mirror of https://lore.kernel.org/lkml/
 help / color / mirror / Atom feed
From: "Jernej Škrabec" <jernej.skrabec@gmail.com>
To: Neil Armstrong <narmstrong@baylibre.com>,
	Andrzej Hajda <a.hajda@samsung.com>,
	Robert Foss <robert.foss@linaro.org>,
	Laurent Pinchart <Laurent.pinchart@ideasonboard.com>,
	Alex Bee <knaerzche@gmail.com>
Cc: Jonas Karlman <jonas@kwiboo.se>,
	dri-devel@lists.freedesktop.org, linux-kernel@vger.kernel.org,
	Alex Bee <knaerzche@gmail.com>
Subject: Re: [PATCH] drm: bridge: dw-hdmi: Fix RGB to YUV color space conversion
Date: Wed, 10 Nov 2021 21:20:46 +0100	[thread overview]
Message-ID: <12887538.uLZWGnKmhe@kista> (raw)
In-Reply-To: <20211106130044.63483-1-knaerzche@gmail.com>

Hi Alex,

Dne sobota, 06. november 2021 ob 14:00:44 CET je Alex Bee napisal(a):
> As per CEA-861 quantization range is always limited in case of YUV
> output - indepentently which CEA mode it is or if it is an DMT mode.
> 
> This is already correctly setup in HDMI AVI inforame, but we always do
> a RGB to YUV conversion which doesn't consider that RGB input can be
> full range as well.
> That leads to incorrect colors for all CEA modes except mode 1 for HDMI
> and always for DVI.
> 
> To fix this, provide additional csc coefficents for converting from RGB
> full range to EITU601/EITU709 limited range and rename the existing
> arrays to clarify their meaning.
> 
> Signed-off-by: Alex Bee <knaerzche@gmail.com>
> ---
>  drivers/gpu/drm/bridge/synopsys/dw-hdmi.c | 24 +++++++++++++++++++----
>  1 file changed, 20 insertions(+), 4 deletions(-)
> 
> diff --git a/drivers/gpu/drm/bridge/synopsys/dw-hdmi.c b/drivers/gpu/drm/
bridge/synopsys/dw-hdmi.c
> index 62ae63565d3a..1cba08b70091 100644
> --- a/drivers/gpu/drm/bridge/synopsys/dw-hdmi.c
> +++ b/drivers/gpu/drm/bridge/synopsys/dw-hdmi.c
> @@ -80,13 +80,25 @@ static const u16 csc_coeff_rgb_out_eitu709[3][4] = {
>  	{ 0x2000, 0x0000, 0x3b61, 0x7e25 }
>  };
>  
> -static const u16 csc_coeff_rgb_in_eitu601[3][4] = {
> +static const u16 csc_coeff_rgb_full_in_eitu601[3][4] = {

I would rather move "full" and "limited" to the end, since RGB always has full 
range and we want YUV to have full or limited range.

Just one observation - no other matrix sets bit 15 in any coefficient, but yours 
do. I can't see anywhere documented if bit 15 is ignored or not. Can you try 
with it set to 0? If it still works, I would set it to 0 for consistency.

Best regards,
Jernej

> +	{ 0x2044, 0x106f, 0x0644, 0x0040 },
> +	{ 0xe677, 0x1c1c, 0xfd46, 0x0200 },
> +	{ 0xed60, 0xf685, 0x1c1c, 0x0200 }
> +};
> +
> +static const u16 csc_coeff_rgb_limited_in_eitu601[3][4] = {
>  	{ 0x2591, 0x1322, 0x074b, 0x0000 },
>  	{ 0x6535, 0x2000, 0x7acc, 0x0200 },
>  	{ 0x6acd, 0x7534, 0x2000, 0x0200 }
>  };
>  
> -static const u16 csc_coeff_rgb_in_eitu709[3][4] = {
> +static const u16 csc_coeff_rgb_full_in_eitu709[3][4] = {
> +	{ 0x2750, 0x0baf, 0x03f8, 0x0040 },
> +	{ 0xe677, 0x1c1c, 0xfd6d, 0x0200 },
> +	{ 0xea55, 0xf98f, 0x1c1c, 0x0200 }
> +};
> +
> +static const u16 csc_coeff_rgb_limted_in_eitu709[3][4] = {
>  	{ 0x2dc5, 0x0d9b, 0x049e, 0x0000 },
>  	{ 0x62f0, 0x2000, 0x7d11, 0x0200 },
>  	{ 0x6756, 0x78ab, 0x2000, 0x0200 }
> @@ -1023,9 +1035,13 @@ static void dw_hdmi_update_csc_coeffs(struct dw_hdmi 
*hdmi)
>  			csc_coeff = &csc_coeff_rgb_out_eitu709;
>  	} else if (is_input_rgb && !is_output_rgb) {
>  		if (hdmi->hdmi_data.enc_out_encoding == 
V4L2_YCBCR_ENC_601)
> -			csc_coeff = &csc_coeff_rgb_in_eitu601;
> +			csc_coeff = hdmi->hdmi_data.rgb_limited_range
> +				? &csc_coeff_rgb_limited_in_eitu601
> +				: &csc_coeff_rgb_full_in_eitu601;
>  		else
> -			csc_coeff = &csc_coeff_rgb_in_eitu709;
> +			csc_coeff = hdmi->hdmi_data.rgb_limited_range
> +				? &csc_coeff_rgb_limted_in_eitu709
> +				: &csc_coeff_rgb_full_in_eitu709;
>  		csc_scale = 0;
>  	} else if (is_input_rgb && is_output_rgb &&
>  		   hdmi->hdmi_data.rgb_limited_range) {
> 
> base-commit: 89636a06fa2ee7826a19c39c19a9bc99ab9340a9
> -- 
> 2.30.2
> 
> 



  reply	other threads:[~2021-11-10 20:20 UTC|newest]

Thread overview: 3+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2021-11-06 13:00 Alex Bee
2021-11-10 20:20 ` Jernej Škrabec [this message]
2021-11-10 21:26   ` Jernej Škrabec

Reply instructions:

You may reply publicly to this message via plain-text email
using any one of the following methods:

* Save the following mbox file, import it into your mail client,
  and reply-to-all from there: mbox

  Avoid top-posting and favor interleaved quoting:
  https://en.wikipedia.org/wiki/Posting_style#Interleaved_style

* Reply using the --to, --cc, and --in-reply-to
  switches of git-send-email(1):

  git send-email \
    --in-reply-to=12887538.uLZWGnKmhe@kista \
    --to=jernej.skrabec@gmail.com \
    --cc=Laurent.pinchart@ideasonboard.com \
    --cc=a.hajda@samsung.com \
    --cc=dri-devel@lists.freedesktop.org \
    --cc=jonas@kwiboo.se \
    --cc=knaerzche@gmail.com \
    --cc=linux-kernel@vger.kernel.org \
    --cc=narmstrong@baylibre.com \
    --cc=robert.foss@linaro.org \
    /path/to/YOUR_REPLY

  https://kernel.org/pub/software/scm/git/docs/git-send-email.html

* If your mail client supports setting the In-Reply-To header
  via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line before the message body.
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox

all inboxes | Powered by JetHome®