Am 07.01.2014 20:49, schrieb David Vincent-Jones:
> It is just a thought but I would be interested to have other opinions.
You got the 'non-destructive'-thing wrong. It means your RAW will
*never* be modified (binary, at storage location). It does not say
anything about the operations and manipula
I think currently this is more reliable than existing open-source
implementations of content-aware fill, such as resynthesizer.
2014/1/7 David Vincent-Jones :
> Spot removal works just fine but I am a little concerned regarding the
> philosophy of the module in that it appears to swim against the
Spot removal works just fine but I am a little concerned regarding the
philosophy of the module in that it appears to swim against the grain of
'non-destructive' procedure.
The module replaces material over a spot from another area and can
easily be used to eliminate major objects in an image.
If I use a circle to remove a spot the 'before' and 'repair' circles are
linked fairly close together ... easy to manage.
If I draw a shape to remove marks the link is impossibly long and often
on cropped images 'right off the map'.
A small adjustment is needed please.
David
--
darktable 1.3+107~g018e4f1 64 bit.
Multiple (even 2) 'curve-shaped' spot removals, within a single
instance, causes dt to go down immediately.
I notice also with 'curve shape' that the clean area can be located
entirely off the screen if the image has been cropped. This makes
grabbing and movin
On 3/3/13 9:18 PM, jeremy rosen wrote:
> I sort of agree except that before cahnging anything we REALLY need to
> make spot-size more discoverable. Ctrl-mousewheel alone won't make it.
99% of all programs I use (DT is the 1%) use border drag to change the
size of a region.
This what I did in equ
very good point AlicVB, I had forgotten about that...
On Sun, Mar 3, 2013 at 10:00 PM, AlicVB wrote:
> Hi all,
> interesting discussion indeed. But please don't change too much things
> now in spot removal, as in a (near ?) future, this iop will use the
> masks tools...
> So we will need a more
Hi all,
interesting discussion indeed. But please don't change too much things
now in spot removal, as in a (near ?) future, this iop will use the
masks tools...
So we will need a more global discussion about that.
FYI, I have already this change working at home :)
thanks
Aldric
Le 03/03/2013 2
On Mon, Mar 4, 2013 at 9:18 AM, jeremy rosen
wrote:
>> yes.. and i would agree to your idea of just keeping it as it is.. but
>>
>> if for some reason most people aren't handy enough with a mouse to
>> hover over the circle and can't take the mental pressure to interpret
>> the highlighted circle.
> yes.. and i would agree to your idea of just keeping it as it is.. but
> if for some reason most people aren't handy enough with a mouse to
> hover over the circle and can't take the mental pressure to interpret
> the highlighted circle.. :) i think ctrl+zoom for the spot size is the
> less intr
On Mon, Mar 4, 2013 at 8:58 AM, jeremy rosen
wrote:
> I understand your point, but let me do a couple of counter argument
>
> * scroll is easier to discover than control-scroll, so it makes the various
> iop manipulations easier to find. (one could argue that zoom is more
> important, but see my n
I understand your point, but let me do a couple of counter argument
* scroll is easier to discover than control-scroll, so it makes the various
iop manipulations easier to find. (one could argue that zoom is more
important, but see my next point)
* ctrl-mouswheel is a common shortcut for zoom, so
Le 03/03/2013 19:54, jeremy rosen a écrit :
> How about using control mouse for zoom as in light table and normal
> scroll for iop specific action?
I would prefer mouse for zomm and control mouse for iop because Lr does
that and because I do far more zooming into the image than changing some
iop s
Am 03.03.2013 19:54, schrieb jeremy rosen:
> How about using control mouse for zoom as in light table and normal
> scroll for iop specific action?
>
I'd prefer it the other way round. There are not so many iops with mouse
wheel control. OTOH zooming in and out of the image is one of the most
fre
How about using control mouse for zoom as in light table and normal scroll
for iop specific action?
Le 3 mars 2013 17:16, "David Vincent-Jones" a écrit :
> I am glad that I am not the only one anoyed by this behavour ... Alt+
> Scroll (for instance) would be a great solution.
>
> David
>
> On 13
I am glad that I am not the only one anoyed by this behavour ... Alt+
Scroll (for instance) would be a great solution.
David
On 13-03-03 04:16 AM, Moritz Moeller wrote:
On 2/3/13 10:11 PM, David Vincent-Jones wrote:
Spot removal really works well but I find that changing the spot size
conflic
On 2/3/13 10:11 PM, David Vincent-Jones wrote:
> Spot removal really works well but I find that changing the spot size
> conflicts badly with the image zoom and can be a tad frustrating. would
> it be possible to lock the image zoom while the spot module is open?
Bad idea imho. I zoom a lot while
definitely not. when I do spot removal I zoom in and out a lot. Very
important to choose the areas for correction.
it needs a little training to make sure your mouse is in/out of the area
depending on what you want your scroll to do, but we can't block zooming in
spot-removal. it's too important t
Spot removal really works well but I find that changing the spot size
conflicts badly with the image zoom and can be a tad frustrating. would
it be possible to lock the image zoom while the spot module is open?
David
--
On Fri, Dec 21, 2012 at 9:35 AM, Pascal Obry wrote:
> A new version of this branch is ready. See pull request.
thx will try to look at it soonish, but if i'm not reactive feel free
to ask on IRC for another reviewer
>
> BTW, what the best procedure to discuss enhancements?
>
> 1. Should this be
A new version of this branch is ready. See pull request.
BTW, what the best procedure to discuss enhancements?
1. Should this be discussed here before a pull-request?
2. Should I pull-request and discuss vis comments on GitHub?
3. Should I pull-request and discuss here?
4. Whatever as long as
21 matches
Mail list logo
