying make sure that I'm not missing any crucial information before I dive too deep. I thought they could be typos too, but I figure it's better to ask than make assumptions.
User avatar
marqs
Posts: 1034
Joined: Sat Dec 15, 2012 12:11 pm
Location: Finland

Re: NES/SNES 240p dejitter mod

Post by marqs »

BuffaloWing wrote:Is there suppose to be a nes dejitter specific firmware in the github? I only see the snes version. I did check the "nes-fix" branch's output_files directory but every file name contain the snes prefix. Am I looking at the wrong place?
The version under nes-fix branch is the one you're looking for.
borti4938

Re: NES/SNES 240p dejitter mod

Post by borti4938 »

It's just an idea, but is it possible to add an input to select between NES and SNES variant (over a pull down resistor + jumper to Vcc combination). I created to commit just for discussion: link to idea. I didn't have this mod installed yet, so I didn't test this.
Bananmos
Posts: 11
Joined: Sun Mar 18, 2018 10:55 pm

Re: NES/SNES 240p dejitter mod

Post by Bananmos »

So after having it on the shelf for half a year, I finally got around to install the de-jitter I got from videogameperfection on my onechip console (SNSP-CPU-1CHIP-01 model).

The good news is that the de-jitter board does seem to do magic to my SNES + OSSC setup in 60Hz. Whereas before my Sony KDL-40BX400 TV would lose sync and turn black for 2-3 seconds about once to several times per minute, I can now play SNES games at length without running into the problem.

The bad news however, is that the random picture dropouts are still not *completely* gone. I still get very rare picture drops. Sometimes it can go half an hour between each picture drop, but sure enough it will happen at some point in my gaming. And in a way, it annoys me more than if the mod had flat-out failed, because the dropouts being so irregular makes it so tedious to debug this problem. I keep playing around with various settings, and convince myself I've now solved it, only to later realize the problem is still present, after playing a game long enough.

So I figured I'd reach out and ask if anyone knows if there's something more to it... are there some key settings for the SNES that may very occasionally affect the TVs ability to sync? I've heard before that the HPLL2x setting being on could cause this, but I can't notice much difference. Nor did I see much difference from "Analog sync Vth" IIRC... but again, it's hard to know for sure if a setting had any effect, as the problem now happens so rarely.

So some expert knowledge on what other than could cause rare picture drops would be much appreciated.

Maybe I should just ditch this old Sony TV for a more OSSC-friendly one... but I am quite impressed with it managing the 5x scaling mode, which appears to be a rare perk among HD TVs. (Though sadly it doesn't manage 4x, and some games do suffer from the zoomed-in look with a cropped top/bottom)

...and speaking of line multiple modes, I just realised that the 3x mode (my fallback when the 5x zoom-in is game-breaking) does make the picture drops go back to being just as bad as before I installed the de-jitter mod. But only if I select the 256x240 optim mode - Generic 4:3 seems to work ok in 3x mode.
Maybe there's a clue in that? (though I'm stumped as to what the reason is)
rama
Posts: 1373
Joined: Wed Mar 08, 2017 3:15 pm

Re: NES/SNES 240p dejitter mod

Post by rama »

Bananmos, do you have some pictures of your install?
Is there anything special about your SNES RGB cable?
User avatar
marqs
Posts: 1034
Joined: Sat Dec 15, 2012 12:11 pm
Location: Finland

Re: NES/SNES 240p dejitter mod

Post by marqs »

borti4938 wrote:It's just an idea, but is it possible to add an input to select between NES and SNES variant (over a pull down resistor + jumper to Vcc combination). I created to commit just for discussion: link to idea. I didn't have this mod installed yet, so I didn't test this.
There's a good chance that a common implementation would work for both if suitable timing can be found and verified. The problem is that that NES(RGB) and SNES have different delays which might also vary a bit depending on model, temperature etc. Basically the data paths are as follows:

NES: MCLK_XTAL_i (launch rising edge) -> MCLK_o (some logic delay on CPLD) -> CVBS (MCLK->CVBS delay of PPU) -> delay of comparator on NESRGB PCB -> MCLK_XTAL_i (capture falling edge).
SNES: MCLK_XTAL_i (launch rising edge) -> MCLK_o (some logic delay on CPLD) -> CSYNC (MCLK->CSYNC delay of PPU) -> CSYNC_i related logic on CPLD -> MCLK_XTAL_i (capture rising edge).

On SNES the loop seemed to operate well under 1 clock cycle (~46ns). However, the same code had stability issues with some NES installations as the delay got too close to 1 cycle with variance, potentially causing timing violations. The nes_fix firmware was designed to be stable when loop delay stays between 0.5 and 1.5 cycles. It's possible that nes_fix firmware could work perfectly well on SNES, but I don't have enough models to verify that. Alternatively, the code could be changed so that CSYNC_i is directly registered on CPLD (i.e. no combinational logic driven by the raw input signal) which could possibly reduce the delay enough to make 1-cycle loop stable with NES too. Apparently NESRGB dejitter implementation does that, but it has a different CPLD so the results are not fully comparable. Note that your proposed code also changes loop timing characteristic in SNES mode:

Code: Select all

CSYNC_sel <= NES_FIX ? CSYNC_i_L : CSYNC_i;
That should relax the loop timing a bit since there's less logic driven by raw CSYNC_i input. I posted a proposed change on the commit which uses a registered version of CSYNC_i in both modes.
Bananmos
Posts: 11
Joined: Sun Mar 18, 2018 10:55 pm

Re: NES/SNES 240p dejitter mod

Post by Bananmos »

rama wrote:Bananmos, do you have some pictures of your install?
Is there anything special about your SNES RGB cable?
Here's a photo of the de-jitter board installed on the bottom side of the circuitboard in my PAL 1-chip, on top of a SuperCIC install.
Image

And here's a photo of the top bit.
Image

For the mod, I followed the advice given in this thread: https://www.videogameperfection.com/for ... chip-snes/
"Remove X1, C3, R28 and D1 from SNES motherboard.
Leave JP2 open
Leave JP3 open
Connect 5v and GND to regulator output or any other 5v and GND points.
Connect MCLK_EXT_i to SCIC clock output
Connect CSYNC_i to pin 7 of S-RGB chip or any via connected to that pin.
Connect MCLK_o to left hand through hole of X1 (TC1+)
Connect CSYNC_o to top pad of R28
Connect CLK_SEL_i to SCIC mode pin"

In case it is at all relevant, I have not done the ghosting fix mod. I would like to as I believe I've observed it in certain games (though it could just be the LCD of the TV being slow). But the most concise fix I could find only appears to list the part for an NTSC 1chip console: https://www.njglover.com/fixing-ghostin ... es-1-chip/
Assuming the part number for capacitor C11 is numbered the same, then it looks like I really should have done it prior to the SuperCIC mod, as C11 is right under the board... but as other part numbers seem to vary, I'm not sure it is C11 on a PAL 1CHIP. Could anyone confirm?

The cable is a "Nintendo Super SNES RGB SCART cable SYNC on LUMA for PAL console", bought from retrogamingcables.co.uk in December 2017. I suppose it is "special" in the sense that when I now check retrogamingcables.co.uk, they appear to only be selling the more advanced "PACKAPUNCH" cable (which appear to be out-of-stock anyway). I didn't think at the time that the extra shielding for the more expensive packapunch cable would be necessary... and still doubt that could be the culprit?
rama
Posts: 1373
Joined: Wed Mar 08, 2017 3:15 pm

Re: NES/SNES 240p dejitter mod

Post by rama »

The video cable should be fine.
I see a very long, unshielded CSync wire. It may be picking up noise from the oscillators it crosses by.
This noise could contribute to system jitter, then ultimately cause the random dropouts.

Try moving the wire out of the way of the active components.
Also, make that MCLK_o loop wire short and direct. This is your fast master clock, you don't want it radiating around.
borti4938

Re: NES/SNES 240p dejitter mod

Post by borti4938 »

Bananmos wrote: In case it is at all relevant, I have not done the ghosting fix mod. I would like to as I believe I've observed it in certain games (though it could just be the LCD of the TV being slow). But the most concise fix I could find only appears to list the part for an NTSC 1chip console: https://www.njglover.com/fixing-ghostin ... es-1-chip/
Assuming the part number for capacitor C11 is numbered the same, then it looks like I really should have done it prior to the SuperCIC mod, as C11 is right under the board... but as other part numbers seem to vary, I'm not sure it is C11 on a PAL 1CHIP. Could anyone confirm?
Confirm - it's C11 also on PAL systems!

By the way:
I'm up to integrating the dejitter mod into the multi-region mod. It has been a while since I announced it, ...

Image
User avatar
Harrumph
Posts: 368
Joined: Tue Jan 19, 2016 10:06 pm
Location: Sweden

Re: NES/SNES 240p dejitter mod

Post by Harrumph »

Yay borti! Great to hear news about this, really looking forward to it!
Bananmos
Posts: 11
Joined: Sun Mar 18, 2018 10:55 pm

Re: NES/SNES 240p dejitter mod

Post by Bananmos »

borti4938 wrote:
Bananmos wrote: In case it is at all relevant, I have not done the ghosting fix mod. I would like to as I believe I've observed it in certain games (though it could just be the LCD of the TV being slow). But the most concise fix I could find only appears to list the part for an NTSC 1chip console: https://www.njglover.com/fixing-ghostin ... es-1-chip/
Assuming the part number for capacitor C11 is numbered the same, then it looks like I really should have done it prior to the SuperCIC mod, as C11 is right under the board... but as other part numbers seem to vary, I'm not sure it is C11 on a PAL 1CHIP. Could anyone confirm?
Confirm - it's C11 also on PAL systems!

By the way:
I'm up to integrating the dejitter mod into the multi-region mod. It has been a while since I announced it, ...

Image
Oh man, I've seen that not only does your new board integrate the dejitter mod, but it also leaves C11 fully accessible. I really ought to have waited... but oh well :)

On the bright side, I think I've gotten down to the problems with the OSSC and my TV. I followed rama's advice about shortening the M_CLK output (and input, even though it's just for PAL), but saw no improvement from this. Don't think the CSYNC would have mattered, because even if they were long they never touched the oscillators while playing, as I'd been testing the SNES with the PCB standing on the side.

But after playing around more with the settings in the totally unstable 3x 720p mode, it turns out increasing the post-coast PLL setting to 3 would totally get rid of all the picture drops in 720p. And so far, I haven't been able to observe it in 5x mode either, even after hours of leaving it running. So if anyone else is struggling with loss of sync on a Sony TV, it might be useful to know that post-coast seems to be the silver bullet here, at least on the KDL-40BX400 model.

So now it's just a question of whether to solder up the SNES and be happy, or undo all the SuperCIC soldering points to get access to that C11. I'm a bit on the fence here, as I'm not totally sure whether the ghosting I've seen is the SNES or just the TV having a slow response (something I have observed even with the HDMI output from my Hidef NES).

Can anyone recommend some games and/or test ROMs which make the ghosting very obvious?
User avatar
lettuce
Posts: 1334
Joined: Wed Jun 22, 2011 7:10 pm
Location: Bedfordshire, England.

Re: NES/SNES 240p dejitter mod

Post by lettuce »

Kind of a tangent on this, but as your replacing the oscillator crystal could this in theory turn a PAL 50hz NES into an NSTC 60hz NES??
Bananmos
Posts: 11
Joined: Sun Mar 18, 2018 10:55 pm

Re: NES/SNES 240p dejitter mod

Post by Bananmos »

lettuce wrote:Kind of a tangent on this, but as your replacing the oscillator crystal could this in theory turn a PAL 50hz NES into an NSTC 60hz NES??
Not possible. Both the CPU and PPU come in NTSC / PAL variants, with all the timing differences (scanline cycles, vblank time, CPU audio etc) being hard-coded in the hardware. Thus, making a multi-system would require cannibalizing one system, and putting its CPU / PPU in the other one, with very complicated circuitry to "cut off" the other one from the bus.
To the best of my knowledge, no one has ever successfully performed such an intrusive mod (though a few have tried). The effort and hardware parts required (and the fact you'd need two systems to make one multi-region one) makes a multi-system NES with original chips non-viable, even if it's theoretically possible.

The SNES was a very different beast, in that the same chips are used for different regions, with just a few configuration pins changing (and the oscillators being different, particularly for the 1CHIP models).
User avatar
marqs
Posts: 1034
Joined: Sat Dec 15, 2012 12:11 pm
Location: Finland

Re: NES/SNES 240p dejitter mod

Post by marqs »

Bananmos wrote:But after playing around more with the settings in the totally unstable 3x 720p mode, it turns out increasing the post-coast PLL setting to 3 would totally get rid of all the picture drops in 720p. And so far, I haven't been able to observe it in 5x mode either, even after hours of leaving it running. So if anyone else is struggling with loss of sync on a Sony TV, it might be useful to know that post-coast seems to be the silver bullet here, at least on the KDL-40BX400 model.
Which kind of RGB cable do you use? That sounds like you're using a cable wired for sync on luma/cvbs instead of a cable using c-sync.
Bananmos
Posts: 11
Joined: Sun Mar 18, 2018 10:55 pm

Re: NES/SNES 240p dejitter mod

Post by Bananmos »

marqs wrote:
Bananmos wrote:But after playing around more with the settings in the totally unstable 3x 720p mode, it turns out increasing the post-coast PLL setting to 3 would totally get rid of all the picture drops in 720p. And so far, I haven't been able to observe it in 5x mode either, even after hours of leaving it running. So if anyone else is struggling with loss of sync on a Sony TV, it might be useful to know that post-coast seems to be the silver bullet here, at least on the KDL-40BX400 model.
Which kind of RGB cable do you use? That sounds like you're using a cable wired for sync on luma/cvbs instead of a cable using c-sync.
I am indeed using a "Nintendo Super SNES RGB SCART cable SYNC on LUMA for PAL console", bought from retrogamingcables.co.uk. IIRC, I got it because I've read that cable wired for CSYNC might suffer from the checkerboard pattern.

Are you saying that the OSSC / dejitter mod is less reliable for sync-on luma, or something has been configured incorrectly? Everything does seem to work better than ever before as of now...
User avatar
marqs
Posts: 1034
Joined: Sat Dec 15, 2012 12:11 pm
Location: Finland

Re: NES/SNES 240p dejitter mod

Post by marqs »

Bananmos wrote:I am indeed using a "Nintendo Super SNES RGB SCART cable SYNC on LUMA for PAL console", bought from retrogamingcables.co.uk. IIRC, I got it because I've read that cable wired for CSYNC might suffer from the checkerboard pattern.
You're mixing c-sync with composite video, there no extra information besides sync in c-sync signal. Anyway, that being a PAL machine, c-sync doesn't have a dedicated pin on multi-AV connector so luma should be easier to use.
Bananmos wrote:Are you saying that the OSSC / dejitter mod is less reliable for sync-on luma, or something has been configured incorrectly? Everything does seem to work better than ever before as of now...
It depends on how CSYNC_o is connected. If CSYNC_o is not inserted on luma path, then the output will have a short scanline followed by a long scanline. Since that occurs right after vsync, it's still possible to mask the unevenness by post-coast option as you did.
User avatar
marqs
Posts: 1034
Joined: Sat Dec 15, 2012 12:11 pm
Location: Finland

Re: NES/SNES 240p dejitter mod

Post by marqs »

borti4938 wrote:By the way:
I'm up to integrating the dejitter mod into the multi-region mod. It has been a while since I announced it, ...
That looks nice and compact! Does the PCB have a PLL which generates PAL+NTSC MCLKs from that 27MHz XTAL?
borti4938

Re: NES/SNES 240p dejitter mod

Post by borti4938 »

Thank you :)
Yes, it uses the CDCE913 (know from micro's DFO design) to generate PAL and NTSC MCLKs. So, configuration for that IC depends on SNES version - either 1Chip or 3Chip consoles.
I want to test the design in a 3Chip console, too, and try to use the PLL + CPLD also to switch between color subcarrier.

If everything goes well, I will give the design to some testers. It would be a pleasure if you want to test one, too :)
Afterwards, PCB design goes open source for everybody.

Just a question:
Is there any reason why to not feed CSYNC_o to the S-RGB?
Spoiler
Image
User avatar
marqs
Posts: 1034
Joined: Sat Dec 15, 2012 12:11 pm
Location: Finland

Re: NES/SNES 240p dejitter mod

Post by marqs »

borti4938 wrote:Thank you :)
Yes, it uses the CDCE913 (know from micro's DFO design) to generate PAL and NTSC MCLKs. So, configuration for that IC depends on SNES version - either 1Chip or 3Chip consoles.
I want to test the design in a 3Chip console, too, and try to use the PLL + CPLD also to switch between color subcarrier.

If everything goes well, I will give the design to some testers. It would be a pleasure if you want to test one, too :)
Afterwards, PCB design goes open source for everybody.

Just a question:
Is there any reason why to not feed CSYNC_o to the S-RGB?
Spoiler
Image
Yes, that's an option for those who prefer not to replace luma/cvbs with c-sync. Speaking of subcarrier, I've made an update to nes-fix branch (not in the repo yet) that adds clean NTSC subcarrier output (MCLK_XTAL/6) to CPLD pin 11. A few people reported artifacts on NESRGB composite / s-video output with the mod since cvbs/s-video encoder subcarrier input (generated by NESRGB CPLD) was no more periodic with the modified MCLK, so clean SC has to be generated on snes_dejitter. I wonder if the same issue appears with SNES cvbs/s-video output as well since there subcarrier is generated by PPU (at least on 3-chips).
Post Reply