Discussion of dv.com results

Barry_Green said:
Horizontal (as measured in the vertical trumpets):
MTF:---100% 70% 30% 10%
H720:--450... 480. 500. 520
H1080:-530... 550. 600 720

Vertical (as measured by the horizontal trumpets):
MTF:---100% 70% 30% 10%
H720:--525... 550. 630. 650
H1080:-540...600. 630. 660
Click to expand...

I don't even understand how that's physically possible.

Now if I understand correctly, 720p should have a max of 720 vertical lines of resolution, correct?(If I'm wrong, then the rest of what I say is bunk. :grin: )

At the 720 setting, the HVX exhibits a maximum resolution that is far below what the 720p codec is capable of(720x960).

So how is it possible that the HVX can have a higher resolution on the 1080 setting? If 525x450 lines is the maximum the HVX can pull off, then 525x450 should BE the maximum resolution of the HVX, period.
If the resolution of the HVX is below the specs of both the lower and higher resolution formats, then shouldn't the resolution chart exhibit the same numbers in both 720 and 1080?
How can the highest resolution of both the 720 and 1080 formats be different when it's coming from the same CCDs and lens?
 
I can't, I'm sorry! :-(

There are plenty of nice clips out there though.

The things I love are the colors, there very rich... The HVX has a great film like image.

Now I can't confirm this, however in this footage I didn't see any ccd smearing. There were car lights in the frame and it didn't do it. Not saying that it doesn't, however it didn't in these clips.

I used to own the DVX100, sold it off to my friend Barlow. That was my favorite of the small form professional cameras. The HVX is this, however now HD and all the goodies that come with going to HD. So knowing I love the look of the DVX, it's a look I like in the HVX as well.

Also I can't explain it, the 24 temporal rate seems more on spot with the HVX. I don't know, I have not been able to put my finger on it. More research needed.


Michael Pappas
Arrfilms@hotmail.com
PappasArts & Arrfilms Main site

CONTACT VIA AOL INSTANT MESSENGER
AT { PAPPASARTS2 }


XLH1 and HVX200 frame grabs and news here:
http://www.pbase.com/Arrfilms

http://www.PappasArts.com
http://www.Myspace.com/PappasArts



Kholi said:
I guess you aren't going to say anything about those seven awesome clips tha tmake you want to go ahead and order the HVX200?
Click to expand...
 
The only possible thing I can think of is that the 1080 mode of the HVX is employing the spatial offset system with the green CCD and the 720 mode isn't. :huh:
 
Mr. Blonde said:
I'm much less a fan of the numbers game than I am of what my eyes can see in an image.

Comparison footage pleeeeeeease!
Click to expand...
I know, I know most people are. Unfortunately, what we got won't satisfy that request. The entirety of the footage is charts, except for the one dynamic range test. No motion, no dramatic setup, just torture testing. And that's not what people are going to want to see. I think eyes are going to turn toward The Texas Shootout, the one that Mike Curtis and Chris Hurd are putting together and that Adam will also be administering. In that one they have three days to put things through their paces. And they'll have plenty of advance notice to get a qualified Canon expert there, and they'll have time to set up a variety of shots. I knew we weren't going to get a ton of stuff done, but I'm disappointed we got as little as we did; I hoped to have at least an outdoors scene and an indoors scene done as well, but it just wasn't possible.

If I had one suggestion for the Texas Shootout guys, it'd be to drop the VariCam and the F900 from their comparison. Slim down to the four cameras that people are actually looking at buying. It may pique the intellectual curiosity to see how close (or not) these cams come to the big boys, but saddling the testing with two extra cameras is going to seriously inhibit their ability for getting a wider variety of tests. I'd love to see some footage come from them, some footage that really helps people evaluate these products in the type of circumstances that forum readers anticipate using them in.
 
FatBird19 said:
How can the highest resolution of both the 720 and 1080 formats be different when it's coming from the same CCDs and lens?
Click to expand...
Because the codec subsamples to 960 pixels across when recording 720p.

The vertical resolution is quite comparable, because the total image fits closely to within the full vertical height of both formats.

But in horizontal the HVX has 1280 pixels to encode (in 1080 mode) but only 960 pixels to encode (in 720 mode). It's the prefiltering/subsampling that is causing the resolution difference between 720p and 1080p. The live uncompressed footage is a lot closer, but in the recorded footage there's a noticeable difference.
 
Well, I want the HVX because of its variable framerates in 720p. We knew this version was not the ideal 1080p camera yet. But in 720p I think it holds up very well. There is also more to a camera than the resolution charts in the studio: motion, colour, sensitivity.
A point&shoot digital still cam can match a Canon SLR in resoltion maybe, but what image does look better in print?

I think these two quotes are exactly what I expected from this HVX200:
"If we agree that the Panasonic, like the Sony, is a 550 TVl/ph camera, then it's very nicely matched to 720p DVCPROHD recording, but a bit on the coarse side for 1080i recording. In other words, it'll be a closer match, resolution-wise, for the Varicam than it will be for 1080i cameras."

and
"The HVX200 was not the resolution winner, but it did a very good job on skin tones, handling highlights cleanly. It showed less noise than the Canon by a slight amount, and the noise was more colorful: more film-like according to Barry, more like color-under analog to my eye. This is one of those subjective things where taste is as important as numbers"

So if it holds up against the Varicam in 720p and brings the DVX-look in a P2-DVCProHD workflow, then that is exactly what I personally expected.
 
AshG said:
How about some real world perspective? I have been playing with both the XLH and the HVX for over a week. While I agree in general with most of Adam's findings I can tell you that the XLH, like the XL2, does not perform its best with the stock settings which were used. Setting it to NR2 and a low knee dramtically reduces noise and increases the range. It has less noise than the HVX in all equal environments, period endstop.

The resolution seems on par with what I am seeing. 24F looks exactly like 24P, no difference in motion. While a rez chart does show a reduction in rez in 24F, this is not percieved while watching... weird... it still is as sharp or sharper as an HVX.

The HVX is still pretty awesome, great colors (prolly the reason for the chroma noise) great range and very portable.

As I predicted before touching either camera or seeing a single test, we are back where were were with the DVX/XL2 argument... Both are good enough, one has notably more resolution, one a nice color curve. One is more expensive but has a longer slightly better lens. One is a shoulder cam, the other a truly anywhere anyhow cam...

After messing with both I think either is good enough for most.


ash =o)
Click to expand...


basically it is as Barry said in the beginning: the H1 is a grown up XL2 and the HVX is the grown up HVX. Well, they are still juvenile, but they are grown a little....
 
I just did a comparison of the frame grabs (Scene, not chart) from the XL- H1 and the HVX. I sharpened both and the H1 certainly has more fine detail, but it's not a huge difference, in fact I'd say it's less than the difference between a 6 megapixel and 8 megapixel digital camera. And that difference is only noticeable in extremely controlled environments and very large prints. I think, Adam is right, the cameras are much more silmilar than it seems from the charts. There are parts of the HVX grab that seem to show as much detail as the H1, like the camera and tripods in the background, but other parts like the chart letters, that seem less. It's hard to say with certainty but there may be other issues like heat that have an effect on focus, I know this was true when I used the HVX for my Argent shoot. The focus would slip very slightly when the doors were opened and the set cooled off Vs. when we were shooting with the 5K not ten feet away. I will be getting my HVX in this next shipment so I will be doing my own subjective, real world resolution test to determine what I can and can't do with this camera.

Out of curiosity what is the chat rez of the DVX?
 
After 17 pages of replies it's time I threw in the my wrench

The mojo. The magic vibe. Surely one camera has it over the others ?

The DVX100 has the mysterious mojo that everyone see's; the 24p, the latitude, the contrasts, something.............................................

Who cares what camera has the sharpest rez; isn't that negilble at best anyway's ?

What camera has the mojo ?

No, the XL2 cannot be tweaked in post to look like the DVX100 and forget about the FH and FX's and the XKLL100's.....

So tell us; wat camera has the mojo magic ?
 
That's the real question, isn't it? I can tell you this, the H1 looks sharp, so does the JVC but the JVC has real problems with blown highs, they roll off into a strange mauve tone.
The H1 just clips too harsh for my taste, maybe not any sooner than the HVX but definately harder, that's hard to quantify but at almost twice the price it's not twice as sharp. I'd have to say the results stand just the same...
Sharpness... XL-H1
DR... HD100
Noise... Z1U

Mojo... HVX200
 
The perfect way to get your Mojo Camcorder is to learn to use the one you have. My opinion is that we don't get any Best camcorder before the next generation of little camcorders. This FX, Z1, XL-H1, HD100, HVX200 are just a better one but not the final solution for those with a tiny budget.
 
John_Hudson said:
After 17 pages of replies it's time I threw in the my wrench

The mojo. The magic vibe. Surely one camera has it over the others ?

The DVX100 has the mysterious mojo that everyone see's; the 24p, the latitude, the contrasts, something.............................................

Who cares what camera has the sharpest rez; isn't that negilble at best anyway's ?

What camera has the mojo ?

No, the XL2 cannot be tweaked in post to look like the DVX100 and forget about the FH and FX's and the XKLL100's.....

So tell us; wat camera has the mojo magic ?
Click to expand...


go for dvx100b !! :)
you will not be disappointed.
and HD? David Lynch shot his last movie with a poor PD150.....
 
John_Hudson said:
My question was purely subjective in nature

I'm sticking with my DVX100 (Original).
Click to expand...

i understood that ( i m reading this forum for a while ;) ).
i was recommanding this to those who are going to spend a lot of money in a technology that is not mature.
 
Adam Wilt comments

Adam Wilt comments

Adam doesn't participate in forums in general, but after seeing the reactions and suppositions that have been made, he forwarded an e-mail and told me it would be okay to post it. So, without further adieu...

----------------------------------------------------------
AJW: We did get native-format footage from all the cameras, but I have no
way to play back the Canon footage until I get access to another XL H1
(the footage may be in 24f mode; my Sony decks refuse to recognize the
tape), so I thought it best not to go into detail about it. Looking at
the 10-bit uncompressed footage and the HDV- or DVCPROHD-native footage
for the other three cameras, there was nothing in the native footage
that changes any of the numbers.

Furthermore, as has been noted, we didn't have *time* to test for
motion and time-dependent artifacts; we only had time for static-image
tests. As HDV is long-GOP and the performance of the codecs is
extremely dependent on motion and scene complexity, there wasn't really
any value to be derived from comparing off-tape pix for a few static
setups; native recording comparisons will simply have to wait for more
real-world testing on complex, moving scenes.

As a result of these factors I really wanted my article to focus on the
camera heads more than the recorders; I wanted to describe how well
each camcorder did *as a camera*, since the camera itself is a very
important link in the production chain.

All the stills are pulled from the 10-bit uncompressed files (except
the two side-by-side Z1 resolution trumpets, which were pulled from the
HDV footage). If folks see some blocking in the shadows, that's a
result of my compressing the pix for the web. Blame me and my settings
in Photoshop, not the cameras.

We *did* tweak the Canon; it wasn't "just pulled out of the box" as
some are saying. We just didn't tweak it as much as we would have, had
we more time with the camera.

No, we didn't dive into the XL H1's matrix settings (unparalleled on
cameras of this class, and a very strong reason to consider the Canon),
but then, we didn't tweak colors specifically on *any* of the four
cams. Folks trying to pick winners on color rendition based on the
sample frames should give up: all the cameras (even the Z1) have
considerable ability to modify their color palette, but that was
something we didn't even try to do: if we had, we'd still be there, two
weeks later, A/Bing pairs of cameras, running to tweak 'em, arguing
about the vectorscope and screen images...

Of course, picking winners wasn't the purpose of the test at all,
something I tried to point out (with, apparently, mixed success). And
even then, we looked at only a very few factors determining the
pictorial outcome of using these cameras--no ergonomics; no
practicality for a given situation; no frame-rate flexibility; no
how-wide-does-the-lens-go; no audio quality tests; no
can-I-handhold-it-steadily-for-90-minutes; no low light tests; no
complex motion rendering; no flare levels; no how easily can a 1st AC
pull focus on it; no how easily can I pull focus on it while running
'n' gunning; no codec quality--so how can anyone say that resolution
numbers or noise levels or a stop more or less sensitivity makes or
breaks a camera? That's just silly. We captured four or five data
points out of the hundreds you need to think about. There are still
hundreds of data points to go.

Let's put it this way: by my reading of the images, the HVX200 "loses"
in resolution and in the visual appearance of its noise. If it's such a
loser, why do I still have one on order? Mightn't it be the case that
other factors offset the numbers? Might the phrase "variable frame
rates" have some resonance amongst the current and former film shooters
in the crowd?

By all means, look at the stills and footage posted (and yes, I'd be
happy to supply all the 10-bit uncompressed files and not just
one-second DV-compressed excerpts: just tell me who wants to volunteer
the 200 Gigabytes of online storage to host it, and I hope everyone has
a very, very fast Internet connection to view it all, grin). Don't just
debate my interpretation and Barry's interpretation: see for yourself.

And by all means, do your own tests. Get together the best array of
cameras, subjects, and operators you can, capture as much information
as you can, and put it up on the 'net for all of us to see. I'm not
challenging folks ("we worked really hard and if you don't like it, see
if YOU can do better"), I'm encouraging folks ("we worked really hard;
you work really hard too, and we can all learn from each other").
Barry's test was never meant to be the be-all and end-all test, it's
just *a* test. The more tests we have, and the better-run and more
comprehensive they are, the better off we'll all be.

But better yet, grab any one of these cameras, go out, and shoot
something beautiful and persuasive and thought-provoking and outrageous
and emotionally resonant with it. Comparative numbers are fine for
geeks, but viewers don't care; they want *stories*. Never forget that.

Cheers,
Adam Wilt
 
Thanks to Adam for responding! (As well as for all the testing, of course.) Interesting that compression wouldn't change the numbers, even for the still images... a lot of stills on the web seemed seemed to have shown significant softening. Ah, to have one to play with...
 
ullanta said:
Oh, let the fools buy their HVXs, how else will the technology mature?
And how will we?
Click to expand...

it was just a personal recommandation ..... for those who are wondering what to buy.
if you want to buy a pink ferrari go for it, i don t care.
but a lot of people were defending the hvx 6 month ago when no footages had been seen.
there is some kind of consumption crazyness. "I NEED a hvx200!!!!".
i just say "don t forget the dvx100".
this technology will be mature when sensitivity, noise, lenses, size, chroma fringing, workflow, ..... prices, will be improved. i m just waiting for the next generation.end of this year or next one.
 
Barry_Green said:
By all means, look at the stills and footage posted (and yes, I'd be
happy to supply all the 10-bit uncompressed files and not just
one-second DV-compressed excerpts: just tell me who wants to volunteer
the 200 Gigabytes of online storage to host it, and I hope everyone has
a very, very fast Internet connection to view it all, grin). Don't just
debate my interpretation and Barry's interpretation: see for yourself.
Click to expand...

I could host it and provide pretty fast download, but would have to charge 50$ per download ;) Yep, bandwidth is not free :(
 
You must log in or register to reply here.
Back
Top