How To Shoot Razor Sharp Super 8: Use Deductive Reasoning Rather Than Nostalgia -- A Newbie Primer

[Guest santo]

:blink:

 

You asked for a source to demonstrate that photographic primes of the 1970's-80's were 50% sharper than zooms of the era. That's documented in a mountain of photography magazines from the era. It's a fact beyond questioning, but you're questioning it. So I suggested you find a reference to a single test anywhere on the internet of zooms from that era from magazines like, I don't know, Popular Photography, where a zoom tests out at better than 70 lp/mm. I'm pretty certain you won't be able to do it. But even if there is one, I'll be able to pull out a bunch of tests with prime lenses that can beat it by 50%.

 

That's all I was suggesting as documentation to back up my generalization -- which is what you said was lacking and hard to buy. It can easily be generated. It's a pretty reasonable generalization. Today, with the gigantic strides in zoom lenses, there is not nearly as much of a gap. My 20% figure is probably a little too much, but we do have a pretty good source online on the photodo site which should back that up. There's got to be a 20% rating difference between the zooms and primes on the mtf based tests. However, zoom lenses today are a completely different ballgame with aspherics and exotic glass materials more common place, along with advanced manufacturing and design since the begining of the 90's.

 

Anyways, your test is perfectly decent and one you should carry out. And you should throw in the 24mm Nikon in there. I mean, you only have to shoot for a couple of seconds to get some good stills later that you can examine. And try the cameras at a few f-stops. That will give you a clear picture of what you're looking at. Hopefully your Beaulieu's c-mount lenses have been adjusted for correct backfocus for optimum results with the Angenieux 5.9 as it's going to be set at infinity, obviously. The 6-66 was no doubt checked by Bjorn, so that should be right on. The 24mm prime will not have been, but it can't hurt to do your best ground glass focuing and see what happens. Should turn out okay as it's not going to be focusing at infinity. Nikkor wide angle primes are not very sharp in the big picture of things, but it would be interesting to see how it pans out against the R10.

[steve hyde]

:blink:

 

You asked for a source to demonstrate that photographic primes of the 1970's-80's were 50% sharper than zooms of the era. That's documented in a mountain of photography magazines from the era. It's a fact beyond questioning, but you're questioning it.

 

No - I'm questioning your theory that there is a significant prime lens advantage with Super 8. This is a super 8 forum. I think there probably is a bit of an advantage to use a prime, but you make it sound like some kind of revolution.

 

Let me be clear here.

 

Prime lenses on Leicina and Beaulieu cameras makes good sense if as you say, the lens collimation works with the camera. This is probably the best way to make a sharp super 8 image.

 

Transferring Super 8 film to HDcam will likely yeild a grainier image than an SD transfer to DigiBeta. This means the bennefit of transferring to HD is acheiving 4:4:4 color, that a Super 8 negative records. Transferring Black and White Super 8 to HD therefore gives you 4:4:4 color :blink: and a grainier looking image.

 

In addition, you might be interested to know that HD -> SD downconversions cause video artifacting making the logic that one can always downconvert to SD without loss of picture quality problematic. (This is what Eric at FSFT told me) So if you are going to master your picture on HD you and better plan to show your picture on HD too. People working in Super 16mm are transferring to HD 24p as sort of a poor mans digital intermediate so they can print out to film. I don't thing HD video projection has really come of age so if you plan to make downconverted dubs to DigiBeta for film festivals you might be shooting yourself in the foot.

 

Steve

[Matt Serrins]

The real comparison is between prime lenses on a lens interchangeable super 8 cameras and the best super 8 zooms, not a comparison between primes and the fixed zooms on basic cameras, since they obviously aren't good and can't be changed anyway.

 

I don't think the prime lens claim has been proven theoretically or empirically. Just because prime lenses from the 70s resolve 50% more than zooms from that era, that doesn't necessarily mean that they will show 50% more resolving power in super 8, or that there will be a gain at all. Theoretically, if a zoom such as the schneider 6-66 resolves to max level of the film being used, what's the point of putting a higher resolving prime lens. I know that this may not be the case with vision2, but I think that this is the deductive flaw that people were pointing out. And just because a modern lens is 10x sharper than the best super 8 zooms, will putting this lens on a camera necessarily will produce a 10x sharper image? What's the limit of the film? Are there other limiting factors? Does the frame size matter (maybe I'm wrong, but it seems like the smaller information area of super 8 would make the differences between the really good and extremely good lenses less noticeable, since super 8 tops out around HD lets say)? It's not that the claim is necessarily incorrect or not intuitive, but without identifying the upper level resolving power of the film being used, and the resolving power of the lenses in question, it's hard to know whether the prime lenses are useful or overkill vs the best zooms. Maybe they can pull more sharpness out of super 8, but with such a small frame, does it justify the investment?

 

In addition, I don't think there have been any empirical tests that prove this, including the pictures that have been posted. My guess is that primes show some advantage with vision2 over even the best zooms, but I'm not sure how much and whether its worth it. If we knew the resolving power or had some shots of charts or what steve's proposing maybe this would be settled.

 

Also, what's up with the fetishization of sharpness in super 8, of all things. Of course I want my films to be a sharp as they can be, and for that reason I use a beaulieu with a 6-80 and neg, but just because built in zooms on most cameras aren't necessarily the sharpest, that doesn't mean they can't make successful films. Look at the dude who got into Cannes. Sharpness can be measured objectively, but image quality/beauty are subjective and may vary project to project. I'm not sure sharpness will really make anyone's film better, especially in super 8, which is never gonna look as sharp (or maybe I should say resolution)... as 16 to begin with. I really like super 8, and I agree appreciate your efforts to bring it to the limit, but I'd like to have some charts and figures to know the cost benefit before I declare you the winner ;) . I think you'll probably be proven right to some degree, but I wonder if it's kinda like putting lipstick on a pig, to borrow an expression from the south.

Edited December 14, 2005 by MStone

[Sam Wells]

I don't think the prime lens claim has been proven theoretically or empirically. Just because prime lenses from the 70s resolve 50% more than zooms from that era, that doesn't necessarily mean that they will show 50% more resolving power in super 8, or that there will be a gain at all.

 

Well there is a gain, but it can't be as much as the difference between the two lenses alone when seen on film.

 

A blanket prime lens vs zoom lens satement isn't useful, I mean if "all else being equal" but it never is.

In 16/S16 I'd take a Canon 8-64 over some of the 16mm format prime lenses discussed here.

 

I mean it's an interesting question how you can "maximize" Super * image; but I know what you mean about lipstick on pigs.

 

Anyway, I mean I brought up some of the nuances of rear focus only to suggest if you're going to get fanatical you'll have to consider being consistently fanatical :D If you want to explore the virtues of optics resolving 200 + lp/mm (and resolving power figures @ 1000:1 contrast are NOT everything) then you don't want to turn around and transfer your negative on an old noisy MK III or even an Ursa Gold...

 

As for rear focus issues, again, high resolution with careless attention to that is just going to cancel things out eventually.

 

-Sam

[Guest santo]

Transferring Super 8 film to HDcam will likely yeild a grainier image than an SD transfer to DigiBeta. This means the bennefit of transferring to HD is acheiving 4:4:4 color, that a Super 8 negative records. Transferring Black and White Super 8 to HD therefore gives you 4:4:4 color :blink: and a grainier looking image.

 

You're confusing yourself and listening to salesmen too much. You need to get your facts straight and think for yourself. That's what this thread is about: getting the facts straight and thinking for one's self when it comes to super8 to get maximum results.

 

First, HDCAM which FlyingSpot transfers to is 4:2:2. That is unless they're using HDCAM SR which is an mpeg4 format and then it's 4:4:4 -- but it doesn't say that on their website.

 

Second, Cinepost is only using 4:2:2 as well. There is nothing wrong with 4:2:2, and it is a hell of a lot better than DV 4:1:1. However, they don't even offer HD. Did you notice that, Steve? I've seen examples, and their work is good, but let's be realistic: are you going to listen to a car dealer who tells you BMW is overrated when they don't sell BMW? That's what I'd tell you if I worked at such a car dealership. That's not a lie or being dishonest in the strictest sense. That's business.

 

There is no extra grain which is going to be created with an HD transfer over an SD transfer, Steve. Think about it. All you're going to get extra is detail and pretty much all of the colour -- which you lose out on with a DV transfer. What you end up with, Steve, when transfering black and white to HD, or even the 10 bit SD transfer I showed on here was all the subtle gray tones in the shadow. It results in a silky-smooth gradation scale. Black and white benefits even more than colour does from a 10 bit treatment.

 

:lol: There is no extra grain created. What you're going to get is everything that's there in the film. It looks almost exactly like it does projected in fact. Even with SD 10 bit Blackmagic codec transfers. If there's "less grain" in a DV transfer, that's because you're not getting all the information off the frame. That's all. However is it really less grainy with a DV transfer?

 

Here's your Plus-X transfer on DV hot with a home movie camera:

 

 

Here's another uncorrected still (other than the text block], pulled from the short I'm working on, shot with a prime and transfered 10 bit 4:2:2 direct to harddrive (cost me a whopping extra 100 bucks over miniDV).

 

 

Now, Steve, which is grainier? Which captures more shades of gray? Keep in mind also, that mine is shot indoors with only a couple of lights and not in the bright sunshine. It doesn't have the pleasant subject matter, either. :)

 

In addition, you might be interested to know that HD -> SD downconversions cause video artifacting making the logic that one can always downconvert to SD without loss of picture quality problematic. (This is what Eric at FSFT told me) So if you are going to master your picture on HD you and better plan to show your picture on HD too. People working in Super 16mm are transferring to HD 24p as sort of a poor mans digital intermediate so they can print out to film. I don't thing HD video projection has really come of age so if you plan to make downconverted dubs to DigiBeta for film festivals you might be shooting yourself in the foot.

 

Steve

 

HD looks fantastic knocked down to SD. Films shot on super16 transfered to HD look fantastic knocked down to SD. Sorry, but I've never noticed any artifacting in any examples, nor on any DVDs for that matter -- though that's kind of a different situation as it involves even more compression.

 

Steve, you're going to have to do better than "he said..." "she said..." "this guy told me this" and do your homework. There's always a lot of if, ands, or buts and it's pretty hard not to take things out of context when "quoting" people.

[Sam Wells]

What advantage of 4:2:2 over 4:1:1 if there's no color ?

 

I've even gone off a D-beta with Y out only for B&W, that's 4:0:0 D

 

 

-Sam

[Guest santo]

What advantage of 4:2:2 over 4:1:1 if there's no color ?

 

I've even gone off a D-beta with Y out only for B&W, that's 4:0:0 D

-Sam

 

There is a big advantage over DV 4:1:1, Sam. As I think you must realize, digibeta is a 10 bit system. Although some NLE output to 8 bit which is still far ahead of the shadow crunching DV compression codec. Makes all the difference in the world with black and white.

 

If you think treating super 8 like any another motion picture format, which is what Kodak and a whole pile of other businesses are doing with what is basically "what 16mm used to be", and trying to maximize the results you can get is a waste of time, then that's fine. Sam, you're the guy who admited he hasn't shot super 8 in a decade, and recently raved about borrowing somebody's "sanyo" and shooting K40 and remembering how sharp you thought it was. That's the context you're approaching this from. It's as if the past 2 or 3 years never happened.

[Guest santo]

Thanks for posting my image and your own image side by side. This is a good reference. Now members of this board can compare the differences between my image that is hand held, transferred to DVcam and shot with a Japanese zoom camera. And yours: shot with a prime lens and a Leica on a tripod in controlled light conditons and transferred to HD. You paid one hundred extra dollars for your transfer. I paid one hundred dollars for my transfer.

 

My shot is not HD, as I've explained before. If it was, it might look a little better, actually. It's standard definition, just like yours. It's just that it is 4:2:2 10 bit, rather than 4:1:1 DV. And it, too, is handheld. It is there as an example to demonstrate that you were wrong in claiming that there is no gain to be had by upping a super 8 transfer from 4:1:1 as you contended with regards to grain or shading in black and white. You argued that increased sampling only increases grain and does not improve the shading of black and white. That's silly. Of course it doesn't increase grain, and only helps black and white look its best.

[David Mullen ASC]

I'm not sure with a transferring a b&w image if there's any resolution gain from using 4:2:2 instead of 4:1:1, since all the grayscale information is in the luminence channel, but maybe I'm wrong. There may also be a resolution difference in the tape formats of course.

 

Reading this thread, I see that the problem is more of a philosophical dispute. I mean, it's hardly news that using sharper prime lenses shot at optimal stops using sharper film and then using a better telecine and tape format will produce sharper results! That's true even with 35mm.

 

The philosophical dispute is over the very nature of the Super-8 format; is it being used to emulate professional 16mm photography but on a budget, or is it being used because it has a unique look that partially derives for its years of use as a family home movie format, so it carries certain associations of that, more of an ephemeral, nostalgaic quality?

 

If the latter, there is nothing wrong with using an old zoom lens and Kodachrome stock, for example -- if anything, you don't want it to be mistaken for modern 16mm photography.

 

If the first, then you have to stop and consider where the price point is where you are spending almost as much to improve the technical quality of Super-8 as you would by shooting in 16mm. I mean, if you really want a sharper image, using a larger negative format makes more sense. And if you want it to be a grainy yet sharp image, there are ways of achieving that in 16mm.

 

So while it's interesting from a technical standpoint to see just how much quality can be pushed out of the Super-8 format, just as with any format, I wonder about its practical applications when 16mm already exists. Unless you really are still saving a lot of money by sticking to Super-8.

 

Anyway, let's stop putting down people who want to shoot in Super-8 using cheap zoom lenses and old film stocks because they might be getting exactly the sort of images that they want to get.

Edited December 14, 2005 by David Mullen

[Sam Wells]

There is a big advantage over DV 4:1:1, Sam. As I think you must realize, digibeta is a 10 bit system. Although some NLE output to 8 bit which is still far ahead of the shadow crunching DV compression codec. Makes all the difference in the world with black and white.

 

Sorry, Sankyo. I'm not talking about how sharp I "thought" it was I'm talking about how sharp it actually was. I know what sharp is.

 

I would be very curious to see a difference between Luminence resolution in D-beta vs DVCam. I bet you would see little difference. (It's actually higher in DVCam than Beta SP). In any case, my point was chroma compression is in a sense irrelevant; I've taken Y only from D-beta which is in effect 4:0:0.

(note you can't just feed that to anything; a VTR that wants colorburst will not be happy with it...)

 

It is like any other motion picture format, and I've used all the materials you're talking about in 16mm and the ONLY difference is in picture area. Same with 35mm. (Or 65mm)

 

Any approach that optimizes S8 can be used to optimize S16, etc. Best lenses, transfer, et.

 

I'm not objecting to your quest to get the most out of S8 per se, and although it's true I haven't shot S8 since forever I have actually considered it in a possible situation where 16mm would cause "political" issues and I'd prefer not to shoot DV, etc.

 

(I also seem to have made Kodachrome 40 look good in both S8 and 16mm even though everyone here says I "can't").

 

But as David points out, not all goals are the same. If I prefer 7274 to 7217, so what ? I'm not making films about test charts, and if I've given away 5 or 10 lp/mm at some contrast ratio I don't care.

 

I mean I saw "Far From Heaven" and I didn't write a letter to Ed Lachman saying "You used an EXR stock and that's OLD. I paid $8.50 for my ticket and I want MORE resolving power than you gave me, please refund my money."

 

-Sam

[Guest christoph]

for what it's worth, with B/W footage, it doesnt matter if you use 4:2:2 or 4:1:1... also the theoretical Y resolution of Digibeta and DV is the same (note that DVCAM is not higher quality than normal DV, just a better tape format).

the benefit that you will get from a digibeta transfer is slightly better sharpness (DV is compressed), more headroom for correction (Digibeta is 10bit, DV is 8bit) and better long term storage savety (although you should be reasonably save with DVCAM).

 

if this is worth the extra money and inconvenience or not depends a lot on personal preferences and budget i guess.

personally i think DV delivers amazing quality and workflow benefits for the money.

++ christoph ++

[Sam Wells]

I used DVCam generically. (easier than to type DVCam / miniDV)

 

Christoph - do you think I'd really see a difference in Luminence only between D-beta & DVCam ?

 

(I've got D-beta dubs on DVCam here, but since I don't have an A 500 around the house :D

 

Certainly when making those B&W dubs I'd have to look pretty hard to tell which was which)

 

-Sam

[Guest santo]

for what it's worth, with B/W footage, it doesnt matter if you use 4:2:2 or 4:1:1... also the theoretical Y resolution of Digibeta and DV is the same (note that DVCAM is not higher quality than normal DV, just a better tape format).

the benefit that you will get from a digibeta transfer is slightly better sharpness (DV is compressed), more headroom for correction (Digibeta is 10bit, DV is 8bit) and better long term storage savety (although you should be reasonably save with DVCAM).

 

if this is worth the extra money and inconvenience or not depends a lot on personal preferences and budget i guess.

personally i think DV delivers amazing quality and workflow benefits for the money.

++ christoph ++

 

What you get is, 10 bit 4:2:2 vs. DV black and white:

 

10 bit: a compression ration of 2:1 instead of 5:1, no more artifacts in the shadows and video noise, and you're working with 1024 shades of possible gray per pixel instead of 256 shades of possible gray with DV. You can see the difference extremely clearly in footage by doing DV proxy files and playing them back to back. The extra cost for a short film of 75 minutes raw footage was a $100 premium plus an external harddrive. Worth every damn cent if you're going to be putting a short in festivals, as it is certainly in your best interest to show in digibeta as it looks a lot better than DV does up on the screen. Just the fine tuning potential picture correction at home at your leisure outside of an expensive post house environment, is worth who knows how much? A thousand bucks?

 

But you're right on one thing, some people prefer to spend more money for less.

[David Mullen ASC]

Yes, 10-bit versus 8-bit, and compression levels, would make more of a difference for a b&w image than 4:2:2 versus 4:1:1, which refers only to color subsampling, not compression.

[Guest christoph]

Sam,

 

the theoretical Y resolution between DIGIBETA and DV is the same (720x480 for NTSC, 720x576 for PAL), i say theoretical because you're bound to loose some detail in the DV compression. personally i find it hard to see the difference on a normal monitor, but it softens the grain pattern slightly (which strangly enough in some cases looks better than the uncompressed footage)

 

 

Santo,

 

i know the benefits of DIGIBETA, and i'm using them most of the time myself.. but the extra cost is not only in the transfer but also in the postproduction. not everybody has a computer with SDI inputs, and not everybody has a digibeta player. direct to hard disk transfers are a nice option but most typical s8 shooters dont have the knownledge to cope with proxies ect (and most wont see the difference ;)

also you can't see the difference between 10bit and 8bit on a normal monitor.. the 10bits just buys you more headroom for correction, which shouldnt be necessary on a good transfer.

 

for me super8 is a low cost tinkerer format, once i start spending spending the money for leica lenses (not to speak of the fcal range problem) and transfering to HD or something, i'd rather get an old bolex or an arri st and shoot 16mm on outdated stock ;)

 

++ christoph ++

 

ps: i recommend this read for those interested in DV compression and chroma subsampling differences:

http://www.nattress.com/Chroma_Investigati...omasampling.htm

note the differences (or rather, the not so differences) in the luminance channel Y.

Edited December 15, 2005 by christoph

[steve hyde]

ps: i recommend this read for those interested in DV compression and chroma subsampling differences:

http://www.nattress.com/Chroma_Investigati...omasampling.htm

note the differences (or rather, the not so differences) in the luminance channel Y.

 

 

....this is a great link. Thanks. Very informative.

 

Cheers,

 

Steve

[Michael Ryan]

Hello All,

 

This is interesting, but I'm going to add a slightly differnet perspective. I would rather watch a really deeply moving story shot on a Kodak M2 Super 8 (really basic camera), than a borefest that was shot with a 65mm Panavision camera.

 

The technical side is important, don't get me wrong, but it's not the only thing.

 

Mike

[Rick Palidwor]

Hello All,

 

This is interesting, but I'm going to add a slightly differnet perspective. I would rather watch a really deeply moving story shot on a Kodak M2 Super 8 (really basic camera), than a borefest that was shot with a 65mm Panavision camera.

 

The technical side is important, don't get me wrong, but it's not the only thing.

 

Mike

 

Bravo Michael

I find that too many people spend too much time wondering if X device is X percent sharper than Z device. I applaud people who try to get the most out of their gear, which is what we were doing with super-duper 8, but there is a point at which it can become a distraction. To reiterate, it's not the gear, it's what you do with it. Welcome to the board Michael.

 

I also agree with David Mullen above:

"So while it's interesting from a technical standpoint to see just how much quality can be pushed out of the Super-8 format, just as with any format, I wonder about its practical applications when 16mm already exists. Unless you really are still saving a lot of money by sticking to Super-8.

Anyway, let's stop putting down people who want to shoot in Super-8 using cheap zoom lenses and old film stocks because they might be getting exactly the sort of images that they want to get."

 

In this particular case, it's not how "sharp" the gear can get, it's the look you are trying to achieve. I can applaud Santo's efforts and achievements but I get tired of his put-downs of anyone who is not trying to do what he is trying to do.

 

Rick

[Anthony Schilling]

but the extra cost is not only in the transfer but also in the postproduction. not everybody has a computer with SDI inputs, and not everybody has a digibeta player. direct to hard disk transfers are a nice option

 

The deck cost and availability is my problem, still haven't even found one in town to rent... I agree, direct to hard disk might the best option for quality and convienance.

[A.Oliver]

A post for newbies, I guess.

 

The best lenses cost the most when it comes to zooms from the 1970's. If that's what you're most interested in. The best lenses ever made for super 8 zooms include the Angenieux 6-80mm f1,2 and the Schneider 6-66 f1,8. The former NASA project developed lens appeared in C-mount on the Beaulieus as did the latter, but the latter also was redeveloped to ridiculously high criteria with the addition of extra corrective elements and coating by Leica for the legendary Leicina Special.

 

i wonder if leitz opted for the schneider as the quality of the mechanics within the zoom were better than those springs with the 6-80 lens. I was advised some years ago to avoid angenieux lenses with springs inside the lens. Over time the springs weaken and dont pull the moving element back uniformly ( any truth in that? ). I notice canon lenses have springs. Schneider optics dont have springs, they appear to be a much better constructed internally. The quality of images thru my 6-80 ( recently serviced ) is stunning. i would rate the 6-66,6-70 and 6-80 as the best zooms for the super 8 format, the 6-80 being the only true professional lens as the aperature is measured in T-Stops.

[Bob Last]

i wonder if leitz opted for the schneider as the quality of the mechanics within the zoom were better than those springs with the 6-80 lens. I was advised some years ago to avoid angenieux lenses with springs inside the lens. Over time the springs weaken and dont pull the moving element back uniformly ( any truth in that? ).

I can tell that one of the main reason Leitz chose the Schneider because the Schneider is less front heavy. The Angenieux are too long for the Leicina.

 

The quality of images thru my 6-80 ( recently serviced ) is stunning. i would rate the 6-66,6-70 and 6-80 as the best zooms for the super 8 format, the 6-80 being the only true professional lens as the aperature is measured in T-Stops.

When I saw the 6-80 on a 1.5m screen projected by a Beaulieu 708EL, I was amazed. There are so many details that you cannot see in other lenses or projectors. Then I projected it in a 2m screen in Elmo GS-1200 Xenon and I was stunned. It is unfortunate that to really appreciate the beauty of Super 8, it has to be in these high cost equipment.

[Guest santo]

but it softens the grain pattern slightly (which strangly enough in some cases looks better than the uncompressed footage)

 

yes, it does in some cases. at the expense of everything else. in most cases a DV transfer actually makes the image appear "grainier" than it is as it imparts video noise.

 

...i say theoretical because you're bound to loose some detail in the DV compression. personally i find it hard to see the difference on a normal monitor, Santo,

 

i know the benefits of DIGIBETA, and i'm using them most of the time myself..

 

most typical s8 shooters dont have the knownledge to cope with proxies ect (and most wont see the difference ;)

also you can't see the difference between 10bit and 8bit on a normal monitor.. the 10bits just buys you more headroom for correction, which shouldnt be necessary on a good transfer.

 

(...)

 

ps: i recommend this read for those interested in DV compression and chroma subsampling differences:

http://www.nattress.com/Chroma_Investigati...omasampling.htm

note the differences (or rather, the not so differences) in the luminance channel Y.

 

Your link is an interesting one, but what we have is some guy trying to sell us on what he thinks are his great solutions to improving DV images, rather than a stark full frame example showing us how things really are. So I'll provide a simple example. I decided to just take the still I used earlier in DV proxy form. I used maximum quality settings all around. In a way, it does indeed show some smoothing of super 8. Unlike most DV comparisons, it looks less gritty and grungy compared to digibeta black and white. DV looks not only softer, but dirtier most of the time. This is, ironically, a rare time when it doesn't. Some egg on my face :) ? Maybe, maybe not, but carrying on to really have a look...

 

 

 

Look at the detail differences in the shirt. The sharpness of the edges in the knife. Most of that has to do with the DV's 75% reduction in possible gray tones for pixels. Plus the 250% increase in compression. The difference for black and white work between digibeta and DV is striking. It is even more obvious in motion, and really disappointing on the big screen at film festivals for those that chose DV (being an informal party to a few black and white shorts finished in both formats). It's even shocking how much of a difference it makes doing a DVD copy! With black and white there is no question about it. With colour you have a little more leeway, but it can be seen instantly with only a little scale.

 

I don't know what kind of monitor you use, christoph, but checking in on this site/thread tonight I was amazed you would write something like that and claim regular experience in such matters. I even pulled out an old beat up crappy Dell laptop from a closet and had a look and I could certainly see the difference just with a still! So I decided to post.

[steve hyde]

Santo,

 

I'm using an old monitor (nine years old). Maybe you have a 30" Apple cinema display ( I don't know?) Anyway, I do see a difference between these images although I have to look very carefully. For me the knife in the 4:2:2 image looks sharper. The rest of the differences are indistiguishable on my old monitor.

 

Your workflow sounds very interesting. How does this work?

 

You have your footage transfered to hard disk at a bitrate that is comperable to DigiBeta - then you generate proxy files for off-line editing and conform your EDL to the uncompressed files and take your hard disk to the posthouse for a DigiBeta or HD master tape?

 

Is your online in real time?

 

or

 

Will you have to do a lot of rendering during the online?

 

Steve

[Guest santo]

It is unfortunate that to really appreciate the beauty of Super 8, it has to be in these high cost equipment.

 

Although we have different goals with the format -- you being a home filmmaker who demands and appreciates the best, and me a public filmmaker -- this is part of the core philosophy I developed in my super 8 efforts.

 

It's a philosophy that is lost on many, probably most. It's something like those car tuners/builders creating all those "rice rockets". Racism and racist terms really bother me, but that's what everybody seems to call those things, most of them Asian who use the term. Anyways, the concept is to take the least and make the most of it. Sure, one could argue that these guys would be way ahead of the game building a newer generation Corvette or Porsche 911 into a custom super car. It would cost a little more, but you'd have to do a lot less to get more speed and better handling out of it. Those people arguing that are missing the point. Just like people saying, "why not just shoot 16mm?" are missing the point here.

 

Like all good philosophies, the maximization of super 8's potential for narrative filmmakers as a goal can be summed up in one sentence:

 

Super 8 is the format to choose when you really want your film to look as much as possible like it was shot on film because it accomplishes that better than any other format, but the goal is for it not to look like bad film -- which is what you get unavoidably using the traditional super 8 hobby approach to short filmmaking and the methods and means inherent in that.

[Guest santo]

You have your footage transfered to hard disk at a bitrate that is comperable to DigiBeta - then you generate proxy files for off-line editing and conform your EDL to the uncompressed files and take your hard disk to the posthouse for a DigiBeta or HD master tape?

 

Is your online in real time?

 

or

 

Will you have to do a lot of rendering during the online?

 

Steve

 

With both the HD project and the SD project, the workflow is the same. It's on hard disk via the Blackmagic 10 bit codec available on their website. It's transfered directly to a hard drive and that's how I get it via Fed Ex.

 

The location sound was recorded on a miniDV aimed at the scene and kept running over a series of shots and takes, then matched up with the dialog in the various takes which it is edited from.

 

Editing is using a proxy file. I made DV files of the 13 seperate mov files (in the black and white project) and edit from those. The reason being, even with the best computers commonly available, adding real time effects in 10 bit makes things "go wonky" if you will. In 10 bit SD you can pull off basic edits without transitions, but that's about it.

 

When the project is edited, a good solid edit, a DVD is struck and it goes out to the talent for voice dubbing. Because it's shot on super 8, the camera is too loud, and I'm doing it '60 euro auteur style (sounds cooler than it is) and have them dub in their lines themselves using a miniDV cam and a low-end Sennheiser in 16 bit. Nothing fancy. Don't need it. They just need to do it in a closed, quiet room with the DVD playing and the mike aimed at them on a tripod.

 

These are actually both off today, so after a Christmas/New Years vacation, I'm back in town and hopefully will have the talent's best runs at the dubs within a week or two of that. If there's problems, I'll drop by and give them a hand. If needed, I'll bring them into a studio.

 

Then I synch up the dubbed sound and room tone I have them record and get into the sound mix. Dialog is kept short and economical in something shot with super 8 -- at least it should be -- so it's not a big deal. Final Cut or even Vegas -- I switch between both and actually think Vegas has better sound features -- makes it super easy to synch things up. Just get it close and nudge the sound track a frame at a time until it's there and apply a tiny bit of stretch if needed.

 

Once everything is just right -- and this involves a few DVD dubs, too, so that it can be seen what the submission to film festivals will look like, and played on a regular old TV also -- DVDs are made for festival submissions. A final uncompressed mov and/or avi file is created with all the scenes put together and transitions created if needed and DVDs are created from that. When accepted (hopefully) the finished project on harddrive will be sent out to a post house for a digibeta dub using Avid quite likely. That's for the black and white SD. With colour HD, obviously it is posthouse for the final assembly with the edit master file., and colour correction.

 

The black and white workflow described is not in the best financial interests of post houses, and pretty experimental at this stage, quite frankly. With colour, you likely would be better served in some ways keeping a digibeta tape reconstruction at a post house, because you need to have a little better handle on how the project looks when it comes to colour correction. Although the truth is is that there are a lot of colour films which make it into festivals all the time which never went through that. Doesn't mean it isn't in your best interests for best results. With black and white, what are you really dealing with? brightness and contrast. You can get a pretty good final correction without a post house with a little cross format referencing and tweeking.

 

I am sorry if I have been sharp with you on this thread, Steve. It is the nature of the broadcasters, producers, and film studios and whatnot that I have to contend with which has made me not the most pleasant person in addressing people countering what I have to say. You are a sincere person, and don't merit a harsh treatment.