AF-100 Wishlist/Suggestion Discussion Thread

Status
Not open for further replies.
Can we get a cup holder, and maybe the kitchen sink?

(Sorry... this is the 1001 post in this thread, and it seems like everything but the kitchen sink has been asked for a 10 times over :) )
 
Barry_Green said:
That's because the 7D is designed for stills, and has a video mode grafted on. Its stills resolution is the priority. The video mode delivers barely 720p of resolution. If you used an appropriate OLPF for the 7D, you would end up with a 720p camera, that has to record in 1080p mode to get that 720p.

With the AF100 we're talking about something that's taken into the design shed by the video engineers, starting from the perspective that stills are irrelevant. This unit will be tuned to deliver the best video image they can make. I fully expect HPX370/EX1 sharpness out of it.


Nobody knows, outside of the engineers inside Panasonic Japan. You'll get people making guesses, even educated guesses, but that's all. Nobody actually knows.
Click to expand...

They may very well be building the system based off on video, that doesn't mean there aren't some physical limitations to the sensor design and OLPF that aren't still going to have to be decided if they are using the same silicon as the GH1 is. Sure, you can put a stronger OLPF on the front and get rid of all aliasing, but then, you make the image so soft, it defeats some of the gains.

This is the issue I am worried about with this binning approach. Even if you pixel bin, over a relative repeatable area, you still get a TON of area that is not used for sampling. Like the example above. Let's say the sensor has 9 pixels and you use 2 for red. You are throwing out 80% of the red information in that area. That means, whatever OLPF you use, to minimize aliasing and moire, you've got to spread out the spacial content into those sampled areas just the same.

In other words, given you know how the GH1 sensor looks, if you increase the OLPF for the target resolution, it's going to look softer because the false detail that provides false sharpness is going to essentially blur out a bit more, it has to.

Even the RED ONE still has aliasing, that is easy to see. However, when I bring the footage down to 1080p it is perfect. We won't have that option with the AF100. Every piece of footage from the GH13 I have seen with sharp lines has a lot of aliasing in it. What I am worried about it that the AF100 is going to look great downsampled to 720p, but will have an OLPF that is either to much that softens too much, or not enough, and the aliasing is still going to be bad.

From a technical standpoint, it would seem that the better thing to do would be build pixels that do not require binning that are larger or, even better, if you want to maintain the noise rejection capability, just change your filter array so that all pixels around each other are the same color, then you can get the noise rejection capabilities that multi active areas provide. So you'd have blocks of 9 pixels the same color, and of course, it still being a bayer pattern when all binned.
 
Well, let's keep in mind that the best and brightest minds from the best engineering schools in the country are working at these companies. I am pretty sure that dedicated sensor engineers who do this for a living are thinking through all the possible ramifications and we'll just have to see what they come up with.
 
Hehe, well yes, I to a point I agree.

Just remember, engineering doesn't get to decide the world, unfortunately. Marketing will sell you the most amazing product that ever existed, and finance will smack down engineering when things take too long, or start running high in the cost. In the 15 years of microelectronics work (as I work with CMOS everyday) I can tell you I see this all the time. Especially when you don't want to miss your marketing window!

Things like this (analog) take trial and error. Mainly because fab practices are anything BUT accurate. When you have gate thicknesses being a few atoms thick coming along, it goes without saying there is A LOT of variance in processes. Many times, companies are relegated to the modeling the fab provides, and if that is not accurate, issues can come about. Most designs take this into consideration, I've seen variances that were WAY off and caused huge half year delays on product deployments. Mainly because it's not just refabbing a part, it's determining what is wrong that can also take time as it's hard to analyze CMOS designs.

What I hope is that they took whatever they have learned from the GH1 sensor and reworked it somewhat into a whole new sensor. Any gains will be surely welcomed (especially read/reset speed).

Nevetheless, we shall see! At least after 1000 posts they know what we all care about! :) Haha.
 
Barry will be there to put the camera through hoops when the prototype comes out.. It better be as resolute as an ex1 and nothing less
 
Is it possible that with a 4/3 sensor and binning (or line skipping) Panasonic could do what fuji did with their SuperCCD sensor, which pairs high- and low-sensitivity photodetectors in each pixel there by increasing dynamic range. Instead of skipping every other line or binning photodetectors to arrive at 1080p out of a 4K sensor. Why not assign one photodetector to be optimized for highlights and the other "Binned" photodetector to be optimized for for shadows, this could create a image which could closely replicate or surpass the dynamic range of film.
 
Yak said:
Is it possible that with a 4/3 sensor and binning (or line skipping) Panasonic could do what fuji did with their SuperCCD sensor, which pairs high- and low-sensitivity photodetectors in each pixel there by increasing dynamic range. Instead of skipping every other line or binning photodetectors to arrive at 1080p out of a 4K sensor. Why not assign one photodetector to be optimized for highlights and the other "Binned" photodetector to be optimized for for shadows, this could create a image which could closely replicate or surpass the dynamic range of film.
Click to expand...

I'm pretty sure this is what Genesis and F23 / F35 do to achieve their dynamic range.
 
Erik Olson said:
I'm pretty sure this is what Genesis and F23 / F35 do to achieve their dynamic range.
Click to expand...

From what I have seen on the Genesis design was that two separate gain stages are used from each pixel and that the results of the two different gain stages AND A/D converters are added back together in the digital domain. If they do what you say here, it sure makes a lot more sense to me now, as it never made sense before taking it from the same pixel.
 
Mike Harvey said:
Can we get a cup holder, and maybe the kitchen sink?

(Sorry... this is the 1001 post in this thread, and it seems like everything but the kitchen sink has been asked for a 10 times over :) )
Click to expand...

My apologies in advance if this is the 11th time someone asks this one, but: how will the Olympus 4/3 lenses that are servo assisted work in manual focus mode (like my loved 14-54 2.8-3.5 ASPH 1st version) with the AF-100?
 
I think it is safe to say that with the EX-1, HPX370 & the new Canon 300 series, no camera that gets put out after is going to blow these exisiting models away. They are full raster 1080p CMOS chips. Any gains are going to be incremental with detail, light gathering and image quality.

It all depends if this new type of compact video camera will be a fit in your toolbox. Imho, we have already passed an image quality barrier that will stand for quite some time compared to delivery options.
 
Last edited:
There is no one camera that can fit everybody's needs.

Yes, if people need a video lens for video style shooting, the EX1 / EX3 or HPX370 will be a much better deal than the AF100.

But those are not the people using the HDSLRs. They are different cameras for different shoots.

The AF100 is a digital cinema camera.
 
agcohn said:
The AF100 is a digital cinema camera.
Click to expand...
Tell that to the wedding videographers or people who shoot low lit concerts.
This camera basically have an ENG type body for crying out loud. I'll be used for a lot of different things. I don't like repeating that but it's true.
 
It is being marketed as a digital cinema and professional video camera. Frankly, I would expect it to be quite popular with wedding shooters too, who need the functionality of a video camera but want the artistry a digital cinema camera provides.
 
agcohn said:
There is no one camera that can fit everybody's needs.

Yes, if people need a video lens for video style shooting, the EX1 / EX3 or HPX370 will be a much better deal than the AF100.
.
Click to expand...

No, none of those cameras suit my needs. I see the AF 100 as the next step up from a 150. I'm hoping they offer a lense with a power zoom at a good price, but it's not a deal breaker. I'll be using it for broadcast.
 
As much as I want this camera to be the next step up from the HMC-150, I do not know if that is what it will be. This role is largely based upon a servo driven lens with a wide aperature. What I have used my HMC-150 for could not be done with prime lenses.

I think there is a real opportunity here to get more sales with a servo driven lens as this would be a nice upgrade to the HMC-150. This camera could really shine for all kinds of event and corporate work.
 
Camera Expert said:
Tell that to the wedding videographers or people who shoot low lit concerts.
This camera basically have an ENG type body for crying out loud. I'll be used for a lot of different things. I don't like repeating that but it's true.
Click to expand...
I'm sure people will use the camera for ENG, just like how many people are now shooting video on still cameras, but the AF100 is clearly not being designed for ENG work.

If it were, they would have used 2/3" chips to allow people to use HD Video lenses.

ENG cameras really do require a one-size-fits-all servo-zoom lens, and that lens for 4/3" chips is going to mighty expensive when and if it exists.
 
As said before, Panasonic makes something like a dozen ENG-appropriate cameras. This one doesn't have to have anything to do with ENG, to still be a success in the marketplace.
 
Carlos_Nieto said:
My apologies in advance if this is the 11th time someone asks this one, but: how will the Olympus 4/3 lenses that are servo assisted work in manual focus mode (like my loved 14-54 2.8-3.5 ASPH 1st version) with the AF-100?
Click to expand...

m4/3 is backwards compatible with the 4/3 standard. All lens functions that work on one 4/3 camera will work on any other. One nice feature of an open standard. m4/3 has the same electronic pin set in its mount plus a couple of new ones. Just takes a mount adapter.
 
Status
Not open for further replies.
Back
Top