AF-100 Wishlist/Suggestion Discussion Thread

Status
Not open for further replies.
Hey Barry, I'm not trying to beat a dead horse here but I (always dead last to find out new info) just read the rumor that the Sony NEX-7 is supposedly going to have 1080p 60 and 120fps @ 720. I would love to hear your thoughts as to why implementing this in the AF100 would make it more costly. I also realize, as you rightly pointed out, that the SRW9000 is a much higher priced camcorder, B-U-T, I don't understand how a higher frame rate translates into a higher price, since so many cheaper cameras can do higher frame rates. If you're not too busy, could you elaborate more on this facet of camcorder design? It's always puzzled me.
 
... I don't understand how a higher frame rate translates into a higher price, since so many cheaper cameras can do higher frame rates.

One reason that "cheaper cameras" can have higher frame rates has to do with the size of the sensor, a smaller sensor does not overheat as fast as a larger one. Dissipating the heat generated a high frame rates was even a problem on the first RED cameras. Ask some one with a 5D about overheating when the Camera has been used to record video for long periods of time, this is one Big advantage that a 4/3 sensor has over "full frame" 35mm camera like the 5D.
 
One reason that "cheaper cameras" can have higher frame rates has to do with the size of the sensor, a smaller sensor does not overheat as fast as a larger one. Dissipating the heat generated a high frame rates was even a problem on the first RED cameras.
True, but the Sony NEX line uses APS-C sensors, which are larger than micro 4/3.
 
If Sony has figured out how to have a high frame rate with a larger sensor then I hope that Panasonic will be able to match it, because Panasonic does have great engineers.
 
Hey Barry, I'm not trying to beat a dead horse here but I (always dead last to find out new info) just read the rumor that the Sony NEX-7 is supposedly going to have 1080p 60 and 120fps @ 720. I would love to hear your thoughts as to why implementing this in the AF100 would make it more costly. I also realize, as you rightly pointed out, that the SRW9000 is a much higher priced camcorder, B-U-T, I don't understand how a higher frame rate translates into a higher price, since so many cheaper cameras can do higher frame rates. If you're not too busy, could you elaborate more on this facet of camcorder design? It's always puzzled me.
AFAIK, the SRW-9000 doesn't even do the higher frame-rates by default. It needs a $12,000 upgrade card to be able to overcrank: http://www.bhphotovideo.com/c/product/661863-REG/Sony_HKSR9002.html
 
Hey Barry, I'm not trying to beat a dead horse here but I (always dead last to find out new info) just read the rumor that the Sony NEX-7 is supposedly going to have 1080p 60 and 120fps @ 720.
Where was this info? And, rumors are rumors. However, I can point out that there are other cameras that have high frame rates too, such as Panasonic's own TM700 which does 1080/60p, and some Casio model that does something crazy like 300fps.

I would love to hear your thoughts as to why implementing this in the AF100 would make it more costly.
Because the sensor is already likely being taxed to supply 1080 at 24fps. It doesn't have a Red Mysterium X in there (and the Red sensor upgrade alone costs as much as an AF100). If you want a 35mm-sized sensor that can handle full-res 1080/60p and full-res 720/120p, the Mysterium X is the only one out there that does it. And it, by itself, costs $5750 to put in a Red One.

Then, on top of it, as said before, AVCCAM doesn't support anything higher than 720/60p or 1080/60i. So they'd have to implement another format to do it. I personally don't think that's ANY reason to not do it; after all the forthcoming HMC80 has both AVCCAM and DV in the same unit; the HPX300 has DV, DVCPRO, DVCPRO50, DVCPRO-HD, AVC-Intra50, and AVC-Intra100 all in the same unit; and the TM700 has AVCHD and h.264 28mbps modes in the same unit.

So I'm not saying they can't or shouldn't do it; instead I very much hope they do it. I'm just pointing out some reasons why they might not, some roadblocks that may pop up.

Lastly -- if there's truth to the NEX rumors, will it accomplish it by windowing the sensor? Other cameras that get high frame rates (including the Red) do so by windowing the sensor. And doing so will cause a huge drop of res on a pixel-binning line-skipping sensor like the NEX7 is likely to employ (being, as it is, DSLR-based technology).

So, let's see what it actually does. Even if it offers it, will it do it WELL? The HTC EVO 4G cell phone offers 1280x720p video, HD in a cell phone ... but it's awful, horrible video. Nasty. So does it matter that it's in a larger frame size? Heck no. So if the NEX7 offers 1080/60p that looks better than its 720/60p, that will be an accomplishment; if its 720/120p actually exists and looks as good as its 720/60p but with twice as many frames, that will be an accomplishment. Let's judge it on its accomplishments, not on rumors. Heck, as far as I know, there hasn't even been an announcement that there even will BE an NEX7... the very existence of such a product has only ever been talked about on rumor sites, hasn't it? Has Sony confirmed that there is a forthcoming NEX7?

I also realize, as you rightly pointed out, that the SRW9000 is a much higher priced camcorder
Over 10x higher, yes. The SRW9000 is a $67,500 product! And then you add a $12,000 option board and it becomes an $80,000 product, about 13x as expensive as an AF100. And, the SRW9000 is a 2/3" product. It's a lot easier to do higher frame rates on smaller chips than on larger chips.

B-U-T, I don't understand how a higher frame rate translates into a higher price, since so many cheaper cameras can do higher frame rates. If you're not too busy, could you elaborate more on this facet of camcorder design? It's always puzzled me.
There's a difference between being able to do it, and being able to do it well.

I have no doubt the SRW9000's high-frame-rate footage is impeccable. Haven't seen it, but I have no reason to suspect it's anything other than worthy of the HDCAM-SR badge. But it's nearly $80,000 and it's a 2/3" chip.

Find some high-frame-rate footage from a cheap camcorder, and it'll probably answer all your questions. :thumbsup:

And, again, even though Panasonic has a unit that does 1080/60p (the TM700) and it's very low cost, it's also a 1/4" chip. The AF100's chip is twenty-six times larger than the TM700's!
 
Last edited:
It's a lot easier to do higher frame rates on smaller chips than on larger chips.


Ahhh...that's the kind of answer I was looking for :thumbsup:

Although...it does puzzle me why the Varicam 3700 can't do 1080 60p, since it has smaller chips, but I digress. No matter which way it's sliced, I'm very excited about the AF100! Thanks Barry.
 
Although...it does puzzle me why the Varicam 3700 can't do 1080 60p, since it has smaller chips, but I digress.
Varicam 3700 uses interlaced CCD chips. CCDs have much higher power draw, more heat, and pretty much all the high-speed cameras out there are CMOS for those reasons. Plus, again, AVC-Intra and DVCPRO-HD don't support 1080/60p, so there'd be nothing for it to record to.

I think everyone is puzzled as to why the Varicam 3700 doesn't have 60p. There are reasons, but the fact remains that it's a Varicam -- it should do 60p, and they should have done whatever it takes to make it be 60p.
 
Was on a shoot with the new 2700 Varicam with fujinon lens. I just can't explain the image, it's so creamy n beautiful. It's not hvx200 it's not hpx with (name yr 35mm adapter), it's not 5D ... It's Varicam and I simply love the pictures that camera shoots.
 
I think everyone is puzzled as to why the Varicam 3700 doesn't have 60p. There are reasons, but the fact remains that it's a Varicam -- it should do 60p, and they should have done whatever it takes to make it be 60p.


THANK YOU! I couldn't agree more, I think this was one of Panasonic's biggest mistakes.
 
The 3700 does have variable frame rates so all it misses is slow motion. But how important is slow motion? When it's 10x it allows you to shoot special effects in scale or capture an action when the emphasis is on the physics (sports, etc). When it's only 2x slow motion, most hollywood films use frame duplication instead of real slow motion, it works better because it lets you see each frame for more time, and, as I would put it, it puts less emphasis on the motion and more on the emotion. True slow motion is like watching physics in action, it works well for a car ad or anything that emphasizes technology, but frame duplication works better for drama. Frame duplication actually stops time for 2x instead of showing time with 2x detail. A single photograph or even a duplication of a frame for 5 seconds can work dramatically very well, while a 1000fps shot will easily become a caricature of the subject. When the brain is given enough time it focuses on a face or an object/form as a whole. When slow motion is present it focuses on every single insignificant detail, dust, particles, wardrobe in motion, anything that moves fast from one frame to the next. Impressive but usually inappropriate.
 
WOW. This was news to me. Then again I don't have the time to read everything. Can anyone elaborate or point me to the discussion about this?

Also remember, this is at full resolution, once you bring that 3K image down to 2K or lower, your effective color sampling resolution goes up. (on the Scarlet 2/3)
 
WOW. This was news to me. Then again I don't have the time to read everything. Can anyone elaborate or point me to the discussion about this?
The Bayer sampling pattern is incapable of delivering more than 4:2:0 color resolution at full image size. If it's a 3K sensor, and you're trying to get 3K out of it, it will have at maximum 4:2:0. Problem is, the Bayer cannot deliver the nominal resolution anyway -- 4K delivers 3.2K of res, 3K mode delivers 2.4k of res, etc.

Scarlet's recording format is capable of storing 4:4:4, but the imager cannot deliver 4:4:4 so it's kind of moot at that point.

Because of all this, you can't really use an x:x:x notation to describe what it does. If you took a full-raster 4;2:0 image, you'd have 4 luma samples for every 1 chroma sample, for a ratio of luma to chroma of 4:1. In a 4:2:2 system you have 2 luma for every one chroma, a ratio of 2:1. In 4:4:4, there's one chroma for every luma, or a 1:1 ratio.

The Bayer system isn't easy to quantify. I believe Graeme said that he can extract 78% of the nominal resolution as luma, and 50% of the nominal res as chroma. Now, if you wanted to talk about it "as 3K", then its luma to chroma would be 4:1 (same ratio as 4:2:0). But if you refer to the 3K mode as "2.4k", then in 2.4k mode the relative chroma sampling is a lot higher; 2.4k of luma to 1.5k of chroma results in an effective color sampling ratio of about 2.4 to 1. That's better than 4:2:0, but not as good as 4:2:2.

But can you really extract 100% chroma resolution from a Bayer pattern, which means that for any given area of four pixels, there's one blue, one red, and two green? if the red and gblue were quad-sized pixels it'd be easy to say "yes" but they're not, and the color is synthesized by using neighboring pixels... it's kind of hard to quantify.

I don't know. And I will be glad to be corrected by Graeme if he wants to comment. By my speculation, I think the actual effective chroma sampling is somewhere between the 2.4 to 1 on the high end, and the 4 to 1 on the low end. If we split the difference, we get a ratio of 3.2:1. A little better than 4:2:0, not nearly as good as 4:2:2 -- which is what I said in the initial statement.

Note: this has nothing to do with Red or the Scarlet inherently, it's an issue with any and every Bayer-pattern sensor. If you're trying to get the full res off the sensor, this is all you're going to be able to get. We don't know what that means for something like the AF100, which is going to be reading much less than the full res, trying to get a 1080p signal off a 12-megapixel sensor. It remains to be seen what it can deliver.
 
Last edited:
Thanks, Barry. I'll read this again in the morning when my brain is fresher. Nice to get an answer that challenging. I can't wait to read Graeme's answer, but I suspect it may be awhile as he is occupied with the Red Epic firmware bug.
 
We don't know what that means for something like the AF100, which is going to be reading much less than the full res, trying to get a 1080p signal off a 12-megapixel sensor. It remains to be seen what it can deliver.
It should be possible to calculate the theoretical maximum color sampling given the roughly 6-pixel binning the AF100 will likely use.
 
Status
Not open for further replies.
Back
Top