4:4:4 12-bit Uncompressed DVX100

Laurence Maher · June 15th, 2004, 05:46 AM

By the way, Juan, WOW! Just checked that clip of the ext. night shot you uploaded. REALLY NICE!!!!!! How many stops of lattitude are you getting on that would you say?

Helen Mets · June 15th, 2004, 10:10 AM

actually Juan, on the last cap you posted (cap10), the blue ch. alignment was off by 1px to both down & right on 1240x794 scale i.e. move up & left to get it right.

Juan P. Pertierra · June 15th, 2004, 10:34 AM

Thanks Helen...i will re-do it when i get home today.

Laurence, I'm not sure how many stops that is but i'm going to take some shots on film for comparison. Problem is, last time i tried i used ISO400 film because i thought that was closest to a CCD, but I couldn't get a low enough exposure, they all came out very washed out.

Will try again tonight....

Anamorphic countdown timer: 24 hours <g>

Juan P. Pertierra · June 15th, 2004, 07:29 PM

Here is another HD test. I took 35mm stills of this one using 100ISO film this time. I'll have them developed tomorrow.

Only did an auto-levels and an up-rez with base settings.

http://expert.cc.purdue.edu/~pertierr/cap11_HD.tif

Juan P. Pertierra · June 15th, 2004, 07:53 PM

At last! I just hooked up the exposure control, so I now have iris control during RAW capture.

If anyone has any suggestions of what tests can be done with this important aspect of control, let me know. I still have the scene from the previous post setup.

Juan

Justin Burris · June 16th, 2004, 01:51 PM

Juan,

Responding to your request for suggestions:

One thing that I noticed about most of the images that you have posted is that when I adjust the levels in photoshop, I notice that they have most of their picture information in the dark through mid range, and very little picture info in the highlights. I was starting to wonder if maybe the CCD just didn't go that high, but this last one, cap11 HD, seemed to have at least some info at the high end.

So, I am interested to see how the raw images deal with the upper end of brightness; how it transitions to pure white. Now that you can control exposure, we can finally see just how far you can push this thing.

Juan P. Pertierra · June 16th, 2004, 01:55 PM

Thanks Justin....

Here's what I had in mind...setup a similar scene to the last post, and take RAW and DV frames in small exposure interfvals.

I'll try and put together a collage of the DV and raw frames so we can see how much more you can push it.

Haven't gotten the anamorphic yet, any minute now.

Gonna get a gyro, then i'll start the tests.....

Juan

Juan P. Pertierra · June 16th, 2004, 05:13 PM

Here's a thumbnail of the images i'm getting with he anamorphic adapter. I did some color correction on this one and applied a log curve...

http://expert.cc.purdue.edu/~pertierr/cap12_pre.tif

the few dots you see are speckles. I've been trying to wrap my head around the myriad of aspect ratio changes involved, i think i got it close but it might not be perfect.

Juan

Obin Olson · June 16th, 2004, 05:53 PM

Juan how are you going to deal with 12bit files? what can edit a 12bit file? I know after effects can open 16bit files but 12? I am having the same issue with the 4:4:4 10bit camera

Juan P. Pertierra · June 16th, 2004, 05:57 PM

Put the 12-bit data in the upper 12 bits of 16 bits. Set the lower 4 bits to 0. It's the same thing. The data is 16-bits in size but 12-bits in precision.

Juan

Obin Olson · June 16th, 2004, 06:09 PM

so then you waste disk space ? or does the 0 mean nothing in the bottom=16bit file that is the size of a 12bit file?

Juan P. Pertierra · June 16th, 2004, 06:12 PM

At processing time, having 16-bit data actually helps because even though you start with 12-bit precision, all color corrections/FX are done in 16-bit precision.

If you are concerned about storage, then you need to PACK the data.

Computers works in 8-bit bytes, so if you want to use the maximum amount of space, you have to pack the 12-bit data across byte boundaries, which could require some basic error correction to make sure you don't loose data somewhere.

Obin Olson · June 16th, 2004, 06:17 PM

do you know a good fast codec for packing? I will need it BAD as a tiff file at HD is about 5 megs! that will never edit in a computer unless its a SMOKE or some fancy box..do you know anything about BitJazz the makers of the SheerVideo codec?

Juan P. Pertierra · June 16th, 2004, 06:36 PM

35mm vs DVX100 RAW test
~~~~~~~~~~~~~~~~~

35mm Fujifilm ISO100 image:
http://expert.cc.purdue.edu/~pertierr/35vsDVX_35.jpg

DVX100 image:
http://expert.cc.purdue.edu/~pertierr/35vsDVX_RAW.tif

These were taken yesterday, without the anamorphic adapter. The scene is exactly the same in both images, the only differences being the camera/medium, and the film camera was located higher and to the left since the DVX is not very easy to move right now.

Juan

June 16th, 2004, 08:33 PM

Juan, great example! The DVX image looks excellent. Will the speckles in the RAW DVX image ever be gone completely? I've noticed those in previous images as well. I recall you saying that would be fixed with a different capacitor, I could be wrong though.

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

June 15th, 2004, 05:46 AM	#811
Laurence Maher Join Date: May 2004 Location: denton, texas, usa Posts: 416	By the way, Juan, WOW! Just checked that clip of the ext. night shot you uploaded. REALLY NICE!!!!!! How many stops of lattitude are you getting on that would you say?

June 15th, 2004, 10:10 AM	#812
Helen Mets New Boot Join Date: Jan 2004 Location: Tartu, Estonia Posts: 11	actually Juan, on the last cap you posted (cap10), the blue ch. alignment was off by 1px to both down & right on 1240x794 scale i.e. move up & left to get it right.

June 15th, 2004, 10:34 AM	#813
Juan P. Pertierra Major Player Join Date: Jul 2003 Posts: 479	Thanks Helen...i will re-do it when i get home today. Laurence, I'm not sure how many stops that is but i'm going to take some shots on film for comparison. Problem is, last time i tried i used ISO400 film because i thought that was closest to a CCD, but I couldn't get a low enough exposure, they all came out very washed out. Will try again tonight.... Anamorphic countdown timer: 24 hours <g>

June 15th, 2004, 07:29 PM	#814
Juan P. Pertierra Major Player Join Date: Jul 2003 Posts: 479	Here is another HD test. I took 35mm stills of this one using 100ISO film this time. I'll have them developed tomorrow. Only did an auto-levels and an up-rez with base settings. http://expert.cc.purdue.edu/~pertierr/cap11_HD.tif

June 15th, 2004, 07:53 PM	#815
Juan P. Pertierra Major Player Join Date: Jul 2003 Posts: 479	At last! I just hooked up the exposure control, so I now have iris control during RAW capture. If anyone has any suggestions of what tests can be done with this important aspect of control, let me know. I still have the scene from the previous post setup. Juan

June 16th, 2004, 01:51 PM	#816
Justin Burris Regular Crew Join Date: Jan 2004 Location: Olympia, WA Posts: 54	Juan, Responding to your request for suggestions: One thing that I noticed about most of the images that you have posted is that when I adjust the levels in photoshop, I notice that they have most of their picture information in the dark through mid range, and very little picture info in the highlights. I was starting to wonder if maybe the CCD just didn't go that high, but this last one, cap11 HD, seemed to have at least some info at the high end. So, I am interested to see how the raw images deal with the upper end of brightness; how it transitions to pure white. Now that you can control exposure, we can finally see just how far you can push this thing.

June 16th, 2004, 01:55 PM	#817
Juan P. Pertierra Major Player Join Date: Jul 2003 Posts: 479	Thanks Justin.... Here's what I had in mind...setup a similar scene to the last post, and take RAW and DV frames in small exposure interfvals. I'll try and put together a collage of the DV and raw frames so we can see how much more you can push it. Haven't gotten the anamorphic yet, any minute now. Gonna get a gyro, then i'll start the tests..... Juan

June 16th, 2004, 05:13 PM	#818
Juan P. Pertierra Major Player Join Date: Jul 2003 Posts: 479	Here's a thumbnail of the images i'm getting with he anamorphic adapter. I did some color correction on this one and applied a log curve... http://expert.cc.purdue.edu/~pertierr/cap12_pre.tif the few dots you see are speckles. I've been trying to wrap my head around the myriad of aspect ratio changes involved, i think i got it close but it might not be perfect. Juan

June 16th, 2004, 05:53 PM	#819
Obin Olson Trustee Join Date: Jan 2003 Location: Wilmington NC Posts: 1,414	Juan how are you going to deal with 12bit files? what can edit a 12bit file? I know after effects can open 16bit files but 12? I am having the same issue with the 4:4:4 10bit camera

June 16th, 2004, 05:57 PM	#820
Juan P. Pertierra Major Player Join Date: Jul 2003 Posts: 479	Put the 12-bit data in the upper 12 bits of 16 bits. Set the lower 4 bits to 0. It's the same thing. The data is 16-bits in size but 12-bits in precision. Juan

June 16th, 2004, 06:09 PM	#821
Obin Olson Trustee Join Date: Jan 2003 Location: Wilmington NC Posts: 1,414	so then you waste disk space ? or does the 0 mean nothing in the bottom=16bit file that is the size of a 12bit file?

June 16th, 2004, 06:12 PM	#822
Juan P. Pertierra Major Player Join Date: Jul 2003 Posts: 479	At processing time, having 16-bit data actually helps because even though you start with 12-bit precision, all color corrections/FX are done in 16-bit precision. If you are concerned about storage, then you need to PACK the data. Computers works in 8-bit bytes, so if you want to use the maximum amount of space, you have to pack the 12-bit data across byte boundaries, which could require some basic error correction to make sure you don't loose data somewhere.

June 16th, 2004, 06:17 PM	#823
Obin Olson Trustee Join Date: Jan 2003 Location: Wilmington NC Posts: 1,414	do you know a good fast codec for packing? I will need it BAD as a tiff file at HD is about 5 megs! that will never edit in a computer unless its a SMOKE or some fancy box..do you know anything about BitJazz the makers of the SheerVideo codec?

June 16th, 2004, 06:36 PM	#824
Juan P. Pertierra Major Player Join Date: Jul 2003 Posts: 479	35mm vs DVX100 RAW test ~~~~~~~~~~~~~~~~~ 35mm Fujifilm ISO100 image: http://expert.cc.purdue.edu/~pertierr/35vsDVX_35.jpg DVX100 image: http://expert.cc.purdue.edu/~pertierr/35vsDVX_RAW.tif These were taken yesterday, without the anamorphic adapter. The scene is exactly the same in both images, the only differences being the camera/medium, and the film camera was located higher and to the left since the DVX is not very easy to move right now. Juan

June 16th, 2004, 08:33 PM	#825
Guest Posts: n/a	Juan, great example! The DVX image looks excellent. Will the speckles in the RAW DVX image ever be gone completely? I've noticed those in previous images as well. I recall you saying that would be fixed with a different capacitor, I could be wrong though.