FS100 vs AG20

dop16mm said:
What I am trying to get at, and it is all speculation on all fronts, is that the mp numbers are marketing bs. There has been much speculation that the famous f3 chip is not really 3.4 mp pixels but rather photosites which can also be called pixels. These photo sites are made up of 4 actual on silicone pixels in a diagonal rgbg pattern. This results in a 16x9 effective area of roughly 14mp on a 16mp 3x2 chip. You need this much detail to get true 10 bit 444 color space that the f3 is capable of.

Now there is no question that the $15k F3 has better processing than the $5K fs100 which is likely better again than the $1500 vg20, but the differences are much closer than they were just a few years ago. This is as big of a watershed as when the original vx1000 mini dv, put a lot of $50,000 betacams out of work because the average end viewer could not tell the difference. Now of course productions with money will use higher end equipment, because it is more appropriate to the work conditions, but there is no reason why an indie filmmaker with talent and no money need to step up to the next level, and still be able to produce a product that is more than watchable.

rant aside it may very well be the same chip design, but we will never know because sony won't tell us. 16mp plays in the consumer market, 3.4 for pro video. The true video only cams probably have the olpf to go with it, among other things, but panasonic managed to work some mojo with the GH2 chip, and I have no doubt that if they really want to take the consumer convergence market, sony has the resources to do it.
Click to expand...

Ok, tell me if I have this straight, cause I am trying to understand here. Your speculation, is that the VG20 chip and the F3 chip may be 'very similar' if not the same, because the F3 chip is
just calling a 'group' of 4 photosites, a pixel? What I am wondering, is if this is true, would it be possible for the F3 to take 16MP stills? I mean it obviously is NOT possible right now,
but would it be possible in theory? In other words, how does this work....how does the VG20 (assuming you are right) use the actual 'pixels' to get 16MP for stills, but then only use a
group of 4 'photosites' for a pixel for video?......and yet the F3 can't do this? To me, it doesn't make sense, that this would be the case, but I've certainly been wrong before. Not saying
you are wrong at all, I am just wondering.....In every case before, I have seen still cameras video be compromised, because they have to deal with the totally different requirements that still
folks have. But maybe technology has advanced to the point where this is not the case? I don't know, I guess we will see once the VG20 is released. I'd still put my money on 'motion' being
compromised to add still features, but after hearing what you are saying, I wouldn't bet as much as I would have before.....
 
I think Canon will compete with the FS-100 and the F-3. But, There is room for Sony to respond with one more camera between the FS-100 and the F-3. If Canon gets close to the F-3, Sony could offer slog at a better price.
 
If Canon puts a large sensor and interchangable lenses in the body of the XF300/305 or even the XF100/105, they've got a winner. It'll have the best codec of them all with the best assortment of lenses available. Big win all around.
I think eventually they'll have to do so, as the new large sensor video cameras are going to keep shrinking the video market for HDSLRs.
 
VG-20 also does 30db of gain and if the quality of noise is the same as on the FS100 then we'll know that they basically use the same sensor. We'll se once the samples hit. I kinda doubt that it's similar but you never know.
 
nex 5n uses same chip as vg20, and ships in september. Real world samples should be out soon, and it should put speculation to bed. If I was sony I would develop a chip that can be used across multiple platforms so they can be stamped out by the millions and pay for themselves quickly. Then just put different names on them for different markets. Exmor super-35 for digital cinema, Exmor aps Hd for consumer, come on. Now certainly a nex 5n is not an F3, The F3 contains many feature that are tailored to pro use, and priced accordingly, it is much more than a chip in box. But the development of that chip is expensive, and if it can be massed produced it only makes sense. That sort of business plan allows the F3 to be relatively cheap for what it is, say compared to an Arri Alexa or Red which use propriety chips in only one product. The biggest clue is that the sensor in the F3/fs100 are not 16x9 shaped, they are photo aps-c shaped. I'm sure a lot of shooters would love to be able to address the full height of the sensor for anamorphic, but it's not in the firmware.
 
"and from all counts the 5n is havin´some serious high iso glory..."

Exactly why I'm very curious how this all plays out. Here's one review reporting that the NEX 5N is beating both the Mark II and A77 in low light tests:
http://www.eoshd.com/content/3802/which-is-best-for-low-light-sony-nex-5n-or-nex-7-and-a77

I do have one question - how do we know it's the same chip in the VG20/NEX5H? Is that an educated guess based on the listed stats, or something Sony has said?
 
jdv said:
"and from all counts the 5n is havin´some serious high iso glory..."

Exactly why I'm very curious how this all plays out. Here's one review reporting that the NEX 5N is beating both the Mark II and A77 in low light tests:
http://www.eoshd.com/content/3802/which-is-best-for-low-light-sony-nex-5n-or-nex-7-and-a77

I do have one question - how do we know it's the same chip in the VG20/NEX5H? Is that an educated guess based on the listed stats, or something Sony has said?
Click to expand...

However the article unscientifically goes back to the point some of us is making "Will Sony’s decision to put 24MP in the NEX 7 and A77 backfire? It certainly seems so."
 
maarek said:
VG-20 also does 30db of gain and if the quality of noise is the same as on the FS100 then we'll know that they basically use the same sensor. We'll se once the samples hit. I kinda doubt that it's similar but you never know.
Click to expand...

The fact that it has 30db of gain is not an indication at all that it will have the similar noise-free characteristic of the FS100.

Firstly the VG10 had up to 27db of gain and comparing that to the 27db of the FS100 it is a no contest, not even close. Therefore it comes down to processing and the encode process.

I believe that the VG20's 30db will be similar to the VG10's 27db noise-wise. This would be enough to give Sony the justification to call this an improvement while not intruding on the FS100 and F3 space.
 
My impressions are that Canon's photography and camcorder divisions are highly compartmentalized than Sony's (ironically as Sony is known for that overall).

Based on this assumption, Canon would have to address organizational issues before their internal departments can willingly share technologies with one another.

As for Sony, the fact that a Cinealta technology (S35) and XDCAM (codec) technologies were inserted into the NXCAM line shows there is synergy between each camcorder level. Perhaps they are even under one roof?

Only time will tell if Canon can do what Sony unselfishly has done with their technology.
 
-Sandro- said:
...Who the hell cares about stills and RAW stills if I buy a videocamera!
Click to expand...

Me :-)

In still mode, I want high megapixel RAW images for maximum flexibility in post-processing. In video mode, I want progressive scan and modern image processing to get rid of moire. I also want a headphone jack and full control of audio -- and I want it all in one camera body because I am tired of carrying separate cameras around.

I don't care if it is wrapped in a "camcorder" or a "DSLR" form factor (or even a FS-100 style "cube"). This is the camera I've been waiting for. If they've solved the moire problem with the new processor, I will sell my GH2 and buy one.

Cheers,

Bill
 
Here are several short 1080p YouTube clips shot with a preproduction VG20 by videoaktiv in Deutschland: http://www.videoaktiv.de/Praxis+Technik/CAMCORDER-Hintergrundinfo/Sony-NEX-VG20-Testszenen-mit-dem-Vorserienmodell.html

The original files seem to be available for download. I didn't have time to download them, but I did look at them on a 42" plasma at 720p resolution and didn't see any moire on the shingled roofs or the water.

Not sure my GH2 could have done any better :-)

I'll have to see more clean video and try to get my hands on one, but I am pretty sure that this is the camera for me.

Cheers,

Bill
 
Last edited:
Hey Bill, thanks for posting that German footage! I was dead set on getting either the FS100 or the AF100, but now I'm not too sure. Why? Three things: The release of the VG20 and NEX 5N, the announcement of the $8,000 Chinese 2K camera (and of course, the Mark III is around the corner...), and the cost vs performance ratio of the FS100/AF100.

My thinking has always been to wait to the last possible second to buy my cams (I hate renting - I beat the hell out of my equipment). The AF100 continues to shrink in cost (esp used models), while the FS100 is holding on for now. Both cameras seem a wee over priced (by about a grand) given that you can buy a refurb Mark II directly from Canon for $2,000 - but I do want a true video camera form factor.

Now if the VG20 is remotely close to the FS100 for image quality, it's hard not to be interested in it, esp given it's $1,500 price tag. Sure the Chinese camera might be vapourware, and who knows, Canon might screw the pooch with the Mark III (if and when), but if those companies come through with good products in the next 6 months, then the FS100 and AF100's pricing will seem even more out of wack to me.

I know the VG20 will only have mini HDMI out, no XLR inputs, and won't have built-in ND filters - but come on, it's only $1,500. And in to me, that's a small price to pay to be able to shoot on a Super 35 sized chip via an E mount (which is very cool), while also getting "true" video camera form. The ability to take a RAW picture is even cooler - it allows a movie maker to get behind the scenes pics w/o having to have a DSLR as well as giving a wedding/event shooter a "back up still cam" in his video camera.

If it's close to the FS100/Af100 in image quality (and that's the big if right now).
 
Thank you. I've been waiting for original samples form this camera.

I'm really sorry to disappoint you (and myself) but it looks like the NEX-5n :( Better resolution than Canon but still not THAT good (GH2) and yes I see moiré.
Look:




 
Last edited:
Well, four things about that pic and the German footage:

1. Doesn't seem like too bad of moire to me - esp considering it's just about perfect conditions for the effect to happen (although granted, having used Canons for the past year and half, I might be a bit jaded). There were several shots of the blue sky which would have potentially turned my T2i's brain into jello.

2. It's a pre-pro camera - meaning there might be both physical and software changes made to the final model (which would hopefully improve image quality).

3. We have no idea what settings the shooter used. It's very rare - perhaps impossible - for a new camera right out of the box to look good. Most of these pre-pro cameras must be returned within hours after receipt - meaning a very short time to learn all the ins and outs, even for experienced shooters. Normally it takes about a month in the field for pros to zero in the best settings for any one camera. The AF100 in particular had a rough start out of the box, but in time people learned the proper tweaks and since then that cameras produced many fine images.

4. The footage came from the internet. No explanation should be needed there.
 
1- true. I've been shooting basically every week with a Canon for 1 straight year and I can see aliasing and moiré wherever its visible. I can tell from a mile which footage comes from a gh2/real HD camera, "the others" and from these new sonys. I just have eyes for resolution :) I think that for me the presence of moiré and aliasing is automatically proof of a company not taking a product seriously with cheap downscaling.
Anyway the resolution is BETTER but the moiré imho is the same, the one on the NEX-5n is pretty bad!
Look at the buildings far away on the nex-5n pic how soft they are. Typical image with details missing. Seems upscaled (and the image is not). Aliasing is visible in every line produced by buildings.

2- I REALLY hope you're right!

3- I don't really agree, if the camera has bad downscaling there's nothing you can tweak...doing something would be called a wordaround (unless point #3).

4. I don't get it :| what does it change if the original file comes from the Internet or from my camera?
 
kurth said:
and from all counts the 5n is havin´some serious high iso glory
Click to expand...

The low light capability of the NEX-5N is great for a $600 body but it's no FS100. It appears to have some aliasing and moire. Rolling shutter appears to be under control. I've been evaluating RAW 1600 ISO samples from the NEX-5N and the A77 comparing them to identical shots from the 5D, 7D, GH2, D700, and D3s. The NEX-5N sits between the 5D and 7D/GH2 for noise. The A77 was noisier than the 7D and GH2. From what I've seen the FS100 has high ISO performance well above the 5D and in D3s territory. The NEX-5N and VG20 use the same sensor as the NEX-C3 which has been tested by http://www.dxomark.com/index.php/Cameras/Camera-Sensor-Ratings
 
You must log in or register to reply here.
Back
Top