L.V. Boulevard stuff

[Barry_Green]

Wanted to see how the HVX would hold up against the FX1 in the blacks and dark environments. People keep asking about noise in the shadows and whatnot... So I picked up my friend Jay's FX1 and strapped 'em both to a Matthews monitor platform, and carried the rig down the Las Vegas Strip. I figured the low light and black sky of night should make for some interesting discovery.

For settings: the HVX had all my normal favorite settings; cine-like-V and cinematrix, master ped -4 and chroma level +2 and coring +3. I like harsh contrast and a deeper black level, and these settings really deliver it.

On the FX1, I really haven't used it since a year ago, so I don't know the settings all that well. Accordingly I used it in its defaults (no CinemaTone gamma or anything). All settings were zero except for Sharpness, which was at the default of 11. Factory defaults across the board. I put the HVX detail level at 0 to make it roughly comparable.

For the most part both cameras were running at wide open, and I made a halfhearted effort to match the zoom and framing to make them at least vaguely compositionally comparable (although that's no easy feat; keeping two cams balanced and level and pointing the same way on a platform strapped to your chest by a neck strap, while fighting the L.V. Boulevard crowd on a Saturday night... well, let's just say that some of the footage isn't exactly level, and sometimes the white balance switch got moved when I didn't want it to... it was a crazy thing to try to do!) I did have both irises in auto-iris, because with flames going off and whatnot, and no way to ride both irises on the fly when the cams were held at arm's length above my head to get above the crowd, well... might as well let autoiris have a shot at it.

The images end up looking so wildly different, there's almost no point in comparing them, but hey, it's what I got so here's what it is. All I had to go by was the LCDs, so I didn't know how differently they were set until I got back and imported the footage.

Please note: no attempt was made to match the cameras at all. Only thing I did was shoot CineFrame 30 on the FX1, and 30P on the HVX, so that they would cut in the same timeline and post clips without interlace artifacts.

Absolutely no modification of the footage was done in post, except for to convert their aspect ratios to make the final .mp4 and .wmv files. No color correction of any type, etc.

I shot a few things that were completely useless, but I've pulled together three clips that show what I was looking for -- black handling. There's a shot from the Mirage in front of their main hotel, just a dolphin statue.

HVX on the left, Sony on the right.
http://www.fiftv.com/HVX200/HVX-n-FX1-Dolphin-Small.jpg

Looks like the Sony has its black level lifted some, whereas I specifically lowered the blacks on the HVX. Full-size extraction is here:
http://www.fiftv.com/HVX200/HVX-n-FX1-Dolphin.jpg

The video clip is the first half of this file (the pirate ship explosion is the second half):
http://www.fiftv.com/HVX200/HVX-vs.-FX1-on-LV-Strip.wmvx
or
http://www.fiftv.com/HVX200/LVStrip1.mp4x

Those are both the same footage, just a different format for PC vs. Mac users (remove the "x"; it's there to keep people from streaming the files and using up my bandwidth).

Then there's a shot from the Treasure Island hotel's "Sirens of T.I." show, where they do some pyrotechnics for the cannon fire. (which is the second half of the video clip above). This shot is really interesting because when the explosions went off, the Sony's HDV compression just puked. Totally punted; it looks like maybe 320x200 res at the most. Check out this extraction:
http://www.fiftv.com/HVX200/HVX-vs-Sony-Ship-Explosion-Extract.jpg

Look at the macroblocking and the low res on the Sony shot. That's what happens when HDV runs out of bandwidth. I'll freely admit that under most circumstances the Sony (and HDV in general) looks great, but there's no way in :evil: that I'd want to have that happen to my footage. I just don't trust HDV; your mileage may vary.
Full-res JPG from the uncompressed timeline is here:
http://www.fiftv.com/HVX200/HVX-vs-Sony-Ship-Explosion.jpg

As for the color difference -- man, I don't know -- I'm thinking the white balance got toggled on the HVX which made everything go redder. There was a crowd of literally 2,000 people there, and I was holding these silly cameras over my head when the shot went off, so... sorry for the non-thoroughness of the testing, it was hardly scientific! :shocked: Just grabbing what I could...

These last shots are of the volcano at the Mirage.
http://www.fiftv.com/HVX200/HVX-vs.-FX1-on-LV-Strip2.wmvx
or
http://www.fiftv.com/HVX200/LVStrip2.mp4x

[hvx_germanboy]

thanks barry!
exciting as always when you post something!
Great!

[Stevet]

Barry, the HVX200 looks great here.

Interesting, I do not see any noise or compression artifacts in this low light example.

Thanks for posting

[Ed Kishel]

why not set the HVX at the default factory settings like you kept the FX1 to better level the playing field?

[Emanuel]

Ah Barry, you are unique...

[Jim Arthurs]

Thanks Barry, more "Bob Diaz Data Points". :)

The noise looks good, and the artifical edge sharpness is much better on the HVX (i.e. gone)... who knows about the black level on the FX1, so I'm not factoring that into the mix. I do like the feel of the colors of the HVX.

One point that stands out for me... the FX1 is using a single 540 line field stretched up to full frame for its fake progressive CF30. The HVX is using true full-frame progressive. If you compare any frame (I like the pirate ship windows, or rope rigging) you don't see much if any resolution gains from the HVX. Very, very similar in overall detail.

Again, as I've said before, lack-luster 1080 performance from the HVX. This is a good 720p camera, not a 1080 one. From strickly a resolution point, if the HVX was on the rig with the Canon in its fake progressive, you'd be gob-smacked by the difference in rez between them.

I'm an HVX guy for all the reasons listed elsewhere, but will consider it an SD/720 camera only.

Regards,

Jim Arthurs

[Rodrigo]

hmmm I see a good deal of rez difference in the first image

[Emanuel]

Again, as I've said before, lack-luster 1080 performance from the HVX. This is a good 720p camera, not a 1080 one. From strickly a resolution point, if the HVX was on the rig with the Canon in its fake progressive, you'd be gob-smacked by the difference in rez between them.

I'm an HVX guy for all the reasons listed elsewhere, but will consider it an SD/720 camera only.OK I need a cam to 35mm film-out@big_screen, therefore 1080 is a preferential prerequisite. Forget the price range and I have three options:
1) Panasonic DVCPRO-HD's HVX
2) Canon HDV's XL-H1
3) Sony XDCAM-HD's PDW-F330

What will it produce@big_screen better image?

[Justin_Kirch]

very good experiment Barry. thanks for posting.

[Jim Arthurs]

OK I need a cam to 35mm film-out@big_screen, therefore 1080 is a preferential prerequisite. Forget the price range and I have three options:
1) Panasonic DVCPRO-HD's HVX
2) Canon HDV's XL-H1
3) Sony XDCAM-HD's PDW-F330
What will it produce@big_screen better image?

I've seen no footage from the new PDW-F330 to compare with the above cameras. Of the big three 1/3" cameras, the Canon will pack the most pixels on the screen. I believe it to be the only real "HD" camera of the three (Sony Z1/FX1, HVX, JVC).

Yet... if the choice was really that simple... then consider your overall usage of the camera. Do you need really really good SD? Do you need slow motion? Do you need the best codec for the highest motion quality on-board the camera with no add-ons? Is the Canon zoom lens easy for you to use, or is the HVX better for your tastes? And on and on.

For my own use, I'm really not conflicted about which to buy. I need good SD for green screen material, and the HVX is very, very good for that. I'm okay with 720p being "good enough" for most of my clients who need HD. I want some stop-motion/time-lapse options. Viva-HVX!

Now, If I wasn't buying soon, but in six months, and there's some sort of Firestore sized HD-SDI flash/harddrive recorder for the Canon that is inexpensive and records to Cineform directly, and the option to buy just the body and pick up a manual lens, I would think much harder about the Canon.

But for now? The HVX is the perfect bridge to the promise of Red, and will even serve beyond that for SD work. It makes pleasing, luminous pictures.

What I'd like to see, is a resolution chart compare of recorded HVX 720p to 1080. We haven't seen that yet, at least to my knowledge. I don't think 1080 has enough extra rez to off-set the increase in bandwidth over using 720.

Just my two cents. I've been thinking long and hard about all this since the big camera shoot-out, and didn't want to post before spending some time thinking...

Jim Arthurs

[Jim Arthurs]

hmmm I see a good deal of rez difference in the first image

The situation is less clear if you view the movies (pardon the pun). Yes, there's more detail in that one still, but for the vast majority of the side-by-side comparision, they are just about equal. Maybe the auto-focus on the FX1 hadn't quite kicked in yet at that point in the recording?

I'm pulling for the HVX when it comes to resolution, as it SHOULD have more than the crappy 1 field worth of 1080i that the Z1/FX1 has.

Barrys points about motion handling with the codecs, and the noise levels are smack on, though... I have no more concerns about the noise floor on the camera. NBCShooter did a test in another thread with some excellent answers to this whole situation.

Regards,

Jim Arthurs

[Barry_Green]

why not set the HVX at the default factory settings like you kept the FX1 to better level the playing field?
Because like I said, there was no attempt to level the playing field at all. I was interested in the blacks/dark shadow performance.

I just brought the FX1 along because, well, I had it, and why not?

[mochouinard]

I like that you actually showed the HDV at it true form. It is unacceptable, but for home user it is ok. The huge problem with those artifact, is once they get to a video station to be edited and then recompress for DVD or Digital TV broadcast, it will only get worst.

I understand why station do not want HDV. But I do not understand why canon made the XL H1 targeted for TV broadcast(IMHO) with HDV on them.

Ok there is always the SDI-HD available, but we arnt in the under 10k anymore...

DVCPRO HD is good ennuf for me. I am very happy Panasonic made a camera with affordable DVCPRO HD on them. And when I say DVCPRO HD, I only mean frame independant codec. I would still preferer a loseless codec.

Thanks barry for those footage.

[penfever]

Well, I don't care about all of those crazy things - I'm just happy there are more pretty pictures to stare at. Ah, the IQ is fried after a day in the City.

[Ed Kishel]

Because like I said, there was no attempt to level the playing field at all. I was interested in the blacks/dark shadow performance.

I just brought the FX1 along because, well, I had it, and why not?
then why put footage side by side, as if to imply "see how A compares to B". This is obviouis to show a darks/black shadow comparison, they are shooting the exact same footage. Yet tweaks were done to the HVX. This put the Sony at a disadvantage.

Sure you may not have used the FX recently and therefore did not make adjustments, but the HVX should be held to the same standard, IF you are putting footage side by side.

(oh a quick side note: I saw Zombie Hunter here at the Phoenix SciFi Horror Festival a short time back and spoke with the film's lead- that was a lot of fun :)

[Barry_Green]

then why put footage side by side, as if to imply "see how A compares to B". This is obviouis to show a darks/black shadow comparison, they are shooting the exact same footage. Yet tweaks were done to the HVX. This put the Sony at a disadvantage.
I laid out the reasoning of what I was attempting to do. It's at the beginning of the first post.

The Sony is the acknowledged leader in low noise performance. People have been asking how the HVX does in low light, and if there's a noise issue. So I put the HVX in the same scene as the Sony. We know the Sony is the lowest-noise of the units. How does the HVX perform against the Sony? Especially when I pushed it for how I would actually set it up? That's one thing I thought would be useful to find out. What purpose would it serve to artificially adjust the Sony? It was serving as the benchmark. It is a "known quantity."

The point is, if you like how the Sony performs, you should be gob-smacked at how well the HVX performs. I was.


Sure you may not have used the FX recently and therefore did not make adjustments, but the HVX should be held to the same standard, IF you are putting footage side by side.
I know what you're saying, but that totally doesn't apply. I wasn't trying to make a point of "oh look how great the HVX is and what a piece the Sony is." I was using it in its default configuration as a "benchmark", to see how the HVX performs against that established, acknowledged benchmark. Could the Sony have been made to perform better? Almost certainly. But then it loses its status as a benchmark, it becomes an unknown/tweaked quantity.


(oh a quick side note: I saw Zombie Hunter here at the Phoenix SciFi Horror Festival a short time back and spoke with the film's lead- that was a lot of fun :)
A lot of fun to work on too, Chris and Geoff are tremendous and I hope to do more stuff with them. That was a crazy pair of weekends -- the killer bees were the worst! :cheesy:

[Ed Kishel]

I see, now I understand your rationale for this setup.

Ed