Re: SANE_FRAME_IR

Benjamin Gordon <[email protected]> Wed, 22 Jul 2026 17:20:19 -0600
Newsgroups gmane.comp.graphics.scanning.sane.devel
Message-ID <CADaYBq3LeEU=oaE--mVM1bb8Unv1Cjj1E1DmodtOHQEp1pv_LA@mail.gmail.com>
--00000000000007b83806573b63bb
Content-Type: text/plain; charset="UTF-8"
Content-Transfer-Encoding: quoted-printable

On Wed, Jul 22, 2026 at 11:33=E2=80=AFAM Ralph Little <[email protected]> =
wrote:

> Hi,
> We have had a question from a developer regarding support for IR from a
> scanner, namely the LS-5000 supported by the coolscan3 backend.
> This device can supply IR but there are a few different ways in which
> this can be delivered to the frontend. The brief discussion is here:
> https://gitlab.com/sane-project/backends/-/merge_requests/937
>
> As I said in that discussion, I believe the SANE_FRAME_IR option (or
> some variant) is the better one to explicitly support it from the SANE
> standard. I assume that this was a proposed extension at some point.
> However, this would mean we would need to ratify a new standard and ask
> frontend developers to support the new features, something that we might
> struggle to do ourselves what with the backlog we currently have.
>
> There are mentioned some other options, such as delivering the IR frame
> as grey in the next frame, which is done currently by another backend,
> and has the advantage of being supported right now.
>
> Views? Any appetite for an update to the SANE Standard? I'm sure there
> are a few things that people would like to add.
>

If there's the possibility of re-opening the standard, I for one would like
to consider a more general handling of frame formats.  All the scanners
that support eSCL these days can return JPEG directly, and some of them can
even return a whole PDF at once.  It would be great for the frontend to be
able to negotiate some sort of passthrough mode instead of decoding these
and then re-encoding in the frontend.

Thanks,
Benjamin

--00000000000007b83806573b63bb
Content-Type: text/html; charset="UTF-8"
Content-Transfer-Encoding: quoted-printable

<div dir=3D"ltr"><div dir=3D"ltr"><span style=3D"background-color:transpare=
nt">On Wed, Jul 22, 2026 at 11:33=E2=80=AFAM Ralph Little &lt;<a href=3D"ma=
ilto:[email protected]">[email protected]</a>&gt; wrote:</span></div><div=
 class=3D"gmail_quote gmail_quote_container"><blockquote class=3D"gmail_quo=
te" style=3D"margin:0px 0px 0px 0.8ex;border-left:1px solid rgb(204,204,204=
);padding-left:1ex">Hi,<br>
We have had a question from a developer regarding support for IR from a <br=
>
scanner, namely the LS-5000 supported by the coolscan3 backend.<br>
This device can supply IR but there are a few different ways in which <br>
this can be delivered to the frontend. The brief discussion is here: <br>
<a href=3D"https://gitlab.com/sane-project/backends/-/merge_requests/937" r=
el=3D"noreferrer" target=3D"_blank">https://gitlab.com/sane-project/backend=
s/-/merge_requests/937</a><br>
<br>
As I said in that discussion, I believe the SANE_FRAME_IR option (or <br>
some variant) is the better one to explicitly support it from the SANE <br>
standard. I assume that this was a proposed extension at some point. <br>
However, this would mean we would need to ratify a new standard and ask <br=
>
frontend developers to support the new features, something that we might <b=
r>
struggle to do ourselves what with the backlog we currently have.<br>
<br>
There are mentioned some other options, such as delivering the IR frame <br=
>
as grey in the next frame, which is done currently by another backend, <br>
and has the advantage of being supported right now.<br>
<br>
Views? Any appetite for an update to the SANE Standard? I&#39;m sure there =
<br>
are a few things that people would like to add.<br></blockquote><div><br></=
div><div>If there&#39;s the possibility of re-opening the standard, I for o=
ne would like to consider a more general handling of frame formats.=C2=A0 A=
ll the scanners that support eSCL these days can return JPEG directly, and =
some of them can even return a whole PDF at once.=C2=A0 It would be great f=
or the frontend to be able to negotiate some sort of passthrough mode inste=
ad of decoding these and then re-encoding in the frontend.</div><div><br></=
div><div>Thanks,</div><div>Benjamin</div><div><br></div></div></div>

--00000000000007b83806573b63bb--