What's the attraction?

[christophernigel]

Nice to know there alot of filmmaker that have alove for super 8 ,even if it means I am porselytizing or so I have told ! been called a lot worse

[Justin Lovell]

 

For me, I love the aesthetic look of film and the organic magic you can create in camera. (that said, I have a background in VFX and creating film looks in post) so I do understand the value in each approach.

 

Most recently though...

We upgraded our scanner to the lasergraphics scanstation 5k model (8/16/35). We thought our previous 2k raw data scanner was good, but this is incredible.

 

Really makes the choice to shoot film even easier for me as a cinematographer.

 

WWW.FRAMEDISCREET.COM

(we haven't made the scanstation announcement yet, just telling select people at the moment).

 

All the best!

[Carl Looper]

The title is pretty clear, but what makes you guys choose to shoot on Super 8? Last week, I saw a real cheap Canon 814 at my local thrift store and bought it on a lark, thinking it'd be fun to experiment with. Of course, once I got online and started pricing everything out, I was stunned at the cost. I mean, between the cartridge, processing, and telecine, it's like $25 a minute. Not to mention that I'll have to ship the film twice to opposite ends of the country before I know if it turned out any good.

 

I mean, it seems absurd to go through that process all to be able to shoot on the 70s equivalent of miniDV. So, I'm really interested to hear what attracts you to shooting Super 8. I'll admit it's got an interesting aesthetic quality to it, but outside of film nerds I can't imagine most people care.

 

I suspect John's question is more of a rhetorical question, ie. where he's not so much interested in any answers to the question, but in heading off any possible answers. John advances the following:

 

"I mean, it seems absurd to go through that process all to be able to shoot on the 70s equivalent of miniDV. So, I'm really interested to hear what attracts you to shooting Super 8. I'll admit it's got an interesting aesthetic quality to it, but outside of film nerds I can't imagine most people care."

 

The tenor of the lines suggests that John is open to an alternative position from the one he is proposing. For example, it only 'seems' to John that Super8 is absurd. As if the proposed absurdity might be an illusion. As if there might be some other reality behind this illsuion. But the interest he goes on to express is not so much in what that reality might be, but rather: an interest in what might reinforce the proposal of absurdity. Or in what would fail to do otherwise.

 

But in case that isn't enough - there's a second card played in advance of any answers. There will be a certain "aesthetic" to Super8, against which he feels he won't have much of an argument. How to maintain the position that Super8 is absurd? The way he does so is to frame such an argument as that which belongs only to "fim nerds".

 

Indeed as subsequent posts will reveal, what John really means by "interesting aesthetic quality" is "poor image quality":

 

"My comment about it being the "70s equivalent of miniDV" was more a analogy to how it was a popular, low-cost format that delivered poor quality images, much like miniDV." - John Sherman

 

The otherwise affectionate "film nerds" moniker starts to sound as if offense is intended. So on the basis of this cursory analysis we can rewrite Johns real answer as:

 

"It is absurd to go through that process, all to be able to shoot on the 70s equivalent of miniDV. I'm not really interested to hear what attracts you to shooting Super 8. It's got poor image quality to it, which outside of film nerds, nobody cares."

 

And simply hope that John comes back and demolishes the above proposition.

 

C

[Scott Pickering]

I grew up in the 80s when Super 8mm was dying away for video formats. Im glad I stuck with film because the video of the time looks horrible, yet the film scanned today holds up rather well other then scratches and the grain it produces. Super 8 looks beautiful when projected. I agree with others in that a telecine doesn't hold up the image quality you get when projected. Its sharper by telecine, but the grain is more noticeable. That said I still have my Super 8 cameras and have a couple rolls of reversal for the day when I decide to break them out again. I shoot 4K video now because its easier to work with and image quality is excellent, but I still like the look of film over video. I had a 16mm Bolex so many years ago in the 90s, and sold it. I wish I kept it. I love 16mm too. I have been peeking on Ebay at 16mm cameras, but Im just looking. The only reason to shoot on film now is for its look, which Super 8 gives. That is unless you shoot 65mm, which can't be beat by anything right now. Even 35mm holds up well today in resolution.

Edited January 19, 2015 by Scott Pickering

[Carl Looper]

All technology (we might propose) works according to the laws of physics. And the laws of physics are completely indifferent to notions of better or worse. For example, in what way can one argue that the physics governing gravity would be any 'better' than, or 'worse' than, the physics governing electromagnetism, or nuclear forces? By what criteria can one judge the merits or otherwise of any physical process?

 

And therein is partly the answer. Criteria. Until one proposes criteria there is no criteria. And if criteria is properly proposed, there is no longer any question.

 

For example, if the criteria for judging the merits or otherwise of gravity, were limited to work in which only gravity were explored, there would no longer be any question. Or if the criteria is limited to the understanding that "the weak nuclear force is better", or "is worse", then again there is no question.

 

Or if the criteria is limited to "the number of people in the universe who give a poop" there is again no question.

 

Equally, without any criteria there is no answer.

 

In terms of "better" or "worse" there is either unanswerable questions, or unquestionable answers, both of which are quite objectionable situations. Or just plain empty. Instead there is a much more interesting process: the elaboration of insights, be it at a physical level, a technical level, an artistic level, and so on. And using any physical, technical or artistic material one likes.

 

The interesting questions are not so much why one might use Super8 (or any other material), but what one might like to do with such material, or how one does in fact use it, or what was the result obtained using such, or what mioght be said of such a result. And so on.

 

The loaded question of why becomes the buzz of a fly at a picnic, given these other far more interesting questions.

 

C

[Justin Lovell]

Ultimately for me, it's the story that dictates the tools that I use to create an image.

 

Be it super 8 grain and organic film flares/texture or the slick crisp plastic texture of 5k raw images..

 

That and we just acquired the latest 5k lasergraphics scanstation 8/16/35. The images it produces have blown my mind.

 

Justin Lovell

Cinematographer Assoc CSC

 

WWW.JUSTINLOVELL.COM reel

WWW.FRAMEDISCREET.COM transfers/grading

[Carl Looper]

Hi Justin,

 

sounds great the 5K scanning. Does the scan station do Super8 at full 5K, or only 35mm at full 5K. Do you know what the prices will be for a full dynamic range scan (to support off premises third party grading).

 

In any case, congrats on your system (belated xmas present?) and look forward to what sort of results you are able to produce. I love it that frame discreet are themselves artist/filmmakers as much as providing digital transfers for others.

 

C

[Justin Lovell]

Hey Carl!

 

Thanks for your kind words.

 

We got the latest upgrade to the 5k model that will do 8/16/35 all in 5k resolution!

We're working on a new price list. Contact us using the contact page on our website for more details.

 

It was a belated xmas present for sure, we've been in R&D with all types of scanners for the past 2 years before finally finding something that really blew us away.

[Justin Lovell]

double post

Edited January 19, 2015 by Justin Lovell

[Will Montgomery]

We got the latest upgrade to the 5k model that will do 8/16/35 all in 5k resolution!

Very cool. So it's an optical/lens blowup to make the 8mm fill up the sensor? What amazing times we live in.

[Perry Paolantonio]

Very cool. So it's an optical/lens blowup to make the 8mm fill up the sensor? What amazing times we live in.

 

Hi Will - that's correct. In the ScanStation, the sensor/lens assembly pulls back from the film and refocuses (optically) on the film to fill the frame, if it's in 5k mode. If it's in 2k mode, the assembly is closer to the film and a 2.5k window inside the overall 5k sensor is used to create the image, allowing the scanner to run at a faster speed. Past a certain point (and there's no single answer for where that point is, since it depends on things like the film stock, the lens used, and the exposure), you're not gaining much by going with that high a resolution from such a small frame. But if you need to integrate Super 8 in to a 4k production, for example, it's a better way to do it than to scan at 2k and blow it up in post.

 

Also, you can scan to a 2k file in the ScanStation, with the scanner in 5k mode. It's a lot slower this way, but the results are better: http://www.gammaraydigital.com/blog/case-super2k

 

-perry

[Carl Looper]

I've said this often but I've done 5K scans of Super8 on my own DIY setup and playing such back on a 4K (UHD) screen and I can tell you it looks a lot better than a 2K version. It's not just resolution that is an attribute of an image, but tone and/or colour as well. Even if there were no increase in the frequency response of Super8 at 4K, there is definitely an increase in the tonal/colour response. Because there are just more bits per area of film being scanned. All else being equal, if a 2K scan did 24 bits per area of film, a 4K scan would do 96 bits per same area of film.

 

The statistical correlation between film (of any gauge) and light does not arbitrarily stop at some particular image magnification, as if film suddenly decided it will now exhibit a statistically neutral signal, just because the size of a sensor pixel decided not be any smaller. No, it's biased all the way down.

 

C

[Will Montgomery]

I've said this often but I've done 5K scans of Super8 on my own DIY setup and playing such back on a 4K (UHD) screen and I can tell you it looks a lot better than a 2K version.

Sounds like the discussions we had a few years ago on SD vs. HD scans. HD certainly won out unless it was just a horribly soft image on the Super 8... so I'm sure the lenses and stock used make a big difference as to whether or not an 8mm scan looks better in 2k or 4k or I should say how much better.

[Carl Looper]

Sounds like the discussions we had a few years ago on SD vs. HD scans. HD certainly won out unless it was just a horribly soft image on the Super 8... so I'm sure the lenses and stock used make a big difference as to whether or not an 8mm scan looks better in 2k or 4k or I should say how much better.

 

Yeah - it's a pet project of mine - exploring the attributes of incredibly tiny areas of film.

 

In terms of how much better it looks: it's looks 4X better! Not in terms of image sharpness of course (without further digital voodoo), but in terms of tone and/or colour. I mean even if you shot something with a soft lens, or just out of focus full stop, the result will still look 4X better simply because the scan is providing 4X more information.

 

But there's a prevalent misconception regarding film that it's quantification in terms of resolution (lines/mm) is all that's required to characterise the quality of the image - as if image sharpness (frequency response) were the only attribute an image possessed! But the amplitude of the image is also an attribute of the image. An MTF chart is intentionally neutral in terms of amplitude, in order that the amplitude can be used as the channel with which to express the frequency response of the film (it's sharpness). If it wasn't neutral you couldn't quantify the frequency. It would be entangled with variations in the amplitude. But when not shooting MTF charts (in order to quantify frequency response), the amplitude of the signal is no longer neutral. It goes back to being a statistical corollary of all the biases in the light exposing the film, ie. all of which are transferred as biases in the statistics of the film.

 

These biases don't just switch off at some particular scale in film. But in digital sensors there is a scale at which the sensor switches off. For example, if the sensor were an 8 bit sensor, it could only see 256 different shades of grey. Any shades of grey finer than that are lost. A way to alleviate that problem with sensors is to increase the number of bits per sensor cell. But another way to do that, when scanning film, is to use more sensor pixels per area of film being scanned. The grain in film is not a barrier. If, at some scale, the grain went from being a function of biases in the light, to being neutral, then there would be an argument for drawing such a line. But it doesn't. The biases go all the way down.

 

Just look at a projected film side by side with scan of the same film. The projected film always looks better, whether it's 65mm film, or 6.5 mm film.

 

Carl

Edited January 24, 2015 by Carl Looper

[Carl Looper]

To put it another way.

 

Even if all you shot on Super8 was a grey card, a 4K scan of that looks 4X better than a 2K scan of that. The grain is so much more subtle. Soothing. Pleasant. Beautiful.

 

Expose a more complex image and the result is even better.

 

C

[Carl Looper]

Here's another way of putting it.

 

Suppose you were to project 35mm side by side with Super8. You will notice a difference between the two. If we could put a figure on that difference we might call it X. We would say the difference between 35mm film and Super8 film is X.

 

Now lets suppose we scan the 35mm at 4K, and the Super8 at 2K, and looked at those side by side. Well the difference is no longer X. The difference becomes 4X.

 

Why increase this difference in quality, between scanned 35mm and scanned Super8? One is just making the Super8 look worse. It's completely silly.

 

C

[Will Montgomery]

To put it another way.

 

Even if all you shot on Super8 was a grey card, a 4K scan of that looks 4X better than a 2K scan of that. The grain is so much more subtle. Soothing. Pleasant. Beautiful.

 

Expose a more complex image and the result is even better.

 

C

This is good to hear. This summer I hope to have some time to shoot some 50D in Super 8 and try these 4K scans. Something I never thought I'd say.

[Jose luis villar]

I think I disagree with this, according to my tests the super 8 with a Vision 3 you can get a good scan and good lens on the camera, to close 720p. I scan 1080p 10bits then reduce it to 720 and looks fine. Anything that goes higher than this do not see sense, all you get is higher file sizes. For instance, a roll of 15 meters 1080p 10bits scanning Ochoypico occupies 34 gigabytes.

[Josh Gladstone]

I'm with you, Carl. I want each frame to be like a still photo, and I want each photo to be saved at a very high resolution, just like with my still photography. I wouldn't scan stills at a lower resolution than 1600 or 3200 dpi just because people are going to look at them online. I don't think that's overkill, it's archiving. Why shouldn't the same apply to movies? Just because the frames are being played back at 24fps is no reason to accept a lower resolution as 'good enough'.

 

Hard drive space is cheap, get the highest quality possible. That's just my opinion.

Edited January 24, 2015 by Josh Gladstone

[Carl Looper]

For sure.

 

Scanning Super8 at 4K does not create file sizes any larger than scanning 35mm or any other film format at 4K.

 

Now this doesn't mean you have to distribute a work at 4K.

 

One can scale it down to 1080p, or 720p for distribution. The very interesting thing is that when scaling a 4K scan down to 2K one will find the result can still have an improvement over one that was scanned at 2K. It depends on what you do prior to scaling it down.

 

Having a large definition scan allows you to do a lot of great things, prior to any scaling required for distribution. In particular is colour and tone. One can pull out stunning tones or colours if you know what you are doing.

 

Its not about resolution. Or rather, it's not just about resolution. There are subtle statistical biases in the grain that constitute the image. The grain is not random. It may look random if looking at just a small area. And certainly if one removed just one small part of an image that operation might, in itself, go unnoticed. But an image is in the cooperation of each and every part of an image, acting in parallel. The image is not in any single part of an image.

 

Correspondingly, increasing the scan resolution of just one small area of an image, to see if there is any improvement, can appear to have a negligible result on that small area, but one doesn't know how this small area interconnects with every other area. It's in all the interconnections that an image emerges. In their co-operation with each other, and the light to which they sing.

 

It's to do with the difference between a statistical image, which film engages, vs the cartesian image that digital engages, and how to mediate between them.

 

C

[Carl Looper]

Another interesting thing is that there's a digital technique called "dithering" which allows a large bit image to be transferred to a small bit image without contouring occuring as a result. Gradations in tone are maintained

 

Another term for dithering is "fake grain".

 

Now in the digital domain only the dithering (or fake grain) is according to a neutral randomness. Sometimes an ordered pattern is used, but either way there is no image signal in the dither pattern. It is image-neutral. However, it doesn't need to be image-neutral. One could derive a pattern from a biased signal, but the result, if applied, would be skewed by this bias and upset the image. So an image-neutral dither pattern is used.

 

If, however, one created a dither pattern, as a function of the image, while the dither pattern is no longer neutral, it is also no longer arbitrary with respect to the image. The dither pattern interlocks with the image providing a perfect dither pattern for the translation of a high bit image to a low bit image. It varies in sync with the image.

 

C

[Carl Looper]

I should add to that last post, that the image-centric dither pattern is not derivable from a given digital image. The dither pattern, in order to work it's magic, must be either statistically neutral (as can be artificially generated) or derived from a higher resolution version of the same image. If the image is from a digital camera one can't obtain any higher resolution image-centric dither pattern for the image. One can only use fake grain - the least effective of all possible dither patterns.

 

But in a film scan, such a higher resolution image-centric dither pattern (that one might have liked to apply to the digital version of the same image), has already occured in the film itself. It is already dithered by just such a pattern, by analog means rather than digital means.

 

The higher the resolution of the scan, the more powerful is this image-centric dither. The more amenable becomes the image to down-conversion, in either bit-depth and/or pixel-depth.

 

In other words there is really big difference between real grain and fake grain.

 

Carl

[Carl Looper]

To elaborate this further, there is a practice whereby one might apply an emperically derived dither pattern, by simply scanning the grain pattern of neutrally exposed film and dithering a digital image with this.

 

But this dither pattern is not image-centric - it remains statistically neutral (because derived from neutrally exposed film) - so it remains no different from a synthesised pattern in that they are both image-neutral, ie. both remain the least effective dither pattern. However, if one is forced to choose between these two dithering strategies, an argument favouring the emperically derived one, is that it obviates the need to otherwise synthesise one. However, there's nothing in principle that would make it impossible to synthesise the same pattern..

 

They are both as equally ineffective as each other.

 

C

[Juan Carlos Montero Tudose]

Hi all,

 

I run my own video production company and a WebTV to distribute our work.

 

While working with digital video for clients is cheap and fast (technically speaking) just like going to MacDonalds, it's my job doing wedding films where I really take my pride and joy -just like enjoying a homemade meal.

 

I'm also a guitarist and I feel the same when switching from a Stratocaster to a Martin: That's when I start to breathe (and grin) again. :)

 

Shooting Super 8 film is like doing things the way they should be done (I also type on a typewriter usually).

 

I have nothing against digicams, in fact I use them everyday, but just like doing math without a calculator, sometimes I need to go back to our roots.

 

Film is organic, it conveys memories from the past and it's just the way to when you want ahem... well... film look.

 

What else better than the real stuff?

 

 

 

 

Juan Carlos Montero Tudose

=======================
One Happy FilmMaker™

T+F -- +39 0381 325 522
M -- +39 339 297 0384
E -- info@tvitaliaweb.tv
Skype -- tvitaliaweb

www.unmatrimoniodafavola.tv
www.tvitaliaweb.tv

[Will Montgomery]

What else better than the real stuff?

 

That's why I buy Cokes imported from Mexico where they still use real cane sugar and not corn sweeteners. Plus I like the old heavy bottles!

 

Spoke to a filmmaker yesterday who loves his Sony F55 and Blackmagic 4k cameras but when we spoke about film he got teary eyed about the depth and feeling of 35mm. Plus he owns a color shop and all the colorists agree. I don't think it's just nostalgia (although it may be on small formats) and I don't think the numbers necessarily make film better, it's something we can't quite put our finger on (Carl may have a physics argument for film ready to go) and probably a culmination of many small things, but we love it so rock on shooting film.