Why upgrade for more MP

[Stagecoach]

That addiction has lead me to realize that a high MP camera isn't necessary.

Why would anyone upgrade when you could accomplish the same thing in a more inexpensive way?

Seems to me William that you have answered you own question or are you trying to find yourself a reason not to buy another toy

Now on a more serious note, and I have not spent time considering the full technical implications but;

How about if it was to allow you to achieve the same results in less time?


Now that's enough from me as I'm off to find a subject to convert some mega-pixels into something hopefully worth looking at

[chauncey]

How about if it was to allow you to achieve the same results in less time?

That Grahame. is a valid point.

[IzzieK]

That Grahame. is a valid point.

Life is too short at times, William...upgrade now or wait till day before you breathe your last breath and you will have less fun, less fun because you bought your bigger megapixel yet can't use it day after your last breath...and beyond. That'll show your kids your value your photographic endeavor more than leaving the money for them to spend it on some useless things they thought they know how to take care of...

[xpatUSA]

How about if it was to allow you to achieve the same results in less time?

Grahame, that almost makes me wish that I lived in the conventional world of CaNikon/PS/LR . . .

A couple of years ago, I upgraded from 3.4MP to 15MP (on that occasion, Sigma SD10 to DP2M). Was on Windoze XP 32bit, single core Pentium processor, 4GB ram, at the time and was forced to go from Sigma Photo Pro (SPP) 3.5.2 to SPP 5.5.3 because Sigma changed their raw format yet again, as is their wont.

SD10 raw files loaded in 10 sec. DP2M raw files loaded in way over a minute! Not only that but many adjustments in SPP result in all current adjustments being reapplied so that even a slight saturation adjustment caused another minute's-worth of thumb-twiddling. I sold that DP2M as fast as possible and went back to sane editing with the SD10 files.

technical note: The embedded JPEG in SD10 files is 1/4 size. All subsequent models embed a full-size JPEG. So a Merrill raw file is a whopping 50MB+ for a 15MP resolution!

Now I'm on Windows 7 with everything faster and more, so it ain't quite so bad with the recently-acquired 15MP Sigma SD1 Merrill. But it still ain't speedy and I still shoot mostly in low resolution raw (truly binned at the sensor, not sRAW) - about 3.6MP.

In other words, speed-wise, more MP on a Sigma was a big down-grade.

Occasionally, I wish nostalgically that I only had that earlier camera. Some quirks - but simplicity personified.

[chauncey]

Aside from the fiscal reasons for not upgrading Izzie, there is the ego gratification thingy...
proving to myself that I can accomplish something that I set my mind to.

[tclune]

When my D5000 died, I upgraded to the D5200 which has twice the number of pixels. I was a bit worried because I thought my lenses might not be up to the task of dealing with that many more pixels. But my lenses were fine (they're all quite good mid-priced lenses.) The real weak link was me. I had to up my game a bit to deal with the added resolution. Once I made the adjustment, though, I found the higher res genuinely useful. I can crop noticeably more aggressively with the new camera's photos. But the really big win was that the base ISO is even cleaner than the D5000 was -- which was already very good. The D5200 is actually 14-bit at 100 ISO. But I wouldn't have updated for any of that if the D5000 hadn't broken.

[george013]

Aside from the fiscal reasons for not upgrading Izzie, there is the ego gratification thingy...
proving to myself that I can accomplish something that I set my mind to.

That's what I call a professional.

George

[Tronhard]

I think there is a trap in simply upgrading to get more pixels. The pixel count is only part of the story. The more pixels your sensor had the more the processors in the camera have to work to turn them into an image. Essentially once the pixel "pots" fill and report their contents a processor had to take the millions of bits of digital data and create a range of tones and colours to generate the image. Some of the pixels in a sensor are actually used just for this process - hence the difference between total Pixels, and Effective (light-gathering) Pixels in sensor specs. So we come to the issue. If a manufacturer ups the pixel count on their sensor, they must make an appropriate upgrade to the processors that handle the upsized data. Failure to do so will result in significant lags in processing time and storage, and lost data - with resultant image degradation or loss.

There is no doubt we can be pressured by the manufacturers touting their newest technology and I know people who have sold off whole systems to switch brands because of what they perceive as a technological breakthrough. When buying equipment and choosing a brand I personally look at the glass, which doesn't change very often rather than the bodies that DO change frequently.

As mentioned before by anther contributor, the weak point is rarely the gear, it is us... The real challenge is using what we have and not hoping that technology will overcome our own shortcomings - it rarely will!

[Steaphany]

The real challenge is using what we have and not hoping that technology will overcome our own shortcomings - it rarely will!

This is precisely why I took the path of digital back to film. Digital has it's place and there are many situations where it is the only way to go, but as my photography roots are from the 1970's and most of my subjects can easily be handled by a brain-less film camera, I chose absolute basics.

Yet, even my Mamiya RB67 ProSD, produced in the 1990's and far newer than my original Yashica Mat-EM, has proven to carry a learning curve.

[george013]

.... Essentially once the pixel "pots" fill and report their contents a processor had to take the millions of bits of digital data and create a range of tones and colours to generate the image. Some of the pixels in a processor are actually used just for this process - hence the difference between total Pixels, and Effective (light gathering) Pixels in processor specs. So we come to the issue.

I don't think that's right.
A processor doesn't contain pixels, it's the sensor that contains pixels. And all those sensor pixels, or sensels, are light gathering. But not all of them result in an image pixel. By example the borders, they don't have surrounding sensels.

http://cpn.canon-europe.com/content/...oto_sensors.do

George

[chauncey]

When i see your superb shots I always have this nagging suspicion you could get as good a result with a cell phone camera.

Another reason offered from I'm falling deeply and madly in love ....

[Tronhard]

Hi George:

You are absolutely right! I was obviously having a brain fade... Of course processors don't have pixels they deal with data bits. The sentence should have read: " Some of the pixels in a sensor are actually used just for this process - hence the difference between total Pixels, and Effective (light-gathering) Pixels in sensor specs".

Despite that, I hope you agree with my point!

Thank you for finding that, good to know someone reads what I wrote!

[george013]

Hi George:

You are absolutely right! I was obviously having a brain fade... Of course processors don't have pixels they deal with data bits. The sentence should have read: " Some of the pixels in a sensor are actually used just for this process - hence the difference between total Pixels, and Effective (light-gathering) Pixels in sensor specs".

Despite that, I hope you agree with my point!

Thank you for finding that, good to know someone reads what I wrote!

Change "used for this process" in "omitted in this process". The sensels on the border of the sensor don't have surrounding sensels so they are not transformed to an image pixel.

George

[Tronhard]

I think you misconstrue my meaning here.

Having said that Pixels sense the light, I am saying that the pixels that are not used directly for this purpose are used to reconcile the data thus generated. I think you are seeing the word process as being the sensing itself, which is not my intention.

http://www.sensorland.com/HowPage090.html article reprinted from Canon.UK

[DanK]

I am saying that the pixels that are not used directly for this purpose are used to reconcile the data thus generated.

I don't think that is correct. Pixels just collect light; the processor does the reconciling. As George put it, the ones on the edge can't become pixels in the image because they lack other pixels on one side. However, they collect light, and the information about that light is used by the processor to turn adjacent, non-edge pixels into image pixels. The end of the Canon article that George linked has a good explanation of this.

[Steaphany]

Lets go to a typical source on Imager Architecture, below is the layout of the On Semiconductor KAI−16050

The KAI−16050 Image Sensor is a 16−megapixel CCD in an APS−H optical format. Based on the TRUESENSE 5.5 micron Interline Transfer CCD Platform, the sensor features broad dynamic range, excellent imaging Performance, and a flexible readout architecture that enables use of 1, 2, or 4 outputs for full resolution readout up to 8 frames per second. A vertical overflow drain structure suppresses image blooming and enables electronic shuttering for precise exposure control.

Surrounding the central array of 4896H x 3264V 5.5um x 5.5um Pixels are Buffer Pixels, Dummy Pixels and finally Dark Reference Pixels. The data sheet describes the Dark Reference as:

There are 22 dark reference rows at the top and 22 dark rows at the bottom of the image sensor. The dark rows are not entirely dark and so should not be used for a dark reference level.

Use the 22 dark columns on the left or right side of the image sensor as a dark reference. Under normal circumstances use only the center 20 columns of the 22 column dark reference due to potential light leakage.

and for the Dummy Pixels:

Within each horizontal shift register there are 11 leading additional shift phases. These pixels are designated as dummy pixels and should not be used to determine a dark reference level.

In addition, there is one dummy row of pixels at the top and bottom of the image.

Lastly, the Active Buffer Pixels:

12 unshielded pixels adjacent to any leading or trailing dark reference regions are classified as active buffer pixels. These pixels are light sensitive but are not tested for defects and non−uniformities.

Yes, these are real pixels, they are just used to set a dark reference level or simply ignored.

[Tronhard]

Without getting into too much technology! To quote Steaphany from her very detailed response: ", these are real pixels, they are just used to set a dark reference level or simply ignored."

I reference my comments from Canon.UK's article to which I provided a link:

"The actual number of pixels on a sensor is the total number of pixels. However, not all of these are used in forming the image. Those at the edge are ignored by the camera in forming the image, but their data is used by those further from the edge. This means that every pixel used in forming the image uses the same number of pixels to create its colour data."

However, we are getting away from my point:

Just putting a bigger sensor in a camera doesn't necessarily make it produce better pictures without a corresponding upgrade to the Digital Signal Processors. It may also require an upgrade in the system bus, cache and the data storage device.

Why don't we just debate on that one please?

[Stagecoach]

However, we are getting away from my point:

Just putting a bigger sensor in a camera doesn't necessarily make it produce better pictures without a corresponding upgrade to the Digital Signal Processors. It may also require an upgrade in the system bus, cache and the data storage device.

Why don't we just debate on that one please?

I wonder Trev if there is actually anything to debate here? Would we not expect the manufacturer to automatically upgrade these peripherals in the same way as a car manufacturer fits upgraded suspension, brakes e.t.c to his 'biggest' engined model?

[Manfred M]

Without getting into too much technology! To quote Steaphany from her very detailed response: ", these are real pixels, they are just used to set a dark reference level or simply ignored."

However, we are getting away from my point:

Just putting a bigger sensor in a camera doesn't necessarily make it produce better pictures without a corresponding upgrade to the Digital Signal Processors. It may also require an upgrade in the system bus, cache and the data storage device.

Why don't we just debate on that one please?

Internal to the camera, for sure. External to process the images, not really. I use the same computer to process my 36MP data files from the D800 (36MP) that I used on the D90 (12MP). Image processing is not all that memory / processor intensive, even when I do non-destructive work in Photoshop and have lots of layers in an image file.

Storage needs - well perhaps, but that would have grown anyhow but perhaps not as fast. My wife and I took some 15,000 images last year when we did our two-month tour of South Asia and that is my single largest "collection" of image files. The only time I needed to upgrade my processing capacity was for HD video editing. Now that is really resource hungry.

[xpatUSA]

Without getting into too much technology! To quote Steaphany from her very detailed response: ", these are real pixels, they are just used to set a dark reference level or simply ignored."However, we are getting away from my point:

Just putting a bigger sensor in a camera doesn't necessarily make it produce better pictures without a corresponding upgrade to the Digital Signal Processors. It may also require an upgrade in the system bus, cache and the data storage device.

Why don't we just debate on that one please?

With all due respect, Trev, how can we discuss it without getting into technology to the necessary depth?

And, since it is written in "SPV" (Standard Photographic Vague), how can we discuss it at all?

As always, please pardon my pedantry and no offense intended.