Imamacuser
Veteran
Since the Nikon D3100 isn't full frame will the video not be 1920x1080?
Will the Nikon D3100 shoot in actual 1920x1080?
- Thread starter Imamacuser
- Start date
iHartVideo
Well-known member
Yes it will but I have no idea what full frame has to do with it.
The sensor resolution is 14mp, 1920x1080 is just over 2mp so its got more than enough capacity for HD.
The sensor resolution is 14mp, 1920x1080 is just over 2mp so its got more than enough capacity for HD.
seejay1031
Well-known member
Yes and No... but not for the reason you are suggesting. Full Frame and 1080P are totally unrelated.
The camera will shoot actual 1920 x 1080 in the same way that the 7D (not full frame) or 5D (Full Frame) do... that is, their will be that many pixels but the actual resolved resolution will be much less, and there will likely be stairstepping and artifacting that comes along with on the fly downrezzing that DSLRs all do (for now)
The camera will shoot actual 1920 x 1080 in the same way that the 7D (not full frame) or 5D (Full Frame) do... that is, their will be that many pixels but the actual resolved resolution will be much less, and there will likely be stairstepping and artifacting that comes along with on the fly downrezzing that DSLRs all do (for now)
steventimothy
Well-known member
That's speculation until someone does a resolution test - they may not be line skipping like the D90 or Canon, they may be using their ASIC to re-res intelligently on the fly.
I'm guessing Sony had a heavy hand in how the 3100 sensor was developed - even if they technically don't make them anymore - so it wouldn't be any big surprise if the quality of the 3100 and upcoming 7000 looks similar to the A33/55 or VG10 video (hopefully not the NEX-5, it has pretty bad color aliasing). Their new line of chips is better designed to be able to generate 1080 video. As mentioned many times, ALL these sensors use binning or skipping routines to get their video, though how they do it varies among manufacturers. If you're interested in the topic there are a few links out there that either show or, more commonly, guess at how the video mode is implemented (few sensor makers actually reveal their secret recipes to the public). If you're interested I can point you to some links on the topic
i think the problem is often that all debayer, nr and other manipulation is done after the scaling, using the same algorithms that are used on full sensor data, or even worse an even simpler one designed for live view purposes only. there are cerainly ways to at least debayer as part of the skipping algorithm for example.
Matt
OK, this is turning out much longer than I expected, hope I don't ramble. Some things I'm making up names for (like a superpixel, which is just a pixel that has a full set of RGB values). When I refer to subsampling it just means skipping over pixels, or binning, which I'm taking to mean the process of sampling several pixels of the same color and generating a single value for all of them before that reading gets transferred to the data pipeline
One place to look at a way chips use subsampling is by looking up a paper by Cypress for their full frame CMOS sensor, it's at:
http://www.cypress.com/?docID=18524
Pages 5-8 show the different kinds of subsampling routines available for this chip and I would think there are quite a few other chip makers that use a roughly similar routine for generating live view/video. The paper is slightly techy but the pictures are pretty self explanatory, the sensor pretty much reads out a square of 4 pixels (1R 1B 2G) then skips over a predetermined number or rows before sampling another group. This seems like a pretty basic subsampling routine to me, nothing fancy at all. This method easily produces a new smaller resolution pattern, where each new pixel in the layout comes with a full set of R,G and B values (a superpixel) so no demosaicing/interpolation is needed. Note that, with the Cypress chip at least, subsampling instructions are hardcoded onto the chip (Table 2 shows the codes required to activate a particular subsampling mode). I'm guessing that this has probably been one of the most common methods to sample sensors for video/LV in the past.
But that's just one way of sampling a sensor for video. Another method appears to involve scanning just one pixel (instead of a 2x2 group) then skipping to another single pixel farther along. This is apparently what Samsung does, and was deduced by the brilliant Falk Lumo when he was studying the Pentax K-7's video mode. By experimenting he concluded that the Samsung chip was reading every 6th pixel in video mode. A nice write-up of it is here:
http://falklumo.blogspot.com/2009/06/k-7-as-movie-camera-part-i-technical.html
Since neither Pentax nor Samsung published their methods at the time for subsampling he had to guess at the pattern being used so he came up with something like this based on the color shifts he saw in the videos:
http://falklumo.smugmug.com/gallery/8729393_oDRh3/1/#577245567_vwSxZ-A-LB
Now Falk is very smart and concluded Pentax was probably not optimizing its sensor based on the color aliasing he saw and actually proposed a subsampling pattern of his own that would make more sense:
http://falklumo.smugmug.com/gallery/8729393_oDRh3/1/#577316092_vLWjd-A-LB
Patterns like these (especially Falk's proposed improved pattern) would give you the 1536x1024 resolution image that would essentially be a checkerboard pattern; basically you'd have one superpixel with a set of full RGB values followed by a blank space of nothing, where colors would presumably have to be interpolated from the 4 neighboring superpixels. Samsung on their website has since advertised that their 14mp chip has a read 1 skip 2 mode and a read 1 skip 4 mode, which would be very close to what Falk hypothesized and Pentax may well have had a read 1 skip 5 mode custom done for their chips. Subsampling like this seems especially prone to color aliasing.
And then there's Canon, the company that has attracted the most attention WRT its sampling routines. From what I've read on the web it uses a different implementation than either of the above (the entry on this thread is a good place to dive in: http://forums.dpreview.com/forums/read.asp?forum=1000&message=33265391). It's a little complicated so I'll try a simple explanation followed by a picture (thanks to DSPographer for helping me out with the pixel spacing especially). It appears Canon scans every third line on its sensor when in video mode. On these rows every pixel is read but it appears that 3 pixels of the same color get binned together. So horizontally pixels are getting binned in trios while vertically they're getting skipped over (this explains why color moire is worse on the horizontal axis than the vertical). What this does is create a new smaller bayer pattern 1/9 the size of the original sensor, so there are no "superpixels" at all with this method. Here is an illustration of what I think is happening:
http://i73.photobucket.com/albums/i202/richtrav/Canonbinandskip.jpg
If you do the math you'll see that even the 5DII wouldn't have enough pixels to form a 1920 wide pattern, so they're probably upscaling on all their cameras to get to "1080p". I don't know how they implement their 720 mode (maybe every 5th line???) but obviously fewer pixels are getting scanned. Their VGA mode seems to be based on a downscaled version of their 720 mode (except for the special crop mode on the T2i and 60d which appears to be a full scan of the center portion of the sensor)
This makes it difficult to compare video "resolutions" of different cameras because they're all using different shortcuts to get there. The company that's probably doing it the best is Panasonic. Lord only knows how Panasonic is getting their video but it's a pretty good method, while the black and white resolution isn't much different from other cameras (which means still not close to true 1080) the GH1 can pick up fine details and textures missing from Canon and Pentax cameras.
Sony seems to be picking up more detail as well with their new sensors, if you download the video here: http://vimeo.com/14730325 you'll see at 1:02 a comparison of the VG10 with the Canon 550. The 550 certainly does curvy lines and fonts more smoothly but the VG10 is picking up detail completely missed by the Canon. UPDATE: No need to download the whole video, you can see still grabs here for the canon: http://www.point8cam.com/blog/wp-content/uploads/02.jpg and here for the sony: http://www.point8cam.com/blog/wp-content/uploads/01.jpg. I don't know how Sony is implementing their video either, you can get papers for some smaller Sony chips on their website that may offer some clues, like here: http://www.sony.net/Products/SC-HP/cx_news/vol55/pdf/imx032cqr.pdf. I would also guess that Nikon may have relied on Sony to help develop the 1080 mode for their new cameras.
Well that's it, I'm tired - time for bed - hope this didn't bore you too much. I have tried a few very informal experiments with some of these cameras, they're on my photobucket account.
Richard
OK, this is turning out much longer than I expected, hope I don't ramble. Some things I'm making up names for (like a superpixel, which is just a pixel that has a full set of RGB values). When I refer to subsampling it just means skipping over pixels, or binning, which I'm taking to mean the process of sampling several pixels of the same color and generating a single value for all of them before that reading gets transferred to the data pipeline
One place to look at a way chips use subsampling is by looking up a paper by Cypress for their full frame CMOS sensor, it's at:
http://www.cypress.com/?docID=18524
Pages 5-8 show the different kinds of subsampling routines available for this chip and I would think there are quite a few other chip makers that use a roughly similar routine for generating live view/video. The paper is slightly techy but the pictures are pretty self explanatory, the sensor pretty much reads out a square of 4 pixels (1R 1B 2G) then skips over a predetermined number or rows before sampling another group. This seems like a pretty basic subsampling routine to me, nothing fancy at all. This method easily produces a new smaller resolution pattern, where each new pixel in the layout comes with a full set of R,G and B values (a superpixel) so no demosaicing/interpolation is needed. Note that, with the Cypress chip at least, subsampling instructions are hardcoded onto the chip (Table 2 shows the codes required to activate a particular subsampling mode). I'm guessing that this has probably been one of the most common methods to sample sensors for video/LV in the past.
But that's just one way of sampling a sensor for video. Another method appears to involve scanning just one pixel (instead of a 2x2 group) then skipping to another single pixel farther along. This is apparently what Samsung does, and was deduced by the brilliant Falk Lumo when he was studying the Pentax K-7's video mode. By experimenting he concluded that the Samsung chip was reading every 6th pixel in video mode. A nice write-up of it is here:
http://falklumo.blogspot.com/2009/06/k-7-as-movie-camera-part-i-technical.html
Since neither Pentax nor Samsung published their methods at the time for subsampling he had to guess at the pattern being used so he came up with something like this based on the color shifts he saw in the videos:
http://falklumo.smugmug.com/gallery/8729393_oDRh3/1/#577245567_vwSxZ-A-LB
Now Falk is very smart and concluded Pentax was probably not optimizing its sensor based on the color aliasing he saw and actually proposed a subsampling pattern of his own that would make more sense:
http://falklumo.smugmug.com/gallery/8729393_oDRh3/1/#577316092_vLWjd-A-LB
Patterns like these (especially Falk's proposed improved pattern) would give you the 1536x1024 resolution image that would essentially be a checkerboard pattern; basically you'd have one superpixel with a set of full RGB values followed by a blank space of nothing, where colors would presumably have to be interpolated from the 4 neighboring superpixels. Samsung on their website has since advertised that their 14mp chip has a read 1 skip 2 mode and a read 1 skip 4 mode, which would be very close to what Falk hypothesized and Pentax may well have had a read 1 skip 5 mode custom done for their chips. Subsampling like this seems especially prone to color aliasing.
And then there's Canon, the company that has attracted the most attention WRT its sampling routines. From what I've read on the web it uses a different implementation than either of the above (the entry on this thread is a good place to dive in: http://forums.dpreview.com/forums/read.asp?forum=1000&message=33265391). It's a little complicated so I'll try a simple explanation followed by a picture (thanks to DSPographer for helping me out with the pixel spacing especially). It appears Canon scans every third line on its sensor when in video mode. On these rows every pixel is read but it appears that 3 pixels of the same color get binned together. So horizontally pixels are getting binned in trios while vertically they're getting skipped over (this explains why color moire is worse on the horizontal axis than the vertical). What this does is create a new smaller bayer pattern 1/9 the size of the original sensor, so there are no "superpixels" at all with this method. Here is an illustration of what I think is happening:
http://i73.photobucket.com/albums/i202/richtrav/Canonbinandskip.jpg
If you do the math you'll see that even the 5DII wouldn't have enough pixels to form a 1920 wide pattern, so they're probably upscaling on all their cameras to get to "1080p". I don't know how they implement their 720 mode (maybe every 5th line???) but obviously fewer pixels are getting scanned. Their VGA mode seems to be based on a downscaled version of their 720 mode (except for the special crop mode on the T2i and 60d which appears to be a full scan of the center portion of the sensor)
This makes it difficult to compare video "resolutions" of different cameras because they're all using different shortcuts to get there. The company that's probably doing it the best is Panasonic. Lord only knows how Panasonic is getting their video but it's a pretty good method, while the black and white resolution isn't much different from other cameras (which means still not close to true 1080) the GH1 can pick up fine details and textures missing from Canon and Pentax cameras.
Sony seems to be picking up more detail as well with their new sensors, if you download the video here: http://vimeo.com/14730325 you'll see at 1:02 a comparison of the VG10 with the Canon 550. The 550 certainly does curvy lines and fonts more smoothly but the VG10 is picking up detail completely missed by the Canon. UPDATE: No need to download the whole video, you can see still grabs here for the canon: http://www.point8cam.com/blog/wp-content/uploads/02.jpg and here for the sony: http://www.point8cam.com/blog/wp-content/uploads/01.jpg. I don't know how Sony is implementing their video either, you can get papers for some smaller Sony chips on their website that may offer some clues, like here: http://www.sony.net/Products/SC-HP/cx_news/vol55/pdf/imx032cqr.pdf. I would also guess that Nikon may have relied on Sony to help develop the 1080 mode for their new cameras.
Well that's it, I'm tired - time for bed - hope this didn't bore you too much. I have tried a few very informal experiments with some of these cameras, they're on my photobucket account.
Richard
Last edited:
very interesting. from what i've seen it doesn't seem like the d90 uses "superpixels" but rather just scales down the raw using skipping, then debayers that. bad idea for video but probably good enough for the live view it was designed for. grabbing "superpixels" and skipping will cause aliasing, but should at least give you the full sharpness and resolution.
RieniO
Well-known member
Maybe it's because of this pixel overkill being downsized that DSLR video looks so good. For example when I watch an HD satellite channel on my old widescreen TV it looks sharper than when watching a normal widescreen channel even though both channels will end up projecting the same amount of pixels on my TV screen. So AFAIK, pixel on pixel doesn't give the best image.
Emanuel
Cross-Examiner
Very interesting info. If you don't mind, I'll post it @ reduser.net and cinematography.com obviously referring the sources.
Oh gosh, thanks I guess, I hope all that info is correct! By the way there is now a comparison out between the VG10 and 7D on a Chinese website. 100% frame grabs are available to compare the two, for resolution they're very close but Sony looks to have better shadow detail out of the box and handles sharpening better. It also seems to pick up detail in fine textures just slightly better. The 7D does better with slightly angled lines
http://dcdv.zol.com.cn/190/1905828.html
http://dcdv.zol.com.cn/190/1905828.html
Samuel H
Major Contributor
ok, that explains why here the phone showed much more detail on the sony, and the magazines on the canon
on the better-conducted chinese tests it is always the canon that shows a sharper picture
Yes the focus thing occurred to me after I started looking more closely and saw that the Canon had focused just slightly closer than the Sony. But I'd have to disagree about the Sony not being as sharp as the Canon, there is an article out now that does a much more thorough analysis: http://dcdv.zol.com.cn/190/1905828.html
Resolution is virtually a dead heat, I prefer the Sony images though mainly because of their better sharpening routines and much better shadow detail out of the box (though I realize you can tweak the Canons somewhat to make up for that). Here are two comparisons between the Sony:
http://2a.zol-img.com.cn/product/50/310/ceyVEUgISscZE.jpg
and Canon:
http://2e.zol-img.com.cn/product/50/308/cechFxMj2U692.jpg
Unfortunately the framing isn't exactly the same but still, I don't see where one camera is producing more significant detail than the other. But to me the Sony image is just more pleasant looking. Hopefully Nikon's new cameras will give video as good as this mode
Resolution is virtually a dead heat, I prefer the Sony images though mainly because of their better sharpening routines and much better shadow detail out of the box (though I realize you can tweak the Canons somewhat to make up for that). Here are two comparisons between the Sony:
http://2a.zol-img.com.cn/product/50/310/ceyVEUgISscZE.jpg
and Canon:
http://2e.zol-img.com.cn/product/50/308/cechFxMj2U692.jpg
Unfortunately the framing isn't exactly the same but still, I don't see where one camera is producing more significant detail than the other. But to me the Sony image is just more pleasant looking. Hopefully Nikon's new cameras will give video as good as this mode
here is a test i found between the nex-vg10 , gh1 and 5d
unfortunately it is a very bad test , even the lenses and iso doesnt match but , you can clearly see the diffrence
http://www.youtube.com/watch?v=wdMvJBUlEW4
IMO sony had a good idea , but only sony mount with sony crap lenses and 1080i is a deal killer for me
unfortunately it is a very bad test , even the lenses and iso doesnt match but , you can clearly see the diffrence
http://www.youtube.com/watch?v=wdMvJBUlEW4
IMO sony had a good idea , but only sony mount with sony crap lenses and 1080i is a deal killer for me