More news...

[Michael Schatz]

Our OVF will use a spinning mirror (not a prism) in a completely new way. While we think the new concept is quite unique, the things we are concerned with are:

1. The expense of development
2. The low volume of sales
3. The high price we would need to charge (excess of $20K) to work within our modular system
4. Maintenance
5. Its inability to view and focus in low light
6. In ability to add a viewing LUT.

One of the things we have talked about is just to drop development and let those who need an OVF buy another camera brand. But then they lose the value of being in the RED modular system... and 5K at 100fps.

We would love to hear your thoughts.

Jim

I don't think it justifies the investment- not many people will be able to afford this "luxury". As the Rolling Stones put so perfectly 'You can't always get what you want."

PS. Sweet jet Jim!

 

[Brad Webb]

Our OVF will use a spinning mirror (not a prism) in a completely new way. While we think the new concept is quite unique, the things we are concerned with are:

1. The expense of development
2. The low volume of sales
3. The high price we would need to charge (excess of $20K) to work within our modular system
4. Maintenance
5. Its inability to view and focus in low light
6. In ability to add a viewing LUT.

One of the things we have talked about is just to drop development and let those who need an OVF buy another camera brand. But then they lose the value of being in the RED modular system... and 5K at 100fps.

We would love to hear your thoughts.

Jim

What if you build a version of Epic Brains with the OVF built in? Would that help lower the cost of the OVF? That way people who want this feature just buy the specific Epic OVF brain and you don't lose those customers.

 

[Cüneyt Kaya]

Our OVF will use a spinning mirror (not a prism) in a completely new way. While we think the new concept is quite unique, the things we are concerned with are:

1. The expense of development
2. The low volume of sales
3. The high price we would need to charge (excess of $20K) to work within our modular system
4. Maintenance
5. Its inability to view and focus in low light
6. In ability to add a viewing LUT.

One of the things we have talked about is just to drop development and let those who need an OVF buy another camera brand. But then they lose the value of being in the RED modular system... and 5K at 100fps.

We would love to hear your thoughts.

Jim

i am pretty sure the dinosaurs moving to digital would love an OVF.
but after using red over a year a OVF feels like this:
flying to the moon only by watching out of the window.
doesnt make sense for me. i love the histogram, the frameguides, the overview about my camera settings etc,
i never used the clear evf mode...but some really good dps hate those blinking lamps in the EVF and would love the OVF. but dont know how much of these are really buying cameras...hehe.

i wouldnt buy the OVF for 20k. and i wouldnt buy it for 5000.
maybe i am just too young for an OVF...hehe

 

[Anders Holck]

The viewfinder we use of te Phantom is the Sony VFC35W (viewfinder color 3.5" wide). It has a large LCD screen with a removable eyecup assembly and the LCD is on a nice hinge mount so that it can be used on its own. There are lots of controls on it including two hidden underneath that access a complete menu system of additional controls. Very nice. Native resolution of 1280x720.

Not quite. The Sony HDVF-C35W has a 640xRGBx480 delta layout LCD panel. It is a nice finder though

 

[Mike Smith]

Funny, just yesterday I explained on CML some of the advantages to having a mechanical shutter on a digital camera. This is the same thing. An optical viewfinder that uses a spinning mirror will eliminate the "jello" effect of a rolling shutter such as on the RED One. Don't know the planned read/write time of the Epic-X, but this can be a nice side benefit.

The spinning mirror for reflex viewing is also a mechanical physical shutter to the gate. When the shutter is covering the sensor, the chip can then activate the electronic shutter, and then it really doesn't matter how long the read/write time is as long as it can happen while the sensor is blocked (a very long time). Think of the read/write time to b equivalent to the time it takes a mechanical shutter's leading edge to pass from one edge of the sensor to the other. Then the whole shutter is blocking the sensor for a very long time and finally the trailing edge of the shutter reveals the sensor from one edge to another for another short time window, after which the sensor is exposed to light for a long period. On a traditional 180° shutter at 24fps, the time the sensor is fully exposed and fully covered is just shy of 1/48 of a second, while the read/write time that the edge of the shutter moves across the frame is measured in microseconds. The Phantom HD GOLD does the read/write in just a microsecond, but it is also designed as an extreme highspeed camera and it costs a pretty penn more than the RED One. The RED One takes much longer, but I cannot recall the time at the moment.

ut anyway, if a mechanical shutter is nvolved, the issue goes away. Wait until the physical shutter has blocked the sensor and then perform the electronic read/write. No more rolling shutter artifacts. This is what the D21 does and it works great. Yes there is the limitation of not being able to use a wide shutter beyond 180° (or whatever it is built to), but as on the D21, it is simple to design a way to park the mirror shutter in the open position (no light to the viewfinder) should one wish to alow more light to reach the sensor.

I could go into the virtues of an OVF for the more traditional reasons (judging lighting & contrast & color), but this is a secondary benefit that should not be overlooked.


Sorry if this isn't clear or rambles. I'm somewhere between 10:30pm and 4:30am thanks to jet lag.

I'm a bit late to this thread (just saw it), but I wanted to comment on this OVF improving rolling shutter issue...

The way I understand a CMOS sensor to work would mean that an OVF would have no effect on rolling shutter at all. A CCD sensor stores the effect of the light that hits it capacitively. Once the whole sensor has been exposed to an image as desired (1/48th of a second or whatever), the light from that image is read out a line at a time and processed/recorded. With a CCD, there is no skew because all points of the sensor are exposed to the image at the same time. With a CMOS sensor like that on the RED on the other hand, the light hitting the sensor is read out "live" a line at a time since it cannot store the effects of the light that hits it. Any time the CMOS sensor is covered by a shutter, it cannot function as a sensor and there is nothing to be read from it. It does not "store" the image temporarily as does a CCD sensor.

So a mechanical shutter will not affect this issue at all. But with a 5 microsecond scan time for Epic, this is almost totally moot.

Mike.

 

[Anders Holck]

Mike,
The cmos sensor does integrate (collect light) continuously until you read out and reset the gates.
The origin of the rolling shutter phenomenon is not that the photosites doesn't integrate at the same time, but that the readout/reset of each line is sequential.

Remember that at 24 fps, 1/48 exposure time, each frame is exposed for about 21 milliseconds. If the readout latency from top to bottom is 5 milliseconds, the whole image is still integrating in sync for 16 milliseconds.

(I'm not implying the RED ONE is using 5 MS to read out the sensor, just using that value as an example)

If you block the sensor just prior to the readout, and uncovers it after all the lines are reset of the blanked image, you will hide the readout latency (rolling shutter)

But you will then reveal the shutter blade wipe latency

 

[Jeff Coatney]

I'm a bit late to this thread (just saw it), but I wanted to comment on this OVF improving rolling shutter issue...

The way I understand a CMOS sensor to work would mean that an OVF would have no effect on rolling shutter at all. A CCD sensor stores the effect of the light that hits it capacitively. Once the whole sensor has been exposed to an image as desired (1/48th of a second or whatever), the light from that image is read out a line at a time and processed/recorded. With a CCD, there is no skew because all points of the sensor are exposed to the image at the same time. With a CMOS sensor like that on the RED on the other hand, the light hitting the sensor is read out "live" a line at a time since it cannot store the effects of the light that hits it. Any time the CMOS sensor is covered by a shutter, it cannot function as a sensor and there is nothing to be read from it. It does not "store" the image temporarily as does a CCD sensor.

So a mechanical shutter will not affect this issue at all. But with a 5 microsecond scan time for Epic, this is almost totally moot.

Mike.

The only way a mechanical shutter can interfere with a sensor is to physically block the light from hitting it. An OVF that is in need of repair may frequently block the light hitting the sensor and result in black frames or shadowed frames. The upside is, you'll know immediately if your OVF is f*cked, the downside is, who needs a headache like this?

How an OVF can go bad:

1. Mis-alignment on a micro scale can result in soft focus, if the operator is using the OVF to judge focus without then confirming the focus by looking at the actual 1:1 focus check from the sensor. In order to confirm focus, you will need to check the image from the sensor to ensure the decisions you're making thru the OVF are matching what is hitting the sensor. Remember once you add an OVF, you then have not one. but two optical pathways that must be in PERFECT alignment. "PERFECT" = $$$$.

2. The spinning mirror can ruin a take if the motor and the electronic shutter are out of sync. This could happen frequently, since they will most likely be discreet systems within the same body module. They will most likely have some control signal for syncing the spinning mirror and the electronic shutter, but the spinning mirror is governed entirely mechanically, while the electronic shutter is run by the CPU. Integrating these systems: the camera and the OVF could mean a new boardstack or a new asic or both.

3. Micro vibrations caused by the spinning mirror/ motor assembly can cause the sensor housing to shake, to an imperceptably small degree, yet enough to create visable imaging artifacts in the frame. It may be the perfect "earthquake" effect, however, so we should put that in the plus column.

4. The spinning mirror/ motor assembly could EAT through batteries like you've never seen. Good thing they'll be hot-swappable!

5. The heat-build-up from the motor and the friction of the mirror on its capstan will add to the heat generated by the normal operation of the camera and in some situations, could possibly overburden the temperature management system, resulting in a shut-down or as yet unknown repercussions, such as dropped frames.

6. Frequent servicing could result in more wear on critical set-screw threads that will result in ever-increasing servicing which will lead to... ad infinitum.

7. Has anyone explored the repercussions of an electric spinning motor's effect on the on-board audio?

8. OK, you got your spinning mirror, now how about the rest of the assembly? The eyepiece, the lenses? How could they possibly interfere with the operation of the camera or what it can be mounted to or inside of?

9. Are we really to believe that RED can isolate the development costs on this program to just the people that want it enough to pay through the nose for it? I can't believe that it won't result in at least another grand or two on the pricing of new brain modules. It may even delay the release of new brains, in order to retrofit the capability into the line.

These are just the obstacles I can think of off the top of my head.

The OVF is extinct technology in the digital world, let it die and turn into crude oil like the dinosaur that it is, please. Its like the camera's tail, do we really need it to walk upright?

The actual feedback signal from the actual sensor at 1:1 pixels and all the attendant data that is displayed in the EVF and the LCD is too much info, really? We NEED to have a dumbed-down viewfinder? That is SOooo 20th century.

If it's such a luxurious accessory, why not gold-plate the thing while you're at it?

 

[Mitch Gross]

Mike, thnk about it this wa. If it functioned as you state, then the image would be one big blur.

 

[Mitch Gross]

Jeff, with all due respect, the technology of OVF and spinning mirror shutters has effectively evolved over the past seven or eight decades. All of the issues you bring up are non-issues. I could go through each one, but trust me when state that as I read through your list I thought "no problem, no problem, no problem."

 

[Ryan E. Walters]

While I would still prefer an OVF- if it came in at +$20k, it would be a rental item for me- not a purchase item. And with the modularity of the new cameras I wonder how many of the modules the local rental houses will carry, and if the OVF will even be one of them ...

So realistically, at least for me, it looks like I'll be sticking with the LCD / EVF.

 

[Jeff Coatney]

Jeff, with all due respect, the technology of OVF and spinning mirror shutters has effectively evolved over the past seven or eight decades. All of the issues you bring up are non-issues. I could go through each one, but trust me when state that as I read through your list I thought "no problem, no problem, no problem."

Mitch, with all due respect, they may not be problems in that they are solvable, but they are problems in that they cost money and time to sort out. Which is really the point I want to make. I don't want to pay for the development on a future module.

 

[Mike Smith]

Whoa! I seem to have stirred up a hornet's nest on this one. Not my intention.

Jeff: I deliberately simplified my explanation of the way CCD and CMOS sensors differed to try (unsuccessfully?) to explain their different forms of operation. I did not mean to imply that either sensor could receive light differently to the other. Obviously any sensor exposed to light, whether behind a lens or not, will be bathed in that light all over its surface. It is in the way that each is read that makes the difference in skew. The CCD is read after it is exposed to that "all over" light. It is read without further modification to the light (in the form of electric charge) stored in each pixel. So there is no time difference between any pixel in the stored, then processed, image. The CMOS sensor on the other hand, is read in a sequence while the light is hitting it since it doesn't store charge. That scanning of the live image from frame top to bottom, is what leads to skew since each line of scan is taken at a slightly different time. But if you scan at 5 microseconds a frame as Jannard says the Epic does, this skew is pretty much irrelevant.

Mike, thnk about it this wa. If it functioned as you state, then the image would be one big blur.

Mitch, I have the greatest respect for your contributions to this forum so my bringing up these points are in no way meant as any form of attack. Just discussion. But in this case, I have to ask you what you mean here. Why would a sequential scan make the image one big blur?

 

[Ryan Gaw]

maybe its just me but seeing optically through the lens is a luxury thats worth having. I am very glad that they are giving us this option.

 

[Raphael Varandas]

wow

 

[Jeff Coatney]

Whoa! I seem to have stirred up a hornet's nest on this one. Not my intention.

Jeff: I deliberately simplified my explanation of the way CCD and CMOS sensors differed to try (unsuccessfully?) to explain their different forms of operation. I did not mean to imply that either sensor could receive light differently to the other. Obviously any sensor exposed to light, whether behind a lens or not, will be bathed in that light all over its surface. It is in the way that each is read that makes the difference in skew. The CCD is read after it is exposed to that "all over" light. It is read without further modification to the light (in the form of electric charge) stored in each pixel. So there is no time difference between any pixel in the stored, then processed, image. The CMOS sensor on the other hand, is read in a sequence while the light is hitting it since it doesn't store charge. That scanning of the live image from frame top to bottom, is what leads to skew since each line of scan is taken at a slightly different time. But if you scan at 5 microseconds a frame as Jannard says the Epic does, this skew is pretty much irrelevant.

Mike, I agreed with your initial post, I wasn't disagreeing with you. In fact, I don't disagree with anyone on the topic of why people want an OVF. What I take issue with is the perceived need to make Red Digital Cinema company offer it as an option. Even Arri doesn't take the OVF seriously enough to put into all the cameras they're offering. That alone proves its not critical to making professional digital images.

But I don't run Jim's company, I'm just a customer (#1502). I make my case and tell Jim and his team my two cents here on RedUser and I know they'll do whatever it is they were going to do in the first place.

My only real insight on this topic is that I know how big a job it is to implement the needed engineering expertise and manufacturing resources, and it costs a lot of money. I just don't want those costs creeping into the rest of the product line for when I trade in my Red One and buy my Epic.

 

[Peter Franzen]

Maybe all we need here is a system similar to a TLR.

One lens for the sensor, and another coupled lens for the OVF. Problem solved, and you're only out the cost of another set of lenses.

 

[Mitch Gross]

Mike, what I meant was that if the subject in the frame was moving and the photosites were scanning on and off in the manner you described, then the moving object would smear itself across the frame like a slit scan exposure.

 

[Mitch Gross]

Mitch, with all due respect, they may not be problems in that they are solvable, but they are problems in that they cost money and time to sort out. Which is really the point I want to make. I don't want to pay for the development on a future module.

Yes, but just like the current EVF, they could buy the intellectual property for the technology from someone else and then run with it, thereby skipping the heavy lifting in the development process.

Know what Aaton did for theie new 35mm film camera, the Penelope? They hired P+S Technik to make the optical viewfinder system for them, just like Dalsa had done for the Origin and Evolution. That's an efficient way to get the product out to market at a high quality. In this case it would be expensive to buy (as Jim warned) but not particularly increase costs to anyone else on other modules (as you fear).

 

[Sanjin Jukic]

Yes, but just like the current EVF, they could buy the intellectual property for the technology from someone else and then run with it, thereby skipping the heavy lifting in the development process.

Know what Aaton did for theie new 35mm film camera, the Penelope? They hired P+S Technik to make the optical viewfinder system for them, just like Dalsa had done for the Origin and Evolution. That's an efficient way to get the product out to market at a high quality. In this case it would be expensive to buy (as Jim warned) but not particularly increase costs to anyone else on other modules (as you fear).

Mitch,

I mentioned the fact that Dalsa asked P+S Technik to make 4K optical viewfinder for Dalsa Origin cameras almost year or more ago on this forum.

It's a right time that RED finally start to contact some of European third party camera accessories manufacturers for development of certain parts

for EPIC or Scarlet and P+S Technik is one of the best partners for that sort of things.

But unfortunately they already had "clash" with IMS mount warranty issue and I was one of those kind of a Guinea Pig victim in all that communication "mess" between the two.

And I still hope in a better future for all of us parties involved in that case.

 

[Michael Hastings]

communication generally implies a bilateral flow of information