HV20 vs HV10 vs JVC HD Everio

[Tony Corkovic]

Since my writing is limited to one arm(broken hand) i will ask what is better?
HV10 is great image quality, is HV20 better quality, is the new HD Everio better? i like everio since it uses HDD. opinions?

[Robert Ducon]

Have yet to see any examples.

[Ron Lemming]

Yes, all we can do is wait. Everio has smaller image sensors than HV10/HV20 however (1/5" vs 1/2.7"). We'll have to wait and see if that means something. The HV10/HV20 should be capable of a tad shorter DOF than the Everio thanks to the larger sensor anyway.

[Dave Haynie]

Originally Posted by Ron Lemming

Yes, all we can do is wait. Everio has smaller image sensors than HV10/HV20 however (1/5" vs 1/2.7"). We'll have to wait and see if that means something. The HV10/HV20 should be capable of a tad shorter DOF than the Everio thanks to the larger sensor anyway.

Ultimately, it's "seeing is believing"... but there are some things you can examine. The Everio has three CCDs in an offset configuration... this means that you get three 960 x 540 pixel CCDs, with the green offset 1/2 pixel in each direction from the red and green. This improves light gathering, but also affects color much in the same way a single CCD's Bayer pattern does -- there's some interpolation to do, and potential errors as a result. In HD, the effect of this is dramatically less noticable -- most 3-chip HDV cameras on the market do pixel offset with interpolation (some don't even yield a true 1440x1080 or better even after interpolation, of course).

The light gathering ability in either case is directly related to pixel size, somewhat related to noise in the system. CCDs tend to be noisier than CMOS today, at least if you're running very current CMOS hardware (CCD hasn't changed much lately, it's a mature technology, CMOS sensor technology is still evolving quite quickly). But some CMOS chips leave significant space between sensor sites, so there are unknowable differences between a 1/3" CMOS and 1/3" CCD sensors... until you actually test. CMOS chips do generally have better dynamic range, usually a useful 14-bits, versus 12-bit on a CCD.

So the JVC claims to have a 3.28um pixel pitch, while the HV10 has a 2.75um pixel pitch. Based on that only, I would expect the JVC to have superior low-light performance.. but again, it's also CCD vs. CMOS, and the JVC could be noisier. I know the 3-chip consumer Panasonics, with 1/6" SD sensors have a bad rep for low-light, and the 3-chip Sony HVR-V1 has 1/4" HD (offset) sensors and a reputation for marginally acceptable low-light performance (in fact, ANY HDV camcorder is going to look bad up against something like the Sony VX2000/VX2100 on low-light... as do many SD cams).

Some things are certain.. this one will deliver REAL 1920x1080 HD, since in recording to HD, there's no need to stick to the bounds of the HDV format (which limit 1080i to 1440x1080). They're apparently recording MPEG-2, not AVC, so that's about 5 hours at HDV quality, maybe a bit less at 1920x1080. That's fine for many uses, but not necessarily all that geat for others.

[Robert Batta]

JVC HD7 OIS simple bad
http://www.listvideo.com/details.php?image_id=124

[Rafael Lopes]

It sucks that JVC decided not to go with progressive. They're walking a very thin line with business. Where the heck is panassonic in all this?!

[Tony Parenti]

Originally Posted by Dave Haynie

Ultimately, it's "seeing is believing"... but there are some things you can examine. The Everio has three CCDs in an offset configuration... this means that you get three 960 x 540 pixel CCDs, with the green offset 1/2 pixel in each direction from the red and green. This improves light gathering, but also affects color much in the same way a single CCD's Bayer pattern does -- there's some interpolation to do, and potential errors as a result. In HD, the effect of this is dramatically less noticable -- most 3-chip HDV cameras on the market do pixel offset with interpolation (some don't even yield a true 1440x1080 or better even after interpolation, of course).

The light gathering ability in either case is directly related to pixel size, somewhat related to noise in the system. CCDs tend to be noisier than CMOS today, at least if you're running very current CMOS hardware (CCD hasn't changed much lately, it's a mature technology, CMOS sensor technology is still evolving quite quickly). But some CMOS chips leave significant space between sensor sites, so there are unknowable differences between a 1/3" CMOS and 1/3" CCD sensors... until you actually test. CMOS chips do generally have better dynamic range, usually a useful 14-bits, versus 12-bit on a CCD.

So the JVC claims to have a 3.28um pixel pitch, while the HV10 has a 2.75um pixel pitch. Based on that only, I would expect the JVC to have superior low-light performance.. but again, it's also CCD vs. CMOS, and the JVC could be noisier. I know the 3-chip consumer Panasonics, with 1/6" SD sensors have a bad rep for low-light, and the 3-chip Sony HVR-V1 has 1/4" HD (offset) sensors and a reputation for marginally acceptable low-light performance (in fact, ANY HDV camcorder is going to look bad up against something like the Sony VX2000/VX2100 on low-light... as do many SD cams).

Some things are certain.. this one will deliver REAL 1920x1080 HD, since in recording to HD, there's no need to stick to the bounds of the HDV format (which limit 1080i to 1440x1080). They're apparently recording MPEG-2, not AVC, so that's about 5 hours at HDV quality, maybe a bit less at 1920x1080. That's fine for many uses, but not necessarily all that geat for others.

Actually it's not REAL 1920x1080 if it's achieved via interpolation and the CCDs are 960x540. I'd much rather have the TRUE 1920x1080 CMOS sensor in the HV20, even if I lose 480pixel in downconversion to HDV.