Mailinglist Archive: opensuse (908 mails)

< Previous Next >
Re: [opensuse] apache 2.4 performance issue / processwire.
On 06/09/2016 05:43 AM, Carlos E. R. wrote:


It depends on what you need it for. If you just like the background
image and want to use it on your desktop, copying the original
(Ctrl-I -> Media -> Save as) is the better option, but if you want to
document what it looked like on your screen, well ....

Notice that for me a screenshot is a screencapture, it is something for
which the original is the computer display, for whatever reason (think
of a page explaining how to use certain software). And as such, it has a
fixed number of pixels and colours. It can not be improved, it is what
it is (unless you switch the computer video hardware or choose another
possible resolution).

Yes; I said in an earlier email that I can see a screenshot of the
complete screen with menubars and framing. Certainly if you are
documenting a product or a bug this makes sense.

But if I'm capturing an image, say of a witty cat picture or something
amusing, its the image I'm interested in capturing, and it may well be
that the page its displayed on has down-rendered it, so loosing quality
from the original. I go to the original, not the screen image.


Then. Given a given picture, whatever source, it can be converted to
jpeg. Jpeg is a lossy format. It compresses at the cost of worse
quality. There is a compromise, you can improve the quality, or improve
the size, not both.

Yes, and if the picture you start with is the original rather than the
screen capture, you may well have a higher resolution, higher quality
object to start with, which gives you more options to choose from when
you start playing around with it.

I have the same option with my camera. I choose to shoot in RAW format.
If I shoot in jpeg mode the camera is already degrading the raw
information, probably to one of a number of "scene" settings.
If I have the RAW then I can, in post-processing with Darktable (or
Photoshop if I were on Windows) convert that to any or all of the
'scenes', crop, and otherwise manipulate the image in many ways the
camera is not able to. And still have the high quality original
The RAW original has information that the derivatives does not.

What's more, I'm not stuck with jpeg. I can produce a 16-bit TIFF or a
16-bit JPEG2000 from that RAW. More information that in the jpeg the
camera produced. And I can do that in any one of a number of color spaces.

Yes, Darktable _could_ start with a jpeg or gif, but it cannot recreate
the information that was lost when the image was downgraded.

My nominal 16M P&S camera produces a 4608 x 3456 pixel sized RAW that
contains the 12-bits for each of the 3 colours plus phase information.
Some cameras do 14 or 16 bits, some have even larger sensors, 20M or 24M.


Not only that, my RAW image contains a lot of metadata; not just the
exposure settings, was flash used, shutter/aperture, date and time, but
also the geolocation - my camera has GPS. In fact it has SmartGPS and
can also tell me the name of the location and often what I was
photographing (e.g "Stonehenge" or "Parthenon". If it can't, then
Darktable can make use of Googlemaps of FreeMaps for the same, based on
the Lat/Long.

http://photographyconcentrate.com/10-reasons-why-you-should-be-shooting-raw/


Thus you can do several conversions at different
quality factors (from 1 to 100) and compare visually the results, and
choose which one you like.

Yes, and the example above with shooting RAW and using Darktable I have
all that and much more that I can do ...

... BECAUSE I'M STARTING WITH MORE INFORMATION !

Which is the whole point I'm trying to make here.

Its one thing to have screenshots to document a program, but if you're
saving images from the web as images, then go for quality.


If you still don't understand what I mean, try it.

I do understand. I've been using GIMP and other image processing
software for a very long time.

No other application does this with jpegs.

That is not the case.
There are a number of photo-processing tools that can.


(I hate that cameras compress to jpeg. It should be png. At least a
choice. No, I do not know if Q=100 means no losses)

Indeed, and it would be nice if we could flash the software of older
cameras to do this. And while we're about it lets have them all use the
standardized RAW format from Adobe - DNG.

.... and then we wake up.


Try to compress a screenshot with text to jpeg. As you move the slider
to smaller size you will see the fonts to degrade a lot, get grainy,
undefined. Often it is not worth it and you have to select png or gif.

The issue here is loss of information.
The 'text' isn't text, its just an image, and looses resolution just
like everything else.
This is not HTML, NAPLPS or SVG where the text is a text object distinct
from the image objects or bitmap objects. As long as its a bitmap its
going to degrade in resolution whatever you do because its bits and no
its your brain that's seeing it as text, just as you brain sees the
image is a cat and not a house. The thing is that a degraded image of a
cat (or a house) still looks, yo your brain, like a cat (or a house).
When the text looks like it was carved in stone and then subject to
4,000 years of being exposed to the weather in the Orkneys, then you'll
have a hard time reading it.

The issue here, and the whole point of what I'm going on about, is
ENTROPY. Loss of information in the process you're describing.


--
A: Yes.
> Q: Are you sure?
>> A: Because it reverses the logical flow of conversation.
>>> Q: Why is top posting frowned upon?

--
To unsubscribe, e-mail: opensuse+unsubscribe@xxxxxxxxxxxx
To contact the owner, e-mail: opensuse+owner@xxxxxxxxxxxx

< Previous Next >