HMC150 1080 vs. HPX170 720?

strangways

strangways

Well-known member
I am trying decide if an HMC150 is good enough for what I need, or if I'll have to pony up for an HPX170.

I greatly prefer 4:2:2 color sampling over 4:1:1, and presumably also 4:2:0.
However, because of recording times, I would only shoot 720p with the HPX170, but would gladly shoot 1080p with the HMC150.

So I've been doing some thinking:
4:2:2 chroma subsampling: the chroma resolution is halved horizontally.
4:2:0 chroma subsampling: the chroma resolution halved both vertically and horizontally.

Which means:

HPX170 960×720 luma / 4:2:2 subsampling = chroma 480x720
HMC150 1920×1080 luma / 4:2:0 subsampling = chroma 960x540

Theoretically giving the HMC150 twice the horizontal chroma resolution, but 25% lower vertical chroma resolution.

I've been hearing a lot that people are saying that on the HMC150, shooting 1080 is not giving much of a detail increase over 720... but, is it giving increased chroma resolution if I shoot 1080 and output 720 from my editing software?

Wanted I really want to avoid are those ugly stair-step jaggies I get on my DVX100 om edges with high color contrast, like a bright red shirt on a dark background.
Does anybody have a single frame of HMC150 footage at 1080p that would show that, but disappear when converted to 720p?
 
Last edited:
Which means:

HPX170 960×720 luma / 4:2:2 subsampling = chroma 960x720
HMC150 1920×1080 luma / 4:2:0 subsampling = chroma 960x540
Click to expand...

On the HPX-170, the 960x720 with 4:2:2 has 480 x 720 chroma resolution.

With the 1080 mode, the HPX-170 has 1280 x 1080, with a chroma of 640 x 1080.


Also at the 720p mode, the HMC-150 has a chroma of 640 x 360.


Bob Diaz
 
I completely missed converting the horizontal. I've edited the post and corrected it! Thanks for pointing it out, Bob.
 
I found this from Barry, and it raises a good point I hadn't considered: both 1080 and 720 get compressed to 21 mbps, which could matter far more than color subsampling.

Barry_Green said:
I think the 150 does just fine in 1080p. I just think it excels further in 720p (meaning, the picture is sharper in 1080p, but not dramatically so, whereas in 720p the compression is a lot milder so; viewing it as a 1080p camera, it does well, but viewing it as a 720p camera it stomps.
Click to expand...
 
You have to be aware that the figures obtained by multiplying footage resolution and chroma downsampling factor don't mean much more than... upscaling. The actual chroma resolution is limited by the sensor resolution - 960x540 for both HMC150 and HPX170. Even though pixel-shifting can usually resolve more luma samples, chroma won't benefit from this technique.
In this context - the HPX170 chroma resolution of 640x1080 (when footage is captured at 1280x1080 at 4:2:2) is only theory. You should get similar actual chroma resolution by shooting at native sensor resolution (960x540 - if the camcorder was capable of it) and upscaling to 1280x1080 4:2:2 with proper chroma smoothing (filtering).
 
So Trez, it sounds like you're saying that the chroma resolution of both cameras are limited by the 960x540 resolution of each CCD, which sounds right to me.

If the HMC is using pixel shifting (and possibly upscaling) to get the recorded 1920x1080 luma, and the 4:2:0 chroma subsampling is recording 960x540 chroma, that would be the best it could possibly be getting from the CCDs anyway.

Therefore, in conclusion, I just stop worrying and go get an HMC150 to start shooting with. That sounds good to me!
 
strangways said:
If the HMC is using pixel shifting (and possibly upscaling) to get the recorded 1920x1080 luma, and the 4:2:0 chroma subsampling is recording 960x540 chroma, that would be the best it could possibly be getting from the CCDs anyway.
Click to expand...

Yes, exactly. But the codec bitrate might be insufficient when shooting some fast-moving stuff with HMC at 1080p, just as you mentioned.

By the way - it is now well-known fact that HMC150 outperforms HPX170 at 720p - but this was quite a surprise when first users did side-by-side tests and noticed this.
Of course, AVCHD 720p has better raster resolution, but people thought the limited AVCHD coded bitrate would become a bottleneck. However, they forgot that HPX170's DVCPRO HD codec bitrate at 720p was only 40Mbit/s (compared to AVCHD 24Mbit/s). Considering the fact that AVCHD is much more advanced and efficient and that DVCPRO HD carries a lot of redundant chroma samples in 720p, this shouldn't be a big surprise, actually.
 
Last edited:
I'd certainly be happy with 720p, so that's great news, Trez!
I've been going through the forums, but some of the threads have gotten so huge that I missed side-by-side tests.

I've tried searching, but... you wouldn't happen to have any links handy, would you?

I often work with a company using HVX200 and 200a cameras, and would love to be able to show them that if I got an HMC150, it would be at least as good if I gave them footage transcoded to DVCProHD, and even a little better if I kept it in AVCHD.
 
strangeways

someone says he can tell the difference in color rendition between 170 and 150 (170 being better) - post #5:

http://www.dvxuser.com/V6/showthread.php?p=1480807#post1480807

He says heavy color correction reveals the difference. This is the first time I heard about it and I actually think there would be not difference provided 150's 4:2:0 chroma was properly upsampled to 4:2:2 before cc.
 
trez said:
You have to be aware that the figures obtained by multiplying footage resolution and chroma downsampling factor don't mean much more than... upscaling. The actual chroma resolution is limited by the sensor resolution - 960x540 for both HMC150 and HPX170.
Click to expand...

I must admit this is not true. Jan from Panasonic convinced me that subpixel-shifted sensors of 150 and 170 are capable of resolving more chroma details that 960x540. So having 4:2:2 on 170 definitely makes sense. It can significantly improve greenscreen compositing. This has been discussed here:

http://www.dvxuser.com/V6/showthread.php?t=154436&page=4
 
trez said:
Yes, exactly. But the codec bitrate might be insufficient when shooting some fast-moving stuff with HMC at 1080p, just as you mentioned.

By the way - it is now well-known fact that HMC150 outperforms HPX170 at 720p - but this was quite a surprise when first users did side-by-side tests and noticed this.
Of course, AVCHD 720p has better raster resolution, but people thought the limited AVCHD coded bitrate would become a bottleneck. However, they forgot that HPX170's DVCPRO HD codec bitrate at 720p was only 40Mbit/s (compared to AVCHD 24Mbit/s). Considering the fact that AVCHD is much more advanced and efficient and that DVCPRO HD carries a lot of redundant chroma samples in 720p, this shouldn't be a big surprise, actually.
Click to expand...
Where's the post of this comparison? Would like to see it as well.

Thanks
 
You must log in or register to reply here.
Back
Top