Just Published - Mirrorless wars

[Tony Jay]

I just saw the video by Kevin. We will meet in the war zone and photograph

If my workshops is any measure I can only say that I see few Sony cameras and as many Fuji and Olympus, but all are dwarfed by the number of Canons and Nikons until this day. I had expected 4-5 years ago that this would happen much faster.

So far I have not purchased a mirrorless system and I have looked into many EVF's and thought, OK, one day I will have one, but so far I don't like what I see even though it is what I get.

As stated by many including Kevin and including the recent video from Tony and Chelsey Northrup (see here https://www.youtube.com/watch?v=mYYaQz83etY ) the traditional camera makers like Canon and Nikon would need to cannibalise their existing lineup and market mirrorless as better than what they have. They also would need to invest in both product technologies, so not that attractive. Btw. the same problem car manufacturers have now with the transition to electric. It will be interesting to see how many will survive these transitions. It is not a given they all will. I'm pretty sure they will not all make it.

Your observations do not surprise me one bit!
When it comes to change in photography there is always inertial resistance...
Yes, there have always been early adopters (this forum is probably over-represented in this respect) but the mainstream has always been relatively slow to adapt
Also, it depends on what is changing and how big that change is.

I own Both Canon and Sony equipment.
There is absolutely no doubt that late-model Sony cameras easily outperform their Canon counterparts. However, there is also no doubt that Canon cameras produce stellar images.
Also, the reason that I was able to buy and use Sony cameras was that it was not a case of going cold turkey on my Canon equipment. Every Canon lens, with some limitations at times, can be used on the Sony bodies. Where the Canon bodies hold an edge I use them; when the Sony bodies have the edge they get a run. I still use both systems interchangeably according to the needs of the image...
But, the point is that even if the Sony bodies are demonstrably better in many situations to their Canon counterparts, the Canon bodies are also demonstrably good enough.
I probably would never have changed apart from the fact that I wanted to print big and so resolution and fine detail became particularly important.

If Sony (and other mirrorless vendors) are able to maintain their technological edge over Canon, Nikon, and other latecomers to the mirrorless party, then, over time, more and more photographers will switch and newcomers will increasingly choose Sony (or one of the other mirrorless systems) as their first investment.

None of the currently implemented new technology is truly revolutionary. When it comes to the end-image obtainable the changes are currently evolutionary rather than revolutionary. The delta is not so large that it positively demands immediate change...

I expect the market to continue to shift, incrementally, as the benefits of shooting with mirrorless bodies become more apparent. Yes, the Sony A9 is a definite marker as to the potentialities of mirrorless but most of those relate primarily to how one shoots, and only secondarily to the quality of the end-image.

One thing is apparent to me: the market situation pertaining to mirrorless bodies definitely demands a response from the latecomers to this market. Canon and Nikon are both showing signs of stirring! I for one cannot wait to see what they come up with!

Interesting times ahead...

[vartkes]

I have lived my last forty years of my career in the world of bleeding edge technology. I started with the first generation microprocessors and ended it very recently. I lived, on three distinct occasions (twice at Hewlett Packard and once at Blackberry) the annihilation of a technology leader that relinquishes its leadership. I have also shot Canon since 1985 until three months ago. I walked away from Canon having fully realized that they relinquished their technology leadership after 5D mark II. So the camera market was ripe for a discontinuous change with a new entrant to seize the leadership mantle; SONY has done this. One of the many lessons I learnt through my professional work is that in Technology once the leader relinquishes or becomes lazy to tending to its leadership it can NEVER regain it. There is not a single example that has negated this phenomenon; and please do not bring up APPLE's 'reincarnation' since what Jobs did was to use the name Apple and created a totally different company in a new emerging market. The original Apple is still resting in peace.
So my conclusions: it is almost irrelevant what CANON does now they are done with. As for NIKON with all the technology dependence they have on SONY and other suppliers of core technologies, they were never the leader since cameras went digital. Whit that observation we now have three positions left that will rule the roost for the next decade or so; SONY in full-frame, FUJI and SONY in APS-C and one of Panasonic or Olympus in micro four third product segment. I predict the overall winner will be SONY since they hold the wand over a collection of key technologies that no one else has.
Just in case anyone cares what I replaced CANON with? It is FUJI, because I really like their X-mount optics and they are investing at a good rate and gaining market share big time. They will be around until I am too old to hold a camera in hand.

[D Fuller]

Hi,

With modern sensors, the cost is slowing down everything by a factor of four, one frame per second instead of four frames a second. How hard is that to get?

The pipeline is 16 bit wide. Processors are normally 8, 16 or 32 bit wide. No one builds a processor 12 or 14 bit wide today. That said, it may be possible that some part of the ASICs may be optimized for 14 bits, but, digital technology is normally multiples of 8.

Modern CMOS sensors use column wise conversion and the converters are probably ramp type converters. A reference voltage is ramped up until voltage matches the voltage from the pixel. The value is simple the number of clocks until match is reached.

If you want to measure with 16 bit precision, you need to have 1/4 of the step size on the ramp, increasing conversion time by a factor of form. So, you get a slow conversion.

On older systems, most CCDs and some CMOS, the voltage from the pixel would go trough a preamp to of sensor ADC. Those would be flash type ADCs, so you could get a 16 bit device from Burr & Brown, and that would deliver 16 bit data. But, the input data to the converter would still be limited by pixel noise. So, you have 72 dB of data, corresponding to 12 bits and feed it into a 16 bit converter. So you get 16 bit of data with low four bits representing noise.

Would you do that in engineering school, it would be regarded a serious error. If you have input data with say two decimals, you should never present the result with more than two decimals.

So, the material cost of going 16 bits is zero, but with modern sensors it would lead to significant performance loss and the old sensors don't have accurate data anyway.

Best regards
Erik

This all makes sense. Engineering is task of finding a balance of competing variables. It makes sense that there is no cost if performance loss is acceptable. What is the effect if performance has to stay the same?

[ErikKaffehr]

Hi,

If we had ramp type converters and tried to increase bit depth while maintaining performance, there would be two possible routes:

Increase clock frequency on the ramp four times. But, that would yield a fourth of the sampling time and generate more heat.

As the data is still something like 13 bits accurate, it would not affect image quality in a positive manner.

The other way to that would be to add more ADC, like having sixteen ADCs for each column. That would benefit performance, but still not improve data.

Just to give a few examples from the Sony world.

Sony A7rII uncompressed data is actually 16 bit, but with the two most significant bits fixed at zero. So, it is 14 bit data in a 16 bit format.

But, the data is more like 13 bits, according to guys who know a lot more than me.

All that applies to single shoot only. In serial modes the camera reverts to 12 bit mode, utilizing 11 bits.

So, when FPS is needed the camera is essentially 11 bits only.

The A9 has a stacked sensor with several ADCs working in parallell for each column.

The Fuji GFX may not do those tricks, but it's electronic shutter is limited to something like 1/3 second sweep time.

In short, common sense is that if you have data that is accurate to 14 bits you would sample it at 14 bits. Sampling with more bits never makes any sense.

Best regards
Erik

This all makes sense. Engineering is task of finding a balance of competing variables. It makes sense that there is no cost if performance loss is acceptable. What is the effect if performance has to stay the same?

[BernardLanguillier]

In short, common sense is that if you have data that is accurate to 14 bits you would sample it at 14 bits. Sampling with more bits never makes any sense.

It doesn't make any sense technically... but it has been making a huge marketing sense for the MF vendors throughout the years.

Cheers,
Bernard