Data to Film with SI 2K at DVinfo.net

Sergio Sanchez · August 14th, 2007, 08:18 PM

This is a link to a clip on youtube somebody uploaded, it shows a projection of a Data to film tests for a film that will be shot using the SI2K, its low quality but you can get the picture.

http://youtube.com/watch?v=MVqImnb7x9Y

Jason Rodriguez · August 14th, 2007, 11:34 PM

Whoa!

That stuff looks pretty good!

BTW, is there anything else you need from us, or anything that we're not providing that would make the workflow smoother?

Now we are still working on a path towards film calibration, and I've also created some Cineon reverse LUT's for the camera so that you don't to meddle with the problems you had in After Effects before. But I'd like to hear what the pro's and cons are that you see so far, and how it can be improved in the future.

Bob Grant · August 15th, 2007, 08:40 AM

Looking good indeed!

I've been told but yet to verify for certain that our local lab can go directly from Prospect2K to 35mm. If nothing else that's going to avoid having to shuffle huge amounts of data around.
I guess my best bet is to ask the lab about this, but if anyone can give me any pointers it'd be much appreciated, how to go about calibrating the whole process, so we know that what we see on a monitor is within the ballpark of what'll be seen on the silver screen from 35mm or a 2K DCI print. I understand that the last part is probably way more of an issue than a 35mm print.

Sergio Sanchez · August 15th, 2007, 12:15 PM

Thank you Jason.

I believe that there are a lot of Pro`s in the SI2K workflow, most of the cons are in one hand related to the ignorance of the user, I mean it is a new process, most of the people are used to film and typical broadcast production in wich always is a physical medium involved. The first problems we had were caused by misunderstanding of the post process, you cant postproduce the material as video, you have to do it thinking of it as a DI captured from a telecine.

This Tests were done exporting the shots directly from after effects, to DPX with no curve applied, we used the preset LUT`s from the baselight to calibrate the shots. And the results are very good. You can say that this Data is directly from Prospect to 35mm. We used a Celco Film Recorder, Kodak 5201 negative and Premier 2393 for the print. We went this route because all the shots were done using the Alpha sensors, and we had some issues with 2K material ( In the last sensor this is solved, hurray)the Kodak premier rich contrast curve managed to hide all the issues.

The correct DPX path we`ve been discussing is, I believe the best way to go if your are delivering on 35mm, I`ve been doing some tests for the next Data to film in wich we are going to define the correct look for the film. For that one im using the new sensor, and Im going to finish with a more transparent process, i`ll be using Kodak Vision 2383 for the prints, and Arri laser as a film recorder, and I will treat the material as film, using cineon curves for greater flexibility in color correction, as well as a diferent color space for the DPX exports in AE.

I found a tutorial in the adobe website that is very usefull, I`ve noticed that the ITU709 color space is not apropriate for film work, the color saturation is very flat, so i`ve been using the ProPhoto colorspace instead and the results are amazing. The only issue im having is that somehow Cineform does not support the changes in the colorprofile, so I have to double render using the color profile converter effect. (http://www.adobe.com/designcenter/af...olrmanage.html)

Right know i believe that the reverse LUT`s will be very helpful. The film calibration is very easy, we only did a gamma correction to the material prior to the film recording process, and that was all, if you use a standard curve in your material, one that all the finishing apps would recognize like the Cineon curve you will not have any major troubles.

Sergio Sanchez · August 15th, 2007, 12:20 PM

I believe that the best way to improve all the problems is to show your customers the best way to go. Tutorials and Documentation is the best way to solve the issues. I know that you`re working on that, so i think the workflow will improve with time, and thats preciselly the most interesting part of the SI2K. It can improve with time, new firmware, new LUT`s, patches, etc.

Joseph Kassana · August 15th, 2007, 01:37 PM

Hey I wanna buy one of these units - How do I do so?

Jason Rodriguez · August 16th, 2007, 12:00 AM

Please contact Steve Nordhauser at 518-279-9098, or fill out an order request on our website at www.si-2k.com, and press on the "contact us" button on the home-page.

August 14th, 2007, 08:18 PM	#1
Sergio Sanchez Regular Crew Join Date: Feb 2007 Location: Mexico Posts: 176	Data to Film with SI 2K This is a link to a clip on youtube somebody uploaded, it shows a projection of a Data to film tests for a film that will be shot using the SI2K, its low quality but you can get the picture. http://youtube.com/watch?v=MVqImnb7x9Y

Thread Tools	Search this Thread
Show Printable Version Email this Page	Search this Thread: Advanced Search

August 14th, 2007, 11:34 PM	#2
Jason Rodriguez Trustee Join Date: Mar 2003 Location: Virginia Beach, VA Posts: 1,095	Whoa! That stuff looks pretty good! BTW, is there anything else you need from us, or anything that we're not providing that would make the workflow smoother? Now we are still working on a path towards film calibration, and I've also created some Cineon reverse LUT's for the camera so that you don't to meddle with the problems you had in After Effects before. But I'd like to hear what the pro's and cons are that you see so far, and how it can be improved in the future.

August 15th, 2007, 08:40 AM	#3
Bob Grant Trustee Join Date: Nov 2005 Location: Sydney Australia Posts: 1,570	Looking good indeed! I've been told but yet to verify for certain that our local lab can go directly from Prospect2K to 35mm. If nothing else that's going to avoid having to shuffle huge amounts of data around. I guess my best bet is to ask the lab about this, but if anyone can give me any pointers it'd be much appreciated, how to go about calibrating the whole process, so we know that what we see on a monitor is within the ballpark of what'll be seen on the silver screen from 35mm or a 2K DCI print. I understand that the last part is probably way more of an issue than a 35mm print.

August 15th, 2007, 12:15 PM	#4
Sergio Sanchez Regular Crew Join Date: Feb 2007 Location: Mexico Posts: 176	Thank you Jason. I believe that there are a lot of Pro`s in the SI2K workflow, most of the cons are in one hand related to the ignorance of the user, I mean it is a new process, most of the people are used to film and typical broadcast production in wich always is a physical medium involved. The first problems we had were caused by misunderstanding of the post process, you cant postproduce the material as video, you have to do it thinking of it as a DI captured from a telecine. This Tests were done exporting the shots directly from after effects, to DPX with no curve applied, we used the preset LUT`s from the baselight to calibrate the shots. And the results are very good. You can say that this Data is directly from Prospect to 35mm. We used a Celco Film Recorder, Kodak 5201 negative and Premier 2393 for the print. We went this route because all the shots were done using the Alpha sensors, and we had some issues with 2K material ( In the last sensor this is solved, hurray)the Kodak premier rich contrast curve managed to hide all the issues. The correct DPX path we`ve been discussing is, I believe the best way to go if your are delivering on 35mm, I`ve been doing some tests for the next Data to film in wich we are going to define the correct look for the film. For that one im using the new sensor, and Im going to finish with a more transparent process, i`ll be using Kodak Vision 2383 for the prints, and Arri laser as a film recorder, and I will treat the material as film, using cineon curves for greater flexibility in color correction, as well as a diferent color space for the DPX exports in AE. I found a tutorial in the adobe website that is very usefull, I`ve noticed that the ITU709 color space is not apropriate for film work, the color saturation is very flat, so i`ve been using the ProPhoto colorspace instead and the results are amazing. The only issue im having is that somehow Cineform does not support the changes in the colorprofile, so I have to double render using the color profile converter effect. (http://www.adobe.com/designcenter/af...olrmanage.html) Right know i believe that the reverse LUT`s will be very helpful. The film calibration is very easy, we only did a gamma correction to the material prior to the film recording process, and that was all, if you use a standard curve in your material, one that all the finishing apps would recognize like the Cineon curve you will not have any major troubles.

August 15th, 2007, 12:20 PM	#5
Sergio Sanchez Regular Crew Join Date: Feb 2007 Location: Mexico Posts: 176	I believe that the best way to improve all the problems is to show your customers the best way to go. Tutorials and Documentation is the best way to solve the issues. I know that you`re working on that, so i think the workflow will improve with time, and thats preciselly the most interesting part of the SI2K. It can improve with time, new firmware, new LUT`s, patches, etc.

August 15th, 2007, 01:37 PM	#6
Joseph Kassana Regular Crew Join Date: Jun 2007 Location: London Posts: 92	Hey I wanna buy one of these units - How do I do so?

August 16th, 2007, 12:00 AM	#7
Jason Rodriguez Trustee Join Date: Mar 2003 Location: Virginia Beach, VA Posts: 1,095	Please contact Steve Nordhauser at 518-279-9098, or fill out an order request on our website at www.si-2k.com, and press on the "contact us" button on the home-page.