BBC film look article at DVinfo.net

Frank Granovski · May 23rd, 2003, 02:11 AM

http://www.bbc.co.uk/rd/pubs/whp/index.html

Zac Stein · May 23rd, 2003, 05:45 AM

Wow what an article, anybody care to do an english translation?

Zac

Jason Rodriguez · May 23rd, 2003, 10:16 AM

I can't seem to access the article, if anyone has it, I'd appreciate if they could email it to me at jrod@mindspring.com with the subject "Film Look".

Thanks

Jason Rodriguez

Christopher Hughes · May 23rd, 2003, 10:30 AM

I have just been on the site, around the time you posted your message and downloaded it without any problems. I WOULD send you it, but its too big for an attachment with my email account. If you go to the site to download, its pretty slow and its a big file pushing 2.33MBs so It'll take a while. Sometimes it seems like the screen is freezing as you move to another screen then try to go back to download page. But I found this common with PDF downloads that download as you view page by page. So just be patient I guess.

Nigel Moore · May 23rd, 2003, 11:55 AM

I downloaded without any trouble (which is more than can be said for understanding it!)

Jason, I've e-mailed it to you, but if you have any problems, let me know and I'll try zipping it.

nigel[at]truffy[dot]com

Cheers

Stylianos Moschapidakis · May 23rd, 2003, 12:41 PM

I have been having trouble downloading pdf files and also opening attachments lately. Could anyone e-mail me the article? But please copy-and-paste it. It'd be great if someone could do that.

Thank you.

Rick Spilman · May 23rd, 2003, 01:06 PM

Ahem. Well, I started reading it and quickly got bogged down in the math and the graphs. I have an engineering degree so I am sure if I sat down and worked my way through I'm sure I could make sense of it. Do I have the time? That is another question entirely.

Skipping ahead to the conclusions, they are simplier if not necessarily much more useful. Shoot in progressive mode. Set-up the camera gamma corrrector, black stretch and knee controls in a lab as it cannot be reliably done in the field. Roughly the same comment on MTF and skin tones. DOF, do the best you can. BBC has heard of but hasn't tested the mini-35 adapter.

So an interesting if not necessary practical white paper.

Jason Rodriguez · May 23rd, 2003, 02:57 PM

Hey Thanks Nigel for the email.

BTW, I found it very interesting where he describes how the contrast range of high-end video cameras is equal to film. The only problem he mentions though is since the gamma in video is not distributed linearly as in film, you basically have to underexpose a couple stops and do a color correction to get the same percived exposure lattitude of film. In otherwords, there's a ton of unused lattitude in the blacks, and it's noise that limits the use of that range of the curve. So basically he was saying that if you expose a video camera normally, you aren't using the full exposure range of the camera, leading many people to believe that there isn't that much lattitude in video when there in fact is. At least that's what I got out of it.

Jason

Joseph George · May 23rd, 2003, 07:43 PM

I did not read it fully, but here are some highlights. The paper claims that HD and high-end SD cameras can have the same latitude as film, however the curve is very nonlinear. (IMHO the non-linearity would make that wide a range more or less useless. In actuality only Thompson Viper and CineAlta SR can claim same latitude as film.) They suggest using progressive scan with speed similar to film -- 24p, 25p, 30p too would be OK. They said that 2/3" cameras have the same DOF as Super 16 mm film and suggested using Mini 35 adaptor but have not tested it. They have some suggestions that would be useless, as they would increase noise too much. The government owns BBC and they gave some researchers write the paper. No new conclusions really.

Akos Szemenyei · May 23rd, 2003, 09:09 PM

Highly technical but good info, probably have to re-read it a couple of times. Thanks for sharing.

Jason Rodriguez · May 23rd, 2003, 09:51 PM

According to the April American Cinematography article on the DVX100 though, the CineGamma feature is a pure linear response curve, with no black compression or knee. It's not, as the name suggests, a gamma curve that is applied to the video to make in more film-like. Supposedly the curve is a straight line so that when taken to film, the film will impose it's own response curve on the image instead of trying to counteract the video gamma of the camera. So would this type of Gamma curve offset the nonlinearity of normal video gamma that wastes so much dynamic range?

Jason Rodriguez

May 23rd, 2003, 02:11 AM	#1
Frank Granovski Outer Circle Join Date: May 2002 Location: Hope, BC Posts: 7,524	BBC film look article http://www.bbc.co.uk/rd/pubs/whp/index.html

May 23rd, 2003, 11:55 AM	#5
Nigel Moore Major Player Join Date: May 2003 Location: Switzerland Posts: 358	I downloaded without any trouble (which is more than can be said for understanding it!) Jason, I've e-mailed it to you, but if you have any problems, let me know and I'll try zipping it. nigel[at]truffy[dot]com Cheers __________________ _________________ Time flies like an arrow Fruit flies like a banana

May 23rd, 2003, 12:41 PM	#6
Stylianos Moschapidakis Regular Crew Join Date: Apr 2003 Location: Cambridge, MA/Crete, Greece Posts: 189	I have been having trouble downloading pdf files and also opening attachments lately. Could anyone e-mail me the article? But please copy-and-paste it. It'd be great if someone could do that. Thank you. __________________ Stelios G.M.

May 23rd, 2003, 05:45 AM	#2
Zac Stein Major Player Join Date: Aug 2002 Location: Melbourne, Australia Posts: 730	Wow what an article, anybody care to do an english translation? Zac

May 23rd, 2003, 10:16 AM	#3
Jason Rodriguez Trustee Join Date: Mar 2003 Location: Virginia Beach, VA Posts: 1,095	I can't seem to access the article, if anyone has it, I'd appreciate if they could email it to me at jrod@mindspring.com with the subject "Film Look". Thanks Jason Rodriguez

May 23rd, 2003, 10:30 AM	#4
Christopher Hughes Major Player Join Date: Jan 2003 Location: Sunderland, England (y Merida, Mexico) Posts: 269	I have just been on the site, around the time you posted your message and downloaded it without any problems. I WOULD send you it, but its too big for an attachment with my email account. If you go to the site to download, its pretty slow and its a big file pushing 2.33MBs so It'll take a while. Sometimes it seems like the screen is freezing as you move to another screen then try to go back to download page. But I found this common with PDF downloads that download as you view page by page. So just be patient I guess.

May 23rd, 2003, 01:06 PM	#7
Rick Spilman Major Player Join Date: Dec 2002 Location: Jersey City, NJ Posts: 366	Ahem. Well, I started reading it and quickly got bogged down in the math and the graphs. I have an engineering degree so I am sure if I sat down and worked my way through I'm sure I could make sense of it. Do I have the time? That is another question entirely. Skipping ahead to the conclusions, they are simplier if not necessarily much more useful. Shoot in progressive mode. Set-up the camera gamma corrrector, black stretch and knee controls in a lab as it cannot be reliably done in the field. Roughly the same comment on MTF and skin tones. DOF, do the best you can. BBC has heard of but hasn't tested the mini-35 adapter. So an interesting if not necessary practical white paper.

May 23rd, 2003, 02:57 PM	#8
Jason Rodriguez Trustee Join Date: Mar 2003 Location: Virginia Beach, VA Posts: 1,095	Hey Thanks Nigel for the email. BTW, I found it very interesting where he describes how the contrast range of high-end video cameras is equal to film. The only problem he mentions though is since the gamma in video is not distributed linearly as in film, you basically have to underexpose a couple stops and do a color correction to get the same percived exposure lattitude of film. In otherwords, there's a ton of unused lattitude in the blacks, and it's noise that limits the use of that range of the curve. So basically he was saying that if you expose a video camera normally, you aren't using the full exposure range of the camera, leading many people to believe that there isn't that much lattitude in video when there in fact is. At least that's what I got out of it. Jason

May 23rd, 2003, 07:43 PM	#9
Joseph George Permanently Banned Join Date: Mar 2003 Location: Los Angeles, California Posts: 265	I did not read it fully, but here are some highlights. The paper claims that HD and high-end SD cameras can have the same latitude as film, however the curve is very nonlinear. (IMHO the non-linearity would make that wide a range more or less useless. In actuality only Thompson Viper and CineAlta SR can claim same latitude as film.) They suggest using progressive scan with speed similar to film -- 24p, 25p, 30p too would be OK. They said that 2/3" cameras have the same DOF as Super 16 mm film and suggested using Mini 35 adaptor but have not tested it. They have some suggestions that would be useless, as they would increase noise too much. The government owns BBC and they gave some researchers write the paper. No new conclusions really.

May 23rd, 2003, 09:09 PM	#10
Akos Szemenyei Regular Crew Join Date: May 2003 Location: Saitama, Japan Posts: 111	Highly technical but good info, probably have to re-read it a couple of times. Thanks for sharing.

May 23rd, 2003, 09:51 PM	#11
Jason Rodriguez Trustee Join Date: Mar 2003 Location: Virginia Beach, VA Posts: 1,095	According to the April American Cinematography article on the DVX100 though, the CineGamma feature is a pure linear response curve, with no black compression or knee. It's not, as the name suggests, a gamma curve that is applied to the video to make in more film-like. Supposedly the curve is a straight line so that when taken to film, the film will impose it's own response curve on the image instead of trying to counteract the video gamma of the camera. So would this type of Gamma curve offset the nonlinearity of normal video gamma that wastes so much dynamic range? Jason Rodriguez