On Sun, 8 Jan 2006, sean wrote:
> > Turn off the chroma filters. The luma filters appeared to be OK but
>
> OK. I'll try that.
That cleared up the problem for me. But then a little later on I
moved up to capturing uncompressed with an external TBC, etc.
> Do you see any d
Steven M. Schultz wrote:
On Fri, 6 Jan 2006, sean wrote:
I've captured an old vhs tape via a canopus advc-300 with dvgrab.
The result shows colors lagging action. That is, a large
blue sleeve that moves across the screen is jerky, with
patches of color left behind. (I'm sure there's a bett
On Sun, 8 Jan 2006 16:11:50 +0100
Jean Delvare <[EMAIL PROTECTED]> wrote:
> Hi Ronald and all,
>
> I have been working on a few fixes and cleanups to the i2c chip drivers
> used with the zr36067 driver (adv7170, adv7175, bt819, bt856, saa7110,
> saa7111, saa7114 and saa7185).
Judging by that lis
The bt856 driver has a register cache much larger than needed. We
really only write to 3 registers, so a 32-byte cache is a bit too
much. We can be just as efficient with a 6-byte cache. We could even
do with a 3-byte cache, but at the cost of additional arithmetics
arguably not worth the spared 3
Change all the Zoran (ZR36050/ZR36060) drivers to use i2c_master_send
instead of i2c_transfer when possible. This simplifies the code by a
few lines in each driver.
Signed-off-by: Jean Delvare <[EMAIL PROTECTED]>
---
drivers/media/video/adv7170.c | 17 +++--
drivers/media/video/adv7
Hi Ronald and all,
I have been working on a few fixes and cleanups to the i2c chip drivers
used with the zr36067 driver (adv7170, adv7175, bt819, bt856, saa7110,
saa7111, saa7114 and saa7185). There are 7 patches, I'll post them now
in reply to this message. These patches are against 2.6.15-git4 a
Drop the adv7175 register cache, as it is only written to and never
read back from. This saves 128 bytes of memory and slightly speeds up
the register writes.
Signed-off-by: Jean Delvare <[EMAIL PROTECTED]>
---
drivers/media/video/adv7175.c | 10 ++
1 file changed, 2 insertions(+), 8 de
Explicitely state the number of registers the SAA7111 has, and use
that defined value where relevant. This should prevent any future
array overrun like the one I just fixed in the saa7110 driver.
This patch also saves 8 bytes of memory as a side effect, as the
register cache was larger than needed
Fix a (probably harmless) array overrun in the DECODER_DUMP command.
No big deal as this command is not used anywhere anyway. Also reformat
the dump so that it displays nicely.
Signed-off-by: Jean Delvare <[EMAIL PROTECTED]>
---
drivers/media/video/saa7110.c | 12 +---
1 file changed, 5
Drop support for the ENCODER_DUMP command in the adv7175 driver.
ENCODER_DUMP was never actually defined as far as I can see, so the
code was ifdef'd out, and I suspect it was never used, not even once,
as it includes an obvious array overrun.
The register values of this specific chip can be dumpe
Fix the i2c block write mode of the saa7114 driver. A previous code
change accidentally commented out a local variable increment, which
should have been kept, causing the register writes over the I2C bus
to never be batched, replacing any attempted block write by slower,
individual write transactio
11 matches
Mail list logo
