[monochrome]

Investment is psycho, not real, self fulfilling prophecy. It's enough for investors to think the market is going to go up to make them decide to invest, as they invest new products and services are offered and GDP goes up. So if everyone of us if positive, the GDP will go up. If everyone says like you, GDP will go down. You have to say that the market is going to go up if you want it to go up.

Economic activity - employment, consumption, production, capital investment, Fiscal spending, tax revenue - all these are real. M * V = P * Q1 is theory, and good theory at that.


Investment is rational. Trading ticker symbols in the stock market is something else entirely, governed by Behavioral Finance, which is psycho.

You seem t be talking about the stock market. I'm talking about the Economy. Current circumstances are real - very real. They likely will improve over time and there likely will be relative normality at some point next year, but nevertheless there will be real, permanent financial damage, real demand destruction, real impairment of capital (some of it permanent - see airlines and restaurants) real fear among the population to venture back out, and possibly Coordination2. We need to be very thoughtful and deliberate about how we order our decisions for the next couple years. For a rational investor with staying power and a long term view, this is likely an opportunity to own shares of great companies at once-a-decade prices and put them away for - essentially - ever. I'm actually a cautious optimist longer term but a realist very short term.

At this point I think it best that we move on to a more photography oriented discussion, OK?



1. Money Supply x Velocity of Money = Production * GDP Deflator (1-Inflation). The Velocity of Money always declines substantially in these crises. in 4Q2008 V declined 50%. This time would be worse but for recent Monetary and Fiscal actions.
2. Consumers won't buy (at the margin) until their jobs are secure. Producers won't hire to restart production until they're sure there are people who will buy their products. The Fed's big fear is that we lock up in this "Coordination" at the end of the forced shutdown.

---

Last edited by monochrome; 04-04-2020 at 11:45 AM.

[Rondec]

The stock market will probably bounce back quicker than the economy. It really tanked hard, maybe harder than it should have.

On the other hand, take two months off of any really productive work for most people and then the beginning of June of middle of May tell people they can "return to normal." It just feels like there are going to be plenty of hardware stores and restaurants that just never reopen because there is no money left to do so. Plenty of people will have no job to return to, although of course, plenty will have one as well.

[monochrome]

The stock market will probably bounce back quicker than the economy. It really tanked hard, maybe harder than it should have.

On the other hand, take two months off of any really productive work for most people and then the beginning of June of middle of May tell people they can "return to normal." It just feels like there are going to be plenty of hardware stores and restaurants that just never reopen because there is no money left to do so. Plenty of people will have no job to return to, although of course, plenty will have one as well.

My entire company of 250,000 was told Friday to plan for remote work (or rotating hours in customer-facing branches at full pay) until the end of May. My son has enough cash and credit to keep his business open at 10% of normal revenue until August, but revenue is dependent on high school and college athletic seasons. If schools open normally in September he’ll survive but it will take 5 years to get back to where he was January 1st.

Everyone is guessing because we won’t be released from stay-at-home orders until we see how the virus is behaving and we just don’t know what will happen with that.

[clackers]

The Alpha 9 II reads the AF 60 times per second, this is not the bottleneck.

It only has 24 Mp, WAI, that's also why its frame rate is high.

It doesn't have 36Mp, 42Mp or even 50Mp of other full frame cameras.

[WorksAsIntended]

It only has 24 Mp, WAI, that's also why its frame rate is high.

It doesn't have 36Mp, 42Mp or even 50Mp of other full frame cameras.

It also has a very fast processor.
Even the slow (in terms of compute power) z7 manages 60fps, having double the compute power is not uncommon in high-end bodies now.
The sensors are fast enough these days, they are not "slow stills sensors". In fact, they come from the same wafers as video camera sensors in some cases.
The af pixels of the a9 II are gettinf indepent readouts from the other ones btw, so the total resolution of the sensor does not matter for this topic.

[beholder3]

It also has a very fast processor.
Even the slow (in terms of compute power) z7 manages 60fps

The 2019 Nikon Z7 has a max frame rate of 8 FPS for comparable 14bit raws, which is pretty much the same (less) frame rate of the 2013 Pentax K-3 I.
The Z7 doesnt even have twice the pixelcount of the K-3 I.

A9II (2019): 480 MPx/sec

Z7 (2019): 365 MPx/sec

NX1 (2014): 423 MPx/sec

K-3 I (2013): 199 MPx/sec

In digital world not even processing twice as much data after 6 years of progress is ... underwhelming, to put it nicely.

[WorksAsIntended]

The 2019 Nikon Z7 has a max frame rate of 8 FPS for comparable 14bit raws, which is pretty much the same (less) frame rate of the 2013 Pentax K-3 I.
The Z7 doesnt even have twice the pixelcount of the K-3 I.

A9II (2019): 480 MPx/sec

Z7 (2019): 365 MPx/sec

NX1 (2014): 423 MPx/sec

K-3 I (2013): 199 MPx/sec

In digital world not even processing twice as much data after 6 years of progress is ... underwhelming, to put it nicely.

First of all, the parameter you are looking at is a very bad indicator for cpu power. While the K3 is completly locked up getting those fps, the Z7 does a lot of other computing by the side. It produces images for the evf, it has a complete picture analysis (eye-af) and predictive tracking (granted, at the z7 working more or less badly). Still, the Z7 has a weak cpu compared with a A9, a D5 (and D6) or a 1DxM3. The R5 will propably also outperforme the Z7 in this matter, I cannot see another way for 8k full sensor footage otherwise.

Another indicator: if you look at the pure cpu power of the camera: it does take more processing to create the compressed raw files for 9fps shooting than the 8fps 14bit NEF files. This seems to rather be a limit in buffer bandwith.


Also, in the last six years there was not much more gain in processing power than doubling. Looking at Intel for example we already had 14nm Broadwell six years ago. AMD changed a lot since than, which is mainly because they started at a really bad place. Single thread performance of ARM did not change much either, they only print bigger dies now.

Looking at interconnection bandwidth (which I guess is the bottleneck for the Z7 at what you are looking at) it is pretty much the same. A doubling in six years is the best we had.

Times have changed and without bigger dies, the performance increase is not as it used to be.

[clackers]

Even the slow (in terms of compute power) z7 manages 60fps, having double the compute power is not uncommon in high-end bodies now.

?

Its frame rate is only 9 fps for three seconds, then it slows to 4 fps.

Are you talking instead about video?

It can do 120 fps of just 2Mp output, and only 30fps of 8Mp.

[WorksAsIntended]

?

Its frame rate is only 9 fps for three seconds, then it slows to 4 fps.

Are you talking instead about video?

It can do 120 fps of just 2Mp output, and only 30fps of 8Mp.

I am talking about the frequency of permanent sensor readout. In this case for the evf.

[beholder3]

First of all, the parameter you are looking at is a very bad indicator for cpu power.

Noboy here is talking about "cpu power". You miss the point. The discussion is exactly about Mpx/sec full sensor readout usable for full res stills image per second.

The Z7 uses lame technology far behind cameras from 4 years earlier.

Looking at interconnection bandwidth (which I guess is the bottleneck for the Z7 at what you are looking at) it is pretty much the same. A doubling in six years is the best we had.

And that is why there was a much better performing NX1 many years earlier?

Hint: The Z7 1.2 MPx EVF obviously doesn't need full sensor readout.

[Rondec]

I am talking about the frequency of permanent sensor readout. In this case for the evf.

Clearly, as beholder3 says, you aren't getting a full read out for the EVF. You would hardly need that for such a tiny screen.

[WorksAsIntended]

Clearly, as beholder3 says, you aren't getting a full read out for the EVF. You would hardly need that for such a tiny screen.

It is however a downsample of a lot of pixxels more than displayed. Otherwhise the picture in the evf would get grainy as hell.

---------- Post added 04-07-20 at 03:26 AM ----------

Noboy here is talking about "cpu power". You miss the point. The discussion is exactly about Mpx/sec full sensor readout usable for full res stills image per second.

The Z7 uses lame technology far behind cameras from 4 years earlier.




And that is why there was a much better performing NX1 many years earlier?

Hint: The Z7 1.2 MPx EVF obviously doesn't need full sensor readout.

MPx/s is a useless parameter as it does not determine any function by the camera and use of milc.

The nx1, as based on android, is a special case with a lot more processing power than others back then indeed.

[Jonathan Mac]

The camera market is still in flux and add in the coronavirus epidemic and the result is that it's not really the best time to judge who prefers what with their buying preferences. It seems virtually all new lens announcements these days are for mirrorless cameras and I think these new lenses will close the biggest gap between DSLR systems and mirrorless systems - lens choice. However, new lenses are produced to new standards, which means that people want them sharper and don't care if that makes them much bigger (apparently) but this also means they're much more costly. Mirrorless systems don't have all those older lenses which are a bit hazier at larger apertures and some focal lengths but where that is reflected in their lower prices.

In the west there's also the continuation of the idea that bigger cameras are better than smaller ones, something which will likely change bit by bit.

[WorksAsIntended]

The camera market is still in flux and add in the coronavirus epidemic and the result is that it's not really the best time to judge who prefers what with their buying preferences. It seems virtually all new lens announcements these days are for mirrorless cameras and I think these new lenses will close the biggest gap between DSLR systems and mirrorless systems - lens choice. However, new lenses are produced to new standards, which means that people want them sharper and don't care if that makes them much bigger (apparently) but this also means they're much more costly. Mirrorless systems don't have all those older lenses which are a bit hazier at larger apertures and some focal lengths but where that is reflected in their lower prices.

In the west there's also the continuation of the idea that bigger cameras are better than smaller ones, something which will likely change bit by bit.

Lens sells for mirrorless is much bigger than dslr lenses according to Sigma. That however may just be due to the fact that the mounts are pretty new, so people need to buy new lenses if they want native ones.

There are still small new lenses out there. Nikon z has smaller lenses for now, the 35,50,85 1.8 and also the relativly compact 24-70 f2.8 and f4, the Canon rf 35 1.8, the Sony 35 1.8 and 55 1.8 and 85 1.8 as well as the complete Zeiss Batis lineup. There is demand for small lenses. The Canon rf 70-200, the Tamron 70-180, all of them are smaller than the old mount equivalents.

Not all new lenses are big and according to Sigma their best sellers are the smaller lenses like the 56 1.4.

[beholder3]

It is however a downsample of a lot of pixxels more than displayed. Otherwhise the picture in the evf would get grainy as hell.

EVF images are never "downsampled", just as this would require both full readout plus downsampling on the fly. Just the same as video. The EVF is just a secondary display, no different to the main display on CMOS cameras.


The K3 I sensor doesnt provide 30 FPS full 24 Mpx sensor readout plus downsampling only because you can take and view on the display 30 FPS uncropped video.

Both low res displays on a mirrorless camera only show selective pixel readout.

MPx/s is a useless parameter as it does not determine any function by the camera and use of milc.

Not. Any other parameter you might want to introduce is useless.


MPx/sec is the key compound measure for a cameras ability to capture image data and move it through whatever pipeline and processing into internal buffer before writing it to memory card. Nobody cares what happens in partial steps in between.
A sports shooter doesnt want 60 FPS 1.2 MPx images. They want the full data saved.