CSS tables - more confused

[dorayme]

I argued against your suggestion to habitually record or scan in much
bigger dimensions than what is expected to be needed for the end image,
thus a downsize scenario.

If you know what you are going to do and have good experience in these
matters, fine. I believe my advice is the safest for those who are less sure
of themselves and probably good advice for those who are too sure of
themselves. Now don't go and take this the wrong way, personally, please. I
emphasise different things to you and so far have no reason not to.

I took a look at the site you mentioned in another post -
http://nickyguides.digital-digest.com/bilinear-vs-bicubic.htm

The reason that *both* versions of the reduction in fact look *not very good
at all* may be because the source was inadequate. It may be a deliberate
thing by the author to make a point? But why not a good sharp pic that is
better in one than in the other to make his point? I think because of the
following reason which in no way helps your case: the effect of the arguably
poorer bilinear algorithm is to fuzz things up a bit (reminding me of the
effect PS blur and Guassian filters. Used precisely to improve images with
faults in them. It involves a loss in sharpness).

The Paint Shop pro spiel might be a consumer safe thing they say to do for
the average punter. But I say start with good pics if you can and your rule
should be to preserve sharpness, bicubic is the way to go. Your argument
about starting with what you are likely to need is a risky strategy. Better
to have the bicubic as rule of thumb if you are in the business of dealing
with good pics and adopt other strategies for the minority of other cases.

Your :

This experiment demonstrates that the rule of thumb is correct:
http://nickyguides.digital-digest.com/bilinear-vs-bicubic.htm

is not correct. It does not *demonstrate* this at all. You may still be
right but I am now seriously doubting it. (And, I am sorry to have to add
this, I am not doing this to be combative, I am doing it because my
reasoning about the complex background assumptions and my own eyes and
practice dictate it)


dorayme

 

[Jim Scott]

This should not be a problem that results in such marks. A good quality scan
from a decent pic of that size will result in a good pic that would more
than fill most screens (because the pixels on a screen are very much less
than the equivalent dots bunched in a digital print of the same file. Plenty
enough material to work with). Try not to use such commands as "save for the
web" if you can help it - because you mostly don't know what is happening
and therefore lose control and experience. Scan at a high resolution. Open
in PS. In one or other order (I am awaiting feedback on a slight question on
this), reduce the pic to a size you roughly want and compress choosing a
number from 1 to 10 in JPG levels. Try 5 or 6 to begin. 4 might be ok. I
suspect the faults I mentioned are due too much compression (lower numbers
like 0,1,2). You can keep compressing a bit (don't go the other way though)
and you can adjust the px dimensions down (again, don't go the other way)
till you get a result you want. You will get experienced and not have to
fiddle too much after a while...

Good luck...

dorayme

Thank you
Before you disappear can you recommend a scan resolution for my 6 x 4 film
prints?

 

[Oli Filth]

<SNIP>
I can't do *best* visual tests on the screens in front of me right now but
will later on a very big brand new one I own on another machine. The *quick*
test I have just done on a good sharp regular photo taken with a digital
camera of an outdoor scene results in no difference I can detect but more
importantly for now: no difference in file size. To clarify:

I jpgd at level 5, I resized and then saved as jpg at the same level as my
first jpging. I got a file of 64,667K. I then went back to the original and
resized first, jpgd at level 5 and saved. Again 64,667K. So if information
loss is reflected in file size, this "third loss" is not happening? But I am
not saying here that you have no case. I will look into it further.

File-size is a rough indicator of information content, but it doesn't
map precisely to JPEG quality. I just did a test on a photo-realistic
image (original image 1132x384, resized to 20%, JPEGged at "5"), and the
Compressed->Resized image file-size was actually *bigger* than the
Resized->Compressed image file-size, even though the total error was far
higher (when viewed via the difference method I described earlier), and
therefore had lost more image information.

Will think more on this. In what you say may be the answer to my puzzle
about file size. I did suppose that a gradient was an excellent model for
most photographic pics. But maybe I am wrong. I was thinking it is just the
thing gifs are bad at, they are bad at photos mostly too and I was supposing
it was the feature of gradual tonal changes that are at the heart of so much
photography that was being captured by my gradient

Actually, constant gradients *are* something JPEGs handle very well,
almost perfectly, in fact. This, strange as it seems, is a reason not to
use them as a test. If you are looking to observe and measure loss, then
it's best to test it with something that JPEG doesn't handle so well, so
that there's more of an error signal to observe. A photo with a lot of
sharp edges would be a good example.

If you're interested in the technical reason, it's because (broadly
speaking) JPEG compression works by paying less attention to
high-frequency parts of the image. Edges have lots of high-frequency
information, which is consequently removed during compression. So on
reproduction, the edge looks blurred (typically you'll see "ringing" on
highly-compressed images with lines). A smooth gradient has little
high-frequency content, and hence gets through the compression
relatively unscathed.

 

[dorayme]

Thank you
Before you disappear can you recommend a scan resolution for my 6 x 4 film
prints?

I mostly scan just such sizes (and four at a time on a decent sized flatbed)
at 300 dpi (Sometimes at 600 but I better not provoke the regulars on this
newsgroup too much!) You will be on the right side of things doing this. Set
your scanner software to produce largely uncompressed results like tiffs...
The rule is this, unless you *really* trust scanner-packaged software, use
it only so far as you have to. Use a good quality editor to do the rest. I
prefer to do as much editing as I need on the full file before reducing size
and jpging. Get into the habit of "saving as" to preserve your original
"raw" digital scan as fall back source. Fix faults on the raw file, (you can
significantly reduce file size just by doing this - faults like scratches,
spots, hazes, etc etc are just more bytes! And do not use level 0 or 1 (or
2?) when you come to jpg. Get file size down, but trade height width size
instead of quality. 3 can be ok. But try higher, you may be surprised.

dorayme

 

[dorayme]

Actually, constant gradients *are* something JPEGs handle very well,
almost perfectly, in fact. This, strange as it seems, is a reason not to
use them as a test. If you are looking to observe and measure loss, then
it's best to test it with something that JPEG doesn't handle so well, so
that there's more of an error signal to observe. A photo with a lot of
sharp edges would be a good example.

If you're interested in the technical reason, it's because (broadly
speaking) JPEG compression works by paying less attention to
high-frequency parts of the image. Edges have lots of high-frequency
information, which is consequently removed during compression. So on
reproduction, the edge looks blurred (typically you'll see "ringing" on
highly-compressed images with lines). A smooth gradient has little
high-frequency content, and hence gets through the compression
relatively unscathed.

Yes, I am interested and this makes a lot of sense, and it is well put. I
will now make a better model with lots of sharp edges and also make some
tests on some 80 MB photos (meant for high end printing).

dorayme

 

[Luigi Donatello Asero]

I mostly scan just such sizes (and four at a time on a decent sized flatbed)
at 300 dpi (Sometimes at 600 but I better not provoke the regulars on this
newsgroup too much!)

What about 1200 dpi?

 

[dorayme]

"dorayme" <[email protected]>



What about 1200 dpi?

Well, this may be a bit of overkill for chemist processed 6X4 prints from
film stock. Slow scanning and processing and hugely unnecessary files.

300 should be fine... I am just wary of trying to be mingy and saying, ah, I
want only 72 dpi so I will scan at this or even 100.


dorayme

 

[dorayme]

can assure you there's a difference that one can easily measure:

* Assuming JPEG1 is the Resized->Compressed image, and JPEG2 is the
Compressed->Resized image.

* Load JPEG into Photoshop

* Copy one and paste it as a 2nd layer on top of the other.

* Right-click the top layer in the Layers panel and select "Blending
Options..." for the top layer, then select "Difference" in the "Blend
Mode" drop-down box (i.e. it subtracts it from the bottom layer).

* Select Layer->Flatten Image to create a single layer.

* Select Image->Adjustments->Curves... to get the contrast adjustment graph.

* Bring the low end right up (set the point to Input=4, Output=255). You
*will* see a difference then.


Obviously, this doesn't prove which is superior. To do this, perform the
difference comparison test above twice, firstly between the resized BMP
and JPEG1, and then resized BMP and JPEG2. The result for the second
will be much brighter, proving that this picture is definitively more
inaccurate and therefore has lost more information.

Check out http://olifilth.co.uk/compression/ if you can't be bothered to
do all this yourself! (I assure you that JPEG3 and JPEG4 were created
with the same contrast enhancement settings.)

This thread got a bit complicated, if I recall. Anyway, there was a question
about the difference between jpging a pic and then resizing down or vice
versa. I had not seen any difference (I had thought that the former might
have been superior a method - I now realise that I was influenced less by my
own actual practice (which is mixed on this score), than a related practice,
namely to do a fair bit of retouching and other work on the big file before
resizing down). Oli has kindly referred to a few tests to see the difference
(in favour of resize first). Some comments:

The pics in the url referred to were too small for me to judge the matter.
And, frankly, I get a bit confused by the fact that the jpgs are of a
different compression ratio. In the case under discussion we are talking one
fixed compression (and I prefer 0 for this test as it is the most extreme
and should bring things out into the open).

As for the reasoning above in the recipe of steps, this would be interesting
if it was an indicator of what can be seen in a likely real case. And a real
case is easier to follow. So I did one. I got a tiff of 27 MB (2611x2663
pixels at 300ppi). It was a very sharp scene of a harbour setting with
chairs and tables, water craft, water, palms, flower pots, boats and ropes.
There were gradients and lines, horiz, diag, vert, edges of all sorts, low
and high frequency stuff... I jpged max at O. Then resized to 600 wide. Nice
and big still to get a good eyeful on my screen at 100%.

Conclusion: resize first is superior. One can see a higher order of
sharpness. One has to look a bit carefully (saw it first in a small pot of
flowers! But it *is* there Oli. There was no difference in file size at all
- so info is just arranged a bit differently. The term "info loss" is not as
simple as it seems, it is quality rather than just quality.

So! Anyway, I had to see for myself! Could never see it when using smaller
pics. Thanks for your stuff on this. Most interesting.

dorayme