Updated product pages: Why not pixel density ?

I don't know what's the point in omitting this piece of technical data (and why writing it with capital initials). I believe, in economics if you'd like to know the income per head you don't quote a country's income and let them divide it by the number of inhabitants. I'd love to see this piece of information in the sensor section. I understand it might be difficult to gather in certain situations, but then you could write "unknown".
I agree!!

It was one of the most useful technical data of the entire database. I want it again!!
No it wasn't. Of what possible use is the pixel density figure?
It is an important consideration in the 'reach' you will get and IQ.
hard to see how directly you can get the 'reach' from it without the sensor size also, and if you have the sensor size, then the pixel count (or better, pitch) gives you the reach information more directly.
It compares how many pixels you get covering your subject.

EG take a bird in the center of a 1D4 and a 5D2 and a 7D at same distance and focal length. The 7D gives you more pixels on the subject.

In theory this will enable you to crop more and enlarge more, but IQ will also come into how you then use the result.
 
Yes, pixel. I don't know where sensel came from and of course I would not want to infringe on anyone's right to make up words, but why not use the word that the image sensor community uses? No one there is confused about what it means in the sensor context. And no one I know is confused when it comes to using the word pixel for sensors, image processing or displays.

Well, arguing about this is probably fruitless. But from my point of view it is like car engineers calling the chassis as chassis, and yet a small group of drivers decide among themselves to make up a word and start calling it a conveyance structure or "conture" for short. So, you can imagine the engineers standing around smacking their foreheads and shaking their heads while trying to communicate effectively with the drivers. why why why?!
 
Yes, pixel. I don't know where sensel came from and of course I would not want to infringe on anyone's right to make up words, but why not use the word that the image sensor community uses? No one there is confused about what it means in the sensor context. And no one I know is confused when it comes to using the word pixel for sensors, image processing or displays.

Well, arguing about this is probably fruitless. But from my point of view it is like car engineers calling the chassis as chassis, and yet a small group of drivers decide among themselves to make up a word and start calling it a conveyance structure or "conture" for short. So, you can imagine the engineers standing around smacking their foreheads and shaking their heads while trying to communicate effectively with the drivers. why why why?!
Of course, modern monocoque construction vehicles don't have a chassis. Some engineers call it the 'monocoque'. On the other hand, some even more modern vehicles have substantial non-structural, deformable, bolt-on body structures around the monocoque, so the monocoque has become quite similar in some ways to a 'chassis'. The 'mono' bit is something of a misnomer. Perhaps it's just a 'coque'.
--
Bob
 
Yes, pixel. I don't know where sensel came from and of course I would not want to infringe on anyone's right to make up words, but why not use the word that the image sensor community uses?
I completely understand what you're saying, and don't disagree. However, let me see if I can't make a case for the term "sensel".

A "pixel" is used to refer to both a "picture element" on the computer monitor as well as a "sensor element" on the sensor. What a "sensel" records is different than what a "pixel" displays.
No one there is confused about what it means in the sensor context. And no one I know is confused when it comes to using the word pixel for sensors, image processing or displays.
I concur. English has many words with multiple meanings, and the context tells us which meaning is appropriate. Thus, the word "pixel" is no more confusing than the word "bat".
Well, arguing about this is probably fruitless. But from my point of view it is like car engineers calling the chassis as chassis, and yet a small group of drivers decide among themselves to make up a word and start calling it a conveyance structure or "conture" for short. So, you can imagine the engineers standing around smacking their foreheads and shaking their heads while trying to communicate effectively with the drivers. why why why?!
Like I said, I don't disagree with your position. Myself, I've come under fire for "making up" terms like "total light" to refer to the total amount of light that falls on the sensor during the exposure.

Many also fight Bob and I over the definition of "exposure" -- Bob and I define it as the density of light that falls on the sensor, whereas others feel that ISO figures in to exposure (I use the made up term "apparent exposure" to figure ISO into the mix).

The term "aperture" is even more problematic. Bob feels that "aperture" should refer to the diameter of the entrance pupil (as do I), but I think that common usage of "aperture" to mean "f-ratio" makes that too confusing, so I say "aperture diameter" instead.

All this could be fixed if a qualifying adjective were used with the term "aperture" -- virtual aperture to refer to the diameter of the entrance pupil, relative aperture to refer to the f-ratio (quotient of the focal length and entrance pupil diameter), and physical aperture to refer to the physical diameter of the aperture.

Lastly, the term "noise" is usually used to refer to the NSR (noise-to-signal ratio). Note that I don't say SNR (signal-to-noise ratio) since a photo with "lots of noise" has a high NSR, and a photo with "low noise" has a low NSR. This, in my opinion, is much less confusing than the term SNR which has the exact opposite relationship.

My point is that terminology problems are rife in photography, "sensel" and "pixel" being the least of the issues.
 
I don't know what's the point in omitting this piece of technical data (and why writing it with capital initials).
I would say that they have omotted it for two reasons.
  1. Just as they say - it causes confusion - and also critique.
  2. They lost the crusade against high pixel density.
The last point I guess is also important. DPReview had a crusade against the mega pixel race. They claimed that low pixel count is beneficial for image quality. And as a weapon in that crusade they used pixel density. Unfortunately (for DPReview) and fortunately (for the users) DPReview was wrong. More pixels increase image quality. We see it al the time. Newer cameras have more pixels and better image quality.

This fact is trivially true for bigger sensors (for reasonable ISO values) - such as APS-C and above, but even (in a lesser degree) the slightly smaller FourThirds. For an APS-C camera you will increase image quality if you make it 30 MP instead of 15 MP.

The only cameras where you have to limit the number of pixels are very small sensor cameras and cameras for extreme ISO and/or extreme speeds.

There is always a trade off at some point.

And to get a feel for the current trade off point - the APS-C Sony sensor in the K-5 and D7000 cameras take very good images at ISO 6400 and can be used at ISO 50000 even though its 16 MP.

--
Roland

support http://www.openraw.org/
(Sleeping - so the need to support it is even higher)

X3F tools : http://www.proxel.se/x3f.html
 
It compares how many pixels you get covering your subject.
So do pixel count.

Assuming you dont change the size of the sensor. And if you do - you probably also use another focal length fully compensating - and then you are back at the same pixel count.

So - pixel count is more useful for your example.

--
Roland

support http://www.openraw.org/
(Sleeping - so the need to support it is even higher)

X3F tools : http://www.proxel.se/x3f.html
 
Yes, pixel. I don't know where sensel came from
I also prefer pixel, but I know where the sensel came from.

Its the Foveon X3 confusion regarding pixel counting that is the cause.

Foveon/Sigma choose to count the three color detectors as three pixels even though they are in the same spatial location.

Then to be able to sensible talk about resolving power of a Foveon chip sensel was used. The triple pixel thingie was called a sensel.

I (and others) tried to convince them that the color triple was a pixel and that the individual triples could be simply be called detector elements.

But som wanted the 3 times higher pixel count values as they were fans of Foveon/Sigma and wanted Foveon/Sigma to be able to use three times as high values in their ads.

So ... thats the cause of the horrible word sensel.

--
Roland

support http://www.openraw.org/
(Sleeping - so the need to support it is even higher)

X3F tools : http://www.proxel.se/x3f.html
 
I suspected as much, but the X3 pixel is just a 3-layer pixel in my opinion, or a pixel with 3 outputs. But, I don't want to open that can of (dead) worms again!!!

I wound up doing some work about 2 years ago for Samsung that will finally be published at the 2011 IISW in June. It involves kernels of pixels (e.g. RGBG) except with different filters (like clear and magenta) and some of the pixels in the kernel having two layers. The goal was to try to improve the low light luminance SNR without screwing up the color too much. Net result: Improvement possible but cost-benefit ratio is unfavorable.

It will be on my own website later in June, assuming the afterglow of the meeting in Japan is weak....
 
It compares how many pixels you get covering your subject.
So do pixel count.

Assuming you dont change the size of the sensor. And if you do - you probably also use another focal length fully compensating - and then you are back at the same pixel count.

So - pixel count is more useful for your example.
Please, try reading and quoting all my post before answering, I took the trouble to specify different sensor sizes and same distance and focal length.

"EG take a bird in the center of a 1D4 and a 5D2 and a 7D at same distance and focal length. The 7D gives you more pixels on the subject.

In theory this will enable you to crop more and enlarge more, but IQ will also come into how you then use the result. "

So - pixel count is not more useful for my example ;-)
 
Please, try reading and quoting all my post before answering, I took the trouble to specify different sensor sizes and same distance and focal length.

"EG take a bird in the center of a 1D4 and a 5D2 and a 7D at same distance and focal length. The 7D gives you more pixels on the subject.

In theory this will enable you to crop more and enlarge more, but IQ will also come into how you then use the result. "

So - pixel count is not more useful for my example ;-)
My bad - for Canon users the measure might be interesting. Only Canon has mostly one set of lenses and different size sensors.

For us others the measure might be interesting if you have an FF lens, an FF camera body and an APS-C camera body.

And of course only if you take images that are cropped to a certain physical size and if you are interesting in doing the kind of comparisons you are doing.

I would say its not likely that the DPReview community is all that interested overall.

Sorry.

--
Roland

support http://www.openraw.org/
(Sleeping - so the need to support it is even higher)

X3F tools : http://www.proxel.se/x3f.html
 
To avoid confusion between the terms (and concepts) of physical aperture (or virtual aperture, if you prefer) and "f number" or "f ratio", which don't sound too good, the Astrophotography crowd uses for the later the term "focal ratio", calling the former just "Aperture" (I think that for refracting telescopes this is the same that the physical diameter of the front element)
--
Antonio

http://ferrer.smugmug.com/
 
To avoid confusion between the terms (and concepts) of physical aperture (or virtual aperture, if you prefer) and "f number" or "f ratio", which don't sound too good, the Astrophotography crowd uses for the later the term "focal ratio", calling the former just "Aperture" (I think that for refracting telescopes this is the same that the physical diameter of the front element)
This confusion is due to word economics.

For telescopes the diameter is most interesting and for photography the F-number. So - both those are called something simple - like aperture.

--
Roland

support http://www.openraw.org/
(Sleeping - so the need to support it is even higher)

X3F tools : http://www.proxel.se/x3f.html
 
To avoid confusion between the terms (and concepts) of physical aperture (or virtual aperture, if you prefer)...
The virtual aperture is synonomous with the entrance pupil, which is image of the physical aperture when viewed through the front element of the lens.
...and "f number" or "f ratio", which don't sound too good, the Astrophotography crowd uses for the later the term "focal ratio"...
Well, "f-ratio" is, of course, means "focal ratio", which is the ratio of the focal length and entrance pupil diameter. I guess you could say, "focal / entrance pupil ratio". ;)
...calling the former just "Aperture"...
Yes. Both Bob (bobn2) and Lee Jay (ljfinger) are proponents of this terminology. I don't disagree with that, of course, but feel that the term "aperture" has been so bastardized to mean "f-ratio" that any hope of using the original meaning of the terms, as still used by the astro community, is all but impossible.
... (I think that for refracting telescopes this is the same that the physical diameter of the front element)
For most long lenses as well. For example, the entrance pupil diameter for the 200 / 2.8L is 200mm / 2.8 = 71.4mm, which is a good match for the 72mm FE diameter, and the entrance pupil diameter for the 300 / 4L IS is 300mm / 4 = 75mm, which, again, is a good match for the 77mm filter diameter.
 
To avoid confusion between the terms (and concepts) of physical aperture (or virtual aperture, if you prefer) and "f number" or "f ratio", which don't sound too good, the Astrophotography crowd uses for the later the term "focal ratio", calling the former just "Aperture" (I think that for refracting telescopes this is the same that the physical diameter of the front element)
This confusion is due to word economics.

For telescopes the diameter is most interesting and for photography the F-number. So - both those are called something simple - like aperture.
Actually, for photography, it's still the aperture diameter, but most people don't understand that, unfortunately:

h ttp: www.josephjamesphotography.com/equivalence/index.htm#superquick

(banned by DPR, so you know it's good!)
 
Actually, for photography, it's still the aperture diameter, but most people don't understand that, unfortunately:
Yeah ... I know. But it does not help knowing it. You may be just 100% right - but if very few agree with you ... youd better do things wrong.
When DPR has amended the 50D review to exclude the comments about its apparently excess pixel density, anything is possible. Keep on about it often enough and consistently enough, and people eventually get the message.
--
Bob
 
Actually, for photography, it's still the aperture diameter, but most people don't understand that, unfortunately:
Yeah ... I know. But it does not help knowing it. You may be just 100% right - but if very few agree with you ... youd better do things wrong.
A while ago, in the Army, when putting up camo nets, my platoon sergeant noted me arguing with others about how they were doing it. He called me aside and said, "You're right, but by the time you convince them, you could have just done it their way." I conceded on the spot.

However, here, on DPR, I have the luxury of not arguing the point in the middle of the desert at noon, so I can take the long route, and patiently wait as more and more come to understand what is going on and why. ;)
 

Keyboard shortcuts

Back
Top