OSSC Pro bug reports

[Zacabeb]

I think there might be a need for an official thread for reporting bugs in the OSSC Pro firmware, but for now here’s an unofficial thread.

Firmware v0.74:

1) YCbCr output is flagged as full range. While YCbCr normally is always limited range in BT.709, some displays apply the specified quantization range regardless.

2) Auto phase doesn’t seem to work all the time, often being 180° out-of-phase, especially visible in optimized modes. This might be due a conflict between samples picked for phase detection and output. This was observed with various PSX optimized modes corrected for the PS2’s 54 MHz sample clock when playing PSX games (effectively 3,456 samples per line both output by the PS2 and sampled by the OSSC Pro), and using the official Sony YPbPr cable.

Note: Improved in firmware 0.75

3) Not sure if this happens all the time, but sometimes it seems the last altered non-profile dependent setting isn’t stored (e.g. LED PWM).

4) If setting the OSD to Off and then turning it back to Full again with the OSD button on the remote, it is possible to enter a state where the list item shown on the OLED display is different from the list item highlighted in the OSD.

Note: Fixed in firmware 0.75

5) The advanced timings for the newly added DC/PS2/GC 640col mode show up listing only one of the systems for each set of timings (for instance, the 480p timings is listed as “DC 640×480” instead of say, “DC/PS2/GC 640x480p”).

6) When the input timings change or a different scaling algorithm is selected, the OSSC Pro might not reload the stored phase value for the appropriate mode but seems to use auto phase or default to 0°. A workaround exists by pressing up or down on the phase hotkeys on the remote after a input timing change or selecting a different scaling algorithm, as this makes the stored phase value get reloaded (though adjusted up or down a step from pressing the hotkeys,) allowing the desired phase to be recovered quickly.

Note: Fixed in firmware 0.75

• This topic was modified 1 year, 3 months ago by Zacabeb. Reason: An additional bug listed
• This topic was modified 1 year, 3 months ago by Zacabeb. Reason: Yet one more bug
• This topic was modified 1 year, 2 months ago by Zacabeb. Reason: Notes added about fixes made in firmware 0.75

[triple_lei]

I got one but I’m not sure if it’s really a “bug “(happened with 0.73 as well):

My capture card program (Media Express, used with Blackmagic UltraStudio Recorder 3G) doesn’t see the audio unless I set the OSSC Pro audio to Quad stereo: on (4.0). That is, I’ll only see the proper movement for audio channels 1 and 2 in my capture card this way. (I’ll also get frozen bars for channels 5 and 6, but who cares)

[Jon]

Autophase doesn’t work for me with pc engine (256 mode), mega drive (256 mode) and saturn (352 mode).

[marqs]

1. Where do you see YCbCr flagged as full range? YQ bits should be 0

2,6. There is a fix which is under evaluation

3-4. Will need to look into these

5. Sampling preset names are not 1:1 with the setting names in menu

A test firmware to address the audio issue is linked on this thread. It also include a proposed fix for the sampling phase issues, but not fully tested yet.

[Jon]

I still have the same issues with auto phase with the firmware linked in that thread.

[Zacabeb]

@marqs I’m seeing the YCbCr flagged as full range on my Sony KDL-40W3000 TV. Since it supports sYCC, it accepts full range YCbCr. If the YQ bits are 0, maybe the TV is afffected by the ITC bit? It could be a quirk with Sony TVs of that era (around 2007). Maybe the YQ bits should be set to indicate limited range in YCbCr?

Regarding sampling presets, I think having the names correlate between preset names and advanced timings names would be helpful for the ones that cover multiple consoles.

It’s nice to see fixes for the phase and audio issues being in the works. 🙂

Edit: There’s a bug when altering sample rate that can break active area/backporch adjustment, which also needs to be looked into:

OSSC Pro bug: Positioning can break if sample rate is adjusted

• This reply was modified 1 year, 3 months ago by Zacabeb. Reason: Added another bug

[marqs]

ITC bit is 0 (which reminds I should add an option to set ITC like in Classic). YQ=0 already indicates limited range, and that is separate from xvYCC which is set via different bits. Does the TV report the detected range or is that just assumption based on e.g. visible headroom on luma?

The sampling rate adjustment issue is fixed and the change will be included in next firmware.

[Zacabeb]

The TV doesn’t directly report the detected range, but allows for overriding the auto setting for quantization range to force it to be treated as limited or full. I could see the picture was handled wrong by my TV and verified that its auto setting thinks the signal is full range.

Is limited/full range indicated by Q as well as YQ when transmitting YCbCr? My TV might be responding to the Q bits instead of YQ bits even when receiving YCbCr. 🤔

[marqs]

FW v0.75 fixes (4), and phase setting should be now more consistent (2, 6) but not guaranteed to be always optimal so some work remains to be done there.

[Richard N]

Hi, I’d like to add another issue to the Pro.

• Pro Firmware Version: 0.75
• Pro Setup: AV1 is connected to a PS1 via RGB SCART.  AV2 is connected to a PS2 via Component.  AV3 and AV4 are currently not connected to anything.  SPDIF is connected to the PS2 with its audio routed to AV2.
• Problem Category: Audio
• Problem Description: I sometimes get a picture without sound and the only way to resolve it is to turn the Pro off and on again.  The problem occurs when I turn the Pro on and it cannot be resolved by changing the inputs.  I’ve tested the audio in 3 ways: (i) via SCART on AV1, (ii) via Component on AV2, and (iii) via SPDIF routed to AV2.  Same issue exists in all cases.
• Request: Has anyone else experienced it too and is there an alternative setting / fix available?

[marqs]

Hi, I’d like to add another issue to the Pro.

Which display is in question and are you able to test with another one to see if it is a compatibility problem with the specific TV/monitor?

[Richard N]

Hi, it’s a Samsung NU7470 (here is its specs page in case it helps).  Unfortunately, I don’t have another TV with which to test the Pro.  The Pro is hooked up to the same HDMI port that’s also usable for ARC.  I don’t have any ARC receivers connected, but could it cause a problem?

[Richard N]

Hi Marqs, hope you’re keeping well. Just to check, is there any solution for my problem (reply #61656)? To recap, I sometimes get a picture without sound and the only way to resolve it is to turn the Pro off and on again.  The problem occurs when I turn the Pro on and it cannot be resolved by changing the inputs.  I’ve tested the audio in 3 ways: (i) via SCART on AV1, (ii) via Component on AV2, and (iii) via SPDIF routed to AV2.  Same issue exists in all cases. I’ve tested it with my Samsung NU7470 TV (specs here); I’ve no other TV to test it with. Could the problem be caused by using the HDMI ARC port?

[marqs]

I have not bumped into the issue myself or heard anyone else either. You can indeed try non-ARC HDMI input on the TV to rule out related potential issue on TV side.

[Richard N]

Thanks Marqs. I’ll keep trying it and will come back to you. Hopefully, it’s not a defective unit.

[Author]

Posts