Camera quality of output summary scores for the top 50 as rated by individually detailed DXO Mark Mobile Reviews

On Fri, 04 Oct 2019 09:29:15 -0400, nospam
wrote:

In article , Eric Stevens
wrote:

Either you are way off the rails or you are trying to misdirect.
Sampling theory has *nothing* to do with the digitizing of an analog
signal.

That is the definition of sampling theorem.

https://en.wikipedia.org/wiki/Nyquist?Shannon_sampling_theorem

"In the field of digital signal processing, the sampling theorem is a
fundamental bridge between continuous-time signals and discrete-time
signals. It establishes a sufficient condition for a sample rate that
permits a discrete sequence of samples to capture all the information
from a continuous-time signal of finite bandwidth."

Are you familiar with what digital signal processing is? Hint, it's
about digitizing an analog signal and analoging a digital sample of an
analog signal.

You also might benefit from learning what continuum-time and discrete-
time mean in this field.

Yes I do know about all this.

Obviously not, a you said "Sampling theory has *nothing* to do with the
digitizing of an analog signal" which is spectacularly wrong and shows a
complete ignorance of the topic. How complete? My dog knows exactly as
much about it as you do; which is to say you and he both know nothing at
all.

The difference, of course, is my dog is only spreading his **** in the
yard, not all over USENET, so score one for the dog.

The point is that in the DSLRs under discussion there is no continuous
time signal. The sensor is exposed and for a fixed time and then the
charge of each sensel is digitized. There is no sampling as such.

Absurdly, spectacularly wrong.

For months nospam has been telling me whee I am wrong without ever
telling me the nature of my error. Could you do me a favour and tell
me the nature of my mistake?

it wasn't just me. *several* people told you the nature of your many
*mistakes* (plural) the last time or two this topic came up, including
myself, alan browne, ron and i think one or two others.

i'd link the thread but google broke searching old posts.

That's OK. My allegedly out of date 'broken' news reader can track
them down if it really becomes necessary.

--


Eric Stevens

There are two classes of people. Those who divide people into
two classes and those who don't. I belong to the second class.

rOn Fri, 4 Oct 2019 02:57:09 -0700 (PDT), Whisky-dave
wrote:

On Friday, 4 October 2019 04:52:53 UTC+1, Eric Stevens wrote:
On Thu, 3 Oct 2019 02:20:10 -0700 (PDT), Whisky-dave
wrote:

On Thursday, 3 October 2019 07:19:12 UTC+1, Arlen Holder wrote:
On Wed, 2 Oct 2019 07:10:08 -0000 (UTC), Lewis wrote:

Their test as bull****, their scores are worthless, and they take money
to pump scores.

They are worthless.

While Eric Stevens takes one approach to ask Lewis to responsd with facts
o i.e., "how do you view them?"

But NOT relivant facts.

Of course it's relevant!

We are back to this matter of how little you know about colour
management.

Nothing to do with colour managment.
It's back to you understand almost nothing about technology.

So when he describes images as having a "****ty colour" the problem
will not be (for example) that he is watching them on a 25 year old
CRT television?



The point is that the matters he is complaining of may
very well be artifacts of the manner of his viewing.

or the way you're veiwing thiungs, remmebr you recalibrate whether it's needed or not.

That aspect has
to be cleeared up before accepting his comments.

Of course but it seemed clear to everyone but you.



I use another approach which instantly shows apologists are immune to fact.

Which approach is that then ?


The Apple apologists (see list below) love to deprecate reliable reviews.

What makes them reliable is your view,.

come on I'm waiting....

We have consumer testing in the UK too.

One such is called Which
https://www.which.co.uk/
Those testing and working there are paid from the subscriptions they receive.


o Yet, none of them can name even a single reliable site - that is better!

Why should there be one that is better, and can you tell me what the word better means in this instance.

What does better mean.
But I;d say our which reports are better, they are more honest in testing, but I don't always agree with their money for value points so I've usually ignored that score/rating.
I was watching a consumer programme last night where 3 cordless vacuum cleaners were tested.
I also know that here one memeber of staffs child is 'paid' for reviewing items on amazon.
On facebook you can pay to have things liked.
And I understand that liked isnlt really a review, but some can't tell the differnce.


Not a single one.

can you tell me what the best wine is or the best car.


Hence, while Eric is repeatedly and logically asking Lewis to explain:
o "How do you view them?"

Eric isn't very logical.

and what is meant by o "How do you view them?".

Duh!

Yes duh!.

But why won't you answer any of my points.

--


Eric Stevens

There are two classes of people. Those who divide people into
two classes and those who don't. I belong to the second class.

In article , Eric Stevens
wrote:


It is clear that in the case of the Nikon D70 the data coding has no
direct connection at all with the dynamic range of the sensor. If
Nikon has done this with the D70 it is more likely that they have
used
a similar trick in other cameras, including the ones that have
atracted your ire for their claimed dynamic range. It is much more
likely that you have continued to misunderstand the actual situation
than that DxO takes bribes to report impossible results.

they do that for lossy compressed raw.

they do *not* do that for uncompressed or lossless compressed.

you're *way* over your head on this stuff.

They can do whatever they like between the sensor and the encoded
data.

they can, but they are still limited by the adc.

The output of the ADC is limited by the ADC.

that makes no sense.

Then think about it.

i have, and a lot more than you.

You can feed a much wider dynamic range into the
ADC and the ADC will chop it down to its own limits.

that's what i've been saying all along.

at least you finally figured it out.

But it doesn't
affect the dynamic range of the input device.

nobody said it did.

In article , Eric Stevens
wrote:

The point is that in the DSLRs under discussion there is no continuous
time signal. The sensor is exposed and for a fixed time and then the
charge of each sensel is digitized. There is no sampling as such.

Absurdly, spectacularly wrong.

technically true, as it's not sampling in the time domain.

however, it shows a very major lack of understanding.

Are you prepared to explain what it is?

the last time this came up, *several* people tried to explain it to you
and you kept on arguing, which is what's happening again.

Because you haven't explained. You didn't then and you haven't now.
Asserting is NOT explaining.

wrong. it was explained in detail by myself and others, with numerous
references.

the fact that you think a camera sensor is sampled in the time domain
is proof you do *not* understand what's going on, despite you thinking
you do.

Factual summary scores for detailed DXO Mark Mobile Reviews
(as of 10/6/2019)
https://www.dxomark.com/category/mobile-reviews/

If you wish to deprecate these summary scores of detailed tests
o Simply name a _better_ fully comprehensive smartphone QOR test site

Top 10 smartphones in extremely detailed camera quality of results scores:
01 Huawei Mate 30 Pro (121)
02 Samsung Galaxy Note 10+ 5G (117)
03 Huawei P30 Pro (116)
04 Samsung Galaxy S10 5G (116)
05 OnePlus 7 Pro (114)
06 Honor 20 Pro (113)
07 Huawei Mate 20 Pro (112)
08 Xiaomi Mi 9 (110)
09 Huawei P20 Pro (109)
10 Samsung Galaxy S10+ (109)

All the rest:
11 Apple iPhone XS Max (106)
12 Asus ZenFone 6 (104)
13 HTC U12+ (103)
14 Samsung Galaxy Note 9 (103)
15 Xiaomi Mi MIX 3 (103)
16 Google Pixel 3 (102)
17 Apple iPhone XR (101)
18 Google Pixel 3a (100)
19 LG G8 ThinQ (99)
20 Samsung Galaxy S9+ (99)
21 Xiaomi Mi 8 (99)
22 Google Pixel 2 (98)
23 OnePlus 6T (98)
24 Apple iPhone X (97)
25 Huawei Mate 10 Pro (97)
26 Lenovo Z6 Pro (97)
07 OnePlus 6 (96)
28 Apple iPhone 8 Plus (94)
29 LG V40 ThinQ (94)
30 Samsung Galaxy Note 8 (94)
31 Sony Xperia 1 (94)
32 Xiaomi Pocophone F1 (91)
33 Asus ZenFone 5 (90)
34 General Mobile GM9 Pro (90)
35 Google Pixel (90)
36 HTC U11 (90)
37 Vivo X20 Plus (90)
38 Xiaomi Mi Note 3 (90)
39 Samsung Galaxy S7 Edge (89)
40 Apple iPhone 7 Plus (88)
41 Samsung Galaxy A9 (86)
42 Crosscall Trekker-X4 (85)
43 Nokia 9 PureView (85)
44 LG G7 ThinQ (83)
45 Samsung Galaxy A50 (83)
46 LG V30 (82)
47 Motorola Moto Z2 Force (82)
48 Samsung Galaxy S6 Edge (82)
49 Motorola Moto G7 Plus (80)
50 Apple iPhone 6 (73)
51 Google Nexus 6P (73)
52 Meizu Pro 7 Plus (71)
53 Lava Z25 (70)
54 Samsung Galaxy S5 (70)
55 Motorola Moto G5S (69)
56 Apple iPhone 5s (68)
57 Nokia 8 (68)
58 Samsung Galaxy J2 Pro (2018) (65)

On Sun, 06 Oct 2019 08:26:48 -0400, nospam
wrote:

In article , Eric Stevens
wrote:


It is clear that in the case of the Nikon D70 the data coding has no
direct connection at all with the dynamic range of the sensor. If
Nikon has done this with the D70 it is more likely that they have
used
a similar trick in other cameras, including the ones that have
atracted your ire for their claimed dynamic range. It is much more
likely that you have continued to misunderstand the actual situation
than that DxO takes bribes to report impossible results.

they do that for lossy compressed raw.

they do *not* do that for uncompressed or lossless compressed.

you're *way* over your head on this stuff.

They can do whatever they like between the sensor and the encoded
data.

they can, but they are still limited by the adc.

The output of the ADC is limited by the ADC.

that makes no sense.

Then think about it.

i have, and a lot more than you.

You can feed a much wider dynamic range into the
ADC and the ADC will chop it down to its own limits.

that's what i've been saying all along.

at least you finally figured it out.

But you also long ago denied that it was possible to digitally code an
excessively high dynamic range by scaling it down to suit the ADC?
Aftre all, the ADC is not scaling an actual photon count but only an
analog proxy for the photon count.

But it doesn't
affect the dynamic range of the input device.

nobody said it did.

Then why do you think DxOMark are lieing when they attribute dynamic
ranges of 14+ to some cameras?

--


Eric Stevens

There are two classes of people. Those who divide people into
two classes and those who don't. I belong to the second class.

On Sun, 6 Oct 2019 21:07:34 -0000 (UTC), Lewis
wrote:

In message nospam wrote:
In article , Eric Stevens
wrote:


It is clear that in the case of the Nikon D70 the data coding has no
direct connection at all with the dynamic range of the sensor. If
Nikon has done this with the D70 it is more likely that they have
used
a similar trick in other cameras, including the ones that have
atracted your ire for their claimed dynamic range. It is much more
likely that you have continued to misunderstand the actual situation
than that DxO takes bribes to report impossible results.

they do that for lossy compressed raw.

they do *not* do that for uncompressed or lossless compressed.

you're *way* over your head on this stuff.

They can do whatever they like between the sensor and the encoded
data.

they can, but they are still limited by the adc.

The output of the ADC is limited by the ADC.

that makes no sense.

Then think about it.

i have, and a lot more than you.

You can feed a much wider dynamic range into the
ADC and the ADC will chop it down to its own limits.

that's what i've been saying all along.

at least you finally figured it out.

But it doesn't
affect the dynamic range of the input device.

nobody said it did.

The size of the input is irrelevant once it gets through the ADC,
something he still cannot seem to understand.

I fully understand that. I in effect made that point many months ago
(and I have just made it again in another post) when I said the
dynamic range of the sensor can be sacled to suit that of the ADC.
nospam denied it. I also said at that time that the output of the
sensor can be characterised in some way to suit the range of the ADC.
nospam denied that also. But now it seems that Nikon does both those
things.

--


Eric Stevens

There are two classes of people. Those who divide people into
two classes and those who don't. I belong to the second class.

On Sun, 06 Oct 2019 08:26:49 -0400, nospam
wrote:

In article , Eric Stevens
wrote:

The point is that in the DSLRs under discussion there is no continuous
time signal. The sensor is exposed and for a fixed time and then the
charge of each sensel is digitized. There is no sampling as such.

Absurdly, spectacularly wrong.

technically true, as it's not sampling in the time domain.

however, it shows a very major lack of understanding.

Are you prepared to explain what it is?

the last time this came up, *several* people tried to explain it to you
and you kept on arguing, which is what's happening again.

Because you haven't explained. You didn't then and you haven't now.
Asserting is NOT explaining.

wrong. it was explained in detail by myself and others, with numerous
references.

the fact that you think a camera sensor is sampled in the time domain
is proof you do *not* understand what's going on, despite you thinking
you do.

You were the one who introduced sampling. The opening paragraph of the
message to which I am replying is a quote from me where I explain why
such a thing is not happening.

--


Eric Stevens

There are two classes of people. Those who divide people into
two classes and those who don't. I belong to the second class.

In article , Eric Stevens
wrote:


You can feed a much wider dynamic range into the
ADC and the ADC will chop it down to its own limits.

that's what i've been saying all along.

at least you finally figured it out.

But you also long ago denied that it was possible to digitally code an
excessively high dynamic range by scaling it down to suit the ADC?
Aftre all, the ADC is not scaling an actual photon count but only an
analog proxy for the photon count.

nope.

of course it's possible to compress it before the adc, or many other
things, however, that's not done in any of the cameras tested and
unlikely to be done in any future camera.

But it doesn't
affect the dynamic range of the input device.

nobody said it did.

Then why do you think DxOMark are lieing when they attribute dynamic
ranges of 14+ to some cameras?

because it's *impossible* to get more than 14 stops when the adc is 14
bit (or less).

In article , Eric Stevens
wrote:


You can feed a much wider dynamic range into the
ADC and the ADC will chop it down to its own limits.

that's what i've been saying all along.

at least you finally figured it out.

But it doesn't
affect the dynamic range of the input device.

nobody said it did.

The size of the input is irrelevant once it gets through the ADC,
something he still cannot seem to understand.

I fully understand that.

no, you very definitely do not understand that, nor do you understand
quite a bit more.

I in effect made that point many months ago

no you didn't.

(and I have just made it again in another post) when I said the
dynamic range of the sensor can be sacled to suit that of the ADC.

it can be, but it isn't.

nospam denied it.

nope.

I also said at that time that the output of the
sensor can be characterised in some way to suit the range of the ADC.

it can, but it's not done that way.

nospam denied that also. But now it seems that Nikon does both those
things.

they do not.