
So what this means is we can now conclude that some revisions of the Saturn simply don't have LPF enabled on 352 res mode, yet they do have it enabled for 320 res mode. So it was nothing to do with the FM being weird. The Saturn itself is to blame.
did they chance their adapter design at some point? Just asking since I bought the adapter nearly a year ago and you make it sound a little bit as if it was a new product.HD Retrovision component cables for the Genesis now have a Saturn adapter you can buy from their store.
My apologies if the Saturn adapter was that old. I only recently heard about it and was just excited to share in this thread that Framemeister users finally have a solution to the jail bars.Fudoh wrote:did they chance their adapter design at some point? Just asking since I bought the adapter nearly a year ago and you make it sound a little bit as if it was a new product.HD Retrovision component cables for the Genesis now have a Saturn adapter you can buy from their store.
Yeah I didn't think it would have helped either, but then Nick mentioned their cables have built-in LPF, and I suddenly realized it was the perfect solution to the problem. Then again, we only sort of recently figured out exactly what the problem was thanks to the OSSC. We also now know for a fact that Micomsoft have screwed up their own implementation of LPF over the various firmware updates to the FM. I've heard rumors they know they've been outclassed by the OSSC and are 'moving on' from this type of product. Just rumors anyway.Fudoh wrote:of course - great find. I wouldn't even have considered trying one of these for the jailbar issue we've been seeing.
Code: Select all
Solaris Support (Solaris Japan)
Sep 4, 14:51 JST
Hello,
Thank you so much for your detailed explanation about the problem. We completely understand and agree that it is unacceptable on Micomsoft's side. The jailbars are the screen are a big problem and it is in Micomsoft's best interest to address these concerns. Thank you again for bringing this issue to us and we will definitely forward your message to Micomsoft.
Best regards,
Ranka
Join Solaris Japan on Facebook and Twitter to keep up with special deals, news and more!
I agree 100%, since downgrading the firmware resolves the issue it's safe to say that this is 100% a Micomsoft fuckup.FBX wrote: We also now know for a fact that Micomsoft have screwed up their own implementation of LPF over the various firmware updates to the FM.
Actually you still would want to use the HD Retrovision cable into your Framemeister. The HD Retrovision cables do not upscale at all, and most displays do a horrible job of upscaling 240p content if they can handle it at all. So you would feed the HD Retrovision cable into the Framemeister and use my HDR profiles.lastbronx wrote:
Are there any other known ways to resolve the issue, without tracking down the always sold out Retrovision cable? The irony is I bought the Framemeister mainly for my Saturn and with a HD Retrovision cable I wouldn't even need the framemeister.
zero. End of life product and a bug that was introduced years ago. Still a shame though, since - as the earlier FWs show - it's possible to configure the LPF to work in a much better way.How likely do you think it is that Micomsoft will fix this issue with a firmware update?
Unless those Samsung TVs do two things:lastbronx wrote: Yes the HD Retrovision cables don't upscale, but they still produce a satisfying picture on their own. (Samsung TVs)
That's a shame! Are they working on a predecessor or just giving up on this idea?Fudoh wrote:zero. End of life product and a bug that was introduced years ago. Still a shame though, since - as the earlier FWs show - it's possible to configure the LPF to work in a much better way.How likely do you think it is that Micomsoft will fix this issue with a firmware update?
Yes true!FBX wrote:Unless those Samsung TVs do two things:
1.) Nearest neighbor upscaling.
2.) Handle 240p properly.
It won't look anywhere near as sharp and correctly scaled as a Framemeister profile.
People have made the claim they don't have jailbars on their Saturn, but I've yet to see one that didn't have this problem on the higher res modes. Only 320 mode appears to be universally free of this issue. It could be the other people were checking with 320 mode games, or they had a setup that added in LPF they didn't know about. Whatever the case may be, I've yet to find such a mythical Saturn that doesn't have LPF issues in 352 and 704 modes.lastbronx wrote:
Does anyone know if there are SATURNs out there without jailbars on the Framemeister? I want to know if it's worth it to keep trying to track one down.
Ashura wrote:
I'm currently between living situations, but I actually wanted to see if FBX would recap the bad Saturn at some point. I figure if he's up to that, he could see the difference himself before capping.
(That second Saturn is weird. It will always boot Silhouette Mirage, a 320 game, but will hang if a game tries to use 352 until it's been running a while. After it starts booting things, I've run games for 4+ hours with no issues.)
That's actually my working theory, to be honest. Maybe it's just because of the caps. Or at the very least, it's one thing we can check and mark off our list of reasons. We at least know it's the Saturn at this point.FBX wrote:It does sound like you have dried out caps. And I've done a full cap job on a dead Saturn before, so it is something I can do if you ever get to a point where you want to get it restored. It would be interesting to actually see a Saturn with no LPF issues. Come to think of it, I should probably just recap my Saturn and see if that has an effect. It would be funny if all this time it turned out to be just bad caps.
It could be bad caps. It could be they changed a certain cap in a certain revision. It could be the more likely situation that it's a board revision that does it. Who knows until we try. However, re: FW problems. It works on 1.11 because how the low pass filter on 1.11 and prior operates is different than all the versions post 1.11. The jailbars were always there, it's just that they're smoothed out by the framemeister's LPF, which later got changed somehow. You can see them clearly on the OSSC with the LPF off, for instance.lastbronx wrote:About the caps. I'm not an export on the matter, but I don't think it's the caps, because I tested 4 Saturns and FBX tested 3. Isn't it a bit unlikely that 7 SATURNS all have bad caps? Also if it were bad caps, wouldn't the problem exist on ALL resolutions? And if not, why does downgrading the FW to 1.11 resolve the problem?
I think it's more likely that they forgot to implement the LBF for certain resolutions when they rewrote the newer firmwares.
You're not going to either. Micomsoft has ended production and support for the framemeister.lastbronx wrote: I have not heard back from Micomsoft or Solaris Japan yet.
What could you do with that?rama wrote:The register map of the scaler chip would be most helpful
Micomsoft has no incentive to publish their source code. They would do so at a loss of company value and out of the goodness of their hearts, and well, companies are generally lacking in that.lastbronx wrote:What could you do with that?rama wrote:The register map of the scaler chip would be most helpful
I know Micomsoft has ended production, but that means it would be even more important to give us the source code to their firmware, or can someone "hack" the Framemeister and write a better Firmware? Would gladly donate to that cause.
The only thing that surprises me, is that not more people are reporting this problem, not even MLIG mentions this. They just say, the SATURN is the best example for pristine image quality with the Framemeister, lol. Maybe because they use GSCART, which gets rid of the scanlines?
The Scanlines for me are VERY strong, much stronger than the picture shows if plugged directly into the FM. Especially in motion. I did notice, that if I use a passive SCART switcher, the scanlines become slightly less visible. Maybe 5% less, not much. Makes me wonder if I can get rid of them by 'weakening' the SCART signal enough before it reaches the Framemeister. I do use a SCART adapter with sync stripper, is it possible that this causes the scanlines to be more prominent? Shouldn't be the cause to my knowledge, since it removes just the composite part of the signal. What SCART adapter are you guys using?
If some SATURNs, don't produce Scanlines, is it more likely, that those who don't are earlier revision models or later ones?
How can I add a LPF to the SATURN?
Micomsoft has no incentive to publish their source code. They would do so at a loss of company value and out of the goodness of their hearts, and well, companies are generally lacking in that.
This hasn't been widely supported since it's only on certain resolutions of Saturn games, a console with a pretty small following compared to most others of its time. That and a number of people have moved on to the OSSC which features properly configurable LPF.
You might be able to find or build a small LPF circuit just for the Saturn? That or buying an OSSC are the angles I'd look into- finding a Saturn without the issue seems hopeless. Knowing how badly the Genesis was designed I'm pretty confident it's an outright design flaw on Sega's part.
not when you run it in generic 4:3 mode (which is basically the same as the Framemeister does).The OSSC requires different settings for all different resolution types of the SATURN, tho right?
I was thinking maybe you could build an external circuit with a SCART input and output on it that just adds LPF. If gscartsw can do it, it's possible, but I haven't heard of anyone doing it before.lastbronx wrote:How can I add a LPF to the SATURN?
And with the way profiles work, won't your timings for each input resolution be saved? Or are they all grouped together as "240p" and the OSSC doesn't know what horizontal res a game is outputting?Fudoh wrote:not when you run it in generic 4:3 mode (which is basically the same as the Framemeister does).The OSSC requires different settings for all different resolution types of the SATURN, tho right?
In generic mode you get pixel perfect sampling on the vertical, but kinda interpolated sampling (for the lack or a better term) on the horizontal. With the optimized modes you add pixel perfect sampling on the horizontal and ONLY here you need different settings per input timing.