Neo Geo Framemeister profile wip:

pva

pva

Goal! Goal! Goal!
Joined
Aug 28, 2017
Posts
110
FBX said:
Got an OSSC. And yeah, blows the Framemeister away, but is quite a bit more complex to develop optimal settings for. I'm getting the hang of it though. Check out my AES in 4x optimal integer scaling on the OSSC
Click to expand...

Care to share your settings?

In the spirit of you showing me yours if I'll show you mine first, here are my settings for Lx3/Lx4 320x240 optim. mode on my Japanese rev. 3-6 AES. I came up with the values by using the formula posted by marqs on the Shmups forums, and some framerate and timing information gleaned from the NG dev wiki.

Code: 
H. samplerate	384
H. synclen	28
H. backporch	28
H. active	320
V. synclen	1
V. backporch	6
V. active	256

Note that I had to lower the V. synclen (8 scanlines per the spec) and V. backporch (16 scanlines) values quite a bit, because otherwise I couldn't get an image on my Panasonic VT60 plasma. (The lower values result in the picture being a bit off-centre vertically.) I must also remember to lower V. active to 240 each time after loading the profile to restore the proper AR. If I save the profile with V. active set to 240, no image is displayed before it's increased to 256 again. Both of these are quirks of my display, so YMMV.
 
FBX

FBX

Quiz Detective
Joined
Aug 7, 2017
Posts
89
pva said:
Care to share your settings?
Click to expand...

No problem, but keep in mind my focus has been on H.samplerate and Sampling phase. Obviously porches, sync lengths, and active areas are going to be entirely display-dependant. My settings are meant purely for video capture and/or streaming using Datapath Vision cards (I'm using the E1s). Once I'm confident I've mastered optimal timing functions, I'll then start working on display-compatible versions of them.

At any rate, my AES is a 3-3, and I didn't realize that the RGB SCART cable I have uses composite video for sync. As such, I had to add in a medium amount of sync LPF in addition to 9Mhz of video LPF. So with those conditions, here are the settings for Datapath Vision cards:

Video LPF: 9Mhz

Sampling phase: 315 degrees
H. samplerate: 384
H. synclen: 32
H. backporch: 24
H. active: 320

V. synclen: 3
V. backporch: 21
V. active: 224

Analog sync LPF: 10Mhz

This was the absolute cleanest I could get the RGB image to look with my stock 3-3. It's almost perfect, but there are a few slight anomalies on solid colors. It couldn't be helped though. Here's a screenshot where you can see the sampling errors in the timer countdown numbers:

Neo Geo - OSSC - 4x - 01.png
 
pva

pva

Goal! Goal! Goal!
Joined
Aug 28, 2017
Posts
110
FBX said:
No problem, but keep in mind my focus has been on H.samplerate and Sampling phase.
Click to expand...

Out of interest, how did you arrive at the optimal sampling phase? I've been using the sharpness test pattern from the monitor test tool ROM included with your FM profiles, and while I can rule out the most egregious-looking phases, I've had a hard time deciding on the best setting. Are there any specific things to look out for?

At any rate, my AES is a 3-3, and I didn't realize that the RGB SCART cable I have uses composite video for sync. As such, I had to add in a medium amount of sync LPF in addition to 9Mhz of video LPF.
Click to expand...

Oh, I forgot to mention that my AES is connected through a gscart_lite with its sync cleaning/regeneration feature is turned on, as I initially encountered some sync problems. As a result, I've left Analog sync LPF to Off, because the gscart_lite will strip the AES' sync signal and digitally generate a stable signal in its place.

I've also left Video LPF to Off because I couldn't discern any difference between it and the 9 MHz (SD) setting. Again, are there any visual artifacts I should be keeping an eye out for?
 
FBX

FBX

Quiz Detective
Joined
Aug 7, 2017
Posts
89
pva said:
Out of interest, how did you arrive at the optimal sampling phase? I've been using the sharpness test pattern from the monitor test tool ROM included with your FM profiles, and while I can rule out the most egregious-looking phases, I've had a hard time deciding on the best setting. Are there any specific things to look out for?
Click to expand...

I'm actually going to be releasing video showing how to tune the best possible phase settings for each console. I'll be working on it tomorrow and will likely have it upload on youtube in a couple days. I think you'll like the trick I use, because it takes a lot of the guess work out of trying to find optimal phase timing.
 
pva

pva

Goal! Goal! Goal!
Joined
Aug 28, 2017
Posts
110
FBX said:
I'm actually going to be releasing video showing how to tune the best possible phase settings for each console. I'll be working on it tomorrow and will likely have it upload on youtube in a couple days. I think you'll like the trick I use, because it takes a lot of the guess work out of trying to find optimal phase timing.
Click to expand...

That sounds great (and delightfully meticulous), looking forward to seeing it! You're KarbuncleX on Youtube, right?
 
FBX

FBX

Quiz Detective
Joined
Aug 7, 2017
Posts
89
Sorry for the delay in getting the video uploaded. I had a ton of other things I needed to finish before I could get to it. Anyway, it's recorded in 1080p with the line4x image centered, so please watch in 1080p fullscreen mode for best results. Cheers!

https://www.youtube.com/watch?v=EBStHr4XCTg
 
pva

pva

Goal! Goal! Goal!
Joined
Aug 28, 2017
Posts
110
FBX said:
Sorry for the delay in getting the video uploaded. I had a ton of other things I needed to finish before I could get to it. Anyway, it's recorded in 1080p with the line4x image centered, so please watch in 1080p fullscreen mode for best results. Cheers!
Click to expand...

I finally had the time to watch this and tinker with the timings for my AES and Japanese MD2. The process was dead easy, and both consoles now look absolutely pixel perfect in Lx4/320x240 optim mode. My hat's off to you, sir!
 
You must log in or register to reply here.
Top