Jump to content


Photo

Rushes Tk consistently dark


  • Please log in to reply
7 replies to this topic

#1 Ian Cooper

Ian Cooper
  • Basic Members
  • PipPipPipPip
  • 469 posts
  • Other
  • England

Posted 18 July 2010 - 03:43 PM

Over the last couple of years I've used three different labs here in the UK to process and Tk my 16mm neg film to BetaSP & MiniDV (the only formats I can handle). They've all offered a 'best light' service as standard. The results I get back seem to be consistently dark and look underexposed, irrespective of where I've gone. The obvious suggestion is that I'm the one at fault, that I'm underexposing the negative!

I have had a process & workprint done as a camera test, the print on that occasion came out looking fine, and the report with it seemed to suggest the exposure of the neg was correct as well (30,34,29 , I'd tried overexposing the neg about 1/3 of a stop). With stills photography I almost exclusively use medium format reversal film, so I'm no stranger to using a totally manual camera with a seperate light meter - and needing to get the exposure right!

When I had some more 16mm film processed back around Easter time the results were once again dark. On this occasion I got back in touch with the lab, explained the problem and asked for their help. They suggested I send the neg back down to them and they'd check the density etc. The result once again was that apparently the neg was exposed fine. Having reviewed the Tk the lab did agree the levels on one or two shots were perhaps a little on the dark side, but overall their comment was: "...In general I thought this was a very good rushes transfer (given the limitations in an overnight rushes scenario) We do our best to add as much to the grade as possible overnight but of course there is, more often than not in all projects, room for some improvement in Post Production..."

I can obviously lighten the material myself on the PC to a certain extent, but the results can be so dark I end up with images riddled with noise! I don't think I'm expecting anything special, just neutral colour, average brightness, no special 'effects' - the video equivalent of a workprint really.

This weekend I've just received back some more film and once again I feel the Tk results are very dark and appear underexposed. Here's a quick rough cut of the Tk with no fiddling of my own:

Unadjusted rushes

Bearing in mind the bright sunny afternoon weather, the outdoor shots look like dusk! The sequences in the boiler room I knew were going to be a struggle as the illumination was a single 2k blonde, but everything after 1:11 in the engine room was lit largely by ambient with just some artificial fill. Generally the exposure was sitting between f4 and f5.6, I was taking both incident and reflective light readings and the two were usually comparable.

I'm slightly surprised to see detail outside the windows in a number of the shots - the light readings outside were about 6 stops brighter than I was metering and exposing for inside, so even allowing for the window glass dimming it down a bit I'd expect the windows to be totally blown out and overexposed - they didn't factor in my metering for the scene at all.


I'm at somewhat of a loss: The results I get back from a best light rushes Tk is (to me) too dark to really do much with other than check focus - there's almost no highlight info in it, yet I'm told these results are 'good', and my own experience seems to suggest this is indeed the 'standard' brightness for such a service from more than one company. I'm led to believe my exposure of the 16mm negative is good, and experience would tend to suggest I could get a 'normal' brightness workprint - yet getting an equivalent brightness Tk seems impossible, despite explaining and asking. A camcorder on autoexposure doesn't give similar dark results, a C41 still film printed on 'auto' settings in the high street doesn't give dark results, when I expose reversal film I don't get dark results, and nothing broadcast on TV looks that dark.

Is there something specific I'm not understanding about the process/service? ...and more to the point, what do I need to do or request to get nice bright well exposed Tk?

I've tried shooting grey cards at the start, I've tried explaining I want to be able to just edit the results together to burn to DVD and have similar brightness/contrast to normal TV pictures, I've tried just asking for 'bright' results, I've tried explaining I don't want any futher Tk or grading etc... when I've got back to the lab they've been most helpful, checking the neg for me etc, but ultimately they don't seem to be seeing the darkness problems I am. The labs themselves are all a long distance from where I live, so calling by in person isn't really an option.

Any clues as to what I might be doing wrong, and/or what I need to do next?
Does the Tk above seem dark to anyone else?


This is the same rough cut after I've tried adjusting the gamma and levels on the PC - watch out for the objectionable noise/grain, although the compression has mushed and hidden a fair bit.

PC adjusted rushes
  • 0

#2 Stephen Williams

Stephen Williams
  • Sustaining Members
  • 4708 posts
  • Cinematographer
  • Europe

Posted 18 July 2010 - 04:24 PM

Looks rather bright to me!

Over the last couple of years I've used three different labs here in the UK to process and Tk my 16mm neg film to BetaSP & MiniDV (the only formats I can handle). They've all offered a 'best light' service as standard. The results I get back seem to be consistently dark and look underexposed, irrespective of where I've gone. The obvious suggestion is that I'm the one at fault, that I'm underexposing the negative!

I have had a process & workprint done as a camera test, the print on that occasion came out looking fine, and the report with it seemed to suggest the exposure of the neg was correct as well (30,34,29 , I'd tried overexposing the neg about 1/3 of a stop). With stills photography I almost exclusively use medium format reversal film, so I'm no stranger to using a totally manual camera with a seperate light meter - and needing to get the exposure right!

When I had some more 16mm film processed back around Easter time the results were once again dark. On this occasion I got back in touch with the lab, explained the problem and asked for their help. They suggested I send the neg back down to them and they'd check the density etc. The result once again was that apparently the neg was exposed fine. Having reviewed the Tk the lab did agree the levels on one or two shots were perhaps a little on the dark side, but overall their comment was: "...In general I thought this was a very good rushes transfer (given the limitations in an overnight rushes scenario) We do our best to add as much to the grade as possible overnight but of course there is, more often than not in all projects, room for some improvement in Post Production..."

I can obviously lighten the material myself on the PC to a certain extent, but the results can be so dark I end up with images riddled with noise! I don't think I'm expecting anything special, just neutral colour, average brightness, no special 'effects' - the video equivalent of a workprint really.

This weekend I've just received back some more film and once again I feel the Tk results are very dark and appear underexposed. Here's a quick rough cut of the Tk with no fiddling of my own:

Unadjusted rushes

Bearing in mind the bright sunny afternoon weather, the outdoor shots look like dusk! The sequences in the boiler room I knew were going to be a struggle as the illumination was a single 2k blonde, but everything after 1:11 in the engine room was lit largely by ambient with just some artificial fill. Generally the exposure was sitting between f4 and f5.6, I was taking both incident and reflective light readings and the two were usually comparable.

I'm slightly surprised to see detail outside the windows in a number of the shots - the light readings outside were about 6 stops brighter than I was metering and exposing for inside, so even allowing for the window glass dimming it down a bit I'd expect the windows to be totally blown out and overexposed - they didn't factor in my metering for the scene at all.


I'm at somewhat of a loss: The results I get back from a best light rushes Tk is (to me) too dark to really do much with other than check focus - there's almost no highlight info in it, yet I'm told these results are 'good', and my own experience seems to suggest this is indeed the 'standard' brightness for such a service from more than one company. I'm led to believe my exposure of the 16mm negative is good, and experience would tend to suggest I could get a 'normal' brightness workprint - yet getting an equivalent brightness Tk seems impossible, despite explaining and asking. A camcorder on autoexposure doesn't give similar dark results, a C41 still film printed on 'auto' settings in the high street doesn't give dark results, when I expose reversal film I don't get dark results, and nothing broadcast on TV looks that dark.

Is there something specific I'm not understanding about the process/service? ...and more to the point, what do I need to do or request to get nice bright well exposed Tk?

I've tried shooting grey cards at the start, I've tried explaining I want to be able to just edit the results together to burn to DVD and have similar brightness/contrast to normal TV pictures, I've tried just asking for 'bright' results, I've tried explaining I don't want any futher Tk or grading etc... when I've got back to the lab they've been most helpful, checking the neg for me etc, but ultimately they don't seem to be seeing the darkness problems I am. The labs themselves are all a long distance from where I live, so calling by in person isn't really an option.

Any clues as to what I might be doing wrong, and/or what I need to do next?
Does the Tk above seem dark to anyone else?


This is the same rough cut after I've tried adjusting the gamma and levels on the PC - watch out for the objectionable noise/grain, although the compression has mushed and hidden a fair bit.

PC adjusted rushes


  • 0

#3 Oliver Christoph Kochs

Oliver Christoph Kochs
  • Sustaining Members
  • 323 posts
  • Film Loader
  • Germany

Posted 18 July 2010 - 04:25 PM

Since they don't have any reference for doing the one-light print or telecine it helps to shoot a color bar and maybe also a grey bar chart. Repeat at significant light changes. There are accurate test charts one may not want to buy. Color bars are on many clapboards as well but i don't think these are very accurate. In your case a good inkjet print of the color bars on matte paper might be a start. If you do that you can at least complain about the prints wrong colors or brightness. Hope it helps, Oliver.
  • 0

#4 Ian Cooper

Ian Cooper
  • Basic Members
  • PipPipPipPip
  • 469 posts
  • Other
  • England

Posted 18 July 2010 - 05:13 PM

Looks rather bright to me!


Which, the unadjusted version?!
Oh well, perhaps it is just me then.

I'd have thought this:
Posted Image
would look closer to this:
Posted Image



Likewise,
Posted Image
would seem to represent a polished brass makers plate like:
Posted Image



and this looks rather dull and flat to me
Posted Image
compared to this version:
Posted Image



From the film at Easter there was this frame shot on a bright sunny day:
Posted Image
Yet watching it on the TV it doesn't seem to have any highlight bright areas, it just seems dull.

This version seems more consistent with the levels of broadcast TV channels and commercial DVDs to me:
Posted Image




Thank you both for your comments, I'm even more puzzled if that IS seen as bright though!
  • 0

#5 Ian Cooper

Ian Cooper
  • Basic Members
  • PipPipPipPip
  • 469 posts
  • Other
  • England

Posted 18 July 2010 - 06:11 PM

I realise it isn't a direct comparison, but this was shot on BetaSP a couple of years ago:

Posted Image
Posted Image
Posted Image


To me the frame appears bright and neither over nor under exposed. That's the type of brightness I'm hoping for from Tk'ed film.





This next frame grab is almost the same view shot under almost the same lighting conditions but on 16mm film, with a 'best light' Tk.

Posted Image
Posted Image
Posted Image

To me it appears dark and dull with almost no highlights, the histogram seems to suggest the same to me as well.




If I then adjust the levels I can produce:

Posted Image
Posted Image

The image now seems well exposed, and the histogram seems to look better balanced as well - just a shame there's all but no image data left to fill it smoothly!


What am I doing wrong, not understanding, or not asking for with the Tk??
  • 0

#6 Saul Rodgar

Saul Rodgar
  • Basic Members
  • PipPipPipPip
  • 1682 posts
  • Cinematographer

Posted 18 July 2010 - 08:29 PM

Ian: The unadjusted Tk footage looks fine on my Apple display. The adjusted stuff can get overly bright and soulless IMO. Some of the grabs do look darker than the running footage tho.

The most obvious question is how are you metering your light?

Beyond that, I think your problem may be of expectations more than anything else. A light meter gives you the proper exposure for 18% gray only, but if the light meter is (luminance) ANSI calibrated it seems the exposure will be proper for 12% gray instead.

If you want it brighter (or darker) you have to compensate yourself, and that is the trick, knowing where to measure to get the exposure you expect. I don't know if you are familiar with the zone system, but even somewhat cursory understanding of it can be most useful on the field.

The light conditions you were filming under where rather flat, so it makes sense to me the results look the way they do. But they are not underexposed. They look like a lot of the footage I have seen of England, Northern Europe and the Pacific NW in the US, and that is the result of constant overcast skies more than anything else.
  • 0

#7 Ian Cooper

Ian Cooper
  • Basic Members
  • PipPipPipPip
  • 469 posts
  • Other
  • England

Posted 19 July 2010 - 02:26 AM

Ian: The unadjusted Tk footage looks fine on my Apple display. The adjusted stuff can get overly bright and soulless IMO. Some of the grabs do look darker than the running footage tho.

The most obvious question is how are you metering your light?

Beyond that, I think your problem may be of expectations more than anything else. A light meter gives you the proper exposure for 18% gray only, but if the light meter is (luminance) ANSI calibrated it seems the exposure will be proper for 12% gray instead.


I'm using a Sekonic L358. I came across the 12%/18% issue when I first got it, trying to shoot reversal stills film - all the results were dark! I've now tweaked the 'compensation' function of the light meter so I can get 'correct' exposures every time using Velvia 120 stills film, predominently using just incident metering. On a clear blue-sky sunny summer day (they do occasionally happen in the UK!) the meter will confirm the "sunny 16" guide. In 'cine' mode the meter assumes a 180 degree shutter (which the camera I was using was also set to), and I had the fps of the meter set at 25 to match the camera speed. I get the same readings if I set the shutter speed at 1/50 as well. My stills photos taken on reversal film in the same location (and other similar places) using similar metering techniques don't appear particularly dark to me.

For the pumping station shots I was using both incident metering and comparing it to reflective readings (taking care not to let bright windows in the field of meter view influence the reading).

Generally speaking the reading I was getting was between f4 and f5.6, and my setting of the lens tended to err slightly on the side of over exposure. The meter reading outside was about f45, so 6 stops brighter. To get the outdoor shots on the same film I had to use 2 stops of ND and still have the lens fully closed at F22. In the background of the indoor shots the detail seen through the open door and through the windows is clearly brighter and slightly overexposed, but to me it only seems perhaps 1.5 stops or so over - certainly not the 6 stops brighter than the lens setting that I was metering and exposing the main subject at.






If you want it brighter (or darker) you have to compensate yourself, and that is the trick, knowing where to measure to get the exposure you expect. I don't know if you are familiar with the zone system, but even somewhat cursory understanding of it can be most useful on the field.


In the past I've tried rating the entire film slower, effectively overexposing it all by 2/3 stop and the results came back just as dark as when I don't deliberately 'overexpose'. At the end of this film I shot a couple of seconds in my flat at night lit just by a 100W and a 60W domestic light bulb - the meter reading was just below f1 (at 500asa), whilst the lens was wide open at f1.8 thus underexposing by just over a stop. The Tk result looks dark, but is no darker than other shots where I'd deliberately overexposed about 0.5 stop. To be fair I'd expect a best light Tk to even out exposure variations, it's just what they're being leveled out to that I'm less sure about.

I'm not an expert on the zone system, but I have followed the basic principles and do on occasion use it when I'm exposing, developing and printing my own B&W stills film - I've also used it to adjust the EI I rate B&W film at to match my development techniques and enlarger contrast.


I think you're probably right about expectations, both yourself and Stephen feel the straight Tk results look fine. Personally I'd have expected the shot of the pressure gauges to be similar brightness to the BetaSP frame grab, with the majority of the highlight detail reaching around 80%, and the white notices appearing white up around 100%. The Tk result however is down near 20%-30% with no detail above 40%, and the white drawings appear dull grey (to me). I've tried asking and explaining to the labs what I'm hoping for on previous occasions, but seem to largely get Tk results looking the same. If the consensus of opinion is that these already look bright enough - then that would explain why!

Thank you for letting me know your feelings.
  • 0

#8 K Borowski

K Borowski
  • Basic Members
  • PipPipPipPip
  • 3905 posts
  • Camera Operator
  • I.A.T.S.E. Local # 600 Eastern Region

Posted 19 July 2010 - 10:49 AM

Solve this by shooting a full-frame grey card that you believe you've lit properly, send it to the lab, and, adjusted for that emulsion batch's base fog, it should read .80R 1.20G 1.60B on a densitometer. Unless, of course, there is a problem with the chemistry at the lab, it will tell you if your exposure is off. You can shoot and measure a pure white source (d-max) or just a severely overexposed grey card and d-min, and a bright blue source (or with a deep blue filter a white card) to check on the lab too.


I'd forget about the TK results completely. There it could be anything. . .
  • 0


Opal

rebotnix Technologies

Abel Cine

Aerial Filmworks

Media Blackout - Custom Cables and AKS

CineTape

Gamma Ray Digital Inc

Willys Widgets

The Slider

Paralinx LLC

Ritter Battery

CineLab

Visual Products

Technodolly

Rig Wheels Passport

Wooden Camera

Metropolis Post

Tai Audio

Broadcast Solutions Inc

Glidecam

FJS International, LLC

Visual Products

Gamma Ray Digital Inc

FJS International, LLC

Technodolly

Broadcast Solutions Inc

The Slider

Wooden Camera

Paralinx LLC

rebotnix Technologies

Glidecam

Ritter Battery

Media Blackout - Custom Cables and AKS

Opal

Aerial Filmworks

Metropolis Post

Rig Wheels Passport

Tai Audio

CineTape

Willys Widgets

Abel Cine

CineLab