I have been telling my students and others for the past year or so that if they plan to make the next great independent feature that it will ultimately cost less(for the 35 uprez - release) to shoot in HiDef rather than MiniDV. I think most on this list understand why. However, recent improvements in the capability of Up-rez facilities may be changing this(as has been mentioned here recently). So who knows where this will take us in the next year or two.
A local facility here in Toronto who are doing a lot of work for American clients have demonstrated an amazing ability to take MiniDV and blow it up to 35 with very respectable results. Aside from the proprietary digital smoke and mirrors that they apply; they recommend right off the top that at tapes or selected takes be bumped up to D-Beta(in part because it's 4:2:2) and then normal Post procedures followed. Now, we all know that if you bump 4:1:1 to 4:2:2 you still have 4:1:1, but I am wondering if there is something else going on(obviously there is). Someone here recently suggested that the Varicam to 35 yielded slightly softer results than HD-Cam, but I would think it would still be vastly superior to DV. I wonder if there is something tangible happening here or if the eye-brain somehow accepts a 35mm resolution print as superior in spite of how it was originated (if it was well crafted of course).
I remember 20 or so odd years ago when the NFB was very big on 16mm as origination for 35 release.
They did a lot of research and one of the conclusions they came to was that 16 blown up to 35 and projected to the same width; would reveal much better results in the 35 print even if it went through an extra optical stage. This is not too hard to understand because we would not be observing 16mm grain, but then that was then and it was long before 7245 and current emulsions such as Vision2.
I am also wondering if others have observed this tangible improvement in DV to 35mm in the past year or so and what your sense is of where this may go (scary on one hand but!!).
The other sense I am getting(although my own experience does not support it) is that there may be a resurgence of 16mm as original. I have always felt that 16 has never been explored and exploited enough as a medium for independent production.
Best Wishes, from the great white north.
Lance et al:
Someday I'm going to figure out who "they" are, other than those tiny little voices in our own imaginations that tell us what we want to hear anyway. After all these years, does anyone really believe that "bumping up" or "up-rezing" an image with inherently limited bandwidth or resolution or any image to a recording/capture medium with such characteristics, to a recording/ capture medium with higher bandwidth or resolution, actually increases the image quality of that image???? I hope not, or Geoff may, ultimately, be forced to rename this forum the "Fairy Tale Forum".
While current technology has improved almost all of the chemical and electronic capture techniques, none has, yet, achieved the vaunted state of alchemy that can turn Nickel into Silver or Silver into Gold. None.
For those of you with some basic understanding of Physics, please envision a 1 kilometre length pipe with a diameter of 5 millimetres, connected to another 1kilometre pipe with a diameter of 10 millimetres. If 100 Litres of water is pumped into the 5 millimetre pipe, then measured, 2 Km later at the exit of the 10 millimetre pipe, will there be any more than 100 Litres of water??? Probably not. Will the water's molecular makeup change??? Probably not. If the water is dirty going into the pipe and we filter it, it may be cleaner at the other end, but the volume of water will probably be slightly smaller.
No one has found a way to turn DV into HD or 35mm Film quality. The best we can hope for is that whatever we start out with retains the original quality. The ONLY reason "bumping up" has ever been a commendable or desirable process is for reasons of post-production convenience or for specific distribution requirements. Not to improve the quality of the original product. Never.
GEORGE C. PALMER
HD and Digital Imaging Services
>it will ultimately cost less(for
the 35 uprez - release) to shoot in HiDef >rather than MiniDV.
I think most on this list understand why.
Not me. That would be due to a choice on a post-production route, not on a strict direct cost basis. Sure the results of the DV-originated material will always be lesser in quality to the HD material, but they don't have to be more expensive.
“They recommend right off the top that at tapes or selected takes be bumped up to D-Beta(in part because it's 4:2:2) and then normal Post procedures followed."
Most likely to minimize future loss of image information. DV25 codecs can vary in quality and there can be loss at various stages of post. DigiBeta will hang onto that info better downstream.
"16 blown up to 35 and projected to the same width would reveal much better results in the 35 print even if it went through an extra optical stage."
One very strong reason would be the general quality of 35mm projectors over 16mm ones. And it can be extremely difficult to pump enough light through a tiny 16mm or Super-16 frame to get a properly and evenly illuminated image on screen without melting the print. It's far easier to do with 35mm film.
Most important lesson in all of this is that the system is only as strong as the weakest link in the chain. So often we only consider the shooting medium, but it can also be the various post steps, the travel between those steps, and the exhibition format.
George C. Palmer wrote :
>After all these years, does anyone really believe that "bumping up" or >"up-rezing" an image with inherently limited bandwidth...actually >increases the image quality of that image????
Well you know I guess I do.
I mean I have to believe in "loss less" compression, I keep getting told so anyway, and after all that is just a system that guesses what I don't want and throws it away and then on replay guesses what was thrown away.
Now if that works why shouldn't a system that guesses what was there in the
gaps between the pixels? I'm actually fairly serious about this.
I actually prefer the idea of guessing in post, in a quiet and adjustable way, than guessing in an automated on location kind of way.
Geoff Boyle FBKS
Director of Photography
I saw a demo today which included examples of 16mm and 35mm originated
material transferred to HD and then output to 35mm film, HD originated
material output to 35mm, and SD video originated material upconverted
to HD and output to 35mm.
The 35mm, 16mm and HD material looked pretty good on the smallish screen, and in a blind test I might have had some trouble distinguishing one format from another. (Apparently, getting transferred to HD is a great equalizer.)
However, the SD material did not look as good as any of the other three formats even thought it had been up-converted with a great deal of skill on a state of the art system. However, most independent features which are fortunate enough to get a distribution deal don't get a theatrical release, they're more likely to get a video distribution on DVD and VHS.
I tell independent producers and directors to focus on a video finish, but protect it for a theatrical release just in case.
Los Angeles based Director of Photography
West Coast Systems Administrator, Cinematography Mailing List
>I am also wondering if others
have observed this tangible improvement >in DV to 35mm
in the past year or so and what your sense is of where >this
may go(scary on one hand but!!).
I have also seen astonishing improvements in DV film-outs in just the past year. It's getting about as good as one can possibly expect. Film still has the edge in lack of vertical sampling artefacts, latitude, and high-frequency detail (extended MTF), but the DV xfers are arguably better than many 16mm blowups (pre-7218, at least!) [and I *did* say "arguably" not because I want to start an argument, but because I fully acknowledge that depending on what you're looking at in a shot, you can have a strong disagreement with my perceptions, grin. No flames, please : I don't want to start a DV vs. film war because such things are silly.
Different tools with different strengths and weaknesses.
Aside from the MTF issues (a.k.a. disappointing long/wide shots), the dead giveaway for me on DV or other SDTV film-outs is the residual scan line structure. Even shooting with camera detail down doesn't hide it enough, and while a good DP can shoot around most SDTV limitations, it's mighty hard to avoid *any* scene elements that call this out.
>The other sense I am getting(although my own experience does not >support it) is that there may be a resurgence of 16mm as original.
7218 may indeed drive that. I'm certainly fielding more questions about
Camera / Editing / Engineering
Menlo Park CA USA
>And after all that is just a system that guesses what I don't want and >throws it away and then on replay guesses what was thrown away.
In our drive to better understand the myriad of issues relating to the imaging qualities of digital media we sometimes say things that can lead non engineers to very clear levels of misunderstanding, and there are many reasons for the confusion in a world where we are forced to leave the familiar womb of analog technology for one that runs counter to our observations of the world around us. This is made all the more confusing when signals of different types ( audio and video ) are not handled or processed consistently. Those who are more familiar with chemical based media think in terms of a frame that contains real images -- so the notion of leaving something out of that image seems, on the surface, to be unclean and prone to distortion. But it is not. Most of the information we capture on film is redundant. In real life most of what we see is redundant and so we assume it is necessary to seeing reality. Not so.
The rules for digitalisation and compression of audio and video are not consistent. Our friends at Dolby Labs can better explain why, but in general the ear is not as discerning as the eye. Thus Dolby can compress a 1.56 megabit audio stream into 256 kilobits or less without the listener detecting any change. This is accomplished by applying principles of 'psycho-acoustics' than can fool our ears.
However our eyes, for the most part, will not tolerate such tampering – so video compression requires that we not unduly tamper with those parts of the data stream that define the visible elements of the image. Fortunately, images are comprised of largely redundant data, that is to say, duplicate information. There is no guessing about what's redundant and what is not -- for the digital stream can be examined at high speed -- and the redundant information removed without in any way affecting image quality. This seems very strange to those of us who grew up in an analog transmission world --where every detail had to be transmitted in every frame. In analog, it takes the full channel bandwidth to transmit a black frame with a single white letter of any size in the middle. The bandwidth is required to define the edges of the letter -- so 29.97 times per second the frame and the transients required to define the letter edges must be transmitted. In a digital environment, the frame need only be analysed once -- and transmitted once. If it is to remain on screen for many seconds it never needs be transmitted again, and herein lies the magic of digital compression.
There is no guessing -- on either end of the digital stream. The video compressor easily identifies redundant information from frame to frame. It need only transmit what has changed from the last frame -- for the receiving station already knows what was in the last frame. No guessing on either end.
Geoff Boyle correctly asserts that uprez conversion can indeed improve image quality. While this appears to be more effective with digitally originated images, especially DV at 720x486, it can also enhance some characteristics of analog sourced images. The degree to which uprez conversion improves image quality is largely dependent on the latitude and bandwidth of the higher rez format. Thus the edge of an object on a 720 pixel frame, which may appear uncertain to the eye, is very precisely defined in the source data stream. Once sensed, that transition takes on the characteristics of the higher resolution image. The eye detects this very real increase in resolution and perceives the image as being of higher quality. Of course it isn't. It's still the same image as before with all of the shortcomings of the lens and sensing capabilities of the original camera.
Then why does it look better? Because the uprez processing shifts the image in into a more subtle luminance space, a better defined chrominance environment and a higher bandwidth transition capability. While it is true that the half chrominance resolution of 4:1:1 sampling is no better, the edges of chrominance are sharper after conversion to 4:2:2. So chrominance bandwidth appears to be higher -- even though it isn't.
And how is this possible? Because there is far better image quality in DV images than is visible when they are converted to video using technology that does not seek to improve the image. Engineers make assumptions about how DV images ought to appear based on the limitations of the camera and inherent bandwidth of the source data. The validity of uprez improvement in these images argues that earlier assumptions about data format and usable image information need further adjustment.
One need only look at the film-out of some of uprezzed DV materials to see how well the uprez operations can improve images in theatrical environments. Is this a threat to those creating materials in 1080i, or 720p @24? Of course not, but it makes one wonder what we might find in these images when they are uprezzed into a far larger and higher bandwidth environment.
ImageStream Studios USA
Lance wrote :
>I remember 20 or so odd years ago when the NFB was very big on >16mm as origination for 35 release.
Yes you certainly would be observing 16mm grain, and magnified at that. (I'm all for shooting 16mm, I do shoot 16, but I prefer to be real about it).
The Holy Grail is blowing up the picture while not blowing up the picture grain, but guess what the picture is made of.…..
Jessica wrote :
>The 35mm, 16mm and HD material looked pretty good on the smallish >screen, and in a blind test I might have had some trouble distinguishing >one format from another.
My question is WHY submit 16/S16 to this "equalizer" ? So far, I've been unimpressed with HDCAM as a DI in 16mm blowups. D6, I don't know. D5-HD, don't know.
Lance wrote :
>I am also wondering if others have observed this tangible improvement >in DV to 35mm in the past year or so and what your sense is of where >this may go
I've seen Webcam blown up to 35mm; it worked fine, looked like BIG Webcam
My opinion is, use the smaller vid formats for what they are; use their "guerrilla" aspects, "techno" aspects, use them to create something like impressionism, or do something we don't have a name for yet.
Or use them when they are the only means you have for telling a truth you must tell.
>7218 may indeed drive that.
I'm certainly fielding more questions about >it recently.
Oh come on. I can show you blowups from ECO-7255/7252, 7247, etc that blow any DV blow up out of the water. So to speak.
Sam Wells wrote :
>My question is WHY submit 16/S16 to this "equalizer"? So far, I've been >unimpressed with HDCAM as a DI in 16mm blowups.
HDCAM makes a poor DI, D6 is OK but ALL the video formats suffer from the same disease : less high light latitude than film, decreased color space.
All you have to do is take a print with highlights that you know looks great on a screen and rack it up on a telecine and transfer to any HD format. All of a sudden you have to make choices about highlights that will be compressed or thrown away, that otherwise look fine in the print.
Mark Smith DP
Oh Seven Films Inc.
George Palmer writes :
>After all these years, does anyone really believe that "bumping up" or >"up-rezing" an image with inherently limited bandwidth or resolution or >any image to a recording/capture medium with such characteristics
Bandwidth, TV lines of resolution, or line pairs per millimetre are probably not being increased at all with the various up-rez methods in use today, but, the sharpness of edges (and some other factors) in the image can definitely be increased a great deal. Properly done, this can have a significant positive impact on the "apparent resolution" of the image.
I can not find it in the dictionary but I have always liked the term "acutance" for describing sharpness, particularly in a low resolution environment. An excellent example of high acutance yet low resolution is in cartoon animation. Frequently there may be a very limited number of line pairs in the entire picture width but the image is nevertheless quite sharp.
High resolution images usually have high acutance or sharpness. Very sharp images do not necessarily have high resolution.
John D. Lowry
Lowry Digital Images
There is no such word as "up-rez," nor is there any phenomena
under the Newtonian laws of physics that will magically give you more
resolution than you start with. Up conversion will give you a larger image
but it won't give you more resolution.
Scott "saw enough smoke and mirrors at COMDEX" Billups - LA
Scott Billups wrote :
>There is no such word as "up-rez," nor is there any phenomena under >the Newtonian laws of physics that will magically give you more >resolution than you start with.
I tend to agree. I never saw any image get better because a half pint of DV got transferred to a half gallon jar of HD or a gallon jar of 35 mm film..
On another note, any one have any sense about digital degradation piling up as one codec gets transferred to another. I have seen some of this effect on a project I shot as it goes through the post process.
.......HD DVC pro origination > DV cam Down converts > into avid Component >out of Avid to Digi beta > and so forth. Granted this is just the rough cut, but the shit is starting to pile up.......
Any body else noticed this?
For what it's worth - I've had experience blowing up DV, Beta, Digi beta,
Super 16mm (through Digi Beta twice) and Hi Def to 35mm - to my eye the
best quality result was in the following order:
Super 16mm to Digi beta to 35mm (one grade was during film to video transfer, the other was after the transfer to video). Hi Def to 35mm (graded after online and before the 35mm burn) Digi Beta to 35mm (spirit was used after the online, but I worked with the Digi Box pre-grading whilst shooting) Beta to 35mm (1987 - with a grade after the online but before the 35mm burn) DV to 35mm (material graded once before the cut - and after the online but before 35mm burn).
Yes, I liked the super 16 to Digi over the Hi Def - this list is about the transferring of media through other media so I've not included film to film, and as for DV as a generation medium I've shot a few long form dramas for then transferred to Digi for TX but with a "film look" effect added in post (I'm not a convert); Equally I've shot a few Digi Beta dramas with film look effect in post and it's good - it's a standard here in the UK as opposed to shooting on super 16mm though that's still done.
But in all of this what does best quality mean ? In my case it's where the most "cinematic" image was achieved. Of course - the question then arises what does "cinematic" mean?
Also, given Geoff's recent "questioning" of just how good Hi Def currently is - there is also an issue of how much to settle for. Arguably you should never settle for anything below top quality - but, if you shoot on a format that is below that standard you have to buy into the "lesser" quality as being part of the look. I'm still struggling with the issues - Opinions anybody ?
Terry Flaxton (Brit based DP)
Terry Flaxton wrote :
>Digi Beta to 35mm (spirit was used after the online, but I worked with the >Digi Box pre-grading whilst shooting)
What would the Spirit be used for here? I'm confused, unless the end result was meant to be tape, and the film out was to get a look...
Mark Smith wrote :
>I have seen some of this effect on a project I shot as it goes through the >post process.
Seems like a dumb path with lots of possible concatenation of compression schemes.
Why not DVCPro HD into Avid via SD SDI (not component) and out to Digibeta? Or do work tapes from DVCPro HD onto DigiBeta. A Meridien Avid at uncompressed or 2:1 or even 3:1 to DigiBeta looks quite good. (I'm only going from Digibeta originals from film (some through a D5HD generation) or Mini DV right into the Avid via SDI.)
Nothing wrong with DVCam, but it's not a good intermediate format, as it's 4:1:1 and 5:1 compressed. Think of it for origination only.
Jeff "uncompressed" Kreines
Terry Flaxton writes :
>But in all of this what does best quality mean ? In my case it's where the >most "cinematic" image was achieved.
Cinematic means many things but probably the most important in the context of digital images is to TURN DOWN (OFF) THE APERATURE CORRECTOR. The long slow roll-off characteristic in the modulation transfer characteristic of film is completely different from the "100% depth of modulation almost to the limit of the system resolution" then an extremely fast roll-off to zero MTF of many digital systems. Aperture correction (enhancement) is the prime contributor to the edgy television-like look.
John D. Lowry
Lowry Digital Images
Jeff Kreines wrote :
>Seems like a dumb path with lots of possible concatenation of >compression schemes.
I didn't say it was a smart path, and BTW I didn't design it either. Sometimes things make a few too many u turns for my taste. Any way, I'm just reporting what happened, not why it happened and sure there would be better ways to do it, but sometimes, the editor/system that was planned for wasn't available and other things happened. No worries in this case because there will be an HD conform from the avid list and life goes on without the concatenation of compression schemes. I was simply citing this particular convoluted path as an example of how things pile up and wondering if any one else has seen anything similar.
Mark Smith DP
Oh Seven Films Inc.
Robert Butche wrote :
>There is no guessing -- on either end of the digital stream. The video >compressor easily identifies redundant information from frame to frame.
I realise that this is the theory but the reality is that humans set the level at which there is "no discernable difference" this is where the guessing comes in.
One persons idea of "no discernable difference" is very different from another’s "no discernable difference".
Then we have the effect of layered compression, every time it goes through an up/down conversion it goes through a different set of "no discernable difference" and these add up.
Geoff Boyle FBKS
Director of Photography
George C. Palmer writes :
>does anyone really believe that "bumping up" or "up-rezing" an image >with inherently limited bandwidth or resolution…actually increases the >image quality of that image????
It's not really a matter of belief or theory. Whatever else uprezzing does or doesn't do, it really does reduce aliasing. Compare a direct SD-to-film transfer with an SD-to-HD-to film-transfer and you'll see what I mean.
And once the image is up-sampled to HD, you can even apply a touch of sharpening without paying quite so serious an artefact penalty. You can test this with stills in Photoshop : Start with a 640x480 image. Copy it. Sharpen the copy and save it. Copy it again, then up-sample it by a factor of 2, using cubic spline interpolation, then sharpen it the same amount. Then compare the two images at the same screen size. The up-sampled version will have appreciably more finesse, an less noise/grain as a by-product of sharpening.
Talk to John Carlson at Monaco Labs in SF, and ask him why they bump SD up to HD before outputting to film. It really *does* make a difference. Certainly
not in *all* ways, but in *some* ways that do count.
Marin County, CA
>There is no guessing -- on either
end of the digital stream. The video >compressor easily
identifies redundant information from frame to frame.
This only applies to Inter-Frame compression codecs. "a format that has nothing to do with HD" and, to the best of my knowledge, all professional recording formats use Intra-Frame compression only, which follows a similar logic but within a frame and not between frames.
Also, another reason for converting 4:1:1 "a format that has nothing to do with HD" to 4:2:2 DigiBeta is if any colour correction is happening the 4:2:2 colour space will allow more subtle variations in the graded colours.
Ben Allan ACS
Director of Photography
To All :
The original posting in this discussion thread was about providing those without access to adequate funding for Hi-Def or Film origination, to the creation of their projects. DV is certainly adequate to that task, and yes, if some of the respondents are all to be believed, it may be possible, using a number of processes to "sharpen" the DV image enough to fool some of the people into believing that they have found a magic pill. Aside for advocacy of those processes, I hope that these processes actually work, for the sake of those here hopeful enough to try them; I also hope that these processes don't add enough "fix-it-in-post" processing costs to obviate the original goal of producing low cost masterpieces.
Nothing I have seen here convinces me that most of the presentations have said that it is "possible" to perform sharpening, Lest we forget, regardless of the efficacy of the DV "digital" recording format, the cameras attached to all DV cameras contain :
1) very unsophisticated processing with virtually no useful (even by analog camera standards), professional quality, camera control handles,
2) dramatically small, Standard Definition, IT imaging sensors with comparatively minute pixels and low pixel counts, which yield dramatically reduced (with comparison to larger sensors and film frames) dynamic range and resolution and
3) as a result have few high quality lens's available (NO HD lens's).
So the initial image is seriously compromised when compared to any professional quality HD camera and even most film cameras. In my experience, even the novice film/HD producer/director/writer has pretty lofty cinematic goals in mind when visualizing the telling of their story. I am not arguing against the use of DV technology in such low budget applications, as long as the "up-rezing" they see as the means to that end can (beyond theoretical the "devils advocacy" expressed in most of the respondent posts here) produce a product that actually achieves those lofty visualization dreams.
How many filmmakers, even novices, would admit having "compromise" in their vocabularies? I submit that even if their purses dictate such reality, they want their product to reflect the importance they attach to their original concept. I do accept the notion that "sharpness" can be added, but not resolution, that grading can add some limited expansion of the original dynamic range, but not recover non-existent grey scale values, BUT at what cost. Post processes, even if effective, add costs back to budgets which don't imply the resources for such "fixes". In my experience, most post production processes usually adds as much or more costs to a production budget as using appropriate production tools and techniques in the first place. If anyone out there knows where the next low budget DV project can be truly upgraded via layered compression, Viper-like/sophisticated color grading, or any of the purported (in this discussion thread) appropriate, effective up-rezzing methodologies, at a cost which preserves the initial "low budget" strategy and which actually produces a measurably improved image, please name the place and the projected cost.
To all those who don't have big budgets, no one will look down their nose at your use of DV if you tell your story well; you may even earn broad respect for overcoming your resource limitations. Many stories don't require sophisticated lighting and imagery, just imaginative use of the tools you have. If you succeed you will have done more than some who have all the resources the bank can provide. But that success will never and can never redefine those limited resources (DV) as equal in capability to the higher level resource (HD or film). However, if and when cost effective processing technology actually yields that capability, rendering the use of DV as acceptable as film or HD in human perceptual terms, even the 35mm folks will be forced to consider such lower cost techniques.
Until then, my story is that silk purses and sows ears really are distinguishably different, and I'm sticking to that story. By the way, I thought we had all agreed
some weeks ago that DV was not HD and, therefore, didn't belong here as a matter of discussion?????
GEORGE C. PALMER
HD and Digital Imaging Services
Long time lurker on this list and now I have something to finally contribute. My short film NIGHT LIGHT was shot NTSC mini" a format that has nothing to do with HD" with the JVC GY" a format that has nothing to do with HD"500 and I had to massage it (not up-rez) to HDCam for this year's Sundance Film Festival. Here's what I did.
1) Captured and edited in "a format that has nothing to do with HD"25 in FCP 3.0
2) Magic Bullet -- convert to 23.976P, de-artefact, color correct -- in After Effects
3) Render out all 23,500 frames as Targa D1 Sequences
4) Convert all frames into fractal files (using Photoshop and the Genuine Fractals plug-in)
5) Generate the HD TGA frames based on the fractal files (STNs)
6) Take TGA Sequences and make Cinewave HD 23.976P .moves
7) Drop back into FCP, sync up the audio mix
8) Dump back to HDCam 1080/ 59.94i (Let Cine handle the 3:2 insertion)
It was very, very time consuming, but it is the best looking "a format that has nothing to do with HD" to HD transfer job I've seen. I experimented using bilinear and bi-cubic interpolation, as well as some of the different hardware scaling methods, but the fractal conversion yielded the best results for my project. Some of the shots are beautiful and some are really crap, but that’s the price you pay when shooting "a format that has nothing to do with HD", which in my case, proved much cheaper in the long run.
If any of you are at Sundance, check out my film, its in Shorts Program III (the program guide says Program II, but its a *typo*)
>Even shooting with camera detail
down doesn't hide it enough, and >while a good DP can shoot
around most SDTV limitations, it's mighty >hard to avoid *any*
scene elements that call this out.
Any hints on where an old DP with no "a format that has nothing to do with HD" to film experience could find a primer on what these limits are and how to shoot around them. ie what elements "call these out"
>Any hints on where an old DP
with no...(digital) to film experience could >find a primer
on what these limits are and how to shoot around them.
ie >what elements "call these out"
Any facility that does film-outs from digital sources should have a list of suggestions somewhere on their website for shooting with a blow up in mind.
The problem I run into is that most of these project don't ever get blown up to 35mm, and some of the suggestions place limits on what you can do in camera to create a certain look.
Here's a few links to facilities (no endorsement is intended) which have suggestions posted on their site :
Los Angeles based Director of Photography
West Coast Systems Administrator, Cinematography Mailing List
Thank you Jessica for the website listings of some post house specializing
in tape to film transfers. Their web pages offered almost as much information
that many of the contributors to the CML list, and I could actually understand
Another reason why I have been a faithful member and will continue to do so!
Allen S. Facemire
DoP IATSE 600
Just a note of thanks for the range of expertise and opinion from the world wide CML community. I would be interested in seeing the results of Eric Escobar's complicated post process but I won't be going to Sundance. If it does in fact accomplish good results that's fine but my original contention was that is actually cheaper and less hassle shooting in HD if one has to shoot on tape(a trend that is not going to go away any time soon). It also appears that Eric may work for Apple and has access to processes and/or engineers that many others may not(I'm guessing here). Perhaps they can sell this expertise to the transfer houses.
My other point was that since we don't live in a perfect world it isn't always about the absolute best image quality in any case. Otherwise we would demand 70mm or Showscan or Imax. So, there is a place for these low end formats and they will continue to improve, but if the people who insist on using low cost formats don't recognize the need for a good DOP then the results will continue to be uneven as they have been up to now. As for myself, I'd love to see 35mm quality without 24 fps motion artefacts; which is why I like 60P projection.
BTW, with regard to the Dalsa camera I was tipped off to it several months by a colleague; who may actually be consulting on it's development. There has been much confidentiality surrounding it and understandably so. I really don't know much more than what's on the website but he did hint that it might match or exceed the Viper, and that it will be unveiled at NAB which I will not be attending either.
So some of you will see it before I will. However I may be able to co-ordinate some tests next Spring or Summer and will be as eager as the rest of you to see the results.