This Paper: Compact System Camera Image Quality

[ThomasH_normally]

I got to wonder about the phenomenon: 
We the people of technology are not always good writers, and so some generic advices for formulation of papers, documents and essays are very common. The most basic of them, let me call it the "Prime Directive," is 'Never start a paper with a word "This Paper"'. This should be as known as the meaning of green and red light at the street corner.

And yet, this essay "Compact System Camera Image Quality" starts with:

"In this paper we review..."

and than only a few lines later the next paragraph starts with:

"This paper reviews the image quality factors..."

Of course, this should not diminish the technical knowledge and its generic unquestionable value, however one have to wonder if enough effort was made in selecting and putting together the content.  We know that "a paper" is not a being or a mechanism, it cannot review anything...  Posts in a public forum may of any quality, both in content and in grammatical valor for a variety of reasons.  Not so an essay! May I suggest to re-formulate and re-post?

[pegelli]

And while they're at it they might as well take out the comparison summary table.
For an article claiming to describe a scientific and technically founded comparison and then "lump" all DSLR's vs. all CSC's is a gross simplification. The spread in either of these catagries is larger than the difference they're trying to descrbe.

[viewfinder]

I was excited to see this article because I have been wondering if, and what, the image differences are between APS-C sensor in mirrorless camera and in DSLR, if indeed there ARE any differences.....

What I learned is that 'micro four thirds' are now called "CSC" and the only type of DSLR worth considering is a 'full-frame' one.

I realise that the car load of experts are very emminent, but maybe they should look again,...just saying!

[01af]

... this should not diminish the technical knowledge and its generic unquestionable value ...

Technical knowledge? Unquestionable value? Huh!?

That paper is nothing but a compilation of self-important techno-babble, with lots of inaccuracies and several mistakes interspersed, culminating in an entirely trivial conclusion. It is obvious that the authors are wading a territory that is slightly beyond their comprehension.

May I suggest to re-formulate and re-post?

My suggestion is, take it down, delete it, and forget it.

[douglasf13]

  The graph at the end of this report is absurd.  Is it lumping APS-C and m4/3 together, and then comparing to only full frame DSLR?  FWIW, it doesn't look like they took the NEX-7 into consideration, since it is 24mp and has a faster shutter lag time than DSLRs.

[DiaAzul]

I'm a little confused as to why this article found its way into LuLa. Whilst I might not agree with some of the articles presented on this site, the majority are very well written, intelligent and present a considered point of view. But this one seems a bit squiff.

In my very humble opinion, I have taken better pictures with my compact camera than I have every taken with a more technically capable full frame DSLR. The DSLR produces commercial grade pictures that meet a commercial need, the former produces more interesting and artistic output. Just because it gets more pluses doesn't make it capable of producing better images.

But then again, I don't need to rehash old arguments.

[wirchrick]

I was also wondering about a very important point that the authors of this article overlooked. 

The fact that you are never optically looking through the lens means that you can 'cheat' in the lens design.  That is you can correct for pincushion/barrel distortion in software and also chromatic abberation.  Because of this, it helps the lens design because you are trying to optimize for all these different concerns and if software solves some of them it make it easier to design a lens that has sharp focus, especially in zoom lenses.

I would like to know what the authors think of this aspect?

Thanks,
Rick

[jrv]

There is a problem here in the "Noise" section:

As pixel sizes decrease the number of photons detected at each pixel decreases (the number of photons detected is proportional to the area of the pixel) and as a result the noise variations increase

.

Smaller photosites have *less* of many types of noise, not more, since there is less surface area and volume.  What matters is not noise but instead the signal/noise ratio, and the smaller photosites suffer from less *signal* and an overall lower S/N ratio.

This in turn is why Nikon's D3s has such good high ISO performance: since the photosites are large they get more signal and hence a better S/N ratio.  Reducing noise is important but increasing the signal also matters.

The article does not mention depth-of-field issues with smaller sensors.