From mboxrd@z Thu Jan 1 00:00:00 1970 Return-path: Received: from mail-ua0-f176.google.com ([209.85.217.176]:39991 "EHLO mail-ua0-f176.google.com" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1752654AbeCFDDD (ORCPT ); Mon, 5 Mar 2018 22:03:03 -0500 Received: by mail-ua0-f176.google.com with SMTP id c14so12066427uak.7 for ; Mon, 05 Mar 2018 19:03:03 -0800 (PST) Received: from mail-ua0-f176.google.com (mail-ua0-f176.google.com. [209.85.217.176]) by smtp.gmail.com with ESMTPSA id k17sm422889uaf.44.2018.03.05.19.03.01 for (version=TLS1_2 cipher=ECDHE-RSA-AES128-GCM-SHA256 bits=128/128); Mon, 05 Mar 2018 19:03:01 -0800 (PST) Received: by mail-ua0-f176.google.com with SMTP id q12so2476658uae.4 for ; Mon, 05 Mar 2018 19:03:01 -0800 (PST) MIME-Version: 1.0 In-Reply-To: References: <1519893856-4738-1-git-send-email-zhengsq@rock-chips.com> From: Tomasz Figa Date: Tue, 6 Mar 2018 12:02:40 +0900 Message-ID: Subject: Re: [PATCH] media: ov2685: Not delay latch for gain To: Shunqian Zheng Cc: Sakari Ailus , Mauro Carvalho Chehab , Linux Media Mailing List Content-Type: text/plain; charset="UTF-8" Content-Transfer-Encoding: quoted-printable Sender: linux-media-owner@vger.kernel.org List-ID: On Thu, Mar 1, 2018 at 7:14 PM, Shunqian Zheng wro= te: > Hi Tomasz, > > > On 2018=E5=B9=B403=E6=9C=8801=E6=97=A5 16:53, Tomasz Figa wrote: >> >> Hi Shunqian, >> >> On Thu, Mar 1, 2018 at 5:44 PM, Shunqian Zheng >> wrote: >>> >>> Update the register 0x3503 to use 'no delay latch' for gain. >>> This makes sensor to output the first frame as normal rather >>> than a very dark one. >> >> I'm not 100% sure on how this setting works, but wouldn't it mean that >> setting the gain mid-frame would result in half of the frame having >> old gain and another half new? Depending how this works, perhaps we >> should set this during initial register settings, but reset after >> streaming starts? > > Thank you. > > I'm not quite sure too. Then I try to change gain during capture by: > capture_10_frames.sh & while sleep .01; do v4l2-ctl -d /dev/video4 > --set-ctrl=3Danalogue_gain=3D54; sleep .01; v4l2-ctl -d /dev/video4 > --set-ctrl=3Danalogue_gain=3D1024; done > > The gain setting takes effect for every single frame, not in mid-frame fr= om > my test. Alright. I wasn't able to confirm the exact meaning of this bit myself unfortunately, but if that's the behavior you're seeing, we should be fine. Reviewed-by: Tomasz Figa Best regards, Tomasz