Jump to content

Reduser.com Bans the "word" 7D


Guest Trevor Swaim

Recommended Posts

There is no such thing as "a purely technical standpoint".

 

Ya there is, numbers vs numbers. Does anyone actually think a 5D can go up against a red of film camera in terms of resolution, dynamic range, frame rate, compression ect?

Link to comment
Share on other sites

  • Replies 80
  • Created
  • Last Reply

Top Posters In This Topic

They've used a 16mm GSAP or gun camera for those types of shots for many years. These are gun cameras for WW2 fighters modified to take C mount lens.

 

http://www.alangordon.com/s_filmcam16_minicam.html

 

Very small cam (also look at the Ikonoskop). However, the Take it to the next level shoot had the SI2K mini mounted in front of the face, not on top of the head. The film camera would stick out too far to offer the same perspective. There is an image of the SI2K on top of the head but they did not use this rig for the shoot and instead adopted a face mounted rig that offered a better perspective.

Edited by Andrew Rieger
Link to comment
Share on other sites

 

done.

 

try doing that with the RED ONE. ouch!

 

How was this shot?, I can find no technical info about it anywhere! Thank you for posting this, I love POV stuff!

 

BTW, this was an S12K, of course a Red could not do this. Nor could most film cameras except for an Ikonoskop-like 16mm cam, which is why I am so curious about how it was shot.

Link to comment
Share on other sites

Yes, it was something like an Aaton Minima as far as my research went when I researched POV shot stuff,

 

and the SI2K was mounted on a helmet as well. http://www.onsight.co.uk/showcase/35/scoring-gold

 

Here's what we did:

(using a NOX camera on the shoulder and helmet.)

 

I have seen that picture of the SI2K on top of the helmet but they did not use that rig, they moved the camera down to in front of the face for the final shoot. There was a discussion about it on Reduser and someone who worked on the set in the camera dept stated that the rig shown in the picture was a prototype that the director went with a different design.

Link to comment
Share on other sites

 

Here's what we did:

(using a NOX camera on the shoulder and helmet.)

 

Loved the vid. You say the camera was on the shoulder and helmet. Did the actor shift his hands to adjust for the camera on the shoulder to make it seem like the camera was his eyes? Was the helmet mounted cam on top of the helmet or in front of the actors face? I am trying to do something similar for narrative work. Again, great job!

 

Any pictures of the setup? Was the NOX a full size camera or a mini camera head?

Edited by Andrew Rieger
Link to comment
Share on other sites

Guest Jeremy Hunt

I think this "boys with toys" attidtude towards digital cameras should really stop, its highly unprofessional.

Link to comment
Share on other sites

Ya there is, numbers vs numbers. Does anyone actually think a 5D can go up against a red of film camera in terms of resolution, dynamic range, frame rate, compression ect?

 

Okay, let me put it another way, there is no situation in which decisions are made based purely on technical information. Such things as opinions, personal preference, marketing, handling and anything else you can think of always come into it.

 

 

Two situations where the 5D2 has better resolution than the Red:

 

1) Shooting footage for a special effects shot. where a high resolution locked off plate is required for the background, and some footage of a person running across it is needed for reference to be replaced by a CGI character. On the 5D2 you can shoot a still of the background at 21MP, do HDR if you want, and shoot the video at 1080P. On a red you can shoot the video at higher resolution but it's not relevant as that's only a reference, but you can't shoot the still at as good a resolution as the 5D2.

 

2) A camera is placed behind the steering wheel of a car pointing back at the driver, where there's room fro a 5D2+ short lens, but a red won't fit. In this case the 5D2 resolution is 1080P (you can argue lower) but the Red's is ZERO.

Link to comment
Share on other sites

  • Premium Member

 

Two situations where the 5D2 has better resolution than the Red:

 

1) Shooting footage for a special effects shot. where a high resolution locked off plate is required for the background, and some footage of a person running across it is needed for reference to be replaced by a CGI character. On the 5D2 you can shoot a still of the background at 21MP, do HDR if you want, and shoot the video at 1080P. On a red you can shoot the video at higher resolution but it's not relevant as that's only a reference, but you can't shoot the still at as good a resolution as the 5D2.

 

2) A camera is placed behind the steering wheel of a car pointing back at the driver, where there's room fro a 5D2+ short lens, but a red won't fit. In this case the 5D2 resolution is 1080P (you can argue lower) but the Red's is ZERO.

 

Sort of two extreme examples -- a still photo and a space where only the 5D can fit. And only in the still photo example does the 5D have more resolution. As for the motion efx plate, the Red would deliver a better image for post, 1080P or not. Exceptions don't prove the rule.

 

I don't think there is any argument that can be made where for standard shooting, the 5D would deliver a better moving image in any way you wish to measure quality, from color to dynamic range, resolution, whatever. Maybe noise at 6400 ASA. But certainly there can be creative, logistical, and financial reasons for choosing the 5D over the Red.

Link to comment
Share on other sites

Yes I know they're extreme but that doesn't matter. I'm not trying to prove the 5D2 is better (it's worse for most video things), I'm merely claiming the Red isn't _always_ better - in this case exceptions do prove a rule.

 

Andrew seemed to be saying the Red _is_ always better than the 5D2 because only technical concerns matter.

I'm saying that sometimes, mostly even, choices aren't made purely for technical reasons.

Link to comment
Share on other sites

Yes I know they're extreme but that doesn't matter. I'm not trying to prove the 5D2 is better (it's worse for most video things), I'm merely claiming the Red isn't _always_ better - in this case exceptions do prove a rule.

 

Andrew seemed to be saying the Red _is_ always better than the 5D2 because only technical concerns matter.

I'm saying that sometimes, mostly even, choices aren't made purely for technical reasons.

 

No Red is not always the best. Choose the best tool for the job. The Hurt Locker crew considered Red but was concerned about going digital in a very harsh environment. Aaton film cameras had proven themselves in harsh environment shooting March of the Penguins so they went with a proven system. The District 9 crew however experienced similar conditions but Red worked out fine. Different decisions both with great results.

 

The example you listed is more based on the form factor of the DSLR. The Epic system packs more resolution, higher dynamic range, more sensitivity and higher frame rates all into a package only slightly larger than a DSLR when properly configured so once the camera is released, the small form factor will no longer be exclusive to Canon. also, if small is what you want, the SI2K mini is smaller than a DSLR, has a lot more resolution and has been shooting artifact free long before Canon enabled the 5D to shoot video.

 

When you break it down, the only real benefit the Canon's have over other systems like the SI2K, Red and Alexa is their low light performance, which Arri and Red are quickly nearing, and the small form factor, which both the SI2K and soon to be released REd epic will match. After that, using a Canon will really come down to budget (which is of course a reasonable choice in this industry. Afterall many are choosing to shoot digital because it is cheaper than film even though film is still the gold standard, for now).

Edited by Andrew Rieger
Link to comment
Share on other sites

  • Premium Member

Here are a couple pictures from the shoot:

 

Wow, those are dangerous looking rigs. Believe me, neck surgery is no damn fun at all. I still have limited feeling in three fingers. And these things are long lever arms acting against your neck. If you ever build such a thing, make it a breakaway, so the rig comes off the helmet before it ruptures a disc.

 

 

 

 

-- J.S.

Link to comment
Share on other sites

  • Premium Member

It's entirely subjective, and I could be wrong, but my strong impression is that a 5D has considerably better dynamic range than a Red.

 

But as I say, it's hard to tell, with the lack of noise.

 

P

Link to comment
Share on other sites

  • Premium Member

It's entirely subjective, and I could be wrong, but my strong impression is that a 5D has considerably better dynamic range than a Red.

 

But as I say, it's hard to tell, with the lack of noise.

 

P

 

No, the dynamic range is limited with the Canons, more like standard Rec 709 video. That's what it was designed to be seen on after all, a video monitor. You are talking about 8-bit 4:2:0 at around 40-45 Mbits/sec at best (Nikons & Lumix are more like under 20 Mbits/sec.) The M-X Red camera has more usable dynamic range with a wider color gamut, less compression, more resolution, etc. No real comparison.

 

I think the shallow-focus look of the 5D image hides a lot of sins. Dynamic range doesn't matter much in out-of-focus areas of the frame.

Link to comment
Share on other sites

Yes I know they're extreme but that doesn't matter. I'm not trying to prove the 5D2 is better (it's worse for most video things), I'm merely claiming the Red isn't _always_ better - in this case exceptions do prove a rule.

 

Andrew seemed to be saying the Red _is_ always better than the 5D2 because only technical concerns matter.

I'm saying that sometimes, mostly even, choices aren't made purely for technical reasons.

 

 

So tell me - which has better resolution. A red camera doused in petrol and set on fire, operated by a corpse wearing a tutu, or 15 5D cameras with their lens caps on, shooting from space?

Link to comment
Share on other sites

  • Premium Member

So tell me - which has better resolution. A red camera doused in petrol and set on fire, operated by a corpse wearing a tutu, or 15 5D cameras with their lens caps on, shooting from space?

MR ROBINSON!

I don't know how you can even contemplate doing something like that to even ONE 5D, let alone 15!

All that cosmic radiation, space junk, meteoric dust, the risk of burning up in the atmosphere ... the fact that it could even occur to you to do something like to a 5D, is frankly ... appalling!

I mean, what kind of a world do we want to live in.

A 5D?!!

Is what I fought and died for in WWII...?

People these days ... I don't know

Link to comment
Share on other sites

So tell me - which has better resolution. A red camera doused in petrol and set on fire, operated by a corpse wearing a tutu, or 15 5D cameras with their lens caps on, shooting from space?

 

 

Depends what you mean by 'resolution'.

Link to comment
Share on other sites

  • Premium Member

No, the dynamic range is limited with the Canons, more like standard Rec 709 video. That's what it was designed to be seen on after all, a video monitor.

 

I never quite understood this. When a camera shows more dynamic range in RAW still mode, what's the issue with retaining that in video mode? I understand you won't have the same precision throughout the range, and I can see how line skipping could cause an apparent loss in range, but even aside from the DSLRs, why do the manufacturers' curves in the video realm generally lop the top few stops off when the sensors are certainly capable of capturing them? Much of what we see ends up being 8-bit 4:2:0 anyways, and still represents a wider dynamic range than cameras seem to be willing to put out. Somebody enlighten me!

Link to comment
Share on other sites

  • Premium Member

I never quite understood this. When a camera shows more dynamic range in RAW still mode, what's the issue with retaining that in video mode? I understand you won't have the same precision throughout the range, and I can see how line skipping could cause an apparent loss in range, but even aside from the DSLRs, why do the manufacturers' curves in the video realm generally lop the top few stops off when the sensors are certainly capable of capturing them? Much of what we see ends up being 8-bit 4:2:0 anyways, and still represents a wider dynamic range than cameras seem to be willing to put out. Somebody enlighten me!

 

I think you'll find that once you color-correct anything for Rec 709 broadcast, you have tossed out some of the dynamic range in order to get good whites and blacks. But that's fine -- what a wider dynamic range gives you is more flexibility in color-correction in order to make changes.

 

It's a little like the issue of color reversal film: the gamma (contrast) is designed for direct projection of the original. So why do you need more information than is needed for projection? Because you have no flexibility to color-correct when starting out with an image already at the higher gamma needed for projection -- you want the ability to bring things up or down into the final range, rather than be locked into it already.

 

Look at something like a Log recording in a video camera like the Genesis: rather than have white at 100 IRE, it's only at 70 IRE. What this allows is a couple more stops of "super white" overexposed detail to be recorded. Then in post, even though you have to set white at 100 IRE, you have some flexibility in adjusting the gamma curve or use knee compression to have overexposure roll-off into pure white more gracefully (more like film.)

 

So these Canon cameras would have to do something similar, take the widest dynamic range that the sensor has to offer and compress it with a gamma curve into an 8-bit bucket. But that would pretty much require then that the image be color-correct to look good for a monitor, and these still cameras were designed to create video that played back on a monitor right off the bat and looked good.

 

Now some people are loading gamma curves into the Canon for a faux Log look, it helps somewhat, but so far, these cameras are not really sending all the DR information they could through the HD pipeline because they weren't designed for that not yet. Not to mention, cramming more DR into an 8-bit recording risks banding problems.

Link to comment
Share on other sites

  • Premium Member

...Not to mention, cramming more DR into an 8-bit recording risks banding problems.

 

Ah... Thank you David. I was sitting here with an Andromeda modified DVX in front of me, and kept scratching my head pondering why on earth Panasonic chose to neuter the normal DV output. And the obvious answer I overlooked was of course "because it's DV!"

Link to comment
Share on other sites

  • Premium Member
No, the dynamic range is limited with the Canons, more like standard Rec 709 video. That's what it was designed to be seen on after all, a video monitor. You are talking about 8-bit 4:2:0 at around 40-45 Mbits/sec at best (Nikons & Lumix are more like under 20 Mbits/sec.) The M-X Red camera has more usable dynamic range with a wider color gamut, less compression, more resolution, etc. No real comparison.

Much as I hate to argue, I'm not really trying to make a direct comparison - more just mentioning in passing that the 5D has really rather good dynamic range at the sensor and that leads to quite decent-looking images. I have no experience with the new Red sensor and I understand it goes some way towards solving some of the problems here.

So these Canon cameras would have to do something similar, take the widest dynamic range that the sensor has to offer and compress it with a gamma curve into an 8-bit bucket.

They certainly do that - in fact you have quite a bit of control over how they do that.

P

Link to comment
Share on other sites

  • Premium Member

This is all beside the point really.

 

The core issue appears to be that cheap Canon stills cameras are usurping work that, by all that all that is proper, right and decent (at least in the Redboy universe), should have gone to the RED One!

 

Yes I can certainly see how a lot of Kool Aid drinkers (not to mention Kool Aid Suppliers) might be mightily peeved by this revelation, particularly after plonking down large amounts of cash they don’t actually have, on cameras that are now possibly losing work to systems about 15% the cost. However this is Cinematography.com, not Reduser.net.

 

The best interests of this particular forum (and the film making community generally) are surely best served by a discussion of the rationale behind the use of the D5/D7, the issues that may be faced, and the general level of acceptance of the results, not endless sermons on how unworthy a prosumer product it is.

 

OK, it may well be 8-bit 4:2:0 MPEG4, but if 8-bit 4:2:0 MPEG4 is all that is required then 8-bit 4:2:0 MPEG4 is all that is required. Get over it.

 

Of course, the real issue for many a would-be video-go-to-guy is that it may uncomfortably distinguish the “fix-it-in-post” sheep from “get-it-right-the-first-time” goats.

 

Just sayin'...

Link to comment
Share on other sites

Create an account or sign in to comment

You need to be a member in order to leave a comment

Create an account

Sign up for a new account in our community. It's easy!

Register a new account

Sign in

Already have an account? Sign in here.

Sign In Now

Forum Sponsors

Metropolis Post

New Pro Video - New and Used Equipment

Gamma Ray Digital Inc

Broadcast Solutions Inc

Visual Products

Film Gears

CINELEASE

BOKEH RENTALS

CineLab

Cinematography Books and Gear



×
×
  • Create New...