7D versus Mark IV: Pixels per bird (1 Viewer)

GYRob said:

would you indeed a guy did this to show it all.
Rob.

Click to expand...

The Kingfisher shots are just a graphical representation of pixel density - it has nothing to do with effective 'reach' from different crops as the FOV with any 1.6 cropper will be exactly the same be it the 20D or the 7D.

Roy C said:

The Kingfisher shots are just a graphical representation of pixel density - it has nothing to do with effective 'reach' from different crops as the FOV with any 1.6 cropper will be exactly the same be it the 20D or the 7D.

Click to expand...

yes thats what i was showing as i think it is a real world view rather than just the maths
Having a 1.6 crop with X amount of MPs will not meen a lot to some .
Rob.

There is a point when writing about pixel, pixel density and resolution is pointless. Unless ego is important.

Judging image quality by pixels is misleading, because the defining conclusion has always to be the output device.

This brings into play other factor totally out of the photographer’s control.

You can argue until you’re blue in the face that your 7D image is sharper, cleaner than the Mk III, and it may well be so, but its meaningless on a web press printing at 10,000 impressions a hour on the equivalent of bog-roll. And that ultimately is how the image is judged.

Most people on this forum will not get into what I call high-end printing, where colour accuracy is paramount, the majority of these images are taking on larger format cameras, be it film or digital, then heavily retouched.

Life too short to get bogged down in anorak rhetoric, I would revel in the fact if I had a 7D, that I have not paid £4.5k for a MK IV, just to achieve the same image quality.

The important parameter in terms of capturing detail is linear resolution.

1D MkIV's sensor has a linear resolution of about 170pixels/mm
7D's sensor has a linear resolution of about 232pixels/mm

Ignoring the limitations due to lens resolution, diffraction effects, etc., 7D's linear resolution is about 1.36x that of the 1D MkIV. This ratio is very similar to the linear resolution improvement that is provided by a 1.4x TC.

If the 1D MkIV's sensor has about 1-stop better noise performance compared to the 7D (which is likely), you could achieve the same resolution (and noise performance) as 7D+500mm f/4L IS using a 1DMkIV + 500mm f/4L IS + 1.4x TC (especially since the 1D MkIV would AF at f/8).

I'm following this discussion with interest and I admit I'm getting confused. From my experience pixel per bird is near irrelevant since the quality of pixel varies. If all pixels were the same then compact cameras with tiny sensors and lots of megapixels (some have 15 mp+) would be the tops. Instead they are pretty poor compared to DSLRs. From my understanding more pixels usually means smaller pixels (other things being equal) and more noise therefore lots and lots of pixels in a smaller sensor is far from desirable.

I don't know much about the 7D or how it compares with the 1D Mk IV but my guess is that the 1D is streets ahead in terms of picture quality. Theoretically the extra pixels of the 7D may mean you'll be able to crop more but the extra noise (primarily from the 7Ds smaller sensor) at the level of extreme crop will mean a poorer image.

I'm not a big fan of teleconverters. Again at extreme crops the image looks less sharp. Not sure about the noise situation when using a converter.

SJC said:

From my experience pixel per bird is near irrelevant since the quality of pixel varies. If all pixels were the same then compact cameras with tiny sensors and lots of megapixels (some have 15 mp+) would be the tops. Instead they are pretty poor compared to DSLRs. From my understanding more pixels usually means smaller pixels (other things being equal) and more noise therefore lots and lots of pixels in a smaller sensor is far from desirable.

Click to expand...

I basically agree with this as per my post #9 in this thread BUT I can remember the same argument being made when the 1.6 croppers went from 6mp to 8 and then again when the 40D came out at 10mp. As technology develops 'other things are not equal'.
to quote a review I saw recently from a well know pro:

All else being equal, the more pixels you put on a sensor the smaller they are, and the smaller they are the noisier they are. However, all is rarely equal. The 7D uses a new “gapless” lens system over the pixels which increases their light collection efficiency and so reduces noise. Upgraded electronics (including the dual Digic IV processors in the 7D) can also reduce noise. So the question is whether the increased pixel count of the 7D makes it noisier than its predecessors (40D/50D). I think I can say that no, it doesn’t.

Comparing the 7D with a full frame or 1.3 cropper costing almost three times as much is not a fair or relevant comparison IMO. For example the new 1D4 has been given a RRP of something around £4500 I believe so it is obvious that it will have advantages over the 7D.

I am still shooting with the humble 40D but from everything I have read noise levels (especially at high ISO) from the 18mp 7D is as good if not better than the 10mp 40D - for this Canon should be commended IMO.

Hi all,

I thing that we should all agree on few things:

1. When we discuss the sensor and pixel quality, we should stay focused on that. Maybe one camera is better on many issues (handling, focusing etc.) from another one, but this does not mean necessarily that it gives better image quality sensor-wise. Of course all other issues might, or will allow the photogapher to capture a better image in the first place.

2. Sensor pixel technology changes continuously, so equal or better results can be achieved from newer sensors, even if their individual pixels are smaller.

3. Given the same technology is used for 2 sensors (this, most times, can be related to model release time), the bigger the individual pixel size the better.

4. So, unless a radical new technology is discovered for sensor making, we should expect that gradually, the individual pixel size will be decreased (more megapixels), with better image quality.

5. The critical question is if the rate of increasing the megapixels in newer models is following the technological capability, to achieve at the same time better image quality.

MY VIEW IS THAT:
For the moment, best image quality as regards Canon can be obained from 5D Mark II, 21 megapixels. This is one year old model, so I assume that same quality can be obtained today from a camera with slightly smaller pixel size. I believe that ID Mark IV, is following that, with a slightly smaller individual pixel size (about 27 megapixels full frame equivalent).

7D has pushed things much futher with full frame equivalent of 46 megapixels. I believe that this is way too much to retain a quality even close to 5D Mark II of 1D Mark IV.

I also believe that it is way too much even compared to 40D, and Canon should have gone from 10 mp of 40D (26 mp full frame equivalent), to maybe 12 mp for 50D (31 mp full frame equivalent), and to maybe maximum 14 mp for 7D (36 mp full frame equivalent). This way the image quality would have been quite close to 5D Mark II and ID Mark IV.

As I said before, other things apart from sensor quality are also counting, so I am sure that 7d is better than 40/50D in many ways, but here I discuss just sensor and pixel quality.

Thanks

Ashetos said:

3. Given the same technology is used for 2 sensors (this, most times, can be related to model release time), the bigger the individual pixel size the better.

Click to expand...

Nope. You're ignoring the topic of this thread.

Follow your point through and you're saying that a 1536x1024 FF sensor will give even better image quality than your 5D MkII. Yes, the noise performance would be phenomenal. Yes, 100% crops would be amazingly noise-free. But image quality would be totally crap because you wouldn't have enough pixels per bird to see any detail.

There are two, mutually-exclusive, factors that are in play when designing sensors. First you want each photosite to be as big as possible to maximise signal:noise. Second you want each photosite to be as small as possible to maximise the amount of detail you can capture. Every sensor design has to be a compromise between those two factors.

In the 5D they've gone towards the large-pixel, low-noise, low-detail end. The 7D is set at the small-pixel, high-noise, high-detail end. The 1D is set firmly between them.

What's plain is that the 7D is just about at the limit for pixel density. From experience and reviews, it looks like you really do need a top-quality lens to take advantage of its resolving power. A substantially higher pixel density will just result in the blur (from imperfect lenses, diffraction, AF uncertainties) being spread over multiple pixels.

I think you are trying to convince me for something that we totally agree: more megapixels are more than desirable, especially for bird photographers, for the detail and the crop capability (pixels per bird), provided the pixel quality is acceptable. But, using your example I could say that with current technology, a 12000X8000 FF sensor would also be crap.

So it all comes down to a compromise of pixel size and pixel quality with given technology.

By the way I do not own a 5D Mark II (although I know very well its capablilites from a colleaque which we go bird photography on a regular basis), but a 40D, and actually I might buy the 7D very soon, so no sentimantal links with any of the above mentioned bodies!

Thanks