Pro-photographers, time to reconsider

Paul Luscher said:
Redisburning: Why a push toward smaller sensors? Moore's Law, for one.

And why a smaller sensor? Smaller, lighter cameras, for one. Hey, wasn't that one of the big selling points of the Leica when it first came out? Anything that moves that ball further down the field is all right by me. If you've ever spent a day hauling around a bag of Nikon Fs and lenses, you'll appreciate smaller, lighter cameras.

As to your comments about lenses, I think another commenter has answered that. But taking your position to its logical extreme, then maybe we should all still be using view cameras and lenses, because after all, smaller doesn't necessarily mean that much better.
Click to expand...

moore's law actually makes certain that full frame will never be surpassed by smaller sensor technology of the same kind as long as it is not abandoned.

and really, the M9 is not a large camera. let's compare it to an m4/3 camera:

M9_EP1_back.jpg


so, it's what, 20% bigger? but the sensor is 200% larger. that's because relative to the camera body, the sensor is tiny. if you can reduce the size of the necessary circuitry and computers in the camera, then you will do loads more to reduce the size of the camera than simply reducing the size of the sensor.

as far as your last point, well the thing is that 35mm already has low yield rates relative to these smaller sensors. that appears to be the current best combination of economics and quality for most working professionals right now.

I would be considerably more convinced by this argument if the only full frame cameras still available were the Canon/Sony/Nikon ones. but Leica has already demonstrated that you don't really need to go that much up in size to get a significantly larger sensor. now, if only they could do that with regards to the price!
 
the problem is the 'significantly' larger leica sensor doesnt do one of the main things canonikon FF sensors do: high iso/low light. how is that explained?
 
I like m4/3, but there's no getting around the fact that physical sensor size = different looking images. Same way medium format film looks different to 35mm.

Compared to a 35mm format full frame DSLR, m4/3 can have in comparison: visibly compressed tonal range, less possibility of creative use of depth of field, lesser IQ in dark situations, less chance of recovering blown highlights, etc.
At the same time, the m4/3 cameras have their advantages like smaller/lighter/quieter/lenses always sharp even wide open/good small super zooms.

It's just basically up to the photographer which compromises suit him/her the most. But there's no doubt that larger sensor = technically better overall possible IQ. For me, I wasn't able to work with m4/3 in all the same situations that I was with the 35mm sensor - I really value the attributes that a larger sensor brings.
 
Fat chance!

Maybe for some appications but not in the gloomy world I work in IMO.
 
redisburning said:
moore's law actually makes certain that full frame will never be surpassed by smaller sensor technology of the same kind as long as it is not abandoned.

and really, the M9 is not a large camera. let's compare it to an m4/3 camera:

M9_EP1_back.jpg


so, it's what, 20% bigger? but the sensor is 200% larger. that's because relative to the camera body, the sensor is tiny. if you can reduce the size of the necessary circuitry and computers in the camera, then you will do loads more to reduce the size of the camera than simply reducing the size of the sensor.

as far as your last point, well the thing is that 35mm already has low yield rates relative to these smaller sensors. that appears to be the current best combination of economics and quality for most working professionals right now.

I would be considerably more convinced by this argument if the only full frame cameras still available were the Canon/Sony/Nikon ones. but Leica has already demonstrated that you don't really need to go that much up in size to get a significantly larger sensor. now, if only they could do that with regards to the price!
Click to expand...

I think there is some fuzzy logic being applied here to the concept of Moore's Law. Gordon Moore's observation was that transistor density is doubling every 18 months or so. Meaning that transistors are shrinking, and their resulting VLSI integrated circuits are getting smaller (read: cheaper to manufacture) while simultaneously operating faster, since the charges have less distance to travel between transistors. So with these smaller chips, more can fit onto the same size wafer, reducing their cost, while their increased performance means they can be sold for a higher retail price; a double whammy of profitability. That's the economics of Moore's Law. It's a business model.

But with a size-specific image sensor (full frame, or APSC, or micro-4/3), the amount of silicon "real estate" it takes to make each chip on the silicon wafer is fixed, meaning that the only way for a manufacturer to increase profits over time (which directly drives up the company's stock price) is to increase pixel density; hence the "pixel wars" of the last decade. But now manufacturers are running up against hard physical limits, hence the quest to find new sensor designs that maintain the high pixel count (small cell size) while still having good low-noise performance at high ISO.

The only image sensors that might scale to Moore's Law over time are point and shoot sensors that are not size specified by format; these can be shrunk down over time, making them cheaper to manufacture due to their reduced size, more can fit on the same sized wafer.

I think the search for more pixels is over; if you need more than what the current crop of full-frame cameras offer, go medium format. As for whether the smaller formats are gaining on full frame, in theory any advance in sensor design for small format chips should also be seen for larger format chips, assuming some great leveling process whereby all the major manufacturers end up employing the same basic chip designs. That hasn't totally happened, as evidenced in the case of Sony, who maintain intellectual property rights to their specific chip designs that seem to out class most of the competition.

-Joe
 
rbelyell said:
the problem is the 'significantly' larger leica sensor doesnt do one of the main things canonikon FF sensors do: high iso/low light. how is that explained?
Click to expand...

Kodak CCD vs Canon/Sony CMOS?

CCD is really good at low ISO, CMOS at higher iso. If Leica simply works with Sony on it's next camera you will get about the same performance.

PKR said:
redisburning said:
and really, the M9 is not a large camera. let's compare it to an m4/3 camera:

so, it's what, 20% bigger?
Click to expand...

What's the difference in weight between those two cameras?
Click to expand...

Leica has always been heavier than cameras of the same size due to materials chosen. If they wanted to make the camera out of polycarbonate and use an EVF I suspect it would weigh significantly less than the pen cameras.

my ZM50 weighs about the same as my dad's Canon 18-55 and the 18-55 is at least twice the size. the ZM is mostly metal, the 18-55 mostly plastic. the latter is still pretty tough though, or at least that is what I would assume.
 
JoeV said:
As for whether the smaller formats are gaining on full frame, in theory any advance in sensor design for small format chips should also be seen for larger format chips, assuming some great leveling process whereby all the major manufacturers end up employing the same basic chip designs. That hasn't totally happened, as evidenced in the case of Sony, who maintain intellectual property rights to their specific chip designs that seem to out class most of the competition.

-Joe
Click to expand...

well this was my point.

I assume that the FF sensors lag behind in pixel density due to yields, as I mentioned before. the reality of production is almost always putting a hamper on the idealism of engineering.
 
Only until the 4/3 lens designers point their pencils at the wide-angle end of the spectrum. I hear the Oly 12mm lens is pretty good. How long before we see a 10mm or even 8? And when sensor technology engineers solve the "light well" problem, there won't be anything stopping lens development on the wide end. :)
 
Why not? If you've noticed recently, Voigtlander for example, has been coming up with 4/3 lenses with f0.95 speed. This is equivalent to a f2 35mm-format lens (DOF-wise, that is). Its very early in this 4/3 game. And things are moving forward very quickly.
 
Jamie Pillers said:
And don't forget... 35mm isn't full-frame either. Its not 8x10. ;-)
Click to expand...

I realize people are tying to be funny when they say this but it's really missing the point.

when Kodak introduced the DCS cameras in 1991 the camera used the Nikon lens mount. And Nikon/Canon were the big pro players in the time. Full frame refers to 35mm because that is the sensor size that mount is designed for.

hearing this line over and over again is just like when people call the x100 or an ep1 a rangefinder. not so offenseive as to keep people from saying it, but similarly annoying.

smiley face.
 
gavin, i agree with all your points except that FF performs better in low light/high iso, and has more DR than m4/3.

certainly both have been true up til now, just as they had been true vs apsc up to a certain point on the technology curve. after shooting constantly with the OMD for about two weeks, i can safely say that it is at least as good as my 5d at 800-1600, maybe better at 16, certainly much better at 3200.

while the 5d DR, and the x100 DR, are both superior, imo, to what i get from the OMD, its not as wide a gap as you think. imo, next one or two generations, it'll be a dead heat.

certainly if FF sensor technology advances apace (and i believe that is not a given, as at some point a given technology maxes out) it certainly will stay ahead of m4/3, in the way the 5diii stays ahead of the 5di. but to me, and maybe to many, i dont need much more than i got from the 5d. so at some point that we're really close to with the OMD, many will feel they dont need the marginal IQ increase they get from the much larger, much more expensive FF competitors. if one can shoot IQ the way a paid pro shot 3 years ago, that seems plenty good enough, no?

its kind of like how folks feel about say the IQ of the zeiss distagon 35/1.4 at say $1200, many feel this is the best IQ lens one can get for a reasonable price. the lets-call-it-'fact' that one can achieve marginally better IQ with some leica lens at $3000 is simply not worth it. 99% just dont need/want/or even understand paying that much more when the distagon is by all accounts freaking stunning.
tony
 
TXForester said:
Bingo! (not the dog) I think that is the point of the article.
Click to expand...

Thank you TXFrorester for steering us back into the topic :)

The article did not argue about specific advantages of larger sensors. If that were the main thing for pros, all pros would be shooting 4x5 or larger, and we'd see a 4x5 (inches) digital sensor becoming more common, won't we?

But I would imagine as a professional (and I am using this term to indicate how dependent you are on your gears to make a living), there is a point where the ability to get super-narrow DoF does not translate a $5000 increase in your pay rate.

When that point is crossed, meaning when that's the *only* thing that the smaller sensor can't give you, then the question is:

Which one would you rather lug all day long (or multiple days in many cases) ???

An 8-lbs D3 with lenses or a 2-lbs E-M5 with equivalent lenses?

(that's an arguable example, but you get my point).
 
Well, we can all argue about whether full frame is better than 4/3, but the point is: over the course of time, image recording technologies have become smaller, but the image quality has generally steadily improved. guess that's what I meant about Moore's Law, though somewhat misapplied. Hey, I'm not a computer genius. But to belabor the point, I'm not sure we would agree that an old glass-plate camera beats a 35mm film camera when it comes to image quality, merely because the recording medium of the former is bigger in size.

4/3 is here to stay, I think, and its image quality is sufficient that it will provide real competition to full frame, and with improvements in technology, may become the "full frame" of the future. There is no real reason to be bound to the notion that only full frame will do.

Yeah, if you want to get into heavyweight pixel-peeping, it's likely full-frame might win out. But for most purposes, images from a 4/3 sensor will be quite sufficient (as the author of the piece pointed out). So frankly, a lot of this back and forth is just arguing about "how many angels can dance on the head of a pin".

A little puzzled by the comment the full frame is better for high ISO. I shoot both my E-3 and my Leica M9 at high ISOs, and don't see any significant difference in image quality. In fact , the E-3 seems to be better at getting natural skin tones under tungsten stage lights than the M9 does. Just my experience.
 
If you are a pro shooting m4/3rd its like a soldier going to war with a pocket knife. A bigger sensor equals higher image quality period. The new D800E will set the standard other systems can only hope for. m4/3rd cameras are great cameras aimed at the casual shooter.
There is more to image quality than low noise at high ISO. We all seem to stop grading a camera there. There is also Dynamic range and color depth. This is where Olympus m4/3rds are weak compared to a bigger sensor. Olympus cameras DXO marks have been stuck in the mid 50s for years while the D800 scores a 95. The dynamic range may be good enough to shoot landscapes without HDR or filters.
If I owned a magazine and hired a pro and he showed up with a Pen or any other mirrorless camera I would send him packing.
 
You must log in or register to reply here.
Back
Top Bottom