Small size pictures utilize all sensor pixels?

[george013]

The file would be in PPI (if defined) the printing DPI.

Personally I am far more interested in concepts rather than arguing about semantics.

You don't have your best day.
The file is just pixels. The monitor has a resolution of x ppi.

George

[MrB]

The file would be in PPI (if defined) the printing DPI.

Personally I am far more interested in concepts rather than arguing about semantics.

On the assumption that dpi had been typed in error, ppi was posted as a correction, not as an argument.

Cheers
Philip

[pnodrog]

George, I am having a great day - be made better if people could read properly... .. PPi is a value often stored in the file. I even used the word "if" in the comment to make it reasonably clear that it was not a property of the file.

I admire your determination to try and find fault with postings..

P.S. A file is not just pixels it is just bits - data in a digital form... not hard to play your game.

[Manfred M]

No, the file is not printed, just resized, in the question.
And all the sensor pixels are used for the image.

For a JPEG file, this is not correct and for RAW data even less correct. A jpeg is an image file and has an R, G and B value for every pixel. Raw data for all cameras using a Bayer array (all cameras except for Sigma cameras and the Leica Monochrom series), the data is R, G, G, B and has to be turned into image data with R, G and B values.

The data has to be scaled to fit the output device requirements unless we are outputting to exactly fits the height and width of the file. This rarely happens. On a computer screen, the image data is downsampled and new data is created and displayed; some of the original data is "thrown away". This is because most computer displays only use somewhere between 2MP of data and 8MP of data; depending on the specific screen.

For a small print, the data will be downsampled as well, but for a large one, the data will up upsampled, i.e. new pixel data will be created to fill the paper. For Canon and HP printers, the native resolution is 300 ppi and for Epson it is 360 ppi, but the printer drivers effectively hide this from the end user.

[george013]

George, I am having a great day - be made better if people could read properly... .. PPi is a value often stored in the file. I even used the word "if" in the comment to make it reasonably clear that it was not a property of the file.

I admire your determination to try and find fault with postings..

P.S. A file is not just pixels it is just bits - data in a digital form... not hard to play your game.

I've seen better days of you.
Phillip mislead you with PPI, it's DPI as you stated at first, the wanted printer resolution. And that's a request to the printer driver to correct your image so that it looks like being printed on the wanted print resolution if printed full size.
An image is a queue of pixels, every pixel contains 3 values for the 3 colors R,G and B. And those values can be a byte or a word or just an amount of pixels as in a RAW file. The logical unit is how they're addressed: in pixels.

Back to topic. I just wondered how that magnification came in in this thread.

George

[george013]

For a JPEG file, this is not correct and for RAW data even less correct. A jpeg is an image file and has an R, G and B value for every pixel. Raw data for all cameras using a Bayer array (all cameras except for Sigma cameras and the Leica Monochrom series), the data is R, G, G, B and has to be turned into image data with R, G and B values.

The data has to be scaled to fit the output device requirements unless we are outputting to exactly fits the height and width of the file. This rarely happens. On a computer screen, the image data is downsampled and new data is created and displayed; some of the original data is "thrown away". This is because most computer displays only use somewhere between 2MP of data and 8MP of data; depending on the specific screen.

For a small print, the data will be downsampled as well, but for a large one, the data will up upsampled, i.e. new pixel data will be created to fill the paper. For Canon and HP printers, the native resolution is 300 ppi and for Epson it is 360 ppi, but the printer drivers effectively hide this from the end user.

Manfred, read the original question.

Are all sensors pixels used regardless of picture size or is only a subset of the sensor pixels used as needed to create a picture of the selected size?

My first question was what he meant. Was picture size meant in framing or content as he mentioned later, or was meant the file size. It turned out to be the latter. And that smaller file size can be archieved either by heigher compression or by resizing, or both. But in his example it's resizing with the same content. The original raster image of 3648x2736 has changed to 640x424, with the same content. A thumbnail. All sensor info is used.

George

[pnodrog]

I've seen better days of you.
Phillip mislead you with PPI, it's DPI as you stated at first, the wanted printer resolution. And that's a request to the printer driver to correct your image so that it looks like being printed on the wanted print resolution if printed full size.
An image is a queue of pixels, every pixel contains 3 values for the 3 colors R,G and B. And those values can be a byte or a word or just an amount of pixels as in a RAW file. The logical unit is how they're addressed: in pixels.

Back to topic. I just wondered how that magnification came in in this thread.

George

The word would be array not queue...

[george013]

The word would be array not queue...

No, an array does have an index, a queue not. I first wrote list but changed that afterwards in queue.

George

[Manfred M]

Manfred, read the original question.
My first question was what he meant. Was picture size meant in framing or content as he mentioned later, or was meant the file size. It turned out to be the latter. And that smaller file size can be archieved either by heigher compression or by resizing, or both. But in his example it's resizing with the same content. The original raster image of 3648x2736 has changed to 640x424, with the same content. A thumbnail. All sensor info is used.

George

George - I did read the question and agree with your comments with respect to the original question. As per my response #11, as there is one exception to this and that concerns the way still cameras record video data.

The original question did not get into compression at all, but regardless, the input data is the same in all cases. The usual issues regarding ppi and dpi have crept into the forks in this thread.

ppi - is an outcome and depends on screen pixel pitch of the screen.

dpi - is dictated by the printer manufacturer; Canon and HP printers have a native resolution of 300 dpi whereas Epson printers have a native resolution of 360 dpi.

[pnodrog]

No, an array does have an index, a queue not. I first wrote list but changed that afterwards in queue.

George

An array has a structure.

[george013]

Every variable has a structure. That's defined when creating them. But beside that we're talking about disk files. An array is a memory variable. And yes, in memory one can approach an image as a 2 dimensional array, or as a queue, or list. A matter of programming.

George

[pnodrog]

George, quaint.

"Arrays are used to implement other data structures, such as lists, heaps, hash tables, deques, queues, stacks, strings"

It would be unproductive for us to continue this thread.

[Acorn]

This topic takes me back a long, long time. I can remember during the early 1950's hearing the parents of one of my friends discussing the purchase of their first television set. Their options were confined to 9-inch and 12-inch models in those days (14-inch screens were just about to appear on the horizon). My friend's mother stipulated that they must get a 12-inch model because there would be more of the picture on the screen.

[xpatUSA]

Now whether it is averaging all the intermediate pixels (probably) or just using every 6th one to do this is largely irrelevant because; with memory not being the cost it was cameras were originally developed with this functionality, it serves no purpose in general photography to be skimping - always shoot at the maximum resolution; your "1)".

Good advice - especially if the intended output is a print of more than 8x10" or A4.

As to "averaging all the intermediate pixels", it's not quite that simple these days.

A quick look here will give us an idea:

http://entropymine.com/imageworsener/resample/

http://www.microimages.com/documenta...resampling.pdf

Enjoy.

[pnodrog]

Good advice - especially if the intended output is a print of more than 8x10" or A4.

As to "averaging all the intermediate pixels", it's not quite that simple these days.

A quick look here will give us an idea:

http://entropymine.com/imageworsener/resample/

http://www.microimages.com/documenta...resampling.pdf

Enjoy.

I was hoping to see a note or comment of the effect on the reduction of noise that is sometimes claimed is a possible benefit when downsizing. The chance of highlighting noise by picking on individual pixels rather than doing some type of averaging seems far too risky to me.

[xpatUSA]

I was hoping to see a note or comment of the effect on the reduction of noise that is sometimes claimed as a possible benefit when downsizing.

Indeed, we should also remember that there are many different downsizing methods, all of which affect "noise" differently. Also, "noise" is often mentioned with no definition of the type of noise, making the whole topic subject to misunderstanding.

Recently I was measuring the actual ISO of one of my cameras.

This needed shots outside of a gray card, using infinite focus resulting in a fairly homogeneous image - perfect for measuring image noise.

I downsized a shot three different ways.

Enter the forum Pedant: for this exercise I used SNR to compare the original and the three downsized ones. I used Signal-to-Noise Ratio (SNR) in order to even out the slight luminance variations of my selections in ImageJ.

So, image actual SNRs(not dB):

Original: 94.319
Maker's raw converter: 78.359
FastStone Lanczos3 (default): 71.704
FastStone Lanczos2 (sharper): 66.456

Since a higher SNR implies less "noise" . . the original appears to be "best".

My SNR is (mean signal value)/(standard deviation) both easily available in ImageJ.

No doubt there are many other definitions of "noise" here . . .

The chance of highlighting noise by picking on individual pixels rather than doing some type of averaging seems far too risky to me.

Quite so. Especially with the weighted types of averaging as can be seen in the links I posted earlier, e.g. bicubic, Lanczos, et al.

[katto]

I think yes,

[Ken MT]

Does this answer the question?

"JPEG or RAW? Which is right for you? The answer to this question depends on how serious you are about your photography. Before you decide, it’s good to know the differences between the two formats.
When you capture an image in the JPEG format, the camera processes the image. The camera also compresses the image based on the quality option you specify in the camera menu. You can store more images on a card when you specify a smaller image size and lower image quality. However, you’ll notice the difference when you print your images.
When you choose the RAW format, you have the ultimate in flexibility. The camera sensor transmits the RAW data to your memory card. What the sensor captures is what you get. You have to process RAW images with either the Canon software provided with your camera or with third-party software such as Apple Aperture, Adobe Photoshop Elements, Adobe Photoshop, or Adobe Photoshop Lightroom. The software lets you fine-tune virtually everything."

[xpatUSA]

Does this answer the question?

"JPEG or RAW? Which is right for you? The answer to this question depends on how serious you are about your photography. Before you decide, it’s good to know the differences between the two formats.
When you capture an image in the JPEG format, the camera processes the image. The camera also compresses the image based on the quality option you specify in the camera menu. You can store more images on a card when you specify a smaller image size and lower image quality. However, you’ll notice the difference when you print your images.
When you choose the RAW format, you have the ultimate in flexibility. The camera sensor transmits the RAW data to your memory card. What the sensor captures is what you get. You have to process RAW images with either the Canon software provided with your camera or with third-party software such as Apple Aperture, Adobe Photoshop Elements, Adobe Photoshop, or Adobe Photoshop Lightroom. The software lets you fine-tune virtually everything."

Where is this quote from, Ken? I ask because it's a bit simplistic ...

I see that only Canon gets a mention. Also, I understand that Nikon is not above messing with the raw data before it is written to the camera's memory card ...

[Ken MT]

https://www.dummies.com/photography/...-canon-eos-6d/