Canon talks about this on their website yet these's little information as to how it actually works, how the cam implements it or what the effective amount of ND filtering that actually occurs.

I shot a lot outdoors so I was planning on getting an ND .6 to start but I'd like to hear about the above if anyone knows exactly what Canon is up to with this feature.

Never heard of that !?

Do you have a link ?

They talk about it here:

http://www.usa.canon.com/consumer/controller?act=ModelFeaturesAct&fcategoryid=177&modelid=14059&pageno=2

Wow. Nice sleuthing, Rich. I don't recall seeing a word about ND gradation in the HV10 owners manual.

I really wish they would write a "pro version" of the manual which fully documents the exposure parameters for every shooting mode, rather than providing vague descriptions for Joe Soccer Mom and leaving it to the hapless users to "reverse engineer" what the camera is doing via exhaustive testing.

ND gradation (and gain) isn't going to show up in the video data, so it may be tough to winkle out such obscure details except under laboratory conditions. Hopefully some of this information will turn up and be posted here.

Canon's site has support by email. Maybe someone should ask them for details.

So what does everyone think this is? Maybe some kind of negative gain setting the sensor goes into? Is this tied to that exposure adjustment dial? Canon doesn't exactly say what get's adjusted with that. If that exposure dial didn't adjust shutter (in AV) or aperture (in TV) but literally just gain that would be good to know.

But when Canon talks "ND" some kind of actual filtering comes to my mind but I'm not aware of any consumer/prosumer systems that have automatic ND filtering (then again I'm not up on all the latest stuff).

My guess is this has got to be electronic and if it can't be controlled it will have to be accounted for by trial and error (as you aptly point out Mike)

On the Sony A1/HC1/HC3 forum, Evan Donn describes internal ND filters in the light path:

I just checked my HC1 with a flashlight. The iris is wide open in the first 7 positions (starting at the far right of the onscreen exposure bar) of the exposure lever. From 8-12 the iris closes until it is about halfway closed (it's a 4 blade diamond shaped iris, by the way). From 13-19 it appears as if 2 ND filters slide in - each one stops at 4 different positions. From 20-23 the iris continues to close (with both ND's fully in place) and at 24 the iris closes completely.

... The NDs are used to allow the widest range of exposures at f.4 probably because that is the sharpest spot for this lens. R.P. Cuenco mentioned above that the service manual references 2 NDs so I think at this point we can consider it confirmed. ... Just to clarify though, if you lock shutter speed and exposure the ND's are not added automatically to prevent overexposure - the camera is perfectly happy letting you screw up your exposure completely if you wish, so you do have the ability to completely stop any automatic exposure. The NDs correspond to repeatable positions on the exposure bar.

http://www.dvinfo.net/conf/showthread.php?t=60555&page=2&highlight=filter

I'm guessing this is what the Canon HV10 is doing also -- adding steps of ND filtration mechanically, in order to hold aperture and shutter speed within a reasonable range in bright light. Since Canon has already fessed up to using some kind of ND gradation, perhaps they will provide further details about it.

I'd bet if Sony doing it Canon has a similar systems since the HV is aimed squarely at the HC1/3 from a competition standpoint. I'm glad it's in the optical path and not some sort of electronic system. Good job in finding that, hopefully we'll find out more as this cam get's wrung out. If the Canon does have a similar system I guess it's safe to assume that exposure dial has some control over it?

Some of Canon's publicity for the HV10 camcorder refers to its "patented Gradation ND system." I searched the US Patent Office for the term "Gradation ND" and found Canon's US patent no. 7,113,318 ... filed April 30, 2004 and granted on Sep. 26, 2006 (around the time the HV10 went on sale).

http://patft.uspto.gov/netacgi/nph-Parser?Sect1=PTO1&Sect2=HITOFF&d=PALL&p=1&u=%2Fnetahtml%2FPTO%2Fsrchnum.htm&r=1&f=G&l=50&s1=7113318.PN.&OS=PN/7113318&RS=PN/7113318

Haven't read all the text yet. But in the images link, Figures 9B and 9C illustrate a control scheme. Starting at 1/30th shutter speed, the camcorder reacts to brightening light by stopping down the iris from f2 to f8. Then steps of neutral density filtration are added, as the iris first dilates and then stops down again. Finally, with increasing brightness, the shutter speed is then quickened in steps from 1/30th up to 1/250th, 1/500th, etc.

I hope that someone more technically proficient than I can read this patent, and spell out for HV10 end users how to maximize mechanical control of the camcorder. For instance, I would assume that in bright light, it might still make sense to use an external ND filter, if one wanted to manually force a slow shutter speed and/or dilated iris while avoiding overexposure.

I took my HV10 out for the first time today; it was a bright, sunny afternoon, about 2:00 pm. I took some photos in Card mode and some of the same scene by snapping a photo during taping in Camera mode. I just compared EXIF info of the various shots; it seemed to keep the same f number, but vary the shutter. I'm certainly not much of a camera expert (never used a ND filter), but I thought I'd share this, in case any of it applies here. BTW, I haven't had a chance to try low light yet, but I was very impressed by my outdoors shots. :)

Jim

I think the red thing (my emphasis) is the ND filter 2 X ND + clear (??).

I think the red thing (my emphasis) is the ND filter 2 X ND + clear (??).

Right. Quoting from the patent text:

Reference numeral 151 denotes a disk-like filter, for example, an ND filter obtained by forming an ND pattern ... on a transparent resin film, for example, a PET (PolyEthylene Terephthalate) film about 0.1 mm thick by inkjet printing. ... The light attenuation pattern of the ND filter 151 is made up of the following three regions. Reference numeral 151a denotes a transparent portion having a transmittance of 100%, in which the optical density is 0 and the number of ND steps (the number of light attenuation steps) is also 0.

Reference numeral 151b denotes a uniform-density portion having an optical density of 0.6 (the number of ND steps is 2); and 151c, a uniform-density portion having an optical density of 1.2 (the number of ND steps is 4). That is, the optical densities of the three ND filters including the transparent portion are set to form an arithmetic sequence.

I took my HV10 out for the first time today; it was a bright, sunny afternoon, about 2:00 pm. I took some photos in Card mode and some of the same scene by snapping a photo during taping in Camera mode. I just compared EXIF info of the various shots; it seemed to keep the same f number, but vary the shutter.
The patent text describes a different control scheme for still shots than for video. One reason mentioned is that shutter speed can go much slower (below 1/30th second) in still mode, as compared to video. Also, it says that image smear can be prevented by using a mechanical shutter in still mode. A more complicated control scheme for still shots using the internal ND filter is described as follows:

Combinations of the densities of the ND filter and the F-numbers of aperture control at individual EVs will be explained below with reference to FIG. 11B. In the region where the EV is 2 (inclusive) to 8 (exclusive), the shutter speed lowers in accordance with each EV, but the F-number of aperture control is fixed to F2 as an open-aperture F-number, and the number of ND steps of the ND filter is also fixed to 0. That is, the transparent portion 151a shown in FIG. 2 of the ND filter 151 is used in this region.

In the region where the EV is 8 (inclusive) to 11 (exclusive), the shutter speed is continuously variable from 1/60 to 1/125 sec, the F-number of aperture control is continuously variable from F2 to F4, and the number of ND steps of the ND filter is fixed to 0.

When the EV is 11 or more, the F-number of aperture control is returned from F4 to F2, and the number of ND steps of the ND filter is switched from 0 to 2. (Referring to FIG. 11B, the switching points of the F-number and the number of ND steps are different. In practice, however, the F-number and the number of ND steps are simultaneously changed.) That is, in this stage, the uniform-density portion 151b having an optical density of 0.6 shown in FIG. 2 of the ND filter 151 is used.

In the region where the EV is 11 (inclusive) to 14 (exclusive), the shutter speed is continuously variable from 1/125 to 1/250 sec, the F-number of aperture control is continuously variable from F2 to F4, and the number of ND steps of the ND filter is fixed to 2.

When the EV is 14 or more, the F-number of aperture control is returned from F4 to F2, and the number of ND steps of the ND filter is switched from 2 to 4. That is, in this stage, the uniform-density portion 151c having an optical density of 1.2 shown in FIG. 2 of the ND filter 151 is used.

In a region where the EV is 14 (inclusive) to 17 (exclusive), the shutter speed is continuously variable from 1/250 to 1/500 sec, the F-number of aperture control is continuously variable from F2 to F4, and the number of ND steps of the ND filter is fixed to 4.

In a region where the EV is 17 (inclusive) to 19 (exclusive), the shutter speed is continuously variable from 1/500 to 1/2000 sec, the F-number of aperture control is fixed to F4, and the number of ND steps of the ND filter is kept fixed to 4. The exposure control diagram described above is applied to single-shot photographing or photographing of the first frame in continuous photographing.

The main difference between this still shot control scheme and the one for video is that the iris is held between f2 and f4, whereas for video it stops down to f8 (said to be the limit owing to small-aperture diffraction problems that develop with tight pixel pitch at higher f-stops).

Another still shot control scheme without ND filters is also described, so it is not 100% clear that the one quoted above is being used. Also, nothing in the patent specifically references the HV10 camcorder. However, Canon's HV10 publicity referring to the "patented Gradation ND" system strongly suggests that one of the schemes mentioned in the patent is being used.

Great stuff!

Nothing against Canon's new system but I think I'll go ahead and get a 37mm ND 6.

Great stuff!

Nothing against Canon's new system but I think I'll go ahead and get a 37mm ND 6.

I agree -- shutter-speed should never go higher than 1/125th nor lower than 1/30th for 60i video.

In fact, shutter-speed should never be used for exposure control!