why is google images so useless to find a good quality version of an image?

In article .com,
Savageduck wrote:

Usually if you want a high resolution, good quality image file you are going
to have to pay for it, and we all know that ³Sobriquet" doesn¹t like to pay for anything.

the words 'like to' are extraneous.

In article , Alfred
Molon wrote:


How would Google (or any other search engine) be able to
automatically detect such a quality difference?

easily, which is then indexed along with other metadata and can be used
as parameters in a search.

On Saturday, January 2, 2021 at 3:54:54 PM UTC+1, Alfred Molon wrote:
In article a7360651-c1f7-4e95-8531-
, says...

On Saturday, January 2, 2021 at 10:23:56 AM UTC+1, Alfred Molon wrote:
In article d4e042b5-bf51-4a66-9907-
, says...
Hi.
Suppose you encounter an image online and you want to find a good
version of it.

https://i.imgur.com/XbGAsus.png

How come google images is such a useless tool when it comes to
finding a good quality version of an image?

Like for instance, if you search for this image via google images:

https://i.imgur.com/QlQtopa.jpg

Google will find countless versions of that image online, but they are all
garbage quality.

https://i.imgur.com/w1Ctu5N.png
Isn't the second search result 1920x1080?

The resolution isn't the only aspect that matters for image quality. It should
be pretty obvious if you compare the two versions I've shown in the original
thread at 100% zoomlevel. On the left side you can even make out detail
like the surface texture of the canvas and on the right side is ugly pixel soup.

https://i.imgur.com/EBKXKG1.jpg
How would Google (or any other search engine) be able to
automatically detect such a quality difference?

Well, just like Tineye does. Provide not just information about the resolution
of the image, but also the filesize of the image and allow users to sort search
results on either or both criteria.
In combination that usually gives a good indication of the quality of the image.

--
Alfred Molon

Olympus 4/3 and micro 4/3 cameras forum at
https://groups.io/g/myolympus
https://myolympus.org/ photo sharing site

On 01/01/2021 21:53, nospam wrote:
In article , Melanie van Buren
wrote:


How come google isn't able to quickly point you to such quality
versions and comes up with countless pages of useless garbage
versions instead?

Are there any useful tools that actually work to find you a *good* quality
version of an image instead of bringing you countless pages of
useless junk?

General algorithmic criteria and SEO at a guess.

it's nothing more than his inability to figure out how to properly do
an image search and then blame it on everything other than his own
ineptness.

I'm not aware that
Google or any other search engine rates images by quality either
technically or critically.

I was discussing beyond algorithms and page ranks. There is no algorithm
for quality i.e. artistic/critical quality.

I'm also wondering why such a **** topic is getting so much traffic.
Post a link discussing photographic method and *crickets*.


--
Melanie van Buren

On Saturday, January 2, 2021 at 4:51:48 PM UTC+1, sobriquet wrote:
On Saturday, January 2, 2021 at 3:54:54 PM UTC+1, Alfred Molon wrote:
In article a7360651-c1f7-4e95-8531-
, says...

On Saturday, January 2, 2021 at 10:23:56 AM UTC+1, Alfred Molon wrote:
In article d4e042b5-bf51-4a66-9907-
, says...
Hi.
Suppose you encounter an image online and you want to find a good
version of it.

https://i.imgur.com/XbGAsus.png

How come google images is such a useless tool when it comes to
finding a good quality version of an image?

Like for instance, if you search for this image via google images:

https://i.imgur.com/QlQtopa.jpg

Google will find countless versions of that image online, but they are all
garbage quality.

https://i.imgur.com/w1Ctu5N.png
Isn't the second search result 1920x1080?

The resolution isn't the only aspect that matters for image quality. It should
be pretty obvious if you compare the two versions I've shown in the original
thread at 100% zoomlevel. On the left side you can even make out detail
like the surface texture of the canvas and on the right side is ugly pixel soup.

https://i.imgur.com/EBKXKG1.jpg
How would Google (or any other search engine) be able to
automatically detect such a quality difference?
Well, just like Tineye does. Provide not just information about the resolution
of the image, but also the filesize of the image and allow users to sort search
results on either or both criteria.
In combination that usually gives a good indication of the quality of the image.

https://i.imgur.com/imSzDmk.png

--
Alfred Molon

Olympus 4/3 and micro 4/3 cameras forum at
https://groups.io/g/myolympus
https://myolympus.org/ photo sharing site

In article ,
sobriquet wrote:

How would Google (or any other search engine) be able to
automatically detect such a quality difference?

Well, just like Tineye does. Provide not just information about the resolution
of the image, but also the filesize of the image and allow users to sort
search results on either or both criteria.

google does that, which can be used in a search query.

as usual, you don't know how to use something and blame it rather than
your own ineptness.

On Saturday, January 2, 2021 at 5:09:47 PM UTC+1, nospam wrote:
In article ,
sobriquet wrote:

How would Google (or any other search engine) be able to
automatically detect such a quality difference?

Well, just like Tineye does. Provide not just information about the resolution
of the image, but also the filesize of the image and allow users to sort
search results on either or both criteria.
google does that, which can be used in a search query.

as usual, you don't know how to use something and blame it rather than
your own ineptness.

As usual you're talking out of your ass.

https://opendata.stackexchange.com/q...h-by-file-size

In article ,
sobriquet wrote:

How would Google (or any other search engine) be able to
automatically detect such a quality difference?

Well, just like Tineye does. Provide not just information about the
resolution
of the image, but also the filesize of the image and allow users to sort
search results on either or both criteria.
google does that, which can be used in a search query.

as usual, you don't know how to use something and blame it rather than
your own ineptness.

As usual you're talking out of your ass.

https://opendata.stackexchange.com/q...h-by-file-size

a link where the answer is 'i *suspect* this is not supported' ?? yep,
that sure is a convincing bit of evidence.

not only that, the answer references google's search parameters, which
includes image size as the first in the list.

now what were you saying about talking out of your ass?

On Saturday, January 2, 2021 at 9:29:47 PM UTC+1, nospam wrote:
In article ,
sobriquet wrote:

How would Google (or any other search engine) be able to
automatically detect such a quality difference?

Well, just like Tineye does. Provide not just information about the
resolution
of the image, but also the filesize of the image and allow users to sort
search results on either or both criteria.
google does that, which can be used in a search query.

as usual, you don't know how to use something and blame it rather than
your own ineptness.

As usual you're talking out of your ass.

https://opendata.stackexchange.com/q...h-by-file-size
a link where the answer is 'i *suspect* this is not supported' ?? yep,
that sure is a convincing bit of evidence.

not only that, the answer references google's search parameters, which
includes image size as the first in the list.

now what were you saying about talking out of your ass?

I've already tested it out (size:500000, which would yield images of 500 kb
or more if it actually worked), and it doesn't work. If you think otherwise,
back up your claims with some actual evidence.

But I already know you never back up your claims with evidence. All you
ever do is talk out of your ass.

On Saturday, January 2, 2021 at 11:28:14 PM UTC+1, sobriquet wrote:
On Saturday, January 2, 2021 at 9:29:47 PM UTC+1, nospam wrote:
In article ,
sobriquet wrote:

How would Google (or any other search engine) be able to
automatically detect such a quality difference?

Well, just like Tineye does. Provide not just information about the
resolution
of the image, but also the filesize of the image and allow users to sort
search results on either or both criteria.
google does that, which can be used in a search query.

as usual, you don't know how to use something and blame it rather than
your own ineptness.

As usual you're talking out of your ass.

https://opendata.stackexchange.com/q...h-by-file-size
a link where the answer is 'i *suspect* this is not supported' ?? yep,
that sure is a convincing bit of evidence.

not only that, the answer references google's search parameters, which
includes image size as the first in the list.

now what were you saying about talking out of your ass?
I've already tested it out (size:500000, which would yield images of 500 kb
or more if it actually worked), and it doesn't work. If you think otherwise,
back up your claims with some actual evidence.

But I already know you never back up your claims with evidence. All you
ever do is talk out of your ass.

Oh and image size in the search parameters only refers to the resolution of
the image. But we were talking about the *filesize* of the images, not
just the *resolution*.

Tineye actually gives you both the filesize and the resolution in the search
results:

https://i.imgur.com/EP3GE0o.png

This whole thread is about the *quality* of images in the search results
and images with a high resolution can still be very ****ty quality, and then
they would typically have a relatively small filesize.

Compare this ****ty quality image at 312 Kb (1920x1080):
https://i.imgur.com/QlQtopa.jpg

and this good quality image at 4.93 Mb (1920x1080):
https://i.imgur.com/TykYAvB.png