|
|||||||||
|
Thread Tools | Search this Thread |
January 19th, 2006, 01:30 AM | #16 | |
Wrangler
Join Date: Oct 2002
Location: Los Angeles, CA
Posts: 2,100
|
Quote:
I've used the HD100/Mini35 combo on 2 jobs now. I find that the groundglass is a little too visible at HD resolutions (which isn't the end of the world), but HDV doesn't deal with all the grain very well...it's too hard to compress cleanly. Not to mention that the HD100/Mini using Cinegamma is 64 ASA, on a good day. One gaffer thought it was closer to 32. In my low budget world, that makes me only want to use it for daylight exteriors. I love it, but I'm not sure how practical it is anymore, like I once did.
__________________
My Work: nateweaver.net |
|
January 19th, 2006, 01:33 AM | #17 |
Tourist
Join Date: Dec 2005
Posts: 4
|
Hi Taylor,
Thanx for all you work in getting information out there. I’m sure you’ll be getting lots of replies to your request for info on how to procced with your tests; there’s lots of people out there trying the same thing. So I'll keep my reply short as possible. We’re in pre pro for a feature that we hope to shoot using the two GY 100 we have, both fitted with the P & S mini 35. We are by passing the tape transport and thus the poor quality Mpeg 2 compression and shooting straight to disk using two Waffian HR-1 recorders and CineForm’s excellent HD Prospect compression as an intermediate. I say hope because there are plenty of issues as yet unresolved. The biggest are: 1) Noise from the CCD 2) Barely acceptable HDV resolution at 720p 3) Poor quality of the relay lens supplied with the mini 35 4) Focus concerns 1) CCD noise. In all areas of the image below a middle to light grey there is a very noticeable pixel dance. It is especially noticeable over lighting gradients on areas of flat colour. Mildly underexposed skin tones have a blotchy look that can resemble great big birth marks. At times it looks like the bit depth of the colour sampling is too low even though we are capturing a 10 bit image straight to disk. My guess here is that the pixel size of an HDV 1/3 inch CCD chip is just physically too small. It is probably something to do with the wavelength of light being similar in size to a single pixel or perhaps the number of individual photons needed to excite the pixel to output each luminance and colour value. My (garden shed) theory is that since a pixel from a 1/3 inch CCD is 1/4 of the size of a pixel from a 2/3 inch CCD it will be hit by a quarter of the number of photons. Perhaps this number of photons is small enough that the essentially random nature of photon strikes becomes statistically significant. Sort of like the difference between doing market research on a group of 25 people or a group of 100 people. Let me re-emphasize that I am just guessing here. 2) Resolution. I’m just not convinced that 720p is really HD. Shooting with detail off gives an unacceptably soft image so we have to resort to the artificial sharpening of a detail function. With that comes all the associated issues such as increased noise and ghosting, issues which are really only noticeable under close inspection but which non-the-less I’m sure contribute to giving that ‘video’ feel. 3) Mini 35 Relay lens. Over all the new mini 35 is a superbly well engineered piece of kit and so it baffles me as to why they would use such a crap piece of glass in their relay lens. I have run resolution test which show an enormous difference between using the lens wide open and stopped down by just one stop. There is a very noticeable colour shift and a foggy softness when wide open that largely disappears at one stop closed. Added to this the are the horrible blue colour aberrations which are present on all strong highlights. These do diminish as the relay lens is stopped down but they are still visible at about 5 stops. Before you do anything with the mini 35 take the relay lens in your hand and have a look at how much glass there is there. It just doesn’t look like enough. I'll be posting some of my test results on a blog that has been set up: www.indiefilmlive.blogspot.com. Let me know if the link doesn't work. 4) Focus. You lose about 2 stops of light using a min 35. Some people say 1.5 but I’m not convinced. It is actually very hard to tell as the Fujinon lens diaphragm seems to be particularly badly calibrated. I read somewhere that the relay lens is only T2.8ish which makes me think that with the ground glass effect added a 1.5 stop loss is impossible. Anyway, you lose light so you have to shoot about 2 stops more open than you usually would, or if you prefer, shooting with 100 ISO stock. Add to that the fact that the relay lens has to be shut down by at least one stop and you might actually be down to 50 ISO. This means you need to shoot high speed primes mostly wide open. Which means you need a really good focus puller. And now remember that the eye piece view finder is basically useless for anything other than operating. How do you judge such critical sharps other than off the ring? There is an answer somewhere but it involves buying some sort of expensive high deff monitor for the operator. We recently saw our first 35mm test blow ups and were all suitably unimpressed. It could have been a poor projector lens and it could have been the blow up process we used (‘LARA’, a south African invention as far as I understand where each frame is grabbed individually of a very high definition monitor). Later this week we hope to be going for a Lighting blow up to be able to compare. There is also quite a good discussion about it the mini 35 and other aspects here - http://www.dvinfo.net/conf/showthrea...t=56502&page=2 I hope this helps. Cheers, Joe de Kadt. DoP. |
January 19th, 2006, 05:02 AM | #18 | |
Trustee
Join Date: Apr 2005
Posts: 1,269
|
Quote:
Have you tried the other alternatives, which claim to lose much less light than the Mini35 like the letus35 and the G35? If the problem is only light, these might help somehow. Do you feel the resolution or sharpness is considerably knocked down in comparison to the stock Fujinon? |
|
January 19th, 2006, 05:28 AM | #19 |
Trustee
Join Date: Apr 2005
Posts: 1,269
|
Joe, very interesting post. Thanks for sharing.
But you should remember you can only expect so much from a $5,500 camera. Sure it isn't Varicam or F900, which by your complains, it seems you might barely be happy even with them, but remember those cameras cost over 15 times the price of the HD100 after they are all ready to shoot the way the HD100 comes out of the box. When you throw an image converter in the equation, it can only “improve” the image as much, before it's trade offs (in this case, in exchange for shallow DOF) start to get in the way. I have been thinking lately if the whole 35mm adapter thing is really worth all the trouble or if having a shorter DOF in close ups only, but an overall sharper and cleaner image is the best way to go with those 1/3” cameras. I’m pretty sure that using a Mini35 in a movie you intend to blow up to 35mm and project is suicide. To be honest, I have my doubts if even a F900 with the Pro35 wouldn’t be suicide. Have you noticed that none of the commercially released movies shot on the F900 used the Pro35? Instead, they used Fujinon HD primes or Zeis Digi Primes for the best sharpness possible. What does that tell you? I have got to the conclusion after long considering to buy a 35mm adapter, that an image converter is a compromise and it should be expected. Hoping for something else is unrealistic. That may be subjective, but one thing is for sure, a $5,500 with an image converter won't hold up against a F900. Big old saying of " You get what you pay for" plays a big role here. I didn’t completely rule out buying a 35mm adapter yet, but I’m not expecting it to improve my image as much as I was a while ago, and sure not expecting it hold up on a 35mm blow up. Have you considered renting a F900 for your shot? You may be much happier than trying to squeeze F900 performance out a 1/3" camera. I have a HD100 and love it, but I know it can't get even close to be a camera made for 35mm blow up, neither can any of the 1/3” HD cameras. I would be surprised if they could even match Super-16 blown up to 35mm. If you throw in a Mini35, I think even normal 16mm would look sharper. Just my opinion. P.S. I'm following all your tests and results very closely. It's all very interesting. |
January 19th, 2006, 06:22 AM | #20 | |
Trustee
Join Date: Apr 2005
Location: Chicago
Posts: 1,214
|
Quote:
|
|
January 19th, 2006, 10:52 PM | #21 | ||
Major Player
Join Date: Jul 2004
Location: Phoenix, AZ - USA
Posts: 300
|
Quote:
Quote:
What about this variation - get a JVC to Nikon mount adapter, throw a Nikon 28mm or 20mm lens on there (with the multiplier becomes telephoto) and then put a G35 or Micro35 in front of that? The Nikon becomes the relay lens and it's a lot less expensive than a Mini35. |
||
January 20th, 2006, 04:30 AM | #22 |
New Boot
Join Date: Jan 2006
Posts: 9
|
Back in 2001 i made a film using a PD150 and a 35mm adapter with prime lenses and i can tell you this is something i'll never do again. We had major problems from vignetting, and rotating ground glass in the footage. Low light was bad and grainy and we ended up lighting everything like we were using 35mm cameras. Overall the experience was bad and large portions of the footage couldn't be used.
As much as i like the 35mm DOF, it not worth compromising your film for it, instead concentrate on storytelling. |
January 20th, 2006, 10:18 AM | #23 | |
Major Player
Join Date: Jul 2004
Location: Phoenix, AZ - USA
Posts: 300
|
Quote:
I think you're right that focus needs to be the story. Before I shoot anything longer form I'm going to have my solution well tested. If 35mm adapters won't get me there then I won't use one. |
|
January 20th, 2006, 10:33 AM | #24 |
New Boot
Join Date: Jan 2006
Posts: 9
|
Joel,
Don't let my experience put you off, back in 2000/01 i believe the mini35 was the only adapter available, it was the pioneer of the products available today. Over the technology has improved and people are getting better results... and yet judging by footage from the JVC and Canon it seems that the extra resolution is bringing those old problems we experienced back. I'm currently evaluating two low budget feature films made with the Sony F700 and F900, and the cinematography is fantastic without the 35mm adapters, in fact for both films it's the story that lets them down. I'm looking at purchasing the Sony HD XDCAM for a project later on this year, i'm hoping the problems will be resolved by then, even so i still may not use one. |
January 23rd, 2006, 05:29 PM | #25 | |
Regular Crew
Join Date: Mar 2005
Location: Los Angeles, CA
Posts: 50
|
What defines 'good cinematography' ?
Quote:
Most importantly, before you buy anything, my advice would be to decide what things you find as nice cinematography because not that the XDCAM could not do nice things, but a 1/2inch CCD might be a hindrence to what you are ultimately looking to acheive, ie 35mm DOF w/out adapters. If I think 1/2 inch, I think about the camera far away from the subject, wide open, and at the very end of the barrel if I want shallow DOF. Where do you want to make the tradeoff? ---------- On a related note, I want to see the 2/3inch GV Infinity/DigiPrime combo. Some call it a 'newscamera' but I've shot narrative on interlaced newscameras and it looked fine. Funny that they used to shoot news with 16mm film. As cliche as it sounds, I look at camera/lens/CCD/cine adapter/film all as items that can not exceed the qualifications of the person using it, or that can excel because someone qualified is using it.
__________________
Taylor Wigton DP, Los Angeles |
|
January 24th, 2006, 04:46 AM | #26 |
New Boot
Join Date: Jan 2006
Posts: 9
|
Taylor,
You're correct, there are currently no 35mm adapters available for 1/2 inch CCD's at the moment, i'm sure by the time i'm ready someone will have a solution worked out, even so i'm actually planning on getting the Sony 2/3 adapter and a set of decent HD Primes, oh and did i mention that i might be getting 2 cameras as this project may end up going 3D. The other reason why the Sony is so important is the variable framerates, currently only the HVX and the Varicam offer this ability in camera. The HVX has fixed lenses and the Varicam is too expensive to buy 2 of them, so this leaves the Sony at a price point that is affordable. Tony |
January 24th, 2006, 10:50 AM | #27 | |
Regular Crew
Join Date: Mar 2005
Location: Los Angeles, CA
Posts: 50
|
There ya go!
Quote:
__________________
Taylor Wigton DP, Los Angeles |
|
January 26th, 2006, 02:52 PM | #28 |
Regular Crew
Join Date: Sep 2004
Location: Atlanta, Ga
Posts: 129
|
Just saw the article, but I have not read it yet (it's long, and I'm at work). So my request may very well be what you are intending to do, but just in case...
I would like to see you guys compare image quality in a couple of ways: the standard HDV compressed quality, and the pre-compression HD output. I know that the Canon has HD-SDI out. I can't recall whether the HVX200 does or not off of the top of my head. The JVC has HD uncompressed out, and JVC has an HD to HD-SDI out box specifically for the camera. I have no idea what the Z1 has, if anything for uncompressed output. I would also be interested in workflow. I know that you were planning, in the last part of the test, to test these cameras w/ various NLEs, but I don't know if your tests will start with the footage in the NLE, or start with getting the info to the NLE. It would be nice to see what any cameras shortcomings or exceptional features are as far as getting everything INTO the NLE (& whether you are going HD-SDI into the NLE or using the minDV tape/Rev-Pro/HDD/Sonys Blu-Ray Pro). It should be particularly interesting considering that you might be testing a GV Infinity, and possibly an HD-XDCAM, given that they use different media, and the GV in particular uses JPEG2000 compressed codec that the others don't use. I don't know what NLEs (other than GV's own NLE) can deal with JPEG2000 either natively or via a 3rd party plugin. Perhaps Final Cut Extreme or FCP 6 will be able to handle JPEG2000 (?) when they are announced (I'm guessing they will be announced at NAB, but that is pure speculation - I've heard nothing other than they will be announced soon, and FCX will run around $10k). I, as well as many others will be watching the series closely, and can't wait for part 2 (and I haven't even really read beyond a quick skim part 1!). Thanks Alex F |
February 1st, 2006, 08:25 AM | #29 |
Tourist
Join Date: Jan 2006
Posts: 3
|
When is part 2?
Hi Taylor
This is fascinating stuff and a great way to test products. When will part 2 be completed? DB |
February 2nd, 2006, 11:34 AM | #30 |
New Boot
Join Date: Oct 2005
Posts: 17
|
Mini 35
I've shot with the Mini 35 a few years back and thought it was too soft thenm with and XL1 (before HDV). While the Mini35 fits the new cameras, I have never heard if the improved the innards for HD work? This could be why it looks even worse on HD footage than SD? I'm currently cutting a f900 show that used the Pro35 with Cooke primes - everyone who sees it assume sit's 35mm, but it was shot by a very experienced DP at a great location.
I'm hoping Cinemek makes a G35 that ataches directly to the JVC body and not to the end of the lens. Footage from the Cinemek appears far superioir to anything I've seen with the Mini35 and it's $1200 vs. $10,000! I too am very interested in the new Sony HD XDCAM gear - though I was disappointed by the 1/2" chip decision. A friend at Sony says the footage so far looks as good or better than the f900. Curious if someone like David Mullen would know how a 2/3" Digiprime or similar would work on 1/2" camera with adpater. I know the focal legth changes but would that prevent you from shooting with it? For instance if you wanted a 50mm look, just use a wider lens that equaled the 50mm on the 1/2" chips? Lance |
| ||||||
|
|