Sony F65 hands on review

[Steve Freebairn]

It's also interesting that when you look at the Sony sensor it IS a Bayer pattern too - it's just turned through 45 degrees. And its not just Bayer pattern sensor that need OLPF's - you either apply a low pass filter or suffer from aliasing artifacts - you can't avoid the laws of optics & physics.

I think this is something that a lot of people don't remember. I pointed this out on one shoot by making the joke that it was very high tech, and the only way to get similar results with other cameras was by shooting at a canted angle

 

[Martin Chab]

Stuart:
Yes, of course the F65 sensor is bayer too and actually the low pass is stronger that i would like to see in that camera.
About the fact that the pixels are no co-sited you have a point.
What it bugs me is why every manufacturer makes so much effort to "hide" the numbers inside "tricky" statements instead of create confidence opening their specifications.
Is the same as when i hear most of the politicians, i wonder...why in the hell they don't say the whole story? Don't they realize that we would trust them much more and because of that we would follow them?
It is like the industry is afraid on showing their so called "flaws". There are no flaws, there are characteristics. I wouldn't use a Mercedes cabriolet to carry bricks and dirt...am i wrong?

 

[Blair S. Paulsen]

The de-bayering algorithms are the key to the final results. They are designed to accommodate the characteristics of the OLPF, angle of the bayer mask, layout of filters (RGB), etc.

Whatever combination of physical factors exist, the algorithm interprets those values to create the resulting RGB image. It is the characteristics of that derived image that matter.

It can be interesting and valuable to know the physics of the process when pushing the electro-optical system. That said, if the derived image can resolve "x" lines on a zone plate then they are in fact resolved, however you got there.

Cheers - #19

Note: If you doubt the power of algorithms in creating images let me cite an example. NASA sent up a space telescope that had a serious focus problem (mirror surface flaw IIRC) and came back with literally millions of soft focus images. By characterizing the error they were able to create an algorithm that actually yielded more resolution that they could have achieved had the original image been in sharp focus. Impossible you say, au contraire, because they were able to reference just how much out of focus an item should be vs how it was represented they could create a more gradual and therefore more granular interpretation. God I love math ;-)

 

[Mark Toia]

The only reason f65 is cheap now is because red epic is cheap. Did not know that they are only 65k now... Last time I was offered one from my Sony dealer was 105k. Well its good to see finally they are trying to be competitive.
The f23 & f35 were over 200k each both gathered dust about 12 month after release. I have Freinds with rental companies with these two bodies sitting on shelves never been looked at in years. Nobody remembers being ripped off by Sony for all those years and never re-cooped there investment.
I do think however the Sony f65 is a great camera. Wonderful pictures from it. But does not tick enough boxes for me. Not even close.

 

[M Most]

Stuart: .
What it bugs me is why every manufacturer makes so much effort to "hide" the numbers inside "tricky" statements instead of create confidence opening their specifications.
Is the same as when i hear most of the politicians, i wonder...why in the hell they don't say the whole story? Don't they realize that we would trust them much more and because of that we would follow them?

If all companies told the "truth" about these things, it would very likely become quickly apparent that most products are far more similar than different. The differences would be obvious primarily to those who are very technically adept and inclined to pay attention to such things. For the largest part of the potential audience, though, there would be little differentiation. Marketing is all about targeting potential buyers as a group and finding ways of creating a message they want to hear and will respond to. Advertising is spreading that message in a way they can't ignore. Exaggerated marketing numbers, numbers that are essentially meaningless, and claims of superiority in specific areas are examples of potential diffentiators that marketing can grab on to.

In the society we live in, when this is done by companies it's called salesmanship. When it's done by politicians, in almost all cases it's simply lying. So there is a difference.....

 

[Mark Toia]

Remember how VHS won over Betamax.. The lessor quality of the two won the race.

Anyways..... Life goes on. All cameras a great. I love them all.

 

[M Most]

my Sony dealer was 105k. Well its good to see finally they are trying to be competitive.
The f23 & f35 were over 200k each both gathered dust about 12 month after release. I have Freinds with rental companies with these two bodies sitting on shelves never been looked at in years. Nobody remembers being ripped off by Sony for all those years and never re-cooped there investment.
.

The cost of anything at any point in time is based on supply, demand, and competitive pressures, which would include the existence of similar product or the lack of it. At the time of the F35 introduction, the landscape was very, different than it is in 2012. The fact that they had a shorter shelf life than either products than came before or than their buyers might have wanted does not suddenly turn the original price into some sort of "rip off" as you want to characterize it. The F35 w a design that was expensive to manufacture and expensive to support. That was not by design, it was because of both the industry and technical circumstances that existed at the time. I would counter that the Red One is very likely to have a much shorter shelf life than its original purchasers might have wanted as well, as evidenced by how many of them have now gone to Epic and Scarlet and how many of them are essentially being given away to production. That's simply what happens with almost any tech these days, the shelf life is shorter because technical advancement is faster. In time some might come to see the Red One as a "rip off" in the same way you're characterizing the F35. The only difference might be the dollar figure involved. A "rip off" means something that isn't worth what it's being sold for at the its being sold. It has nothing to do with whether you as an individual can afford it. I don't know your financial situation, but I would say that most people here - including me - can't afford a Gulfstream G4 jet. That doesn't make it a rip off for those who can.

 

[M Most]

Remember how VHS won over Betamax.. The lessor quality of the two won the race.

.

Yup, good example. Long running time, cheaper cassette - those were the messages the buying public wanted to hear.

Marketing and advertising at its best. Or worst, depending on your point of view.

 

[shashbugu]

by your very definition the F23 and F35 are ripoffs. Not worth the price...

 

[Nick Eason]

Note: If you doubt the power of algorithms in creating images let me cite an example. NASA sent up a space telescope that had a serious focus problem (mirror surface flaw IIRC) and came back with literally millions of soft focus images. By characterizing the error they were able to create an algorithm that actually yielded more resolution that they could have achieved had the original image been in sharp focus. Impossible you say, au contraire, because they were able to reference just how much out of focus an item should be vs how it was represented they could create a more gradual and therefore more granular interpretation. God I love math ;-)

I hope you're not referring to the Hubble Space Telescope. Because if you are, no software image processing was involved in the fix for the mirrors incorrect figure. The solution was an entirely new corrective optical assembly which, on a subsequent service mission, they sacrificed another instrument to install.

http://en.wikipedia.org/wiki/Hubble_Space_Telescope#Flawed_mirror

 

[Martin Chab]

If all companies told the "truth" about these things, it would very likely become quickly apparent that most products are far more similar than different. The differences would be obvious primarily to those who are very technically adept and inclined to pay attention to such things. For the largest part of the potential audience, though, there would be little differentiation. Marketing is all about targeting potential buyers as a group and finding ways of creating a message they want to hear and will respond to. Advertising is spreading that message in a way they can't ignore. Exaggerated marketing numbers, numbers that are essentially meaningless, and claims of superiority in specific areas are examples of potential diffentiators that marketing can grab on to.

In the society we live in, when this is done by companies it's called salesmanship. When it's done by politicians, in almost all cases it's simply lying. So there is a difference.....

I am not sure if i agree with you. The products are not so similar for starters, but also we are talking about pro equipment not consumer level. What i mean is that many professionals would do a good use of a good published specifications sheet, and the ones that can not, well they can have the rest of the sales speech anyway to convince them.
In the other hand when they exaggerate a numberless attribute (like to say: our camera has the best image quality ever), ok you can say that it is salesmanship, but when companies "tweak" the numbers, hide some others, give some more out of context or simply inform the wrong figures....well, in my world that it is called a LIE. Imagine to put together a nuclear power plant with partial specifications and wrong figures
We, consumers or pro consumers have the responsibility to let the industry to know that they are wrong, we shouldn't tolerate that.

 

[Martin Chab]

The de-bayering algorithms are the key to the final results. They are designed to accommodate the characteristics of the OLPF, angle of the bayer mask, layout of filters (RGB), etc.

Whatever combination of physical factors exist, the algorithm interprets those values to create the resulting RGB image. It is the characteristics of that derived image that matter.

It can be interesting and valuable to know the physics of the process when pushing the electro-optical system. That said, if the derived image can resolve "x" lines on a zone plate then they are in fact resolved, however you got there.

Cheers - #19

Note: If you doubt the power of algorithms in creating images let me cite an example. NASA sent up a space telescope that had a serious focus problem (mirror surface flaw IIRC) and came back with literally millions of soft focus images. By characterizing the error they were able to create an algorithm that actually yielded more resolution that they could have achieved had the original image been in sharp focus. Impossible you say, au contraire, because they were able to reference just how much out of focus an item should be vs how it was represented they could create a more gradual and therefore more granular interpretation. God I love math ;-)

I love math too but an algorithm can not create what it it not there, it can only process what it is already stored/captured/whatever

 

[Peter Moretti]

Stuart and Martin, I am almost postive that the F65 does not use an OLPF.

 

[Mark Toia]

" I would counter that the Red One is very likely to have a much shorter shelf life than its original purchasers might have wanted as well,"

" In time some might come to see the Red One as a "rip off" in the same way you're characterizing the F35."

I'm usually a huge supporter of what you say on this forum M Most... ( not sure what your first name is sorry )
So very sorry to disagree with you on this subject.

BUT... $17,500 for a 4k camera body that lasted over 4 years for me. (since I got in early) was far from being RIP OFF-ed. Ive paid more than 50k for DV camera kits & 17k for Mini DV camera kits only a few months before... I just about to pay 330k for a 435es kit.. (Thank God I stumbled on that decision)
So to pay $17,500 for a 4k camera couldn't be more of a gift at the time.
The camera alone in its first month was paid off.... after 3 to 4 years of use, I don't know how many times it was paid for... It to this day has been the most profitable camera (s) We have ever owned, and probably will be for ever more.

So to compare a $220k (HD ONLY) SONY F35... and even worse as 2/3 inche chipped SONY F23... that also only shot HD. (not to mention the price of the HDSR deck that used to drag around behind you) ... Yes.. I term these as complete rip off prices, and my rental friends that still have these cameras sitting in there shelves gathering dust 100% agree with me.
For me to understand why SONY can sell an even better camera now for only $65k.. in the F65 (that is 150k less than its lessor predecessors) is what you need to educate me about for me to think opposite about SONY taking advantage.

Now before you reply all puffed up.. remember, I am a huge fan of your forum inputs, try to be nice in your reply.

 

[M Most]

BUT... $17,500 for a 4k camera body that lasted over 4 years for me. (since I got in early) was far from being RIP OFF-ed. Ive paid more than 50k for DV camera kits & 17k for Mini DV camera kits only a few months before... I just about to pay 330k for a 435es kit.. (Thank God I stumbled on that decision)
So to pay $17,500 for a 4k camera couldn't be more of a gift at the time.
The camera alone in its first month was paid off.... after 3 to 4 years of use, I don't know how many times it was paid for... It to this day has been the most profitable camera (s) We have ever owned, and probably will be for ever more.

I said SOME might regard it that way, not necessarily you. There are a number of people I've heard characterize Autodesk Smoke as a "rip off" at $16K since it now costs $3500 - which conveniently overlooks the fact that the Mac version was itself a drastic price reduction from the existing Smoke product at the time, which hovered around $70K. I've heard others refer to DaVinci Resolve at $1000 as being a "rip off" because most of the features in it are available for free in the Lite version - conveniently overlooking the fact that a less developed version of the same software was selling for almost 100 times the price less than 3 years ago. Tech advancements change things, but the emergence of a disruptive player also often plays a part. That's what Red provided, and yes, it drastically affected the price points of other competitors. What I object to is the term "rip off" because it implies that the company selling the product looked at their potential buyers and said "how can we take money out of their pockets and give them something they'll lose money on and hate us for?" Which would be a pretty ridiculous business decision regardless of who's making it. What usually happens in an established manufacturing company is the company designs a product based on what they perceive the market as demanding, and comes up with a price based on a combination of R&D costs, materials costs, manufacturing costs, shipping costs, marketing costs, support costs, and enough of a margin to make a decent profit based on projected sales figures. Red, as far as I can see, used a very different model, one that involved coming up with a price point first and then designing a product that they could build and sell for that price point. Since they had no factory, no established infrastructure, and didn't plan on using resellers, they went a different path in terms of getting the word out, and that in turn allowed the creation of a user base that didn't care about things like missed milestones, delayed products, changing specifications, operational bugs, and entire product lines that disappeared. I'm not making any value judgements on that, but the fact is that comparing existing companies to a very unique startup like Red - and then claiming "rip off" based on the existing, widely accepted business models - is a bit unrealistic. It is equally as unrealistic to look at products that underwent drastic price reductions for reasons that are not always obvious and claim that at their previous price points they were "ripping off" their users, when in fact they were successfully selling product at those price points. Companies like Red and Blackmagic have looked at the products they're building and obviously concluded that they could greatly increase volume by drastically adjusting pricing, enough so that manufacturing costs would still be covered and volume would ultimately create greater revenues. That was not the way the motion picture business was ever looked at before, simply because such things didn't make sense in a world dominated by film and its associated costs. One could say that Red made the "cheap digital cinema camera" a reality, but one could just as accurately say that digital imaging technology made it possible to create Red. Either view would be just as accurate, so to characterize things that came before that change - a very recent one, by the way - as "rip offs" is to not be paying attention to the fact that we now live in a different technical era. That doesn't make what came before a "rip off." It simply makes it what came before.

So to compare a $220k (HD ONLY) SONY F35... and even worse as 2/3 inche chipped SONY F23... that also only shot HD. (not to mention the price of the HDSR deck that used to drag around behind you) ... Yes.. I term these as complete rip off prices, and my rental friends that still have these cameras sitting in there shelves gathering dust 100% agree with me.
For me to understand why SONY can sell an even better camera now for only $65k.. in the F65 (that is 150k less than its lessor predecessors) is what you need to educate me about for me to think opposite about SONY taking advantage.

You keep bringing up the "only HD" thing, but the fact is that prior to about 4 years ago, that was the maximum resolution available. It also conveniently overlooks the fact that regardless of resolution, F35's and Genesis cameras - as well as the even earlier F900 and F950 - have been used to make dozens of feature films - some very large ones - quite successfully, despite the constant harping here that HD is "unacceptable" for theatrical use. So it was sold as a digital cinema product, and it was used to shoot digital cinema to the satisfaction of those that used it. What difference does it make what resolution it was at the time?

As to the second statement, I would say that what is allowing Sony to sell an even better camera now for less is the same thing that's allowing Red to sell an arguably better camera for less (i.e. the Scarlet vs. the Red One) - technology advancement. Technology advancement is not limited to startups. Sony has been working on 4K imaging for a number of years. This was obvious when they were the first to deploy a 4K digital cinema projector. And it's not only imaging. SSD technology has advanced to the point where they, like Red, can use flash memory for recording, which is by definition a lot cheaper to build than videotape decks, with less to go wrong and less to manufacture. OLED technology has made it possible to make a better viewfinder for a lot less money. In short, what would be surprising is if they couldn't build a better camera for considerably less, not the fact that they can.

I think the real problem in our discussion is that you come to your conclusions based on an owner/operator model. In the world of digital cinema, that's a model that didn't really exist to any large extent until Red basically created it. The cinema market in general, film or digital, was always constructed around a rental model. This was done for a number of reasons, only one of which was the cost of the equipment. Those on the periphery were making their low and no budget projects using things like DV, and later HDV, DVCPro, and other relatively low cost but "acceptable" quality video cameras. Red came along and gave that group what they always wanted, a camera with much higher quality but an "almost affordable" price. So the model changed. But that doesn't make the model that was there before somehow evil. It was what it was because it had to be. In fact, I bought a DVX100 for about $3000 when it first appeared (as did Jarred and a number of others here). That might seem like a high price for something that shot "only standard definition" in today's world, but it wasn't a "rip off" because nothing else at the time could do 24p or anything like it. So technology is relative to the time in which it's first introduced. Now keep in mind I'm talking about "digital cinema" here, not commercials or music videos, in which there has been an owner/operator model in place for some time, even when most music videos were shot on 16mm film. But like a number of other areas of our business, those things used to be able to make money for those that made them. That is, to a great extent, no longer the case. The budgets on a lot of those projects have gone from "tight," to "low", to "zero." So even with a $1500 Canon 7D it's still not easy to make money in music videos these days. And that's the price you pay for having "cheap stuff." Not pleasant to hear. But true.

 

[Blair S. Paulsen]

I hope you're not referring to the Hubble Space Telescope. Because if you are, no software image processing was involved in the fix for the mirrors incorrect figure. The solution was an entirely new corrective optical assembly which, on a subsequent service mission, they sacrificed another instrument to install.
http://en.wikipedia.org/wiki/Hubble_Space_Telescope#Flawed_mirror

Thanks for the link, I never meant to imply that a proper fix was never done or to review the sequence of events surrounding that particular issue. What I thought was worth bringing to the thread was that when presented with the flawed image data they developed an image processing protocol that allowed them to yield far more useable information than one might expect - furthermore, certain image aspects were in fact more subtly defined than they would have been though others, as noted, were quite compromised.

In any case, I stand by my assertion that optimizing the pixel pitch of the sensor, filter pattern, OLPF and the algorithm for demosaicing is the key to turning the raw data set into the best RGB derived image. Whatever the theoretical strengths of REDs bayer implementation, or Sony's, its the derived image that matters.

Cheers - #19

 

[Gavin Greenwalt]

RED affected the price point... but so did the 5D. The reason we have all of these cameras rapidly dropping in price is in part to RED and in part to the larger ecosystem which allowed RED to release a camera using off the shelf FPGAs and a little technical know how. Commodity parts are rapidly increasing in quality and decreasing in price thanks to Moore's law. CCDs don't benefit from production like CMOS chips do. And cinema grade CMOS chips are relatively new. So yes Sony probably saw price pressure to sell the F65 for less, but they also have a lot of technology for pretty cheap at their disposal. Developing a high end tape drive like HD-CAM was a one-off niche business. Developing a SATA3 RAID pack... can be done by a small team and off the shelf parts.

 

[Mark Toia]

Thanks M Most. (still love to know what your first name is, feels weird responding to a M)

That was a very long winded reply, but logical for sure.
Just so we are on the same page. The 200k Sony HD F35 came out one year after the $17,500 dollar 4k RedOne..? Correct.
Im sure it did as I remember looking at it at Nab when they launched it and buy that time I had 3 R 1s.

Look that all siad. I still feel that Sony have been taking advantage of having no great competition for many years... Hence the inflated prices. Now there is solid competition from many angles, they seem to want to cut there prices quite quickly. And being market leaders you'd expect them to be able to make anything competively priced especially with there manufacturing arms. Compared to a company as small as Red.

Lets move on, and agree to disagree a little bit in this discussion. But I still respect your reply none the less. . Im
not saying I'm right or wrong. Like you, it's just our personal views.

 

[M Most]

Thanks M Most. (still love to know what your first name is, feels weird responding to a M)

It's Mike.

That was a very long winded reply, but logical for sure.

Yeah, sorry, I got a bit carried away. That's what happens when you're sitting at home on a Saturday with some time on your hands..

Just so we are on the same page. The 200k Sony HD F35 came out one year after the $17,500 dollar 4k RedOne..? Correct.
Im sure it did as I remember looking at it at Nab when they launched it and buy that time I had 3 R 1s.

That's probably right, but it was essentially an evolution of the Genesis, which made its appearance in 2004, I believe. And at the time, the Red was new, rather buggy, not really taken seriously by the mainstream industry that is Sony's bread and butter, and nobody in the mainstream industry really cared about anything beyond HD. And the pictures on the F35 sure were great (still are..).

Lets move on, and agree to disagree a little bit in this discussion. But I still respect your reply none the less. . Im
not saying I'm right or wrong. Like you, it's just our personal views.

Agreed. I've enjoyed the discussion, though. :cheers2:

 

[Mark Toia]

. Good on you.