OSSC Pro

[Lawfer]

Is there an approximate timeframe for its release yet?

Some time next year, hopefully on the first half. The prototype round already got delayed by 1-2 months due to last minute changes (such as the expansion pinout update mentioned above) and lead time of selected DRAM chip, and there are still various items to resolve before release (not just technical ones). On the price side we've now received some initial quotes which indicate the sale price could settle somewhere on $350-550 range. The final number largely depends on which kind of case is designed/customized and how many units are manufactured at once.

Very much looking forward to it oh and thanks alot also for implementing the feature that takes into account the native 480x272 resolution of the PSP.

[orange808]

Even at ~$500 usd, this promises to be a remarkable bargain. I hope gamers will give this machine plenty of thought before they respond with "sticker shock".

Even if there was a single machine out there with all these features (there isn't), it would cost thousands. I spent years (combing auction/overstock/thrift outlets) and a lot of money putting together a chain of multiple (often used) pro/prosumer machines; the OSSC Pro promises to replace and outperform most of my chain in a single inexpensive box!

So excited for this.

[ASDR]

Yes, the price is a non-issue. Most of us here probably spend more on cables alone

[vol.2]

Price seems fair, but I would also appreciate if there was a bare-bones version that users could roll-their-own enclosures for. I think there's a lot of folks out there who prefer to do that anyway and will just discard the case it comes in.

there are still various items to resolve before release

Is there any software features still up in the air? A big consideration for me is the ability to either have a preset for windowed output to an HDCRT (960x540p) OR to be able to arbitrarily create the windowed resolution and save it as a preset.

Or are the features more or less locked-in and possibly available for consumption somewhere?

[ldeveraux]

Consider me the outlier then, I'm not paying $500 for a Pro...

[NormalFish]

Boutique product, boutique price. Not like you didn't warn us. Price is enough that I'll have to plan for it a bit more but not unexpected.

[bobrocks95]

If the resale value of the base OSSC doesn't plummet and it's closer to the $400-450 mark that isn't too bad. I just worry that I won't need or take advantage of even half the features available, really the only big thing for me is the HDMI input, so I'm not sure how eager I'll be to upgrade right away.

[vol.2]

I just worry that I won't need or take advantage of even half the features available

Right. Which is exactly why I'm looking to confirm a feature that would make it useful for me. It's a more advanced version of an already niche-use product; most people who already have a regular one are happy with it and spent time and money getting it going.

[strayan]

I pity the person who is going to write the manual for this.

[Fudoh]

I pity the person who is going to write the manual for this.

I volunteer

[ldeveraux]

I pity the person who is going to write the manual for this.

Is there a manual for the OSSC? I usually just ask here...

[ASDR]

Is there a manual for the OSSC? I usually just ask here...

http://junkerhq.net/xrgb/index.php?title=OSSC

I just worry that I won't need or take advantage of even half the features available

Right. Which is exactly why I'm looking to confirm a feature that would make it useful for me. It's a more advanced version of an already niche-use product; most people who already have a regular one are happy with it and spent time and money getting it going.

I'd say the general idea is that the original OSSC has fundamental limitations, such as not having enough memory to buffer a few fields / frames. This means it can never do the following:

- Provide high-quality deinterlacing which requires access to previous fields
- Do Yoko/Tate rotation, which requires buffering a full frame
- Normalize any compatible input signal to a standard refresh rate and resolution, which requires buffering a frame, reformatting it and then scanning it out in a different refresh rate
- Saving a screenshot to an SD card etc., requires a framebuffer (guess you could write a static input image line-by-line...)

And so on. The OSSC Pro should be able to do all of the above and has additional hardware features like extra inputs, an expansion port and a more powerful FPGA. This doesn't necessarily mean it'll have all or any of these features on day one, but the limits of what the OSSC hardware can do seems to have been hit and we need a new platform for all these more advanced things to be implemented.

Even though it is the best we currently have, the idea that the regular OSSC is already 'good enough' shows a strong lack of imagination of what a video processor for video games could be. We could apply machine learning to reconstruct the original high-res color signal from composite video, use post-process anti-aliasing techniques to improve early 3D games, add true CRT simulation using 4k HDR displays to really get the look right, we could have the video processor learn a console's defect like jail bars from a series of test patterns generated by something like the 240p test suite and compensate for these errors, and so on. Things that go far beyond even what the OSSC Pro will likely be able to do.

I think there's a huge demand for what the OSSC Pro will offer and even beyond that. I'd love to have a lag free HQ deinterlacer, Yoko/Tate rotation, the option to enable a 100% standard output signal, HDMI input, expansion modules with composite input etc. And looking at what machine learning is doing for video and 3D game upscaling and for improving image quality in smartphone cameras, I hope one day we get a video processor with these types of features as well.

[fernan1234]

The fact that there is still significant demand (relative to the size of this market) for the Framemeister even at inflated prices should tell you a lot, especially considering that besides being composite/s-video ready, it is inferior to what the OSSC Pro will be in too many ways to list.

[vol.2]

No doubt there will be a demand for the Pro "...(relative to the size of this market)," buoyed by all those who have foregone a OSSC in lieu of the Pro when it finally comes to fruition.

But the fact remains that "...the regular OSSC is..." in fact "...already 'good enough'..." for a great number of people.

The more inclusive of features it is when it comes out, the more people's imagination will be stimulated.

[Chacranajxy]

Not deterred by that pricing for a second. I'm willing to pay for quality.

[maxtherabbit]

I think the price is completely fair. The barebones version is the standard OSSC which, if you've been following along, you'd know was never intended to be replaced by this device. It is a different market segment with a different purpose.

[vol.2]

I think the price is completely fair.

I don't think that has been contested? I said it sounds fair. That seems to be the reaction. Or were you just agreeing?

The barebones version is the standard OSSC

The OSSC guy said they were toying around with the complexity of the enclosure. I meant bare-bones as in "board-level," not "feature reduced." Hell, I hope they pack it with all kinds of features. The more the better.

[Gunstar]

I'm lucky I can afford it with the case but if a 'no-case' version is made available so people on tight budgets can get one then I'm all for it. The more people who have these the better for the community.

[XtraSmiley]

He said price can be much lower if they build a lot in the first batch.

I say do a Pre-order that way people can put thier money where their mouth is and hopefully we can drive the price down.

I'd say do the optimitic price and offer a refund if the pre-sale number isn't met to hit that number.

[maxtherabbit]

I think the price is completely fair.

I don't think that has been contested? I said it sounds fair. That seems to be the reaction. Or were you just agreeing?

The barebones version is the standard OSSC

The OSSC guy said they were toying around with the complexity of the enclosure. I meant bare-bones as in "board-level," not "feature reduced." Hell, I hope they pack it with all kinds of features. The more the better.

Yeah I was mostly just echoing the others while also trying to underscore that the pro is not a replacement for the standard. IMO most people would be just fine with the regular OSSC unless they have a use case that requires the extra features

[Drunk_Caterpillar]

Honestly, I feel like HDMI input is a pretty necessary feature. At this point I'm basically only going to use the YPbPr inputs for Saturn and PC Engine—everything else has a digital output mod available.

[Bassa-Bassa]

Sorry if this has been explained already but the thread is too long for me right now: How good as a Mister device for CRT usage can we expect this to be?

Also, 720p into 1440p with scanlines is definitely discarded/not possible?

[Lawfer]

Also, 720p into 1440p with scanlines is definitely discarded/not possible?

Would like to have confirmation for this as well.

[marqs]

Sorry if this has been explained already but the thread is too long for me right now: How good as a Mister device for CRT usage can we expect this to be?

On paper it has all the key blocks to run various emulator cores, but you'll get your answer only after the devs have tinkered with the system. That's assuming they're interested in porting the cores & infrastructure to a new board in the first place.

Also, 720p into 1440p with scanlines is definitely discarded/not possible?

The upcoming prototype can output up to 1920x1440@60Hz, but there are plans to evaluate Sil1136 TX chip before release. Its 300MHz pclk support enabling modes such as 1920x1080@120Hz and 2560x1440@60Hz is tempting, but at the same time it's a major risk due to practically non-existent documentation and support. So even if it appears to work fine on an evaluation board, a lot of verification needs to be done to ensure coverage of all current and potential future requirements before it could be considered as a replacement for the current TX chip. That would mean yet another delay. Another challenge is the +50% performance requirement of such modes, meaning the FPGA model would possibly need to be higher speed grade which equals higher price. In short, 2560x1440@60Hz output might be possible but not without added delay and cost.

[vol.2]

The upcoming prototype can output up to 1920x1440@60Hz

Is that set-in-stone upper-limit for x,y resolution, or is it based on pixel clock? For example, will we be able to do 3840x240p out to a CRT?


If it could, that would bring it in line with the capabilities of the raspberry pi 4 running Retroarch (but of course with the added benefit of scaling original hardware).

[Bassa-Bassa]

Thanks for the answers marqs.

If it could, that would bring it in line with the capabilities of the raspberry pi 4 running Retroarch (but of course with the added benefit of scaling original hardware).

Better yet, with the capabilities of a decent PC running Groovymame. Quite handy to quickly solve overscan issues.

[ldeveraux]

Honestly, I feel like HDMI input is a pretty necessary feature. At this point I'm basically only going to use the YPbPr inputs for Saturn and PC Engine—everything else has a digital output mod available.

Yeah, this right here! I'm sure there are other additions to the Pro I'm forgetting too though. I'm pretty naive when it comes to all the settings the OSSC has to offer, but excited to see what it can do with an HDMI signal if anything. Personally, that's not worth $500 for me, but I could be convinced of the $350.

[kitty666cats]

I'd say the general idea is that the original OSSC has fundamental limitations, such as not having enough memory to buffer a few fields / frames. This means it can never do the following:

- Provide high-quality deinterlacing which requires access to previous fields
- Do Yoko/Tate rotation, which requires buffering a full frame
- Normalize any compatible input signal to a standard refresh rate and resolution, which requires buffering a frame, reformatting it and then scanning it out in a different refresh rate
- Saving a screenshot to an SD card etc., requires a framebuffer (guess you could write a static input image line-by-line...)

And so on. The OSSC Pro should be able to do all of the above and has additional hardware features like extra inputs, an expansion port and a more powerful FPGA. This doesn't necessarily mean it'll have all or any of these features on day one, but the limits of what the OSSC hardware can do seems to have been hit and we need a new platform for all these more advanced things to be implemented.

Even though it is the best we currently have, the idea that the regular OSSC is already 'good enough' shows a strong lack of imagination of what a video processor for video games could be. We could apply machine learning to reconstruct the original high-res color signal from composite video, use post-process anti-aliasing techniques to improve early 3D games, add true CRT simulation using 4k HDR displays to really get the look right, we could have the video processor learn a console's defect like jail bars from a series of test patterns generated by something like the 240p test suite and compensate for these errors, and so on. Things that go far beyond even what the OSSC Pro will likely be able to do.

I think there's a huge demand for what the OSSC Pro will offer and even beyond that. I'd love to have a lag free HQ deinterlacer, Yoko/Tate rotation, the option to enable a 100% standard output signal, HDMI input, expansion modules with composite input etc. And looking at what machine learning is doing for video and 3D game upscaling and for improving image quality in smartphone cameras, I hope one day we get a video processor with these types of features as well.

I wonder, though: what about a regular OSSC or GBScontrol chained with a secondary video processor? I feel that there's many combinations that can pull off a lot of these desirable results for a fraction of $500 - though I assume there is very high potential that no chain out there could do all the upcoming magic of OSSC Pro with quite as low lag...?

Of course, that negates the neatness and convenience of an all-in-one powerhouse of a single device.

[Greg2600]

500!!!!!!!!!!! Come on, a 300% increase for a "Pro" model of anything is vein popping. I wanted a nice case, plus the expansion board for composite/S-video inputs, and didn't expect to have to spend anywhere near that. I could get the current OSSC and a RetroTink for half that. What in the holy Hell components are being put in this?

[orange808]

I'd say the general idea is that the original OSSC has fundamental limitations, such as not having enough memory to buffer a few fields / frames. This means it can never do the following:

- Provide high-quality deinterlacing which requires access to previous fields
- Do Yoko/Tate rotation, which requires buffering a full frame
- Normalize any compatible input signal to a standard refresh rate and resolution, which requires buffering a frame, reformatting it and then scanning it out in a different refresh rate
- Saving a screenshot to an SD card etc., requires a framebuffer (guess you could write a static input image line-by-line...)

And so on. The OSSC Pro should be able to do all of the above and has additional hardware features like extra inputs, an expansion port and a more powerful FPGA. This doesn't necessarily mean it'll have all or any of these features on day one, but the limits of what the OSSC hardware can do seems to have been hit and we need a new platform for all these more advanced things to be implemented.

Even though it is the best we currently have, the idea that the regular OSSC is already 'good enough' shows a strong lack of imagination of what a video processor for video games could be. We could apply machine learning to reconstruct the original high-res color signal from composite video, use post-process anti-aliasing techniques to improve early 3D games, add true CRT simulation using 4k HDR displays to really get the look right, we could have the video processor learn a console's defect like jail bars from a series of test patterns generated by something like the 240p test suite and compensate for these errors, and so on. Things that go far beyond even what the OSSC Pro will likely be able to do.

I think there's a huge demand for what the OSSC Pro will offer and even beyond that. I'd love to have a lag free HQ deinterlacer, Yoko/Tate rotation, the option to enable a 100% standard output signal, HDMI input, expansion modules with composite input etc. And looking at what machine learning is doing for video and 3D game upscaling and for improving image quality in smartphone cameras, I hope one day we get a video processor with these types of features as well.

I wonder, though: what about a regular OSSC or GBScontrol chained with a secondary video processor? I feel that there's many combinations that can pull off a lot of these desirable results for a fraction of $500 - though I assume there is very high potential that no chain out there could do all the upcoming magic of OSSC Pro with quite as low lag...?

Of course, that negates the neatness and convenience of an all-in-one powerhouse of a single device.

Here's how I see it:

About lag and a secondary machine: I still use two DVDO's for different things because they are fast. Of course, they are 4:2:2, so the OSSC Pro still beats them. An Extron is a full frame of lag. Too slow for me. I have a few very nice Extrons and they are not even hooked up right now (I'm good with 1080p output from my chain right now). I've also had good luck with 960p on a DVDO Duo on the launch firmware, but where you gonna source that? If the OSSC Pro can line double and perform video processing with half a frame of lag at 4:4:4, it will beat any OSSC and VP combo I know of. This alone makes the OSSC Pro a great bargain because it's offering something you can't buy anywhere else. (That's right, no 4:4:4 sub-frame lag video processors exist today!)

You'll have to invest time and trouble into the Gonbes. That's fine for some of us but, most people don't build their own machines. It's cheap if you can do the labor. Certainly a machine worth building if you have the ability and time. I understand the upscaling is a frame of lag. Too much for me, thanks.

Want tate rotation and playable latency? I found a few machines at fair prices. I also made a good profit selling them at real market value, because those machines aren't supposed to be cheap. There are high latency options available for less, but who wants to play a vertical arcade shooter PCB with tons of lag? Also, if the OSSC Pro has rotation and padding with one frame of lag, it will best my preferred solution: Calibre HQV Reon machines that specifically support rotation (and those are 4:2:2). I see only one on eBay (the "lite" version Kramer vp-792) and it's $750usd. There's also the Datapath x4, but getting padding is a real pain in the arse. The Calibre ones are also a (much smaller and easier) challenge, because you need to create a "custom warp" to get the lower latency with pan/scan turned off.

Warping features may or may not happen--and many gamers may not like the effect. I hope warp comes to OSSC Pro, because I like the option of CRT curvature. There are a few obscure solutions meant for beamers that do it fast, but they are picky beasts and 100% network controlled (no buttons or remote). I don't recommend them to you. Those are also hard to find and could be expensive. For ease of use and compatibility with signals, the best of the bunch is the tate machine:
a Calibre HQV Reon; they have warp, rotate, padding, and "just okay" 4:2:2 scaling (and physical buttons and a GUI). It's a frame and a half of latency. An OSSC Pro with warping could be faster and it would also be 4:4:4. May or may not happen.

Let's not forget customer service. I've had great experiences with Extron, but many people here have not. Calibre's customer service flat out lied to me. DVDO is basically dead. Aurora is trying to reinvent itself as a video over IP company. TVOne is adrift and I have no idea where they are going; lots of competition in digital signage machines. Kramer machines have too much latency and they seem obsessed with video.

Let's also not forget the OSSC Pro will be a brand new machine and guaranteed to arrive working properly--and backed by people that we trust.

4:4:4, video processing, video game features, low lag, readily available, brand new, with good service. Yes, please.