color correct in house or at home?

[grainy]

Hi folks, any thoughts on the "value" of getting a straight transfer, through a single frame capture machine such as the viperHD, and THEN getting it color corrected, vs. just doing a bit of color correction (using FCP) after the fact? I don't entirely understand the added value of having it done in the transfer house vs. doing it myself at home. I apologise for the ignorance of the question, but I see a lot of (I assume) uncorrected footage samples out there and they look great... just wondering.
thanks
G

[mattias]

the transfer house will most likely not change the scanner settings during a session but do the corrections digitally, so assuming you get at least a 10 bit uncompressed file or tape the result will technically be identical. i think the main reason to have them color correct it is they're better at it than you are. ;-)

[mattias]

it depends on the scanner btw. companies using the flashscan will more likely ride the controls on the scanner for several reasons. first because you can easily, and second because the limited dynamic range would otherwise clip data which isn't recoverable.

[grainy]

the transfer house will most likely not change the scanner settings during a session but do the corrections digitally, so assuming you get at least a 10 bit uncompressed file or tape the result will technically be identical. i think the main reason to have them color correct it is they're better at it than you are. ;-)

Thanks Mattias - so you're saying that with a 10 bit scan what you're getting are improvements on the original?

[mattias]

I'm not sure what you're asking, but what I'm saying is that with a 10 bit uncompressed file or tape you get the exact same source material as the post house would use, so there's nothing except eye and talent that limits your ability to get the same results at home. i don't know the details of the scanner you mention though. it's roger's? ask him.

[carllooper]

Not sure if the following is any help in understanding - but here goes:

The higher the number of "bits" of the scan (eg. a 16 bit scan vs an 8 bit scan), the closer the scan is to capturing the full dynamic range of the original film. You want something close to the full range so you can then adjust things like colour, contrast, levels, etc. Without the full range you'll find that when you adjust things in post that you could get results that are "clipped", meaning that the adjustment you are trying to do exceeds the ammount of information available in the scan.

Carl

[mattias]

the "clipping range" of a scanner is limited by saturation in the sensor though, no number of bits can bring back information that has been electronically clipped. more bits means less quantization noise so there's more info within the range. i know you know this, i just wanted to clarify since your post could be interpreted as if more bits would limit clipping, which it doesn't. the reason i mentioned 10 bits is not because that captures all the available info, but because it's the highest normally used in video. if you scan on a "datacine" machine you might get 12 bit dpx files, which is even better of course.

[grainy]

great info, thanks folks - my lesson: have the pros do it. ;)
thanks!
G

[mattias]

a good lesson to have learned for sure, but if you think you got the eye for it don't let technology stop you. a calibrated monitor and practice is all you need.

[carllooper]

the "clipping range" of a scanner is limited by saturation in the sensor though, no number of bits can bring back information that has been electronically clipped. more bits means less quantization noise so there's more info within the range. i know you know this, i just wanted to clarify since your post could be interpreted as if more bits would limit clipping, which it doesn't. the reason i mentioned 10 bits is not because that captures all the available info, but because it's the highest normally used in video. if you scan on a "datacine" machine you might get 12 bit dpx files, which is even better of course.

Hi Mattias,

yes - good point.

A given scanner will, of course, have a fixed range with respect to what it can see of the original film - irregardless of the number of bits it might othrwise use to divide up that range (or the number of bits a particular file format might use to do the same).

The way in which I described a scanner, in terms of bits, assumes that a scanner that saw a greater range would use more bits to represent that greater range (and a file format with comparable number of bits). But of course it needn't. An expensive HDR scanner, that could see the greatest range, could still just use 8 bits to represent that range - although if the scanner could see a greater range (than another scanner) the manufacturers would typically allocate more bits to represent that range - as in your "datacine" example.

Basically the real reason to let a transfer house adjust the scanning in house (with or without direction from you) is so they can compensate for their scanners inability to see the full range of the film.

However, this is not a digital adjustment (otherwise you could just get the transfer house to give you the exact same data they were using and digitally adjust it yourself at home). Rather, it would be a "physical" adjustment of the sensor exposure or light (to target a particular range of the original film). That physical adjustment is digitally controlled of course.

A third option would be to take the scanner home.

Carl