OSSC Pro: Blue tint to YPbPr if ALC is disabled

[Zacabeb]

I tested disabling the ALC with my PS2 running through YPbPr and found that there is a slight positive bias to the Pb signal. This was still present after disconnecting the Pb connection as well as swapping the Pr and Pb connections, making it possible to rule out the PS2 and its YPbPr cable. The gain and bias values are at their default values and the picture looks normal with ALC enabled.

With RGB-SCART input, there is no equivalent biasing of the blue component when the ALC is disabled. I don’t have any RGsB source with which to test AV2.

Is this normal behavior? I don’t know if I’ll ever have a legit reason to disable the ALC, especially with YPbPr sources, but it would be nice to know.

When I get the time I’ll test my other OSSC Pro to see if it exhibits the same behavior.

I’ve now quickly tested the other OSSC Pro (firmware 0.73) and it does not exhibit the above behavior, although I haven’t been able to test yet how else it behaves.

Edit: I’ve now tested YPbPr over AV1 and it exhibits the same behavior when ALC is turned off. It seems to be an anomaly with this particular ISL51002 or a problem with the board. Again, turning ALC on corrects for the problem.

Edit 2: Lowering Pb offset from 512 to about 418 corrects the offset. Adjusting the Pb gain up (or down) causes the offset to cyclically increase and decrease. Could it be a problem with DC restoration? Again, it doesn’t seem to affect RGB input. I’m stumped.

• This topic was modified 7 months, 1 week ago by Zacabeb. Reason: Fixed a typo in the title... :P
• This topic was modified 7 months, 1 week ago by Zacabeb. Reason: Added more information
• This topic was modified 7 months, 1 week ago by Zacabeb. Reason: More information added to help solve the mystery

[Zacabeb]

@marqs Pinging you since I wish to know if you think it’s normal behavior or if I might need to RMA the affected OSSC Pro.

[marqs]

I’ve seen color unevenness when ALC is disabled, but have not really compared results between different boards. It doesn’t necessarily mean the chip is faulty. The function of ALC is to mitigate offset error resulting from changes in temperature, individual chip characteristics etc. You can read more about that in ISL51002 datasheet (called ABLC there). In normal use there should be no reason to disable ALC.

[Zacabeb]

Thanks, that has helped ease my mind. 🙂

The reason I tested turning the ALC off in the first place was to check how badly the incoming signals needed to be compensated and if there was any clipping with ALC on (there wasn’t). So I just happened to make this observation by chance.

With ALC turned on, everything looks correct and very stable, so I’ll just consider the slight blue bias with ALC off part of the AFE/ADC’s individual personality. 😀

[Author]

Posts