« on: July 31, 2014, 03:34:18 PM »
When I jumped ship from Olympus to Canon, the choice was Canon or Nikon. Canon had the lenses that I liked and when it came to the user interface, there was no comparison...Canon let me do what I wanted easily, Nikon had me diving into menus... I went CanonCanon made revolution few years ago. Currently it is a stagnant company, which still focuses significant aattention to dying P&S market. Product cycle for semi-pro and pro products is very long and shows that Canon does not sufficiently invest in R&D as other companies. Canon started loosing in the following areas:
1) Mirrorless market - they loose to Fuji, Olympus, Sony;
2) Sensor technology - loosing to Sony;
3) Lenses - starting to loose to Sigma and Tamron as these 2 companies started producing high quality lenses, which match or in some cases exceed qulity of Canon lenses (e.g. Sigma 50 mm 1.4 Art, Tamron 24-70 2.8 VC) for affordable price.
Losing is a tough way to put it. Being #1 means you have many mouths to feed -- not just enthusiasts and pros in a forum. I'm truly curious to see what chunk of Canon's business folks like us in this forum (and like-minded shooters not in this forum) actually represent to them.
Also, advances in one small segment of the photography world does not make Fuji, Olympus, Sony, Sigma and Tamron the 'team to beat'. It means that they have had success in one small segment of the photography world. That's all.
And lest we forget, the company arguably most revered for its innovation and 'firsts in the industry' -- Sony -- can't seem to understand photographers well enough to produce a top-to-bottom well thought out camera that is free of non-trivial flaws. They seem predisposed to come up with something cool and useful (that I might want!), shoot it out to the market half-cocked, and under-deliver. What's worse is that they don't seem to learn from this, and they just shoot out another version in record time with similarly iffy results.
So it's more than who is on a roll or has the best team or most innovative pipeline -- I want the company that most consistently satisfies its customers. Bleeding edge innovation isn't my driver. I want a camera/system that does exactly what I want it to do. Canon may be slow, but they have never let me down with what they have delivered.
As things stand today, for my purposes everything about the Canons is superior to the Nikon and Sony offerings except for the sensors and I expect the gap to narrow drastically or even disappear soon.
Look at the clues...
Clue #1: Canon sensors are inferior to Nikon/Sony. Everyone knows that. You can bet that the people at Canon know that too.
Clue #2: Canon executives have hinted that something big or revolutionary is coming.
Clue #3: The 7D2 has been delayed for "production reasons".. We know it isn't just making another copy of the 70D sensor with a few more or less pixels.. it has to be something else.
Clue #4: The delay is NOT DPAF. It is here and in production in the 70D. Being a lower cost camera and stocked and sold in general consumer stores, it is a safe bet that 70D sales will exceed that of the 7D2 AND the entire FF lineup. A bit more for a 7D2 will not matter.
Clue #5: Canon has sensor fabrication facilities that work on much finer lithography than the APS-C and FF sensors of today. Pick up a Canon P/S camera for proof....
Clue #6: P/S sales are declining and this means extra capacity is opening up on those finer lithography production lines...
Clue #7: We know that by going to row or column A/D on the sensor that they could drastically drop noise and increase the DR of their sensors. You can bet that Canon knows this too.
My bet is that the delays in 7D2 production are due to moving the A/D onto sensors with finer lithography. This has to happen at some point and now is the logical time. I would expect a rapid refresh of the FF lineup after this.
Correct me here if I am wrong with this conclusion --- but in all the posts i have read here regarding the internals of a camera body --- heat is a factor in the quantity of noise, which would harm sharpness and overall IQ.
Correct me if I am wrong yet again, but, is it conceivable that canon bodies could improve IQ by finding more ways to reduce heat inside the camera body? Just tossing ideas out there...
Heat can cause two forms of noise: "Amplifier glow", which is really just IR radiation increasing temperature, and dark current noise. Dark current is always flowing through the circuitry of a sensor, and it sometimes causes the release of an electron into a potential well (i.e. a photodiode). The rate at which dark current causes the release of free electrons is dependent on temperature, it effectively doubles every 5.8°C.
Neither of these forms of thermally-affected forms of noise are really a problem for most still photography. The exposure times are usually too short for enough dark current or incident IR radiation to affect the release of additional free electrons. In the case of dark current, CDS units on each column (or possibly each pixel) are reset along with the pixels, and accumulate charge for the same duration as the pixels are exposed. When each column is read out, the CDS unit charge accumulation is subtracted from all the pixels in the given column. For shorter exposures, this can effectively eliminate dark current noise. CDS units usually have a limited capacity, and for longer exposures, or exposures long enough at very high ISO, they may not be able to entirely subtract the dark current noise from the image signal. This is usually only a problem for things like astrophotography, where exposures might be many minutes long.
The use of per-column CDS units (which is usually the case...I believe even Sony Exmor sensors use a per-column digital CDS unit, since it's actually part of their CP-ADC) can actually cause vertical banding. A better approach would be per-pixel CDS units, that are reset along with each of their paired pixels, and accumulate dark current specifically for each pixel. That would eliminate the possibility that CDS units introduce vertical banding (per-column banding, really), but it does increase the complexity of the sensor. I've seen one or two patents that cover per-pixel CDS...it's been a while, but I want to say one of them was a Canon patent. I honestly can't remember.
Anyway, there are sources of thermal noise. On very hot days, your camera will absorb more ambient heat, and that can increase noise a bit, especially at higher ISO. However for the most part, I don't think that significant investment in reducing sensor temperature is going to improve most still photography IQ. The current levels of read noise at lower ISO settings completely swamp any amount of dark current. At higher ISO settings, Canon sensors are already competitive, as they are limited by physics, although a small improvement might be made buy increasing Q.E. from the ~49% average to say ~60%.