HVX vs JVC noise - Panic in Pany town?

[jlames]

As usual filmguy is right on. Numbers be darned in any medium when you're talking sub $10k. When we were directed about a year ago now to make the move to HD from years of SD, I spent 6 months listening to my guys who, collectively, have 90 years in cinema from big to low budget (everybody starts somewhere) and then we rented every single one of the discussion leaders, Canon, JVC, Panny and Sony over a 3-day weekend. In getting ready for the taxman, I had to review the accountant's numbers before signing off on what we owe Uncle Sam and I tallied the cost of that escapade. It turned out that I spent right around $4000 in time and talent one day and then $3000 over the next 2-days in the editing booth with balance, color correction, etc. so that the results were side-by-side and all things were equal. We went from exterior panorama with highly contrasting blue skies spotted with bright, white cumulous clouds, brown mountains and green pastureland to interior sets with controlled lighting and wide to close-up shots. In the end, we collectively viewed the results and discussed our projected distribution needs over the next 18 months when we may very well make another move to true HD (allright, that's a discussion for another time) and, for us, JVC led the pack. But this is very, very subjective and specific determination. Your's may be different. But the point is, the more restricted your budget is, the more important it is to, of course listen to the points made by the many learned individuals who populate this Board, but to compare for yourself the results of cameras in your price range using shooting, editing and printing benchmarks on an identical setup. IMO, you can't make a determination at Video Expo or NAB or anywhere that there is anyone that 1, you're not married to, 2, is contributory to your checking account or 3, working for you. Rent the cameras, shoot the same scene, edit in the configuration you're most likely to use when it comes to distributing the end results and then make a determination as to which camera meets YOUR specific needs. It's money well-spent.

 

[Zack Birlew]

Noise... Noise! Noise!! Noise!!! AAAAAARRRGGGHHH!!!! THAT'S ALL WE HEAR ABOUT NOW!!!!:furious3:

Oh well, that's okay! rogar-BigGrin(DBG)

Yes, there seems to be an issue with many people over the noise issues concerning the HVX200. However, there is noise in the HDV cameras as well, even the XLH1. For filmmaking though, any of these cameras will do a splendid job. All this talk of noise should not affect your buying decision.

 

[Emanuel]

Your's may be different. But the point is, the more restricted your budget is, the more important it is to, of course listen to the points made by the many learned individuals who populate this Board

The film-out tests that you in the US, you are seeing, they are too much important to my evaluation. As well, your points, some of them most informed than mine -- techie background and experience, they are a unique tool that I don't have so easily around me.

but to compare for yourself the results of cameras in your price range using shooting, editing and printing benchmarks on an identical setup. IMO, you can't make a determination at Video Expo or NAB or anywhere that there is anyone that 1, you're not married to, 2, is contributory to your checking account or 3, working for you. Rent the cameras, shoot the same scene, edit in the configuration you're most likely to use when it comes to distributing the end results and then make a determination as to which camera meets YOUR specific needs. It's money well-spent.

Thanks Jonathan for your kindly input.

 

[Emanuel]

OK Emanuel- as far as Canon's 24f mode is concerned- Adam Wilt's 4-cam shootout article at DV.com sums it up pretty well:

[FONT=arial,helvetica][FONT=arial,helvetica][FONT=arial,helvetica][FONT=arial,helvetica][FONT=arial,helvetica][FONT=arial,helvetica][FONT=arial,helvetica][FONT=arial,helvetica][FONT=arial,helvetica]"In 60i, vertical resolution was a solid 700+ lines; in 24f mode, we thought it looked like half-resolution material on the monitors, consistent with what you'd get with field-doubling. Looking at the captured clips, especially the Combi-2.3's bullseyes, I'd have to say that it certainly looks like field-doubled vertical sampling, with about 540 TV lines usable, although certain aspects of the image "feel" sharper."[/FONT][/FONT][/FONT][/FONT][/FONT][/FONT][/FONT][/FONT][/FONT]

If you haven't read through this article yet, you can do so here:

http://www.dv.com/news/news_item.jhtml;jsessionid=JFSYXUOYRHTBCQSNDBCCKH0CJUMEKJVN?articleId=177103305&_requestid=474865

You'll have to register to view material on this site- but it's worth the time, as there's a lot of good information on the site for you.

I already knew this available information. I also posted that one here. And that's why I above-mentioned that I had the same idea before adding more details with Barlow, Billups and Newman sources. But thanks anyway for your care.

With regard to Elton's post about 24f and the DIGIC processor, etc.- there simply is no magic, no tricks to be done here. You can't produce a 24p image from interlaced CCD's without deinterlacing and interpolating your image, and you can't do that without throwing 1/3 to 1/2 of your vertical resolution away- certainly not on the fly.

I appreciate your POV and I agree.

With all that said- I'm not trying to use this space to bag on the XLH-1- not at all. I'm positive that the guys at Cineform are dead-on in their appraisal of the Canon's film-out footage- I'm sure it is quite beautiful. I've seen filmout footage with my own eyes produced from the Sony Z1U that looked amazing.

At big screen? Regarding the 35mm blow-up goal what's your opinion vs. the HD100? And vs. the HVX? And I've seen posted what I above-mentioned that they are saying the HVX footage as better or similar than Super16 and HD100 footage as 16mm. Illya Friedman said that Z1U was under these ones. And perhaps he's biased 'cause as producer he did the film-out, but to him, HVX was so much better than the others.

What's your opinion? -- if there is term of comparision.

Did you know these tests:

http://www.eidomedia.com/test/out_test.htm
http://www.eidomedia.com/test/
http://www.eidomedia.com/test/ottiche.htm
http://www.eidomedia.com/test/dw_up.htm

My idea is that there are CCD differences between 1440x1080 vs. 960x1080. Otherwise, Z1U there isn't 24p -- to go@post from 50i -> 24p will it be the same? According your own words here, it isn't and it makes sense, of course. Maybe, that's why I'm waiting for CineAlta F-330. And perhaps the 24p it is the same or it comes from a similar process as it works with the XL-H1...

But in your earlier post, your main concern seemed to be the fear that 720p wouldn't be a high-res enough format for filmout compared to 1080f, and what I'm trying to get across to you is that those numbers are all imaginary in the first place. The Canon, (as is the case with all 4 of these sub $10k cameras) is not able to remotely approach a measurable res of 1080x 1440- so it's kind of a mute point that it even records at that res.

In interlaced mode, it's res has been measured in the 700-800 line range, and in 24f, it's been measured at around 500-550. That's it- that's what you're getting. It doesn't matter at that point if the format recorded to tape is 720x1280, 1080x1440, 1080x1920, or a 3k Cineon file- the vertical res of your base image is in the range of 500 lines. At the end of the day, that's what you're blowing up to film with.

The 1080 vs. 720 debate for filmout doesn't really mean anything right now, because the Cinealta is the only HD camera on earth that can clearly demonstrate any improvement over 720p's resolution on the vertical axis in a progressive frame environment.

You've got the point. And 'cause aside the pixel shifting I found reasons to understand the HVX noise, precisely, at lower CCD rez and the former fear in to releasing the CCD/DSP specs in order to not aware the navigation. Therefore, I have my doubts concerning all lower specs like the HD100 1280x720=921,600 CCDs pixels vs. the CineAlta F-330 or even the XL-H1 1440x1080=1,555,200. Comparing with the non-stable HVX performance, I find the answer in the pixel shifting that in the JVC is away. And maybe 'cause that we love so much its footage. But according the numbers -- my belief, it's half rez of F900. [here, there wasn't any filmmaker that had guts to go to film-out from Varicam 720p so you can figure out what's the major feeling]

Despite what it may look like- I'm not really trying to push you away from the Canon or toward the JVC- I just hate to see someone with talent and good intentions get caught up in the 1080 vs. 720 numbers game and make a decision as important as this one based on all the wrong factors. What you really need to do if at all possible is get some time in with both cameras- rent them if you have to. See which one feels better to you- see which one says 'Emanuel' the most. At the end of the day- all this number crunching and technical talk won't matter nearly so much as that.

[FONT=arial,helvetica][FONT=arial,helvetica][FONT=arial,helvetica][FONT=arial,helvetica][FONT=arial,helvetica][FONT=arial,helvetica][FONT=arial,helvetica][FONT=arial,helvetica][FONT=arial,helvetica]
[/FONT][/FONT][/FONT][/FONT][/FONT][/FONT][/FONT][/FONT][/FONT]

Unfortunately my personal feeling or taste must be away. I'm prepping a test comparative and this research is already the first and the following steps in order to. Regarding technology I'm completely agnostic and I just believe in the engineering. That's why the numbers count. The question is how they're working...

 

[filmguy1974]

My idea is that there are CCD differences between 1440x1080 vs. 960x1080. Otherwise, Z1U there isn't 24p -- to go@post from 50i -> 24p will it be the same?

Shooting on an interlaced camera for filmout, whether the Sony or the Canon, presents the problem of dealing with going from an interlaced image to a progressive one. My understanding is that 24f is much better implemented than Sony's cineframe 24. That aside- conventional post production wisdom would dictate that shooting 50i with either camera and using a motion-adaptive deinterlacing method will throw away less resolution in the end than using any on-board chip on the camera which operates on the fly.

The question at that point, though, is how willing are you to go through the headache of deinterlacing your footage in post and slowing your audio down by a few percent? Will the naked eye really perceive much difference? Hard for me to say as I sit here with no XLH-1 to perform tests with...

Therefore, I have my doubts concerning all lower specs like the HD100 1280x720=921,600 CCDs pixels vs. the CineAlta F-330 or even the XL-H1 1440x1080=1,555,200.

The published CCD specs for these 1/3" cameras are basically meaningless when compared to measurable resolution. If the Canon gives you a measured res of 800x800 lines in interlaced mode (ballpark) as suggested by the dv.com article, then that's the resolution of your image- period. We're talking about the difference between theoretical performance and real-world performance.

To the best of my knowledge, the real world numbers are as follows:

XL-H1 (24f): 550 (ballpark) lines of vertical res with which to perform your film blowup.
HD100: 700 (ballpark) lines of vertical res with which to perform your film blowup.

But according the numbers -- my belief, it's half rez of F900. [here, there wasn't any filmmaker that had guts to go to film-out from Varicam 720p so you can figure out what's the major feeling

Hey- if you've got access to an F900, by all means- this conversation is over- there's your filmout camera. The thing that's forcing me to do a double take is that I'm getting the impression you would choose to shoot on the XL-H1 over the Varicam simply because it's CCD resolution is theoretically higher.

That's unfortunate, if true, because the Varicam is a 2/3" camera with lower noise, better highlight handling, wider latitude, and better measurable resolution than ANY of the 1/3" cameras currently on the market, the XL-H1 included. No D.P. in their right mind would point you to the XL-H1 if you had a Varicam available. That was my point from the begining really- that there's a whole lot more to a good filmout than raw numbers and published specs. The quality of your recorded image will do a lot more for you than any theoretical numbers can.

It seems you're hellbent on finding the right camera for filmout instead of finding the right camera for Emanuel. In my opinion, the right camera for Emanuel is the one that will ultimately be the right one for filmout, because that's the camera you're going to shoot your best stuff with, whether it's the Z1U, the HVX, the HD100, or the XL-H1, and regardless of CCD specs.

 

[stephenlnoe]

Can you answer me that one unpleasant doubt:

money aside (as you well and wisely mentioned) -- but to my business to spend $6000 or $10,000 or even $20,000 it's not so different, to 35mm film-out, what's the higher option among JVC HD100 vs. XL-H1 vs. PDW-F330?

For sure there are so many variables and Sony's is totally unknown, yet. But what's your rez significance bet?

Hmnn... My thoughts on this.....

The easy part is the Sony. I don't know the first thing about it yet so I can't comment on it.

The Canon. It is a very good camera. It is primarily an interlace camera as it's true nature and the progressive mode could cause the rez to suffer. When you blow up the frame to 1920x1080 the detail of the lower (Frame mode) rez may not be as sharp as you'd want (but maybe it would?). It's all conjecture by me because I haven't actually performed a film transfer with an XL-H1.

The JVC. It is progressive and film is progressive. Hmnn... Makes sense to me that there is less work getting the image to film. No deinterlacing required and the full frame need's only to be scaled. Since the camera holds rez in "still" as well as "camera in motion" scenes this is another plus in my book. You definately want the camera to hold rez whether it's locked down or in motion. The reason? Because the more detail that's held throughout the sequence the better the scaled up image will be.

The JVC is designed for HDTV workflows as well as film print workflows. It's the intention of it's creator that it be used for film print. In my experience, the ProHD picture provided by the HD-100 is worthy of film transfer and luckly we have experts to help with everything that is ProHD to film print. We have "DuArt (New York)" and "I cubed (Chicago)"

I say, get the camera, shoot a short (under 10 min TRT), call DuArt or I cubed and put it on a reel and see for yourself. We have done it and are well satisfied enough to go forward with our digital to 35mm film projects.

In the end, the reality is, the viewer never cares what it was shot on. What matters most as producers of content is, how easy was it to get my product to market and is my bottom line good. In this case it is extremely hard to beat the HD-100 and the ProHD format for film workflow (or HDTV workflow). I haven't found anything easier or more cost effective and that's why I planted my flag in JVC's sand.

S.Noe

BTW: The GY-HD7000 looks very interesting to me and it should follow the same workflow as the HD-100.

 

[ullanta]

It turned out that I spent right around $4000 in time and talent one day and then $3000 over the next 2-days in the editing booth with balance, color correction, etc. so that the results were side-by-side and all things were equal. We went from exterior panorama with highly contrasting blue skies spotted with bright, white cumulous clouds, brown mountains and green pastureland to interior sets with controlled lighting and wide to close-up shots.

(Of course, you know you could easily recoup a lot of that cost by putting the sample footage on a DVD for $10 a pop...)

 

[filmguy1974]

Noise... Noise! Noise!! Noise!!! AAAAAARRRGGGHHH!!!! THAT'S ALL WE HEAR ABOUT NOW!!!!:furious3:

Oh well, that's okay! rogar-BigGrin(DBG)

Agreed. It's amazing how much noise (no pun intended) people are making about the noise issue. Considering the fact that we're talking about 1/3" cameras using very high compression methods, it's amazing that the pictures that come out of them are as nice as they are. We're actually very lucky- but some people can't get past the idea that a $6-9K 1/3" camera just isn't going to get you $60,000 2/3" results. If you look at it from a price to performance standpoint- all 4 of these 'baby' HD cams are pretty miraculous, IMO.

 

[Zack Birlew]

What I don't get is why are people complaining about this now? We've had noisy pictures for years! It's just that they're in HD now!:grin:

Honestly people, you need to be grateful because if we were still using film, you'd have to worry all about over/underexposure, white balance, film types, sound synchronization, lense choices, film costs, ect., ect., ect. But with these cameras, you can go out and film something movie theatre-worthy, do that instead of talking all about limitations with the camera! Nobody is going to release a new camera for sale tommorow that fixes ALL the problems with our current choices, we've gotta use what we've got!

 

[Stevet]

Well, fortuanately, due to technology advances, cameras are getting better for less.

I don't blame anyone wishing for more. Look how much better our footage is with these sub 10K cameras.
This is the drive that keeps technology advancements alive.

I can see how the noise issues arrise, especially since 1/3" CCD cameras exsist, such as Sony that actually fares OK in noise levels.

I happen to think the HD100 is OK on noise level, but just imagine if JVC was able to drop the noise floor another 3 to 6dB !

 

[xray]

Honestly people, you need to be grateful because if we were still using film, you'd have to worry all about over/underexposure, white balance, film types, sound synchronization, lense choices, film costs, ect., ect., ect. But with these cameras, you can go out and film something movie theatre-worthy, do that instead of talking all about limitations with the camera! Nobody is going to release a new camera for sale tommorow that fixes ALL the problems with our current choices, we've gotta use what we've got!

Now you have to worry about SSE, P2 costs, how to edit 25P, Cinema-v or Cinema-D, better coating on lenses, noise here or noise there, archiving, clunck sounds, datarate, FAT32 disk, firewire400/usb2, etc. etc.

See, live is not easy :love4:

 

[harddrive]

What I don't get is why are people complaining about this now? We've had noisy pictures for years! It's just that they're in HD now!:grin:

If I understand what people are posting it's not noise in the conventional sense that people are making a fuss about, but rather what others have described as a sort of blocky artifacting in lowlights and gradients of HVX images.

The point is that whilst true noise is random, this isn't, and that is what makes it so irritating to the people who are disturbed by it. Whilst I haven't had the chance to see how bad (or not) it is, I can understand why it may be seen as a problem.

 

[jimagine]

I happen to think the HD100 is OK on noise level, but just imagine if JVC was able to drop the noise floor another 3 to 6dB !

I don't think it would make a single whit of difference to a viewer of creative content.
You can see so many mistakes in high budget 35mm films today (watch a DVD like Gladiator on your progressive scan player and HD TV, and count the moments of soft focus in some of the steadicam scenes and when actors don't hit their marks in tight CUs) Watch some of the award winning classics from the 60's and 70's and 80's and how grainy and just "dirty" some scenes are, then look at a lot of the non-classics but still money makers and they're anything but pristine on a number of levels.

Geting hung up on a noise floor that's simply of no real consequence in producing good narrative work doesn't make sense to me (not that you're hung up on it <g>). Things have come so far in the last few years - especially now, with this first crop of affordable HD cameras...there are so many advantatges to producing with them versus film for amost all independents, that a perceived noise problem does seem pretty inconsequential.

 

[Stevet]

Things have come so far in the last few years - especially now, with this first crop of affordable HD cameras...

My point exactly.

"a perceived noise problem does seem pretty inconsequential"

Wow... That's a bold statement.

To each his own. Noisy videos looks well.... like noisy video.

I have yet to see any random video noise nail the look of film grain.
Of course it's about picture content. It doesn't mean I should be satisified
grabing a $200 DV video camera from Walmart and be happy with its
level of quality. Of course, I bet noise won't be the only issue here

 

[jdv]

If I understand what people are posting it's not noise in the conventional sense that people are making a fuss about, but rather what others have described as a sort of blocky artifacting in lowlights and gradients of HVX images.

I guess that was what I was getting at in my first post - JVC users don't seem too bothered by the 'quality' of noise produced by the JVC - while there seem to be many (perhaps half?) HVX users who are VERY disapointed by the noise/grain structure on the Pany.

I agree w/ all of those who point out that film has its noise and that many famous well made films have plenty of 'defects.' This fact is what moved our company away from 16mm film and into video in the first place. From that perspective, ALL the newer cameras look to has less noise to my eye (although I have not used the newer Kodak vision stocks).

John