Hi Alexander, > -----Ursprüngliche Nachricht----- > Von: Michael Riesch <Michael.Riesch@xxxxxxxxxxxxxx> > Gesendet: Dienstag, 25. Februar 2025 09:44 > An: Alexander Shiyan <eagle.alexander923@xxxxxxxxx>; Dave Stevenson > <dave.stevenson@xxxxxxxxxxxxxxx> > Cc: linux-media@xxxxxxxxxxxxxxx; Sakari Ailus <sakari.ailus@xxxxxxxxxxxxxxx>; > Mauro Carvalho Chehab <mchehab@xxxxxxxxxx>; Gerald Loacker > <Gerald.Loacker@xxxxxxxxxxxxxx> > Betreff: Re: [PATCH 2/4] media: i2c: imx415: Add missing ADBIT1 register for setup > output format > > Hi Alexander, Dave, > > On 2/20/25 19:47, Alexander Shiyan wrote: > > Hello Dave. > > > >>> The imx415 sensor has one more register for setting the output > >>> format (10/12 bit), which is currently not in the driver. > >> > >> The datasheet states it is the "internal A/D conversion bits" setting, > >> not output format. > >> Output format is set via MDBIT (reg 0x3032). > >> > >> The sensor may well happily truncate 12bit A/D readout to RAW10 > >> output, or left shift 10bit A/D values to RAW12 output. > >> > >> I'll defer to Wolfvision on this one as I would expect them as the > >> original authors to have been given a register list by Sony for the > >> readout modes that they were interested in. Sony may therefore have > >> recommended this apparent mismatch in A/D depth vs output format. > > Cc: Gerald who is already looking into this > > As far as I can gather, the register is not crucial for the correct > operation. The chip works fine for us with the current driver and ADBIT1 > is not mentioned in the register lists for the different modes. > > However, we have successfully set the register in a different project > (MCU + IMX415). > > The patch may not be essential, but it is clearly correct IMHO. > It's correct that this register needs to be set to zero for 10-bit operation according to the datasheet, although there are no visible differences. Reviewed-by: Gerald Loacker <gerald.loacker@xxxxxxxxxxxxxx> Tested-by: Gerald Loacker <gerald.loacker@xxxxxxxxxxxxxx> Best regards, Gerald > Best regards, > Michael > > > > > As far as I can see from the datasheet, the "Operating mode" table (page 48) > > clearly states that the AS-conversion and output bit width values are equal. > > > > Thanks!
