hpx170 or HMC150

A

arrestthisman

Well-known member
I wasn't sure which forum in which to post this. Simply put, I'm in the market to buy. (Low light/noise on the HVX200 is too much for me.)

Going to do mostly personal projects. I know the lowdown and have done a ton of reading on the two, but it's hard to find a summery or comparison between the two, in any meaningful way.

I know pros/broadcasters would benefit from P2, DVCPRO HD, and HD-SDI. I know the difference between 4:2:2 and 4:2:0. I don't need variable frame rates. I don't need SD/backwards compatibility. Cheap media would be awesome.

In terms of image quality, I know they use the same lens/improved CCD chipset. Can anyone speak to the "real world" difference between 4:2:2 and 4:2:0 in terms of quality. Keying, I know, but what about color correction? Will it really make much of a difference? What about the codec itself. I saw a post where Barry spoke very highly of PH 720P/24.

Any little bits of wisdom would help me greatly.
 
"I know the difference between 4:2:2 and 4:2:0."....then later....."Can anyone speak to the "real world" difference between 4:2:2 and 4:2:0 in terms of quality."

i don't know. You were sure you knew already, but now you don't? what do you want to know? have you seen side by side footage? if so, can you tell a difference? if not, you may want to do some more research. if you can see a difference, choose the one you like best.


"i'm not an a$$hole, i just play one in real life." -skinny
 
skinny! said:
"I know the difference between 4:2:2 and 4:2:0."....then later....."Can anyone speak to the "real world" difference between 4:2:2 and 4:2:0 in terms of quality."

i don't know. You were sure you knew already, but now you don't? what do you want to know? have you seen side by side footage? if so, can you tell a difference? if not, you may want to do some more research. if you can see a difference, choose the one you like best.


"i'm not an a$$hole, i just play one in real life." -skinny
Click to expand...


Where contained in this post, which you chose to write, is there any information intended to be helpful?

To be more clear. I know the Academic difference between different chroma sub-sampling rates, and I have seen side by side footage. I cannot tell the difference in most cases outside of keying or magnification of saturated edges. To be fair, and don't laugh too hard, I'm slightly color blind (red green deficient).

Obviously if there were a giant difference then most companies would not bother doing research on subsampling codecs to save space or optimize compression; still I don't know if there are real world differences in things like color correction between the two spaces. I'm not so arrogant to think that nothing could get past me.
 
Have you considered workflow?

AVCHD versus easy to edit DVCProHD?

And for me the color space difference is significant.
 
Thanks for your input David. Yes, I have considered it. I have premiere CS4, which I've used for animation mostly, and FCP. I can just use CS4 for live action now.

That said, I've been looking high and low for comparisons of 4:2:0 versus 4:2:2. That's the big matzo ball for me. All I can find are graphs, and when I do side by side presentations in quicktime, it looks like some smoothing is being applied so it's very difficult to get a real comparison. I'm thinking, now that I'll take a 4:2:0 and a 4:2:2 clip into FCP, and magnify so there's no processing of the footage, or at least very little. I'm going to do that now.

I did notice an hour ago looking at HVX200 side by side with EX1 that there seemed to be a noticeable difference in saturated edges. But the video was encoded in H.264 so even the HVX footage had been taken into 4:2:0.

David, what specifically when viewing the 4:2:2 footage gives away its space compared to 4:2:0. Like I said, I have an academic understanding, but real world manifest is often very different, and I haven't worked in many different color spaces enough to be able to recognize the glaring difference (if any).

Thank you.

UPDATE: OKAY, just took a look in FCP, and I'm feeling like a total moron. There is NO CONTEST. I had always been comparing with viewers, and hadn't really done it in an NLE since the old days in DV. That's a long time.

At any rate, there is no debate in my mind 4:2:2 is beyond superior. It's beyond a night and day difference. I can't believe after all this time... oy. Well... The choice is clear, HPX170, ALL the way. I can see why the HVX has made it into so many theaters. I still can't believe it. It wasn't until like 800% magnification that it was so obvious. Because I do a lot of effects work, this is compulsory.
 
Last edited:
arrestthisman said:
Thanks for your input David. Yes, I have considered it. I have premiere CS4, which I've used for animation mostly, and FCP. I can just use CS4 for live action now.

That said, I've been looking high and low for comparisons of 4:2:0 versus 4:2:2. That's the big matzo ball for me. All I can find are graphs, and when I do side by side presentations in quicktime, it looks like some smoothing is being applied so it's very difficult to get a real comparison. I'm thinking, now that I'll take a 4:2:0 and a 4:2:2 clip into FCP, and magnify so there's no processing of the footage, or at least very little. I'm going to do that now.

I did notice an hour ago looking at HVX200 side by side with EX1 that there seemed to be a noticeable difference in saturated edges. But the video was encoded in H.264 so even the HVX footage had been taken into 4:2:0.

David, what specifically when viewing the 4:2:2 footage gives away its space compared to 4:2:0. Like I said, I have an academic understanding, but real world manifest is often very different, and I haven't worked in many different color spaces enough to be able to recognize the glaring difference (if any).

Thank you.

UPDATE: OKAY, just took a look in FCP, and I'm feeling like a total moron. There is NO CONTEST. I had always been comparing with viewers, and hadn't really done it in an NLE since the old days in DV. That's a long time.

At any rate, there is no debate in my mind 4:2:2 is beyond superior. It's beyond a night and day difference. I can't believe after all this time... oy. Well... The choice is clear, HPX170, ALL the way. I can see why the HVX has made it into so many theaters. I still can't believe it. It wasn't until like 800% magnification that it was so obvious. Because I do a lot of effects work, this is compulsory.
Click to expand...

Arrest, if you haven't read this, you may find some operational information to be helpful once you buy the 170 http://www.kenstone.net/fcp_homepage/hpx_170_evolution_brockett.html

Dan
 
Puredrifting: Thanks for the link. I have been to that site before, but never fully read the article. I'm going to check it out carefully.

Everybody thanks for the input. I guess the only thing left, as a follow up if nothing else; is what I'm seeing accurate? As I said above the difference under magnification is night and day. Is that about right, or am I missing something, or what? I didn't realize the difference was so drastic, and I've never seen that before, so I just want to make sure what I'm seeing NOW is correct.

Thanks.
 
arrestthisman said:
... when I do side by side presentations in quicktime, it looks like some smoothing is being applied so it's very difficult to get a real comparison...
Click to expand...

Comparing the two only makes sense with smoothing applied! This is digital sampling - looking at the footage without chroma filtering (smoothing) is like listening to the digital audio without DAC reconstruction filter. It simply doesn't make any sense. The entire sampling thing assumes that you smooth the discreet values on output. By the way - smoothing should applied to the 4:2:2 footage as well... All the video players, either software or hardware, are capable of smoothing.
The only way of comparing that makes any sense is to look at the footage in its ultimate format (media) - all in all this is what you're shooting for. You don't shoot for FCP, which only provides quick, low quality preview, but for MPEG2 or H.264.
There's more - If you can't tell the difference between 4:2:2 without smoothing and 4:2:0 with smoothing then why waste the codec bitrate for twice as much chroma samples if you can get the same result by simply smoothing in post?
 
Just a quick note.

You will notice a significant different if shooting on a chroma screen between 4:2:2 and 4:2:0. If you are shooting green screen this difference will only really be noticeable by a professional compositor it will make a subtle difference but one that will make the composition a believable one.

However, shoot on blue screen and this is a different story. You have lost so much bit depth from the blue channel that there is very little information with which to do a chroma key properly. With today's keyers this is becoming more and more possible but you will have to do a great deal of noise reduction, thus softening the image considerable, before keying with a multitude of luminance and 3D keyers if available to you, probably some roto and then a sharpen and regrain.

An investment should last - get the best you can.
 
trez:

It's important to me to get the best native signal possible, because on a film-out you need as much color/rez/quality as possible. I'm not working on anything at the moment which will likely require a filmout but I like the idea of having something RTG in the case that I do.
 
arrestthisman said:
trez:

It's important to me to get the best native signal possible, because on a film-out you need as much color/rez/quality as possible. I'm not working on anything at the moment which will likely require a filmout but I like the idea of having something RTG in the case that I do.
Click to expand...

My point was that these cameras don't get enough chroma resolution from the sensor, which is 960x540. So, even if DVCPRO HD raster size can be 1280x1080 (640x1080 chroma) some of the samples it carries are redundant (chroma must be upscaled in camera) - so this is not an actual increase in chroma resolution from original 960x540 and will not result for more precise keying etc. You could get the same in post by simply filtering native sensor (960x540) chroma.
HMC150 working in 1080p has chroma resolution of 960x540 (1920x1080 at 4:2:0) which is the upper limit of the number of chroma samples that the sensor can resolve.

thinkoffline - interesting point on blue vs green chroma keying. not sure what you mean by 'loosing bit depth' from the blue channel, though - does it have anything to do with the weights three color channels contribute to the Y'Pr'Pb' luma (Y) channel (green being 59%, while blue is only 11%) so that keying with green allows for setting more precise luma masks?
 
You must log in or register to reply here.
Back
Top