4:4:4 12-bit Uncompressed DVX100

Peter Plevritis · April 6th, 2004, 12:03 AM

Got really quiet suddenly. Did I miss something and this topic was moved?

Looking forward to seeing the continuous capture.

Run into any problems?

Pete

Stephen van Vuuren · April 6th, 2004, 12:08 AM

I just moved it to it's new home here.

Milosz Krzyzaniak · April 6th, 2004, 09:25 AM

I don't understand. Is there any new location of this thread?

Juan P. Pertierra · April 6th, 2004, 09:28 AM

Peter:

I've reset the chip probe and will hook it all up again tonight to see if I got all 30 bits this time. I've been slowed down a bit by a nasty toothache...

Milosz:
This thread is now in the Alternative Imaging Methods forum, it used to be in the DVX100 forum.

Juan

Rob Lohman · April 7th, 2004, 02:53 AM

Milosz: yes, the thread was moved to another forum. It is now
in "Alternative Imagine Methods". I think it was in the Film Look
forum.

The thread identifier hasn't changed. So if you have the thread
bookmarked for example or have notifications on the thread,
those will still be in place and working! So no problems.

Filip Kovcin · April 9th, 2004, 02:03 PM

juan, any news? any pictures? movies?

filip

Juan P. Pertierra · April 9th, 2004, 02:05 PM

Sorry for the lack of updates, i have been working on a better solution to the probing problem, and i'm almost done with it...now it's just snap-and-go, and hopefully all intermittent bit connections will be fixed.

I have also been working simultaneously on my software which now spits out 16-bit/channel RGB individual frames in a new directory, takes care of flipping over the blue channel, etc.

more to come soon.

Juan

Juan P. Pertierra · April 11th, 2004, 08:51 PM

Allright! So after much time experimenting i designed a new method of easily probing the pins. The good news is that i'm pretty sure I am getting all 30 bits(10-bit RGB). The not so good news is that there is still some noise, which must either be related to the bad cables I am using, or somehow related to the missing lower bits. However, i beleive it has less noise than the images i've posted so far.

Right now the whole thing sits in the same place, so if you'd like to see some boring frames, let me know what you'd like to see(coke can, magazine cover, etc.) as well as lighting.

Tomorrow i'm going to finally move it all to the window and take a decent clip of a car driving by. If it's sunny, it's bound to be a good test, there's plenty of color around. Right now because of my setup i don't have control over exposure(controller disconnected) so everything is at full open, hopefully the ND's will do the trick.

Another note:
While working on the camera I was amazed once again at how much space there is behind each CCD chip, and I realized that an incredibly easy and useful mod would be to mount thermoelectric coolers on each chip! These are only about $15 a piece, and I can adhere one to each chip, run them off a battery. This should have a great effect on the S/N ratio, it's a similar technique to what is used in astrophotography, to reduce the thermal noise to a minimum...i'm not even sure why some manufacturer hasn't included it.

Juan

Juan P. Pertierra · April 11th, 2004, 09:27 PM

Interesting!

Ok, let me see how to explain this...

While looking through the viewfinder in my current scene, there are clearly areas in the scene which are overexposed(allwhite).

Now, i would expect this to be the point at which the CCD's clip, and like so i would expect to get white or close to white in those areas in the RAW feed...

HOWEVER, i am getting values in those areas as if the A/D's actually topped off the 12-bit but instead of giving me the max 12-bit value(i.e.white), it's going around and starting over at the bottom, giving me a dark value, but which really reflects what is in that white area! I'm not sure what is going on, but the way it looks now i think i can use this to pull detail out of the areas that clip out in standard DV footage...

Juan

Obin Olson · April 11th, 2004, 09:28 PM

glad to hear from you again! are you talking about Peltier coolers? how much would that effect the s/n do you think?

keep at it..I am sure you will figure it out....I would love to see a pic or 2 of your setup on the test probe and how it connects to the chips...not that I would ever do it but just to see how it works ;)

Adam Burtle · April 11th, 2004, 09:39 PM

<<<-- Originally posted by Juan P. Pertierra : I'm not sure what is going on, but the way it looks now i think i can use this to pull detail out of the areas that clip out in standard DV footage...

Juan -->>>

Great. Expecting to outfit at least 3 XL's with your mod (whenever finished), and the only remaining problem i have with emulating/duplicating film quality, is the lack of latitude in exposude on digital.. so if you can improve that, all the better!

Milosz Krzyzaniak · April 12th, 2004, 07:25 AM

So, Juan? It seems that my prediction was right? The same as in my friend's Canon D300 the RAWs give more dinamic than standard pictures. And I'll tell you I think this would be the biggest advantage of RAWs among all (4:4:4, no compression).

Suggestion on how to prepare the test situation to shoot. You dont need to move your dvx to the window to obtain some movement, you just may use your hand to move something. Put a doll or something in front of the camera with a lot of colorful things. And the most importatnt: light it with much contrast: put a light to the side of the scene and light to overexposion: we would like to see what is the behaviour of this RAW thing in such conditions.

Just rememer to film not a planar thing (wall, book) but a real 3D something that could give shadows and so on.

Good luck.

Rob Lohman · April 12th, 2004, 07:32 AM

Adam: what makes you think this will work with an XL1S? Juan
is building this with a DVX100 as you can see from the thread
title. I would not assume (automatically) that:

1) the XL1S has 12 bit stuff in it (I don't know it if has)

2) someone knows how to get to it

3) someone has tested whether it will work

4) someone will ever do any of the previous points

etc. A better question might be to ask whether someone knows
if this will work with an XL1S and/or can test that....

Obin Olson · April 12th, 2004, 10:17 AM

wrangle him rob ;)

Obin Olson · April 12th, 2004, 10:18 AM

how much more dynamic range does film still have over the current crop of cmos and ccd cameras/video cameras?

April 6th, 2004, 12:08 AM	#287
Stephen van Vuuren Space Hipster Join Date: Oct 2001 Location: Greensboro, NC Posts: 1,508	I just moved it to it's new home here. __________________ stephen v2 www.insaturnsrings.com

April 7th, 2004, 02:53 AM	#290
Rob Lohman RED Code Chef Join Date: Oct 2001 Location: Holland Posts: 12,514	Milosz: yes, the thread was moved to another forum. It is now in "Alternative Imagine Methods". I think it was in the Film Look forum. The thread identifier hasn't changed. So if you have the thread bookmarked for example or have notifications on the thread, those will still be in place and working! So no problems. __________________ Rob Lohman, visuar@iname.com DV Info Wrangler & RED Code Chef Join the DV Challenge \| Lady X Search DVinfo.net for quick answers \| Buy from the best: DVinfo.net sponsors

April 11th, 2004, 08:51 PM	#293
Juan P. Pertierra Major Player Join Date: Jul 2003 Posts: 479	Update... Allright! So after much time experimenting i designed a new method of easily probing the pins. The good news is that i'm pretty sure I am getting all 30 bits(10-bit RGB). The not so good news is that there is still some noise, which must either be related to the bad cables I am using, or somehow related to the missing lower bits. However, i beleive it has less noise than the images i've posted so far. Right now the whole thing sits in the same place, so if you'd like to see some boring frames, let me know what you'd like to see(coke can, magazine cover, etc.) as well as lighting. Tomorrow i'm going to finally move it all to the window and take a decent clip of a car driving by. If it's sunny, it's bound to be a good test, there's plenty of color around. Right now because of my setup i don't have control over exposure(controller disconnected) so everything is at full open, hopefully the ND's will do the trick. Another note: While working on the camera I was amazed once again at how much space there is behind each CCD chip, and I realized that an incredibly easy and useful mod would be to mount thermoelectric coolers on each chip! These are only about $15 a piece, and I can adhere one to each chip, run them off a battery. This should have a great effect on the S/N ratio, it's a similar technique to what is used in astrophotography, to reduce the thermal noise to a minimum...i'm not even sure why some manufacturer hasn't included it. Juan

April 12th, 2004, 07:32 AM	#298
Rob Lohman RED Code Chef Join Date: Oct 2001 Location: Holland Posts: 12,514	Adam: what makes you think this will work with an XL1S? Juan is building this with a DVX100 as you can see from the thread title. I would not assume (automatically) that: 1) the XL1S has 12 bit stuff in it (I don't know it if has) 2) someone knows how to get to it 3) someone has tested whether it will work 4) someone will ever do any of the previous points etc. A better question might be to ask whether someone knows if this will work with an XL1S and/or can test that.... __________________ Rob Lohman, visuar@iname.com DV Info Wrangler & RED Code Chef Join the DV Challenge \| Lady X Search DVinfo.net for quick answers \| Buy from the best: DVinfo.net sponsors

April 6th, 2004, 12:03 AM	#286
Peter Plevritis Regular Crew Join Date: Mar 2004 Posts: 27	Got really quiet suddenly. Did I miss something and this topic was moved? Looking forward to seeing the continuous capture. Run into any problems? Pete

April 6th, 2004, 09:25 AM	#288
Milosz Krzyzaniak Regular Crew Join Date: Mar 2004 Location: Germany Posts: 64	I don't understand. Is there any new location of this thread?

April 6th, 2004, 09:28 AM	#289
Juan P. Pertierra Major Player Join Date: Jul 2003 Posts: 479	Peter: I've reset the chip probe and will hook it all up again tonight to see if I got all 30 bits this time. I've been slowed down a bit by a nasty toothache... Milosz: This thread is now in the Alternative Imaging Methods forum, it used to be in the DVX100 forum. Juan

April 9th, 2004, 02:03 PM	#291
Filip Kovcin Major Player Join Date: Dec 2003 Location: warsaw, poland Posts: 440	juan, any news? any pictures? movies? filip

April 9th, 2004, 02:05 PM	#292
Juan P. Pertierra Major Player Join Date: Jul 2003 Posts: 479	Sorry for the lack of updates, i have been working on a better solution to the probing problem, and i'm almost done with it...now it's just snap-and-go, and hopefully all intermittent bit connections will be fixed. I have also been working simultaneously on my software which now spits out 16-bit/channel RGB individual frames in a new directory, takes care of flipping over the blue channel, etc. more to come soon. Juan

April 11th, 2004, 09:27 PM	#294
Juan P. Pertierra Major Player Join Date: Jul 2003 Posts: 479	Interesting! Ok, let me see how to explain this... While looking through the viewfinder in my current scene, there are clearly areas in the scene which are overexposed(allwhite). Now, i would expect this to be the point at which the CCD's clip, and like so i would expect to get white or close to white in those areas in the RAW feed... HOWEVER, i am getting values in those areas as if the A/D's actually topped off the 12-bit but instead of giving me the max 12-bit value(i.e.white), it's going around and starting over at the bottom, giving me a dark value, but which really reflects what is in that white area! I'm not sure what is going on, but the way it looks now i think i can use this to pull detail out of the areas that clip out in standard DV footage... Juan

April 11th, 2004, 09:28 PM	#295
Obin Olson Trustee Join Date: Jan 2003 Location: Wilmington NC Posts: 1,414	glad to hear from you again! are you talking about Peltier coolers? how much would that effect the s/n do you think? keep at it..I am sure you will figure it out....I would love to see a pic or 2 of your setup on the test probe and how it connects to the chips...not that I would ever do it but just to see how it works ;)

April 11th, 2004, 09:39 PM	#296
Adam Burtle Regular Crew Join Date: Oct 2003 Location: Los Angeles, CA Posts: 173	<<<-- Originally posted by Juan P. Pertierra : I'm not sure what is going on, but the way it looks now i think i can use this to pull detail out of the areas that clip out in standard DV footage... Juan -->>> Great. Expecting to outfit at least 3 XL's with your mod (whenever finished), and the only remaining problem i have with emulating/duplicating film quality, is the lack of latitude in exposude on digital.. so if you can improve that, all the better!

April 12th, 2004, 07:25 AM	#297
Milosz Krzyzaniak Regular Crew Join Date: Mar 2004 Location: Germany Posts: 64	So, Juan? It seems that my prediction was right? The same as in my friend's Canon D300 the RAWs give more dinamic than standard pictures. And I'll tell you I think this would be the biggest advantage of RAWs among all (4:4:4, no compression). Suggestion on how to prepare the test situation to shoot. You dont need to move your dvx to the window to obtain some movement, you just may use your hand to move something. Put a doll or something in front of the camera with a lot of colorful things. And the most importatnt: light it with much contrast: put a light to the side of the scene and light to overexposion: we would like to see what is the behaviour of this RAW thing in such conditions. Just rememer to film not a planar thing (wall, book) but a real 3D something that could give shadows and so on. Good luck.

April 12th, 2004, 10:17 AM	#299
Obin Olson Trustee Join Date: Jan 2003 Location: Wilmington NC Posts: 1,414	wrangle him rob ;)

April 12th, 2004, 10:18 AM	#300
Obin Olson Trustee Join Date: Jan 2003 Location: Wilmington NC Posts: 1,414	how much more dynamic range does film still have over the current crop of cmos and ccd cameras/video cameras?