RX100: What Did Sony Leave Out to Put 20 MP In?

Started Aug 6, 2012 | Discussions
Dimitri Khoz
Contributing MemberPosts: 943
Like?
Re: RX100: What Did Sony Leave Out to Put 20 MP In?
In reply to nosnoop, Aug 10, 2012

nosnoop wrote:

Dimitri Khoz wrote:

You are totaly wrong.

Sorry, I am correct.

I did not say that RX100 is a bad camera,
it is a good one,
and it would be only better if it had 12Mp instead of 20.

You are the one who is wrong, please educate yourself before writing wrong info on the forum.

Google "aperture equivalence" and read up on it. It will tell you that it is the total amount of photon captured which counts, all others like pixel count is largely irrelevant.

No comments.
Google whatever you want.

If you think that Google has more credibility than scientific publications and university research... hmm...
I am here not to teach you.

But you clearly see DPReview NEX5n vs NEX7 results

showing that same sized sensor will have more noise and less DR with increased MP count.

Reply   Reply with quote   Complain
VarmintCong
Contributing MemberPosts: 529
Like?
Re: RX100: What Did Sony Leave Out to Put 20 MP In?
In reply to trale, Aug 10, 2012

trale wrote:

Dimitri Khoz wrote:

Who cares about D800, RX100 will not perform same as full frame, lol.

Showing your ignorance once again, this time for reading comprehension. No one is comparing RX100 to full frame.

Here let me break this down for you. In a way that even you might understand:
Claim: "More pixels = less DR"
Test: A 36MP camera has more DR than a 22MP camera. Same sensor size.
Conclusion: Claim is false.

A better comparison is the Nikon D4 vs D800, since they both use the latest sensor tech, and one is 36 MP the other is 16 MP, less than half the pixels. The DR range and iso are very similar (D4 is only slightly better), enough that I think you can say there's little downside to more pixels. Nikon primarily chose 16 MP for the D4 to have higher frame rate for things like sports.

Reply   Reply with quote   Complain
trale
Senior MemberPosts: 1,053
Like?
Re: RX100: What Did Sony Leave Out to Put 20 MP In?
In reply to Dimitri Khoz, Aug 10, 2012

Dimitri Khoz wrote:

No comments.

Before you write ANYTHING in this forum, you should stick to "no comment" and stay there.

Reply   Reply with quote   Complain
Dimitri Khoz
Contributing MemberPosts: 943
Like?
Re: RX100: What Did Sony Leave Out to Put 20 MP In?
In reply to trale, Aug 10, 2012

trale wrote:

Dimitri Khoz wrote:

No comments.

Before you write ANYTHING in this forum, you should stick to "no comment" and stay there.

You should not tell other people what to do or not to do.

No one asked you for advice.
And no one really cares.

Reply   Reply with quote   Complain
nosnoop
Senior MemberPosts: 1,691
Like?
Re: Sensor Size or Pixel Size?
In reply to Walt Bilofsky, Aug 10, 2012

Walt Bilofsky wrote:

Suppose we have a sensor. OK, then take four of these sensors, put them in a square, move them twice as far behind the lens and assume it can focus on them.
Same number of photons coming through the lens from the same scene, right?
With (I think) the same amount of noise per pixel - is that right?

So far the discussion on the sensor size has one assumption, that you would use similar sensor technology and an equivalent lens with it.

And you are correct that we have to consider the lens as well; and unfortunately, this adds more complexity to the discussion.

I am not sure if they can make the lens work in your example above. But assuming that's possible, if you keep the total amount of photons the same, there would be no advantage of using a larger sensor. It's like using a very slow lens on the larger sensor, and it sorts of negates the benefit. And actually, when you pixel peep, you would see more noise in the large sensor in your example, as each pixel would receive less photons. However, if you make a same size print from both sensors, the noise would be the same.

And that's the reason why you should not compare different sensor size by pixel peeping; as when you do that, you are effectively comparing the larger sensor at 4x (400%) the magnification level as the smaller sensor in your example.

So as stated before, it is the total amount of light (determined by the lens and sensor size) captured by the sensor which is the single most important factor in affecting the noise and dynamic range (and not pixel count).

Reply   Reply with quote   Complain
trale
Senior MemberPosts: 1,053
Like?
Re: RX100: What Did Sony Leave Out to Put 20 MP In?
In reply to Dimitri Khoz, Aug 10, 2012

Dimitri Khoz wrote:

No one asked you for advice.
And no one really cares.

And no one asked you to troll this forum.

Remember our agreement? (on the last thread from you that was deleted by moderators?)

You stop trolling on this forum, and I will stop trolling you.

Reply   Reply with quote   Complain
oper8nrun
Forum MemberPosts: 56
Like?
Re: Very little is lost
In reply to caterpillar, Aug 10, 2012

Sony left out: a proper instruction booklet!

Reply   Reply with quote   Complain
Dimitri Khoz
Contributing MemberPosts: 943
Like?
Re: Sensor Size or Pixel Size?
In reply to nosnoop, Aug 10, 2012

nosnoop wrote:

So as stated before, it is the total amount of light (determined by the lens and sensor size) captured by the sensor which is the single most important factor in affecting the noise and dynamic range (and not pixel count).

Major factors in limiting the maximum number electrons captured in a semiconductor image sensor are

1) Absorption length and electron densities -

It ranges from about 1 micron for blue light to over 7 microns for red light. RX100 has pixel size of 2.4 microns. The absorption length will also limit the image detail, especially toward red colors, and it also can contribute to photons several pixels away from the target pixel. Dense sensors with small areas per pixel, no matter how many MP they have, are at disadvantage.

2) Fill factor -

The trend is down with smaller pixel size because the non-sensitive portion between pixels becomes a greater part of the pixel area. Even with gapless microlenses, light will be lost at the gap due to scattering. Dense sensors with small areas per pixel will gather less light with increase in MP count.

3) Another factor that affects low light IQ is Read noise (electronics noise) -

Read noise per pixel has little dependence on pixel size. Because all sensors have finite read noise, when one adds pixels together, the total read noise increases. In low light conditions read noise will play much bigger role than photon noise and will degrade the IQ in the sensors that have smaller pixel size.

Reply   Reply with quote   Complain
VarmintCong
Contributing MemberPosts: 529
Like?
Re: Sensor Size or Pixel Size?
In reply to Dimitri Khoz, Aug 11, 2012

Dimitri Khoz wrote:

Major factors in limiting the maximum number electrons captured in a semiconductor image sensor are

You should quote a source when doing word for word quoting:
http://www.clarkvision.com/articles/digital.sensor.performance.summary/

Reply   Reply with quote   Complain
Dimitri Khoz
Contributing MemberPosts: 943
Like?
Re: Sensor Size or Pixel Size?
In reply to VarmintCong, Aug 11, 2012

VarmintCong wrote:

Dimitri Khoz wrote:

Major factors in limiting the maximum number electrons captured in a semiconductor image sensor are

You should quote a source when doing word for word quoting:
http://www.clarkvision.com/articles/digital.sensor.performance.summary/

I made a summary of major points scattered all over that and other articles,
and I mentioned the sources in one of my previous posts.

Thank you, for helping me providing the link for folks who would like to read more about the subject.

Reply   Reply with quote   Complain
trale
Senior MemberPosts: 1,053
Like?
Re: Sensor Size or Pixel Size?
In reply to Dimitri Khoz, Aug 11, 2012

Dimitri Khoz wrote:

Major factors in limiting the maximum number electrons....

You can comfort yourself to sleep with rationalizations you take from other people's work. Does not change the fact that your statement "More pixels = less DR" is false, as demonstrated clearly by two facts:

D800 > Canon 5DIII in DR
D800 > Nikon D4 in DR

A 36MP sensor has more DR than lessor MP sensors in a competitor product, or even another more expensive camera from the same company.

No amount of technical jargon you like to worship will change that fact.

Reply   Reply with quote   Complain
Dimitri Khoz
Contributing MemberPosts: 943
Like?
Re: Sensor Size or Pixel Size?
In reply to trale, Aug 11, 2012

trale wrote:

Dimitri Khoz wrote:

Major factors in limiting the maximum number electrons....

You can comfort yourself to sleep with rationalizations you take from other people's work. Does not change the fact that your statement "More pixels = less DR" is false, as demonstrated clearly by two facts:

D800 > Canon 5DIII in DR
D800 > Nikon D4 in DR

No amount of technical jargon you like to worship will change that fact.

Read that article, may be you will understand more.

D800 already has 14-bit A/D conversion and 16-bit image processing,
plus even with 36Mp it had almost 5 microns pixel pitch.

Huge DR loses are to be expected when pixel pitch decreases to less than 3 micron,
or when image RAW processing will be done in 12 bit as in RX100.

Reply   Reply with quote   Complain
nosnoop
Senior MemberPosts: 1,691
Like?
Re: Sensor Size or Pixel Size?
In reply to Dimitri Khoz, Aug 11, 2012

Dimitri Khoz wrote:

No comments.
...
I am here not to teach you.

I wish you would keep your promise...

Major factors in limiting the maximum number electrons captured in a semiconductor image sensor are

rather than copy and paste text from elsewhere without understanding what it all means to our discussion here.

We know that increasing the pixel count would increase per pixel noise. There is no argument there. The key is to compare the same output size, where larger sensor would be superior, regardless of pixel count, rather than comparing at 1:1 pixel level. This is where most people got confused and got it wrong.

There are many things which can affect the pixel noise, as discussed in that article. But many of them only make small or insignificant difference. The single most significant factor is the total amount of light (photon) captured assuming same sensor technology and sensor efficiency.

Reply   Reply with quote   Complain
DUSTY LENS
Veteran MemberPosts: 3,744
Like?
Re: RX100: What Did Sony Leave Out to Put 20 MP In?
In reply to Walt Bilofsky, Aug 11, 2012

Walt Bilofsky wrote:

What could have been improved on the RX100 if Sony had used a smaller 12 or 16 megapixel sensor with the same pixel density?

Wider or longer lens? Better image quality? Smaller body?

Every camera design is an exercise in tradeoffs. But 20 MP seems more than needed, even for an interchangeable lens camera, let alone a pocketable one.

Did Sony sell out more important features to play the megapixel marketing game?

-- hide signature --

Why so critical of the camera ? There are plenty of 16Mp cameras for you to choose from .

If you really dislike fine resolution so much why not buy a 16Mp or a 10Mp camera ?

Are you tired of cheap already ?

Dusty

Reply   Reply with quote   Complain
rpm40
Senior MemberPosts: 1,937Gear list
Like?
Re: RX100: What Did Sony Leave Out to Put 20 MP In?
In reply to DUSTY LENS, Aug 11, 2012

Right now, Sony is the leader in sensor tech. Canon makes good products, but Sony makes the best sensors available today. Since you love noise charts, see DXO's overall sensor ratings for noise:

1: Nikon D3s
2. Nikon D800E
3. Nikon D4
4. Nikon D800
5. Nikon D700

Who makes those? Hmm..

How about the top rated aps-c size sensors?

1. Nikon D5100
2. Nikon D7000
3. Pentax K5
4. Pentax K01
5. Nikon D3200
6. Sony Alpha 580
7. Sony NEX C3
8. Sony NEX 5N
9. Sony NEX 7
10. Fujifilm X100
11.Nikon D90
12. Nikon D3100
13. Nikon D5000
14. Canon EOS 7D
15. Sony NEX 3
16. Sony Alpha 560
17. Sonly SLT Alpha 55
18. Canon 60D
19. Pentax K-x
20. Canon 550D

What brand manufactures 17 of the top 20 aps sensors ever tested?

Stop trying to imply that Sony isn't in the same class as big bad Canon. It just isn't true anymore, and it hasn't been for 5 years or more.

 rpm40's gear list:rpm40's gear list
Olympus PEN E-PL1 Olympus PEN E-PM2 Olympus M.Zuiko Digital ED 14-42mm 1:3.5-5.6 Panasonic Lumix G 20mm F1.7 ASPH Olympus M.Zuiko Digital 14-42mm 1:3.5-5.6 II R +3 more
Reply   Reply with quote   Complain
Dimitri Khoz
Contributing MemberPosts: 943
Like?
Re: RX100: What Did Sony Leave Out to Put 20 MP In?
In reply to rpm40, Aug 11, 2012

rpm40 wrote:

Right now, Sony is the leader in sensor tech. Canon makes good products, but Sony makes the best sensors available today. Since you love noise charts, see DXO's overall sensor ratings for noise:

1: Nikon D3s
2. Nikon D800E
3. Nikon D4
4. Nikon D800
5. Nikon D700

Who makes those? Hmm..

How about the top rated aps-c size sensors?

1. Nikon D5100
2. Nikon D7000
3. Pentax K5
4. Pentax K01
5. Nikon D3200
6. Sony Alpha 580
7. Sony NEX C3
8. Sony NEX 5N
9. Sony NEX 7
10. Fujifilm X100
11.Nikon D90
12. Nikon D3100
13. Nikon D5000
14. Canon EOS 7D
15. Sony NEX 3
16. Sony Alpha 560
17. Sonly SLT Alpha 55
18. Canon 60D
19. Pentax K-x
20. Canon 550D

What brand manufactures 17 of the top 20 aps sensors ever tested?

Stop trying to imply that Sony isn't in the same class as big bad Canon. It just isn't true anymore, and it hasn't been for 5 years or more.

So why on earth then they can not make good products with the Sony badge on it.
Alpha SLT low light performance is horrible comparing to Canon APS-C products.
S100 has same DR as NEX7 according to the DPReview charts.

Truth is in the real world pictures.

Reply   Reply with quote   Complain
Gwen22
Contributing MemberPosts: 552
Like?
Re: RX100: What Did Sony Leave Out to Put 20 MP In?
In reply to Walt Bilofsky, Aug 11, 2012

The RX 100 is designed to compete with other high end compact and the bigger sensor give to sony the possibilities to have at the same time bigger photosite and more definition.

The photosites of the RX100 have an area 40% bigger than the one of the 10MP-1/1.7" sensor and 70% bigger than the 12MP-1/1.7" sensor, that give, in theory an advantage in terme of dynamique and ISO performance at pixel level.

More MP allowed bigger print but frankly I think that 18.8 MP as it will have allowed a 30x40cm print at 300PPI (without extrapoling data) would have been enough for most of us. It is possible also that the marketing departement think that 20MP is a good thing .......

In the other hand this big number allowed you to crop heavily and compensates the quiet short telephoto. The RX 100 is the only high end compact that give you a 10MP files at 200mm (eq 35mm)

Reply   Reply with quote   Complain
trale
Senior MemberPosts: 1,053
Like?
Re: Sensor Size or Pixel Size?
In reply to Dimitri Khoz, Aug 11, 2012

Dimitri Khoz wrote:

Read that article, may be you will understand more.

Read Aristotle. Maybe you will understand syllogistic logic.

Reply   Reply with quote   Complain
VarmintCong
Contributing MemberPosts: 529
Like?
Re: RX100: What Did Sony Leave Out to Put 20 MP In?
In reply to Dimitri Khoz, Aug 11, 2012

Dimitri Khoz wrote:

If you think that Google has more credibility than scientific publications and university research... hmm...

I work in semiconductors with state of the art stuff, (16nm). When we go to lectures by university experts in the field, they are asking us lots of questions, not the other way around. The problem with university research is they are working on very old technology they get by donations from industry, or buying cheap on the 2nd hand market. So usually their theory and understanding hasn't caught up to the latest thing in industry.

I think you guys are both right - in theory, smaller pixels should give less DR and more noise, but at the state of the art, Sony and others have worked around these issues, so that the difference is small enough that it's not a factor. The D800 sensor for example is maybe a 1 year newer than the D4 sensor - probably Sony (if that's who makes them) was able to improve DR enough in that time to make up for the smaller pixels.

Reply   Reply with quote   Complain
Grobb
Senior MemberPosts: 1,229Gear list
Like?
An RX90: is what Sony Left Out with a 14 MP!
In reply to Gwen22, Aug 11, 2012

Here is an interesting comparison done showing microns per pixel pitch. I do not know how accurate it is, http://www.dpreview.com/previews/sony-dsc-rx100
but it looks pretty close.

It would have been interesting if Sony made an RX90 with only 14 MP in the same size/type sensor (like they did with the nex) for people who do not need 20 MP’s in a P&S.

RX100 = 1.88 (20 MP)
RX90 = 2.6 (14 MP) This is what I would have liked to see Sony make

J1/V1 = 3.86 (10 MP)
S100 = 1.8 (12 MP)
S95 = 2.0 (10 MP)
Canon g1x = 4.16 (14 MP)
Canon G12 = 2.0 (10 MP)
Sony nex7 = 3.9 (24 MP)
Sony nex5 = 5.1 (14 MP) Has lower noise, better detail than the nex7
Panasonic LX5 = 2.0 (10 MP)
12mpx m4/3 = 4.24
18.1mpx APS = 4.29

Reply   Reply with quote   Complain
Keyboard shortcuts:
FForum MMy threads