RetroTINK COMP2RGB YPbPr to RGBs Transcoder

[mikechi2]

Hi Guys, first post here, long time lurker.

I recently picked up one of these COMP2RGB devices but colours look noticeably "off" on my Consumer Trinitron CRT on both my NTSC Nintendo Wii (480i, with official component cables) and NTSC Gamecube (480i, with official component cables). The Wii looks great in 480i RGB (using a CSYNC RGB scart cable from RGC) on the same display. The same applies to the Gamecube, when combining the same RGC scart cable with an Eon MKii and output set to RGB.

RGB scart (good):
https://imgur.com/X6vmQGb
https://imgur.com/jqWUqc7

COMP2RGB (bad):
https://imgur.com/u3NSolx
https://imgur.com/6LaseaL

It appears that the colour conversion to RGB from component is causing issues here, with the Reds and Blues particularly under-saturated.

Any thoughts?

Out of curiosity why are you using the COMP2RGB with these consoles if you already have a SCART cable for them?

Have you tried other consoles and preferable another TV to rule out something weird? Definitely possible that the COMP2RGB is defective, but it seems really odd to have such a well-behaved failure more.

I have something weird with mine. When I get sudden whites on screen they flash up as purple then correct to white. Sometimes when there are whites that are not sudden I get the same purple bleed around the edges.

I think you also e-mailed me. Sorry, need to figure out where my Wii is and dig it out.

[BigSplodge]

Have you tried a different USB cable and USB power supply? The COMP2RGB is more sensitive to problems with the power supply thanthe RGB2COMP.

I just tried a couple of different USB PSUs (including the one supplied with the GSCARTSW) and the same issue unfortunately.

Out of curiosity why are you using the COMP2RGB with these consoles if you already have a SCART cable for them?

I have a GSCARTSW and a GCOMPSW. The GCOMPSW has one ouput connected to the COMP2RGB (which then goes into the GSCARTSW), the secondary output goes to an OSSC. The GSCARTSW feeds the CRT and an OSSC. This means I can supply the CRT with both RGB and component sources up to 576i and then can switch to 480p+ when I'm solely using the OSSC and a 4K panel. I like the flexibility so I'm moving from RGB scart to component on consoles that support progressive video modes.

I already had the RGB scart for my Wii from a while back and I'm simply using it to verify the issue.

I have something weird with mine. When I get sudden whites on screen they flash up as purple then correct to white. Sometimes when there are whites that are not sudden I get the same purple bleed around the edges.

I have EXACTLY this issue too when using my Wii, which doesn't happen with an RGB scart cable (both running in 480i mode). I get it when transitioning between dark and light menus. I haven't tested enough with other consoles to say whether the issue is persistent across all of them, or specific to the Wii.

Have you tried other consoles and preferable another TV to rule out something weird? Definitely possible that the COMP2RGB is defective, but it seems really odd to have such a well-behaved failure more.

Hi Mike! So you make this little device?! Looks like a great piece of kit and this colour issue aside, it solves a problem I've had for a really long time so I'm keen to get it sorted. The issue is clear and present on 2 USA GameCubes and a PAL Wii, unfortunately. I've got a PS2 and XBOX that I can test with it later today and another CRT, so it should be fairly conclusive. I'll get back to you but I'm not hopeful given it has persisted with 3 consoles and multiple PSUs now. I also agree it seems like a very odd failure state: "working, but with oddly de-saturated colours"! I'm hoping it turns out to be something silly that I'm doing

EDIT: Just tested on another CRT, same issue unfortunately. I'll try the PS2 and XBOX next.

EDIT2: XBOX seems to display the same issue. Tested a couple of games and noted that the reds were always muted and under-saturated (much like the Paper Mario image I provided before). I've also taken the GCOMPSW out of the chain to rule it out and I'm connecting directly to the COMP2RGB, so there does seem to be an issue with it of some kind I think.

EDIT3: PS2 also looks very different colour wise over Component (via the COMP2RGB) than either composite or RGB. I believe at this stage that the COMP2RGB is the only common denominator...

[LDigital]

I have something weird with mine. When I get sudden whites on screen they flash up as purple then correct to white. Sometimes when there are whites that are not sudden I get the same purple bleed around the edges.

I have EXACTLY this issue too when using my Wii, which doesn't happen with an RGB scart cable (both running in 480i mode). I get it when transitioning between dark and light menus. I haven't tested enough with other consoles to say whether the issue is persistent across all of them, or specific to the Wii....

Yeah mine is with the Wii also, exactly as you describe. I also use Mister via component and haven’t see this with any other cores so can only state this for Wii personally. I find using a scart cable loses compatibility with some important games so component wii is important to me.

[BigSplodge]

Yeah mine is with the Wii also, exactly as you describe. I also use Mister via component and haven’t see this with any other cores so can only state this for Wii personally. I find using a scart cable loses compatibility with some important games so component wii is important to me.

PAL Wii by any chance?

Managed to get a decent pic of it and it seems reproducible for me moving between home page and settings menu:

[LDigital]

Yeah launch day Pal version

[LDigital]

Is any research being done to rectify this issue? I’m considering trying the beharbros version to see if it occurs there if we can’t get a fix

[BuckoA51]

Happens on every transcoder. I thought Mike had finally nailed it with this one but looks like it can still happen in some circumstances.

[kitty666cats]

Happens on every transcoder. I thought Mike had finally nailed it with this one but looks like it can still happen in some circumstances.

I haven't had any sort of issue with the linuxbot3000 YPbPr to DE-15 RGBHV/RGBS transcoder on my Wii (or anything else). HOWEVER! I have not adjusted the jumper inside it for RGBS, I've just left it on RGBHV (my TM-H150CG takes 15kHz RGBHV, and I've got an Extron 580xi). I'll have to try out the RGBS jumper one of these days and see if I run into any funk...

Just a theory, but it did just dawn on me that the above mentioned transcoder doesn't pass audio. Perhaps the transcoders that use SCART out and pass audio are more finicky?

[Guspaz]

From half-remembered conversations with Ste when we were working on the RGB bypass for the GARO, it's something to do with a bug in TI's LMH1251 chip. Pretty much all transcoders appear to use that chip, the RetroTINK COMP2RGB, the Berharbros GARO, and the Linuxbot3000 transcoder (IIRC he sanded it off but the size/pinout seemed the same). I vaguely recall Ste saying something about the black reference level was floating, and the datasheets didn't provide enough information to fix it externally. I could be totally misremembering all this.

[maxtherabbit]

GBS-C works a treat in pass through mode as a lag free "transcoder"

(it still digitizes the signal and transcodes in the digital domain however)

[mikechi2]

It was easy to fix the 480i <-> 480p/1080i glitch but unfortunately there's no way to reset the chip for issues like the Wii without introducing a horrible drop out.

[BigSplodge]

Hi Guys, first post here, long time lurker.

I recently picked up one of these COMP2RGB devices but colours look noticeably "off" on my Consumer Trinitron CRT on both my PAL Nintendo Wii (480i, with official component cables) and NTSC Gamecube (480i, with official component cables). The Wii looks great in 480i RGB (using a CSYNC RGB scart cable from RGC) on the same display. The same applies to the Gamecube, when combining the same RGC scart cable with an Eon MKii and output set to RGB.

RGB scart (good):
https://imgur.com/X6vmQGb
https://imgur.com/jqWUqc7

COMP2RGB (bad):
https://imgur.com/u3NSolx
https://imgur.com/6LaseaL

It appears that the colour conversion to RGB from component is causing issues here, with the Reds and Blues particularly under-saturated.

Any thoughts?

So it appears to be an oddity with the way some EU consumer Trinitrons handle the sync signal the COMP2RGB puts out. No such issues with 75ohm CSYNC direct from my SFC Jr., PAL Wii, NTSC N64 & NTSC GC (GCHD Mkii) but whatever sync signal the COMP2RGB puts out, neither of my CRTs like.

Interestingly - enabling sync regeneration on my GSCARTSW seems to resolve the colour issue completely (though it seems to introduce some horizontal stability issues). Presumably, enabling sync regeneration when it's not actually required for any of my scart devices is undesirable?