HMC150 HD question

[Ben Tolosa]

Good evening folks,

I read on a review that the HMC150's CCD chips are 960 x 540 and they use a pixel shift arrangement to make a better image. I own a HMC150 and I went to the Panasonic website and couldn't find any information in that regard (Most likely, I didn't know how to search). I love the camera, and I was just wondering if that review is correct or not. And if so, how that translates into HD? Does those 3 chips combined make a HD image?

Thanks for the info!!

Ben Tolosa

[Denny Lajeunesse]

Barry Green has some info on that.

The HMC150, HPX170 and HVX200A all use the same chip and DSP.

Effective resolution after pixelshifting is 1440 x 810.

EDIT: Here's some info. ftp://ftp.panasonic.com/pub/Panasoni...200.CCD-WP.pdf

[Jeff Kellam]

Denny:

The imager has a 1.5 pixel horizontal and vertical offset and this makes the effective resolution 1920X1080.

Jeff

[Denny Lajeunesse]

That's not what Panasonic states in that white paper. See pages 3 and 4.

960 x 540 times 1.5 = 1440 x 810

[Jeff Kellam]

Denny, you are correct of course.

I was thinking about delivered resolution being 1920X1080.

Also, thanks for the link to the white paper. I have seen all the many discussions about the sensor by Barry Green, but never saw the white paper you linked.

Ben, I also have the HMC-150. I formerly shot using the Canon XH-A1 & JVC GR-HD1 cameras. You definately made the best choice currently available at the $3,500 price point. Don't be afraid to tweak the scene file settings. The HMC-150 is capable of great images.

[Ben Tolosa]

Hi Denny ^_^

Thank you for your reply!

Does this means that is not a true (native) 1080 HD camera?

Thanks again!

Originally Posted by Jeff Kellam

Denny, you are correct of course.

I was thinking about delivered resolution being 1920X1080.

Also, thanks for the link to the white paper. I have seen all the many discussions about the sensor by Barry Green, but never saw the white paper you linked.

Ben, I also have the HMC-150. I formerly shot using the Canon XH-A1 & JVC GR-HD1 cameras. You definately made the best choice currently available at the $3,500 price point. Don't be afraid to tweak the scene file settings. The HMC-150 is capable of great images.

Jeff,

Thank you, yes I feel I made the best choice for that price. I love it, it is a very 'cute' camera and it has a very beautiful image!

You also made an excellent choice ;)

Kind Regards!

[Denny Lajeunesse]

Yes, it is not a native 1920x1280 sensor, even with pixel shifting but it looks pretty good regardless. Actually I don't know of any 1/3" cam with a full raster CCD chip. Most only use a vert or Hori pixel shift. I think panny is the only one with both. (good thing). (HPX300 does have full raster but it's CMOS and in another league).

Downside to a small native 1080 chip would be less light sensitivity (really evident in the HMC40 though it is also a 1/4" chip).

The HVX200A and HPX170 use the same chip as the HMC150.

The HMC is the only one that records to a 1920x1080 codec. There is some argument the slight upscaling (1440x810 to 1920x1080) done in the DSP then compressed to AVHCD PH is better than how the HVX/HP cams handle this (1280x1080 codec) though DVCProHD uses less compression and 4.2.2 color and an intraframe codec. Hard to say. Color should go to the hvx/hpx. But res looks pretty close between HMC and the others.

I would argue that the H.264 compression used by AVCHD PH is more efficient but the interframe handling of compression is not as good for movement, etc.. It would be nice to have AVC-I (intra frame version) on this cam but that ain't gonna happen, at least not till the next cams come out. I suspect that's what the next HPX170 like replacement will be.


My head hurts just thinking about it. Barry has probably done the most testing in this regard.

[Jeff Kellam]

Ben:

The HDV codec, which most cameras in the $3,500 price range used until about the last year (when AVCHD came about), supports a maximum of 1440X1080 resolution (the PAR is 1.33). So regardless of if the sensors are pixel shifting or not in the HDV cameras, they are not full 1920X1080, as the HDV codec does not support it.

Beyond what Denny just posted, the only somewhat affordable ($6,100 - $400 rebate=$5,700) full raster 1920X1080 camera is the Sony EX-1R which uses a 1/2" sensor block. It's image quality is pretty amazing. Denny, Barry Green did a comparison of the EX-1R codec vs the AVCHD codec by using the new AVCHD recorder (same codec/rate as the HMC-150) and found the AVCHD codec to be a little better, although they were pretty close.

Despite all the sensor talk and pixel peeping we all tend to do more and more, all the current HD cameras put out a really good image when used properly (exposure & lighting). I think the biggest improvements still need to be in low light sensitivity & the reduction of noise. Every project I edit has no issue with the resolution, but there are always problems with noise & color.

I think there is a little too much focus on video camera resoluton nowdays and not enough on audio resolution. The video is only about 40% of the viewing experience and the audio is 60% of the experience. Few people seem to realize this. However, audio is at least as difficult if not more difficult than learning video, and I don't see the camera manufacturers jumping on audio anytime soon (think pro recorder features; limiters, low cut, filters, different audio resolutions, etc.).

[Andy Tejral]

Originally Posted by Ben Tolosa

Does this means that is not a true (native) 1080 HD camera?

IMO, this is a useless distinction that should be banned. Raster size has little to do with image resolution. Would you really want a Flip-esque device that has a 1920x1080 sensor or the HMC?

The raster size will determine the maximum possible resolution but there are so many other factors as to make that largely irrelevant.

[Ben Tolosa]

Originally Posted by Denny Lajeunesse

Yes, it is not a native 1920x1280 sensor, even with pixel shifting but it looks pretty good regardless. Actually I don't know of any 1/3" cam with a full raster CCD chip. Most only use a vert or Hori pixel shift. I think panny is the only one with both. (good thing). (HPX300 does have full raster but it's CMOS and in another league).

Downside to a small native 1080 chip would be less light sensitivity (really evident in the HMC40 though it is also a 1/4" chip).

The HVX200A and HPX170 use the same chip as the HMC150.

The HMC is the only one that records to a 1920x1080 codec. There is some argument the slight upscaling (1440x810 to 1920x1080) done in the DSP then compressed to AVHCD PH is better than how the HVX/HP cams handle this (1280x1080 codec) though DVCProHD uses less compression and 4.2.2 color and an intraframe codec. Hard to say. Color should go to the hvx/hpx. But res looks pretty close between HMC and the others.

I would argue that the H.264 compression used by AVCHD PH is more efficient but the interframe handling of compression is not as good for movement, etc.. It would be nice to have AVC-I (intra frame version) on this cam but that ain't gonna happen, at least not till the next cams come out. I suspect that's what the next HPX170 like replacement will be.


My head hurts just thinking about it. Barry has probably done the most testing in this regard.

Denny,

Interesting reading. I did not know. Thanks for teaching me ^_^

Kind regards!

Originally Posted by Jeff Kellam

Ben:

The HDV codec, which most cameras in the $3,500 price range used until about the last year (when AVCHD came about), supports a maximum of 1440X1080 resolution (the PAR is 1.33). So regardless of if the sensors are pixel shifting or not in the HDV cameras, they are not full 1920X1080, as the HDV codec does not support it.

Beyond what Denny just posted, the only somewhat affordable ($6,100 - $400 rebate=$5,700) full raster 1920X1080 camera is the Sony EX-1R which uses a 1/2" sensor block. It's image quality is pretty amazing. Denny, Barry Green did a comparison of the EX-1R codec vs the AVCHD codec by using the new AVCHD recorder (same codec/rate as the HMC-150) and found the AVCHD codec to be a little better, although they were pretty close.

Despite all the sensor talk and pixel peeping we all tend to do more and more, all the current HD cameras put out a really good image when used properly (exposure & lighting). I think the biggest improvements still need to be in low light sensitivity & the reduction of noise. Every project I edit has no issue with the resolution, but there are always problems with noise & color.

I think there is a little too much focus on video camera resoluton nowdays and not enough on audio resolution. The video is only about 40% of the viewing experience and the audio is 60% of the experience. Few people seem to realize this. However, audio is at least as difficult if not more difficult than learning video, and I don't see the camera manufacturers jumping on audio anytime soon (think pro recorder features; limiters, low cut, filters, different audio resolutions, etc.).

Hi Jeff,

Again, thanks for teaching me. I really do like the EX-1R. Before I've got my HMC-150 I was between these two. Budget detoured me to the HMC-150. Otherwise, the EX-1R I believe it is 'the' camera. However, many people said the HMC-150 is almost as good, and some people think is a bit better in some shooting circumstances.

Thanks for you post!

Originally Posted by Andy Tejral

IMO, this is a useless distinction that should be banned. Raster size has little to do with image resolution. Would you really want a Flip-esque device that has a 1920x1080 sensor or the HMC?

The raster size will determine the maximum possible resolution but there are so many other factors as to make that largely irrelevant.

Andy,

Well that, I did not know. I appreciate you can teach me.

Have a great 2010!

Regards,

[Robert M Wright]

Originally Posted by Denny Lajeunesse

Downside to a small native 1080 chip would be less light sensitivity (really evident in the HMC40 though it is also a 1/4" chip).

I seriously doubt the lower light sensitivity of an HMC40 (relative to an HMC150) has even nearly as much to do with the higher pixel count, as it simply has to do with the considerable size difference of the chip(s). Surface area is the relevant measure of chip size. When you think 1/3" vs 1/4", it just doesn't seem (intuitively) like a really big difference, but take a moment to calculate the actual difference, in terms of surface area size, and it is quite large. The ratio easily exceeds 2:1.

I've got an HMC40, and while it's a given that the HMC150 is undoubtedly more sensitive to light, for a 1/4" three chipper the HMC40 is no slouch (and the images it produces are indeed tack sharp - probably much closer to an EX1, than an HMC150, in that regard).

[Denny Lajeunesse]

Even if that chip was 1/3" it would probably still be a at least a stop or two less sensitive.

[Robert M Wright]

You know, if that were true, you would think an HMC150 should be roughly as sensitive to light as an EX1, which it clearly isn't.

[Robert M Wright]

Seems we are really all just guessing about this stuff. I'd be real interested in finding out what the hard facts are about pixel count and it's relationship to overall imaging chip light sensitivity (or if there really even is any, for that matter). I've done some Google searching, and can't seem to find any real information on that though. Does anyone know of any solid reference material about the relationship between pixel count and light sensitivity? (A white paper or something?)

[Paulo Teixeira]

This is probably the worse example I can come up with but I’ll post it anyway.
If you put 2 boxes outside with a different size hole in each box, the one with the bigger hole will get more light inside.
MOS/CMOS are obviously made differently than CCD.