RED S35 or Panasonic AF100

[Aaron Newsome]

No, I said it right. The physical shutter helps to eliminate smear.

 

[Rick Burnett]

No, I said it right. The physical shutter helps to eliminate smear.

How? The smear is caused when you overexpose a pixel, and the smearing effect happens when the sensor shifts out that pixel through the rows regardless of the shutter. It has to do with how the charge storing device of the pixel behaves.

 

[Rick Burnett]

Well in doing a little more research, I found a method being used that uses different shutter speeds consecutively to do this. It looks like it is not the fact it is physical, as electronic produces the same results. You do have to sacrifice some of your dynamic range but what I see one company doing is this:

1. Take an image with a REALLY fast shutter speed. The result is JUST the smear.
2. Take an image at the desired shutter speed.
3. Use DSP to essentially subtract out the smear.

What I don't understand is if the line is overexposed, how would there be any information beyond the maxed out pixel values? Of course, this part of the technology is secret so they didn't share that

Interesting nevertheless!

 

[Aaron Newsome]

The way I was led to understand it, was that the smear from other camera designs comes from the fact that not all pixels can be read at the exact same instant in time so if light is still falling on the sensor when the data begins to be read across the rows, smear can be introduced in this way. Since the viper doesn't begin to read the sensor until it is fully covered by the shutter, this can help to eliminate smear.

Look Rick, obviously you are waaay into this technical sensor stuff. Sounds like you know a lot more about it than I do. If you've pointed out something I said incorrect, then congratulations. The rest of the forum appreciates you clearing that up.

I wasn't trying to get in a technical design argument with you. I'll be the first to say I don't know everything. However, I didn't write anything that I didn't think to be true.

You win.

 

[Aaron Newsome]

Also, I have a habit of referring to photosites as pixels too.

 

[Rick Burnett]

The way I was led to understand it, was that the smear from other camera designs comes from the fact that not all pixels can be read at the exact same instant in time so if light is still falling on the sensor when the data begins to be read across the rows, smear can be introduced in this way. Since the viper doesn't begin to read the sensor until it is fully covered by the shutter, this can help to eliminate smear.

Look Rick, obviously you are waaay into this technical sensor stuff. Sounds like you know a lot more about it than I do. If you've pointed out something I said incorrect, then congratulations. The rest of the forum appreciates you clearing that up.

I wasn't trying to get in a technical design argument with you. I'll be the first to say I don't know everything. However, I didn't write anything that I didn't think to be true.

You win.

Hehe, no, this is complete technical curiosity and yes, I have studied CCD design. I am actually a CMOS chip designer so I do get pretty technical quickly sometimes, I gotta watch that, and yes, I constantly call photosites pixels as well I was only asking because I knew how smear was caused, as I've studied it. What I am more fascinated by is another company I found that uses two images to remove the smear. Most CCD (but not all, they are not tied together) use a global shutter. Using the physical shutter means they don't have to add in the circuitry in the sensor, which is a good thing.

Learn something new everyday

 

[Aaron Newsome]

Since you're a lot more technical with this stuff, maybe you can add to what the Thomson/Grass Valley engineers wrote:

"If the mechanical shutter is switched off, generation of electrons continues even during transport of the charged image from the imaging portion of the sensor, to the storage portion. This may (depending upon the highlights in the image) cause vertical stripes across highlights, also known as smear."

For more context, the entire document is very short. You can read it if you like:

http://www.cinematography.net/Files/About_electronic_and_mechanical_shutters.pdf

This document is what I base my understanding of what smear is, and what causes it. Maybe your understanding of what smear is and what causes it will help me expand my understanding of the topic (not that I think I would ever need to know any of this for any reason).

 

[Aaron Newsome]

This document is what I base my understanding of what smear is, and what causes it. Maybe your understanding of what smear is and what causes it will help me expand my understanding of the topic (not that I think I would ever need to know any of this for any reason).

I take that back. The info comes in handy when forum guys challenge you so I guess I better start paying more attention.

 

[Rick Burnett]

Now that I know which CCD method they are using (as there are multiple methods) I see how the global shutter works in this case. My understanding is during the shift, with how the shift works, into the secondary set of charge storage locations, this is where the contamination can occur. Basically, what they have is two sets of charge storage locations. One that can receive incoming light and one that is under an opaque blocker, like aluminum. When they position the shutter over the sensor, they shift all the pixels into the identical charge locations VERY quickly, then, they can take their time measuring the charge in these secondary locations.

I've seen other methods where they shut off the capability of the storage device for the photosite (and they call them pixels in that document, haha) but I believe those aren't as prevalent in the higher end cameras because they use up A LOT of active area. I'd imagine the they use microlens in their sensor to pull all the light from the opaque areas.

In any case, thanks for sharing all this. Very educational. Has nothing to do with S35 or AF100 since they are both CMOS, but actually answered some other CCD questions I have had.

I believe I read that some of the Arri cameras have a mechanical shutter over their CMOS. Ultimately, given the Epic has 5ms, I can only imagine them getting faster in the next few years. Silicon manufaturering is getting more precise which definitely helps with analog circuitry.

 

[David Rasberry]

David, I paid $15k for the body used. Again, I bought it as a hold over til the scarlet gets here but I've already shot a *bunch* of stuff with it. Everyone is generally very impressed with the results. Me too.

Quite a bargain I'd say. I've always liked Thompson cameras in general. Bit of a different look than Sony et al. I like the way they reproduce colors.
Never could afford one though.

Scarlet will be pretty much a hobby camera for me, though I'm sure I will find a few projects to help pay for it. I consider it more of a digital replacement for the Bolex I used to have than a video camera. Very much looking forward to using some vintage 16mm glass on it.

 

[Rick Burnett]

I take that back. The info comes in handy when forum guys challenge you so I guess I better start paying more attention.

Haha! Well I call myself a technocinematographalist I love both the technical aspect and artistic aspect of filmmaking. Have to exercise both sides of my brain.

Many times I do get into the deep technical aspect so I can understand what to expect and why. This forum is littered with people that come from complete technical (especially people who do scientific movie capture) to complete artistic not caring about anything but the image

It also helps me when buying. Had I not had this discussion with you about the physical shutter in that camera, I would not have realized all this information about that camera. Helps me make better decisions so I don't get burned like that time I bought a Nikon D90 to do video with. (And that was my bad).

 

[Rick Burnett]

I plan on shooting with my fixed and saving up enough for the S35 body and Canon lens mount. I just hope switching in and out the fixed brain and the S35 is not too painful. I am thinking I might have to put a card reader on each just to make it faster.

I LOVE the idea of having both 2/3 and S35 since I really need the 2/3 for fight scenes and visual fx work that i do. Are you all going with multi brains? Or just one or the other?

 

[Thomas Church]

^
I plan to do the same.

But in my case I expect to keep Scarlet 2/3 fixed for documentary work.

Pretty much all cases in which you'd use 16 mm in the old days.

Early on, before I get the Scarlet S35 I intend to shoot a few narrative shorts on it, too.

I'm really excited to see what kind of images the fixed lens produces; I just wish it was a bit wider, 28's not wide enough for me.

 

[Aaron Newsome]

Quite a bargain I'd say. I've always liked Thompson cameras in general. Bit of a different look than the Japs. I like the way they reproduce colors.
Never could afford one though.

Yes it was quite a bargain over a year ago. Even back then, they were going for $25k on ebay. Now, the writing is on the wall. Scarlets are coming. AF100 is coming. The DSLR's are good enough for some and a whole lot cheaper, smaller, lighter, less power hungry.

The market for this camera was done a year ago. I've seen a few going for around $10k for the body. Unbelievable really. I had a guy from another forum ask me a bunch of questions because he knew of one being sold for $10k and he was considering it. When I explained that it doesn't do any of the traditional video camera stuff, like audio, or do onboard recording, he decided to go with a video camcorder instead.

 

[Aaron Newsome]

Basically, what they have is two sets of charge storage locations. One that can receive incoming light and one that is under an opaque blocker, like aluminum.

Yes, there's actually SIX 2/3" sensors in the block. 3 are permanently covered. I suppose they would be physically situated directly behind the sensors that gather light. When you consider that there are actually six 4320x1920 sensors, which are being read a reset many many times a second, you can almost begin to understand how this camera uses an ungodly amount of power.

My JVC-100 used to run basically all day on an Anton Bauer Dionic 90. Viper will chew through one in about 45 minutes.

Advantage goes to the single sensor design, on power.

 

[Aaron Newsome]

I LOVE the idea of having both 2/3 and S35 since I really need the 2/3 for fight scenes and visual fx work that i do. Are you all going with multi brains? Or just one or the other?

I'm planning on s35. I already have a 2/3" rig

 

[Rick Burnett]

I'm planning on s35. I already have a 2/3" rig

Haha true!

Yeah, I can see how that would use up so much power!

 

[David Rasberry]

I plan on shooting with my fixed and saving up enough for the S35 body and Canon lens mount. I just hope switching in and out the fixed brain and the S35 is not too painful. I am thinking I might have to put a card reader on each just to make it faster.

I LOVE the idea of having both 2/3 and S35 since I really need the 2/3 for fight scenes and visual fx work that i do. Are you all going with multi brains? Or just one or the other?

I would eventually like both a fixed and a cinema model for concert and theatrical event documentation. If they can earn their keep, I might add an S35 or FF35 brain and still lenses eventually.
There is some experimental fine art motograph shooting I want to do, which needs the higher frame rates of the 2/3" or an Epic, but will be displayed on digital picture frames, so the 2/3" has plenty of resolution for now.

 

[Steve Wake]

OK. fuhgeddaboudit.

 

[Samual Murman]

Thank you, Steve.