Jump to content

Scratch


Roberflowers

Recommended Posts

hello all,

I'm the cinematographer on a graduate film at Chapman University. We have project that is sponsored by the school and out of 150 or so scripts ours was chosen as well as our creative team. We have a GRASS VALLEY 2K scanner ( upgradeable to 4K). Our advisor has contended that we do not need to go to outside facility for our final color grading. That once our film is scanned to files we bring it into this SCRATCH and we can do our own grading there.

 

Now before I get blasted with replies that it is foolish to have a student do their own coloring etc. etc. Let me just say that I KNOW this already. We only have 10000 dollars to make a 15 minute film, and while I have tried to push for using outside facilities for this, I am told that we can't afford it, that money is tied up elsewhere.

 

Those issues aside. I want to know more about this SCRATCH that I've been told about. Does anyone have any info on it?

 

Also I'm going to try real hard to do some tests monday tuesday or wednesday so I can get it back, and maybe run it through our scanner and scratch, just to learn about this process.

 

We as cinematographers are required to keep a log of everything related to the photographic elements of the film that we intend to execute. So I'll send a link of it here soon, any comments and tips and criticisms will be greatly appreciated along the way...Our shoot begins the 19th of jan - 28th. final delivery is May.

 

thanks

 

Robert

Link to comment
Share on other sites

Those issues aside. I want to know more about this SCRATCH that I've been told about. Does anyone have any info on it?

 

www.assimilateinc.com

 

Go to the website.

Read the manuals.

 

Assimilate Scratch is used by many facilities worldwide, from large to small. It is used for everything from playback of full resolution files for visual effects purposes to DI conforming to color correction.

 

What else do you want to know?

Link to comment
Share on other sites

  • Site Sponsor

Assimilate Scratch is used by many facilities worldwide, from large to small. It is used for everything from playback of full resolution files for visual effects purposes to DI conforming to color correction.

 

We have been looking at software color grading systems and I had a Scratch system here for a while, I thought it was a good stable capable system for 2K and although in my opinion it seems to lack some of the more mature color grading tools found in a Baselight or Lustre it is more than useable and has impressive performance. I would think that a properly calibrated viewing enviornment (esp if you are going back to film) would be a more pressing concern.

 

I really wish assimilate would find a better name for this product as it matures, what were they thinking? Here let me put your film on this scratch machine! It detracts IMO from what is otherwise a fine product.

 

-Rob-

Link to comment
Share on other sites

Thanks guys,

I'll check out the sight, and wholeheartedly agree that SCRATCH is a horrible name. My professor, who is not really a cine guy, told me about it in passing, and I thought it would be very strange to pass my film over to a SCRATCH machine.

 

as for the viewing environment being a concern. It is, along with many others. The room that we are to do this in, is still being finished. I really hope they have thought of details like viewing environment.

 

Our workflow will probably follow this line

 

shoot super 16mm

developement at Fotokem with flex file

one light dailies viewed during middle of shoot at our screening room

offline edit on avid media composer or Nitris

scan at 2K resolution the picture lock

bring in to SCRATCH for final grading

Finish on D-5 for 2K digital projection, followed by HD-DVD, DVD, web, Beta Sp, for festivals

AND( and this is only if by some miracle we find more money in April)

35mm FILM OUT

 

I'm oversimplying it, but that is basically what we are doing. I am leaving anything out with regards to the scan and scratch ( wow, that just sounds wierd) ???

 

I keep running across the term LUT, Look Up Table, and I keep thinking that I need to work with that. I read somewhere that someone said it was like a cheat sheet for exposure, sort of taking the information and creating exposure, and color grading. (again, I'm probably oversimplifying)

 

 

Robert

www.reelflowers.com

Link to comment
Share on other sites

I keep running across the term LUT, Look Up Table, and I keep thinking that I need to work with that. I read somewhere that someone said it was like a cheat sheet for exposure, sort of taking the information and creating exposure, and color grading. (again, I'm probably oversimplifying)

 

There's a good thread going on right now about LUTs here.

 

Best of luck to you!

Link to comment
Share on other sites

I keep running across the term LUT, Look Up Table, and I keep thinking that I need to work with that. I read somewhere that someone said it was like a cheat sheet for exposure, sort of taking the information and creating exposure, and color grading. (again, I'm probably oversimplifying)

 

Lookup tables are used to allow a display to emulate anotherr display, in the case of DI work, to let a digital projector emulate a film print. They sit at the end of the processing chain, basically feeding the display itself, and are created by profiling the monitoring device (i.e., the projector), the target device (i.e., a film print), and in some cases some of the intermediate devices (film recorder, negative stock), and combining those profiles to create a table of input values vs. output values so that the monitor can look like the target. This requires a deep knowledge of color science and an understanding of film color space, which is why most companies use a solution provided by companies that have developed them, like Truelight and Cinespace. In the case of a Scratch driven room, those who set up the room are responsible for creating the profiles, creating the lookup tables from them, and installing and maintaining said lookup tables for the colorists' use. Scratch can use Cinespace 3D LUTs directly in software, or alternatively, one can contract with Filmlight (the makers of Truelight) to provide a hardware box and profiling kit to do the same thing in hardware, as a separate device that would sit between the output of the Scratch computer and the input of the projector.

 

LUT's have nothing to do with color correction, other than providing a calibrated emulation of the resulting film print given the images being corrected.

Link to comment
Share on other sites

We have been looking at software color grading systems and I had a Scratch system here for a while, I thought it was a good stable capable system for 2K and although in my opinion it seems to lack some of the more mature color grading tools found in a Baselight or Lustre it is more than useable and has impressive performance.

 

I realize that you're actually praising the product here, but that's a bit like saying that a Boxster lacks some of the more mature performance of a Turbo Carrera - except that the cost differential between Scratch and a Lustre or Baselight is actually greater than that.

 

The fact is that there are really very few markets in the US or elsewhere in which there exists a client base that is willing to pay what it needs to cost to justify the more expensive toolset, given the volume of work. It's understandable that Los Angeles based DI companies may be using Baselight 8's, because in that market, there are large studios that will pay between $300K and $500K for a DI. Those clients, and those kinds of numbers, don't exist anywhere else in the world, with the possible exception of once or twice a year in London. While the volume of work done in DI suites may be growing, the prices being paid for them are shrinking. Cost effectiveness is the name of the game once you set foot outside of the city of Los Angeles. There is practically nothing one can do on a Lustre or a Baslight system that one cannot do on a Scratch system. The differences are primariily in operator convenience (you can also relate this to efficiency) and overall throughput.

 

Besides, the largest part of the cost of a "cost effective" DI is the scanning, recording, negative stock, and print stock - and those items have become such commodities in the eyes of the clients that they're practically given away. So if there is to be a healthy DI business, there must be some area in which a facility can make some money.

Link to comment
Share on other sites

  • Site Sponsor
I hear a lot about the aforementioned color systems. Why does After Effects never get mentioned? Does it stink? Am I missing something here?

After Effects is a compositor and while it has color correction tools they would feel slow and cumbersome if you were accustomed to using a DaVinci or Pogle (RT hardware, nice interfaces) or the above mentioned software based color tools.

 

-Rob-

Link to comment
Share on other sites

  • Site Sponsor
I realize that you're actually praising the product here, but that's a bit like saying that a Boxster lacks some of the more mature performance of a Turbo Carrera - except that the cost differential between Scratch and a Lustre or Baselight is actually greater than that.

 

So true and a Baselight 8 might qualify for even more exotic car status than a mere porsche. And I have driven both and you can keep the boxter, really.

 

The fact is that there are really very few markets in the US or elsewhere in which there exists a client base that is willing to pay what it needs to cost to justify the more expensive toolset, given the volume of work. It's understandable that Los Angeles based DI companies may be using Baselight 8's, because in that market, there are large studios that will pay between $300K and $500K for a DI. Those clients, and those kinds of numbers, don't exist anywhere else in the world, with the possible exception of once or twice a year in London. While the volume of work done in DI suites may be growing, the prices being paid for them are shrinking. Cost effectiveness is the name of the game once you set foot outside of the city of Los Angeles. There is practically nothing one can do on a Lustre or a Baslight system that one cannot do on a Scratch system. The differences are primariily in operator convenience (you can also relate this to efficiency) and overall throughput.

 

While I agree with what you are saying there is market for 2K+ and Pogle products all over the world and a new Davinci or Pogle is certainly in the exotic car price bracket of the Italian variety. Both Assimilate and Filmlight were kind enough to lend us machines to evaluate and I have looked at a Lustre system and taken a look at speed grade as well, the only system I have not yet seen is Film master. My impression as of right now is that the Baselight4 could be a direct replacement for a 2K or Platinum with all of the added benefits of non linear workflow (as could Lustre) where scratch and speed grade could not compete with the rt hardware and the interactivity of the interface. This could easily change in the future as software features are easy to add and computer power grows day by day (along with their consumption of energy) so this market segment is a free running target right now IMO. Also I think that when you compare the cost of a full Spirit suite to a scanner/software suite a northlight or arriscan and a baselight seems relatively cheap :blink: and the quality is miles better.

 

 

Besides, the largest part of the cost of a "cost effective" DI is the scanning, recording, negative stock, and print stock - and those items have become such commodities in the eyes of the clients that they're practically given away. So if there is to be a healthy DI business, there must be some area in which a facility can make some money.

 

Well there is the crux but as a part owner of a lab and a filmmaker working on a feature project I feel I need to look. Furthermore a 35mm photochemical finish will look better than 90% of the DI work out there I really feel that we should not be lowering the bar for quality. Take scanning if it is a commodity what kind of scanner is the price based on, I have seen Northlight scans and Spirit scans and the spirit loses hands down but it is the bench mark right? The same applies in other areas.

 

Don't take this as a business plan :D because I do not know if we can justify getting into the DI business there are many factors and very heavy weights in this segment.

 

-Rob-

Link to comment
Share on other sites

  • Premium Member

Hi,

 

That said, there's practically nothing you can't do in After Effects or even just a nonlinear editor that you can do in Baselight, but there does come a point where it's such a hassle that it's not really usable in that role.

 

You can make AE perform primary grades on a 1K proxy (which is what most Baselights work in) on quite modest hardware.

 

Phil

Link to comment
Share on other sites

  • Premium Member
a 35mm photochemical finish will look better than 90% of the DI work out there

 

You shouldn't really ever have the opportunity to compare, so thats a bit of a bold statement. :D The point of doing a DI is to do things you can't do in the lab - for example grading parts of the frame. So there should never be a point when it is possible to compare the two. Besides, modern film makers can't make their mind up without seeing 50 different options, and thats a bit expensive and time consuming in a lab!

Link to comment
Share on other sites

You shouldn't really ever have the opportunity to compare, so thats a bit of a bold statement. :D The point of doing a DI is to do things you can't do in the lab - for example grading parts of the frame. So there should never be a point when it is possible to compare the two. Besides, modern film makers can't make their mind up without seeing 50 different options, and thats a bit expensive and time consuming in a lab!

 

Oh I don't know David.

 

I've shot "pure film" for 3 decades; now I guess I'm entering, really in the world of all well some of those options.... but I can't pretend I don't want this stuff to act as some kind of ideal optical printer. I'm not saying it will exactly, but I "can't not" make comparisons to the ideally photographic.....

 

-Sam Wells

Link to comment
Share on other sites

  • Premium Member
Oh I don't know David.

 

I've shot "pure film" for 3 decades; now I guess I'm entering, really in the world of all well some of those options.... but I can't pretend I don't want this stuff to act as some kind of ideal optical printer. I'm not saying it will exactly, but I "can't not" make comparisons to the ideally photographic.....

 

-Sam Wells

 

I see what you're saying, but how do you make the comparison if you don't do the exact same thing both ways (optically and DI), in which case you should probably only do it optically because that would be cheaper.

 

I guess my point is that there is no point saying one is better than the other because they are both ultimately intended to provide different things, so there isn't really a comparison to be drawn.

Link to comment
Share on other sites

  • Site Sponsor
Hi,

 

That said, there's practically nothing you can't do in After Effects or even just a nonlinear editor that you can do in Baselight, but there does come a point where it's such a hassle that it's not really usable in that role.

 

You can make AE perform primary grades on a 1K proxy (which is what most Baselights work in) on quite modest hardware.

 

Phil

 

 

Phil there are many things that you can do in a color finishing system (scratch to baselight) which you cannot do in a NLE or are ridiculously hampered in a compositing package. I know of no nle system which can do self calibrated color management from scans to final print stock and account for the viewing enviornment in such a way that it will be consistent.

 

Furthermore all of the color grading systems do work in real time with 2K files and the Baselight will work in real time with 4K files all of these systems will play proxies but if you have a Baselight 8 why would you? esp. if you are working on finishing a 4K project.

 

-Rob-

Link to comment
Share on other sites

  • Premium Member

Hi,

 

Yeah, OK, if you want a calibrated viewing environment that would be an external thing for AE (although there is a truelight plugin for Combustion), but in terms of the actual manipulation tools, the biggest ticket colour correctors look pretty feeble compared to desktop compositors. Different animals for different jobs, of course, and not terribly comparable, but I must admit that I always feel like the toolset on a Da Vinci or Pogle is horribly limited for the money it costs.

 

> Furthermore all of the color grading systems do work in real time with 2K files

 

Only just. And it depends entirely on what you mean by realtime. All of the software systems have a limit to what they will do in realtime (although of course the hardware stuff has a limit to what it will do, period). None of them will do large radius true gaussian blurs in realtime at any resolution.

 

> and the Baselight will work in real time with 4K files

 

Some kinds of Baselight will play them back and perhaps do a few basic grades.

 

> all of these systems will play proxies but if you have a Baselight 8 why would you?

 

So you can do lots of advanced grading in realtime.

 

Phil

Link to comment
Share on other sites

  • Site Sponsor
You shouldn't really ever have the opportunity to compare, so thats a bit of a bold statement. :D The point of doing a DI is to do things you can't do in the lab - for example grading parts of the frame. So there should never be a point when it is possible to compare the two. Besides, modern film makers can't make their mind up without seeing 50 different options, and thats a bit expensive and time consuming in a lab!

 

 

It was late I meant to say that a 35mm photochemical finish will beat out 90% of the "Budget" DI work and I bet if you look into everything as a total package the "straight" 35mm job from stock through prints will be pretty even with either a HD or S-16 or esp. 35 with a Di of some flavor (is a Spirit/2k grade to D5-HD, assemble and filmout a DI??) and furthermore just having every possible option does not mean you will do anything competent with them.

 

That said I have a film we are working on which is 90% S-16 and a good 100,000 feet and the post path is a mess and that is my own fault :angry:

 

-Rob-

 

That said I have a film we are working on which is 90% S-16 and a good 100,000 feet and the post path is a mess and that is my own fault :angry:

 

-Rob-

 

 

Oh and that is a film I am co-producing and shooting I do not want to give the impression I messed up a Cinelab customers film, just my own. Not that there is anything really wrong with it I just could have made it easier on myself if I had put a little more planning into the post path when I started shooting it.

 

:angry: :angry:

 

-Rob-

Link to comment
Share on other sites

It was late I meant to say that a 35mm photochemical finish will beat out 90% of the "Budget" DI work and I bet if you look into everything as a total package the "straight" 35mm job from stock through prints will be pretty even with either a HD or S-16 or esp. 35 with a Di of some flavor (is a Spirit/2k grade to D5-HD, assemble and filmout a DI??) and furthermore just having every possible option does not mean you will do anything competent with them.

 

That's generally true. It also doesn't matter.

 

For the same reasons that "modern" (i.e., inexperienced) filmmakers seem to believe that HD video is at the very least the equivalent of 16mm film and at the very most blows away 35mm, fewer and fewer people even want to consider a photochemical finish, regardless of the quality and regardless of the price. The trend seems to be towards "demanding" a DI finish that beats the photochemical route in price, and the method for achieving this is getting facilities to along with it - and a number of them do.

 

 

but I must admit that I always feel like the toolset on a Da Vinci or Pogle is horribly limited for the money it costs.

 

The last time I checked, you weren't a professional colorist, have little to no real experience with operating either a DaVinci or a Pogle, and little to no experience with doing a colorist's job with a client in the chair next to you and the time constraints associated with real world colorist's work. As someone who has been a professional colorist and has a lot of experience with both systems, I completely disagree with the statement, as would just about every other professional colorist.

 

None of them will do large radius true gaussian blurs in realtime at any resolution.

 

Neither will any compositing system I've ever heard of. But as this has basically nothing to do with color correction, I don't see where this is relevant.

Link to comment
Share on other sites

  • Premium Member

Hi,

 

Michael, I'm talking about the toolset - not the whole workflow of the thing. A Pogle Platinum will do about four very basic things (you could argue that doing it in 10-bit HD in realtime is not basic, but that's not a toolset issue) before running terminally out of steam. It's not like there's the option to go "Oh, OK, I'll wait around a bit so I can have my extra vector". I don't have to have done it full time to know what the damn thing will do, any more than I have to have driven a formula 1 car to know it goes faster than a pushbike.

 

So I stand by the assertion that in comparison to the most basic desktop compositors, hardware colour correctors actually do very little. Yes they do it fast and in a rather particular way, but the actual breadth of capability is simply not that much.

 

And the reason I'm so militant about this is that people just don't seem to see how limited it is, probably because they're age-old film luddites who see anything more than RGB balancing as a panacea.

 

Phil

Link to comment
Share on other sites

I don't have to have done it full time to know what the damn thing will do, any more than I have to have driven a formula 1 car to know it goes faster than a pushbike.

 

So I stand by the assertion that in comparison to the most basic desktop compositors, hardware colour correctors actually do very little. Yes they do it fast and in a rather particular way, but the actual breadth of capability is simply not that much.

 

And the reason I'm so militant about this is that people just don't seem to see how limited it is, probably because they're age-old film luddites who see anything more than RGB balancing as a panacea.

 

Well, once again as someone with a lot of experience on these systems, I would argue that a very basic toolset is all that color correction basically represents. In a digital color timing system, you're just manipulating numbers, so the transforms you create with that number crunching is rather limited in the first place, unless you're really out to destroy the original image in the process. As for flexibility, I would argue that creating more tools to do what someone with experience can accomplish with fewer tools is simply catering to the masses, who want to do these things without actually learning or even knowing exactly what it is they're doing. An experienced colorist goes about the work in a structured way, knowing basically where they want to go and using the toolset in front of them to get there. That's very different than changing a bunch of sliders on an AE plugin until it starts to look interesting. What you see as a lack of depth in the toolset, I see as providing all the tools necessary to do the job, provided you have some degree of experience and expertise when you do it. Any image that requires 6 or 7 isolated parameters is no longer bearing any relation to the origiinal photography, and is now in the realm of a multilayer composite, not a color corrector. Making an original image sing and creating a multilayer composite from different image sources are two difference skills, with two different audiences, two different areas of focus, two different turnaround requirements, and, basically, two different skill sets. One would no more use After Effects for color correction on a DI project with any kind of actual deadline than one would drive a Ford Focus in a Formula One race. Or more to the point, any more than one would shoot "Lost" on an HDV camera.

 

Just because you can do something doesn't mean you should.

Link to comment
Share on other sites

  • 2 weeks later...

The tools can look identical in feature set or even in GUI, but the quality can be different. DSP quality is up to the engineer that wrote the algorithms. There are lots of ways to implement an algorithm, there are custom algorithms especially for anything nonlinear, there is difference in precision of DSP etc. Some algorithms use precalculated look up tables and interpolate in order to provide some type of processing. This is entirely different from a high precision implementation in terms of quality. You image goes through this and it could affect the continuity of tones, especially in shadows and highligghts Just because both systems are digital it doesn't mean they produce equal results.

Link to comment
Share on other sites

Create an account or sign in to comment

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

Create an account

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

Register a new account

Sign in

Already have an account? Sign in here.

Sign In Now
×
×
  • Create New...