Tagged: koryuu n64 s-video ossc sync pal
July 14, 2020 at 11:56 PM #39802
Had the chance to test anything yet? Or better still.. have you found any “solutions” to help with our interference problems?July 20, 2020 at 10:12 AM #39913
Hi leewrigley, sorry for the wait. I have done some tests with the OSSC control. But as I expected I was unable to get the overbright image to the proper levels. When adjusting either the pre-ADC gain or the Y’ input Offset or Gain I can adjust the brightness to my hearts content but it effects the whole picture. The problem with the N64 PAL s-video output is that the highlights are clipping and becoming the same brightness value. I’m afraid I don’t have any tricks to tackle this problem. It seems to only solution is to use a 75Ohm resistor on the N64 Luma output. Maybe we have some OSSC genius around who knows a better solution 🙂
I’m currently still waiting on the RetroAccess cables. I’ve received confirmation recently that they shipped the cables. Now I’m waiting for my overseas US address to receive it at which point I can send it to my home address.October 13, 2020 at 4:14 PM #42071
Okay, I haven’t shared any information in a while. It has been quite busy, so my apologies for that.
I strongly believe I now understand where the interference or “dot crawl” is coming from. I think it has to do with the comb filter of the Koryuu. Testing multiple configurations with different equipment, this is the conclusion that I keep returning to.
* Using High quility shielded S-Video cables from RetroAccess (both original and modified for PAL N64)
* Using multiple revisions of the console
* Using different settings on the OSSC
All result in a noisy image
* Using a Framemeister, Retrotink 2x Pro, Retrotink 2x Pro to OSSC (VGA)
All result in considerable less, but still somewhat noticeable dot crawl. Maybe because the Koryuu has a sharper image/signal?
The “a-ha” moment came as soon as I used the new Retrotink MINI. This device has two options for its comb filter. Auto and Retro. The website states that this device has been built in such a way that it supports N64 PAL consoles over s-video (both normal and PAL modified s-video cables seem to result in the same image). Also saying that if there is any dot crawl present in the image that you should switch to the “auto” comb filter. This proved wonders. The image with this device + s-video with the “auto” comb filter, is the cleanest I’ve ever seen my unmodified PAL N64.
However, due to what the Koryuu is, I believe that with an adjusted comb filter for this device it can look even better than the Retrotink Mini. Especially when connected to the OSSC. Is it possible for the developer to look into this comb filter issue and change it with a firmware revision?October 13, 2020 at 4:25 PM #42072
hello again, i wasnt aware of this feature in the retrotink 2x, i have actually just a couple of days ago purchased a new n64 cable from classicgamestore which is compatible with pal n64s. I am expecting some impovement from the not so great cable from consolegoods. one more upgrade i am still to make however i was the upgrade from the thick but still basic component cables i had bought cheap from amazon – to the hd retrovision component male to male so that i would have the highest quality possible every step in the chain.
now that you have wrote this i do feel less confident however and hopefully you are right and this could be added in firmware maybe we should try and contact the koryuu creator to see if this can be done. anyhow i will update the forum with my results from the classicgamestore svideo cable and maybe another one next month once i buy the hd retrovision cables also.
thanks for the updateOctober 14, 2020 at 8:36 AM #42096
I actually own the classicgamestore.ch PAL N64 S-Video cable. I can tell you that it is by far the best cable I’ve used so far along with the NTSC RetroAccess cable. These two are mostly identical accept for the PAL modification and the mold for the Nintendo Multi AV plug being slightly different.
I assume you mean for the component cables to go from the Koryuu output. This will likely not change anything unless your current cables are garbage. I tried to compare cheap and thin RCA composite cables to professional component cables. The results were hard if not impossible to distinguish from eachother. However, if you’re chasing perfection like me, it could still be worth it.
I strongly believe that the Koryuu’s comb filter is to blame for the dot crawl noise. I’m unsure if the developer reads these forums. BuckoA51 is one of the Keymasters. Maybe he knows if he reads this?October 14, 2020 at 1:09 PM #42102
S-Video should, theoretically, not require comb filtering, as the luma and chroma components are already supposed to be cleanly separated.
Are you sure the effect is dot crawl, rather than a regular interference pattern? If the pattern is very regular, do your S-Video cables have *both* S-Video and composite connectors at the other end? This can potentially cause a regular “netting”/checkerbox pattern to appear on the image due to crosstalk and coupling. I guess this sort of pattern can be eliminated with a comb filter, but it is not supposed to be there in the first place.
I definitely want to get to the bottom of this, as some Googling results in finding reports with similarities to yours pertaining to PAL N64 S-Video output over the last ten years or so.
(An excessively bright image, referred to in an older message in this thread, sounds like S-Video cables wired for the NTSC N64, but apparently cables proper for PAL N64s are now sorted.)October 14, 2020 at 1:42 PM #42103
thanks @megari for oyur replies in both threads, i can confirm that my svideo cable (modded to repair the overly bright pal svideo signal) does have the composite lead attatched also and i am currently awaiting a better cable to be delivered to test if it looks much better. the interference (at least for myself) is mcuh more like a very harsh checkerboarding. i was under the assumption that higher quality svideo and component cables would solve this problem it was only after @noctua tested these things that he discovered the comb filter was solving his issues.
I have not long moved house but in a couple of days i will take some footage of the n64 output and i would be happy to test custom firmware for you to resolve if need be (would need to buy the jtag device thing first)
thanksOctober 14, 2020 at 2:05 PM #42110
Here is some photos I took back when I first encountered the problemOctober 14, 2020 at 6:25 PM #42118
Another thing that came to my mind, although it doesn’t seem to quite fit the symptom, is a cost-cutting measure some cheap CVBS/S-Video combo cables use: instead of having 3 separate wires for CVBS, luma and chroma, only CVBS and chroma are wired, with luma being substituted with the CVBS signal. This completely ruins the S-Video picture, yielding dot crawl, just like in CVBS. What makes this even worse is that combing the luma component of S-Video is not done, as proper Y/C separation is assumed.
But, as said, the symptoms do not quite match that extreme scenario, at least considering the photos @leewrigley posted, which do not look like the dot crawl you get with composite, rather than a mesh pattern more indicative of a coupling effect due to crosstalk.October 14, 2020 at 6:30 PM #42119October 15, 2020 at 8:54 AM #42134
@megari I’ll see if I can get some good footage of the problem tonight and will post it here. I suppose I yelled Eureka! to soon. The fact that s-video shouldn’t need a comb filter makes a lot of sense when you put it that way. Goes to show how much I know.
Noise like this is always hard to properly describe. What I’m seeing is kinda like the pictures leewrigley posted. Diagnoal lines going through the image. However on those pictures it almost looks like it effects the entire image. On my end it’s only showing on specific shades. For example. The red and yellow text on the Pokemon Stadium title screen as well as Psyduck’s head. The other parts of the image look pretty much fine. I should also note that using other s-video sources with the same cable are fine. Using my SNES for example shows a perfectly clean image with the Koryuu. Composite sources also look fine (at least as good as composite can look that is).
Due to this I thought maybe my N64 was faulty, so I tried another one with the same board revision and one with an earlier board revision. Both show the same issue.
I’ve used a bunch of different cables for this. Started with the consolegoods.co.uk PAL modified s-video cable. This one wasn’t to great. Poor shielding and such. I’ve also tried the official Nintendo cable (NTSC), retro-access.com (NTSC), and classicgamestore.ch (PAL). These don’t have the yellow composite RCA lead. For testing the NTSC cables I’ve soldered a 75 Ohm resistor between Luma and ground on the console motherboard. The latter two cables are built really well and perform otherwhise excellent.
Now I’m thinking that maybe it could be a crosstalk problem. But not caused by the cable bu by the console itself? The PAL N64 doesn’t look like it was supposed to even have s-video support. It seems more of a leftover that they wanted to omit but didn’t for whatever reason. The fact that later board revisions and french systems flat out don’t work when you hook up an s-video cable says enough. Maybe the fact that the RetroTink MINI’s “auto” comb filter solves the issue is a necessary evil? Even though s-video normally shouldn’t need it?October 15, 2020 at 3:41 PM #42148
@Noctua What you are describing does sound like some sort of chroma crosstalk. It seems it can have quite a few different variations, depending on the particular hardware.
I did some digging regarding different kinds of chroma crosstalk artifacts. Does any of these look familiar?
Pokémon Stadium title screen on PAL 64 S-Video (improper vs. proper circuit)
Chroma leaking into luma due to bad comb filter when digitizing VHS through composite
Diagonal + checkerboard-like artifacts in “S-Video” capture, due to feeding CVBS into luma
Mesh pattern and washed-out colors on SNES S-Video output
N64 S-Video chroma crosstalk artifacts, with focus on color saturation
You may want to expand the images to full size to properly see the artifacts (or lack thereof).
megariOctober 16, 2020 at 10:23 AM #42168
So I tried multiple times to capture a good example of the problem. The video captures look pointless so I’m not going to bother showing it. The best I was able to capture can be seen with thist link: https://imgshare.io/image/20201015-2120081.NxfSiO
This is the File Select screen from Mario Party 3 which has a dark blue background and is a worst case scenario. You can clearly see the diagnal lines in the blue shade. Now try to image that they are moving as well. The direction and speed can change over time.
I suppose the pictures of the second link are closest. Though it seems to effect the entire screen there. On my end it’s only certain large shades of color.
The S-Video cable I use has proper shielding and most definitly uses a 750hm resistor as there is no brightness issue. I’m unsure if the capcitors are necessary. I’ll see if I can open the AV plug and check what components they added. Maybe that is to blame here. Still think its weird that the Retrotink MINI’s auto comb filter completely solves the problem though.October 16, 2020 at 11:15 PM #42184
@Noctua OK, I see. The picture gives enough of an idea what you’re seeing. That the lines change direction and speed is something I am not quite sure what to make of – could be some sort of a sampling artifact combined with the interference pattern, with aliasing effects. In any case, it definitely does look like a clear interference pattern, and I’ll take your word for how jarring it is.
You mentioned having an S-Video cable from classicgamestore.ch that has specifically been built for the PAL N64. Is it this one? Have you verified that the cable actually contains all the required components for a proper S-Video output from a PAL N64? PAL N64s require not only 75 Ohm termination resistors (to ground) on both the luma and chroma line, but other components as well, specifically, capacitors with specific capacitances, to output a properly filtered S-Video signal.
According to viletim’s research, these components need to be present in a PAL N64 S-Video cable, as they are missing from the console itself:
- Luma line
- 75 Ohm resistor to ground
- 220 µF electrolytic capacitor in series
- Chroma line
- 75 Ohm resistor to ground
- 68 nF capacitor in series
Could you try and make sure that the cable has the components specified above, or some other configuration that achieves the same effect, please?
 Any series resistors should not be needed for proper brightness levels. Otherwise, something is likely wrong.
October 21, 2020 at 1:57 PM #42281
- This reply was modified 6 days, 4 hours ago by megari. Reason: List syntax fixes
@megari I can confirm that is that cable I have ordered and just received today, problem is though now I am only getting a black and white image… To he fair, I think it is actually just the cable and not the koryuu, here is a picture (not good quality sorry 🙁 ) of the PCB inside. I think if you look at the top contact in the picture you can see that it appears that some solder has bridged over onto the 75 ohm resistor ( I think that’s what it is, it’s so small that it’s almost unreadable), do you think this bridge could be causing this issue at all? Otherwise I’m beat!
- Luma line
- You must be logged in to reply to this topic.