On 09.06.2010 20:13, Dave P wrote: > On Tuesday 08 Jun 2010, VDR User wrote: >> I'm not sure what FE_CAN_QAM_AUTO is supposed to represent exactly so >> the problem may not be in VDR, but rather your cards driver not >> accurately reporting it's capabilities. In which case, the driver >> needs to be fixed of course. > > The card is a Twinhan DST DVB-T using the bttv driver. I'm not an expert at > reading kernel source, but in drivers/media/dvb/bt8xx/dst.c it does seem > that FE_CAN_QAM_AUTO is unconditionally set. The card documentation > doesn't specify whether the device can in fact handle all possible QAM > settings. My understanding of the FE_CAN_QAM_... flags is that the driver must tell the application which modulations it can handle. If FE_CAN_QAM_AUTO is set, the application doesn't need to tell the driver which modulation to use, because the driver (or better: the device) can detect it automatically. If a driver only sets FE_CAN_QAM_AUTO, the applications can't know which of the various modulations the device can actually handle. I don't think that FE_CAN_QAM_AUTO is a synonym for "can handle *all* modulations". The simplest fix (if it is unknown which modulations the device can actually handle) is probably to set all FE_CAN_QAM_... flags in the driver. Your patch is not a solution, only a workaround. It would prevent drivers that can actually handle only a subset of modulations, but do have "auto" capability, from being properly handled by the application. Klaus _______________________________________________ vdr mailing list vdr@xxxxxxxxxxx http://www.linuxtv.org/cgi-bin/mailman/listinfo/vdr
