Dubious TK Grade

[Rob Webster]

Hello,

 

Watching the rushes from a short I recently shot (S16, TK to DigiBeta), I was troubled by just how deeply crushed the shadows were, and how abrupt the transition to black appeared. This caused a coarse, rather unattractive picture and, more significantly, a loss of important detail in the shadows.

Now, I know that the low-key lighting, the high contrast of the stock and the intentional underexposure contributed to the sombre tone of it, but I can't help thinking that a more tactful grade would have saved some vital detail. It seems very unlikely to me that the actor's waist in the attached picture, for example, should be so completely black in open daylight, with the dynamic range of a modern stock.

 

I would value a second opinion on the matter immensely, to assess whether a second transfer could potentially resolve any of these issues? Or am I just fooling myself?

 

Many thanks for your time,

 

Rob

 

 

P.S. I'm having trouble uploading more than one pic, please let me know if it would be useful to upload a couple more in a new post...

[K Borowski]

How much did you underexpose, and why?!?!?

 

 

Sometimes making the image to dark is done in an attempt to minimize excessive grain and flat contrast.

 

 

To be blunt: I'm sorry, you may be at fault here and there isn't anything to save in the shadows without excessive flatness and noise.

[Chris Burke]

Hello,

 

Watching the rushes from a short I recently shot (S16, TK to DigiBeta), I was troubled by just how deeply crushed the shadows were, and how abrupt the transition to black appeared. This caused a coarse, rather unattractive picture and, more significantly, a loss of important detail in the shadows.

Now, I know that the low-key lighting, the high contrast of the stock and the intentional underexposure contributed to the sombre tone of it, but I can't help thinking that a more tactful grade would have saved some vital detail. It seems very unlikely to me that the actor's waist in the attached picture, for example, should be so completely black in open daylight, with the dynamic range of a modern stock.

 

I would value a second opinion on the matter immensely, to assess whether a second transfer could potentially resolve any of these issues? Or am I just fooling myself?

 

Many thanks for your time,

 

Rob

 

 

P.S. I'm having trouble uploading more than one pic, please let me know if it would be useful to upload a couple more in a new post...

 

What film stock did you shoot with and what ƒ stop did you use for the picture displayed. On my monitor, I can just barely see some detail on the actor's stomach. That tells me that there is more to dig out. How far under was the shadow area?

[K Borowski]

Chris, there's more detail there, but I think during TK someone decided to leave it under because it looks like a very thin negative.

 

Methinks things'll get milky if they try to pull out more shadow detail (unless of course they do some curve adjustments and Photoshop trickery).

 

 

Even for S-16 it looks too grainy for properly exposed footage; look at the extreme amount of grain in the clouds even printed densely.

 

 

 

Rob, did you shoot a grey card for them to grade to?

[Rob Webster]

Yes, I shot a greycard, naturally. The shot I posted as an example was exposed for the highlight, i.e. the flat background. The actor's stomach was one and half to two stops under. As I said, yes the negative is thin, but I'd expect detail even at three stops under, wouldn't you? The stock used in this instance was f64D, but even in scenes shot with the lower contrast 250T and exposed more towards a split, the issues are the same. It looks like the grader attempted to apply a distinctive look to the footage, and I wonder whether something like a technical grade would help to reveal the detail that's currently lost.

 

Thanks for your time.

[Adrian Sierkowski]

Have you spoken with the facility yet? If not, give them a ring and ask them whats up. Ask if they can pull any of the before/after stills from their framestore... if that's how they work (my lab stores frames from mostly everything they grade) and speak with them about your dissatisfaction with the results and see if you can get them to allow you to throw the neg right up on the TC and look for yourself how it is w/o a grade.

[K Borowski]

Yes, I shot a greycard, naturally. The shot I posted as an example was exposed for the highlight, i.e. the flat background. The actor's stomach was one and half to two stops under. As I said, yes the negative is thin, but I'd expect detail even at three stops under, wouldn't you? The stock used in this instance was f64D, but even in scenes shot with the lower contrast 250T and exposed more towards a split, the issues are the same. It looks like the grader attempted to apply a distinctive look to the footage, and I wonder whether something like a technical grade would help to reveal the detail that's currently lost.

 

Thanks for your time.

 

 

There'd be noisy, muddy detail. The TK guy is probably trying to hide what he interpreted as a mistake. At 3 stops under, you'd not be able to get a "proper" image at all probably. Anything more than a stop of underexposure cannot be completely corrected back to a normal exposure.

 

Again, I don't understand why you went under so much. If the image gets lightened up (at least traditionally, without any curves or contrast correction) you are going to get a very flat image with no true blacks. There's a difference between latitude and proper exposure. You can go probably 4 or more stops over and still be able to correct back, but there is very little room for underexposure that will still look correct, unless of course you WANTED the image to be grey, flat, and low-contrast.

[Rob Webster]

Perhaps I'm not making myself clear: I did not intend the areas that were 3 stops under to be 'graded back up to proper exposure', or anything of the sort. these areas were simply the shadows of the picture. at roughly three stops under the set stop, my concern was at the fact that there was no detail in these areas. that's all. and my question was whether it was reasonable to believe that the grade may have contributed to that.

 

Rob

[K Borowski]

Rob, I'm not adequately explaining myself either, I don't think.

 

It does look dark, but, no I wouldn't expect any detail at 3 stops under. Detail that is there, you don't WANT to see because it will be milky and muddy, unless that is what you are going for.

 

 

So maybe have them lighten it up a little (1/2 a stop or so), but if you wanted to see detail there, I'd say you'll want to reshoot unless you have someone that is a digital wizard with curve manipulation. Even here, it will never look as good as if you had opened up a stop.

 

 

In a situation like this, I would have probalby just shot normally or under one stop and then just "printed" it darker. If it is a static shot, you can even selectively shade in for more of a silhoette effect.

[Saul Rodgar]

It does look crushed, and the negative looks thin. From personal experience, I know some of the detail exposed 2 1/2 stops under is salvageable, maybe 3, but it will generally always look dark and grainy, or washed out if they just crank the brightness up. There are too many variables here to properly judge without looking at the density of the negative. Was the stock fresh? Are you certain the exposure was correct? Are you certain the filters (if any) were correctly taken into account for exposure, etc? I would try doing a densitometry test on the neg or sending the film to another lab and going form there. Calling whoever was present at the TK bay when the film was transferred is always the first thing to do.

Edited April 21, 2010 by Saul Rodgar

[K Borowski]

but it will generally always look dark and grainy, or washed out if they just crank the brightness up.

 

Exactly. And even if they adjust the analog gain on the scanner (actually adjusting the intensity of the scanner light source during imaging), you're in the heel of the curve on the negative, and the high-speed, grainy layer of the film.

 

So without curves, windows, or some sort of other non-generalized correction, lightening up the image enough to see more into the shadows is going to flatten out the highlights.

 

 

 

Unfortunately, a lot of people seem to fall prey to Kodak marketing. A 500T film, in 16mm, with 14 stops of latitude, is NOT the best choice most of the time on a small budget. No matter how advanced it is, it is going to look grainy, and is still going to perform at its best rated at 250 or 320 EI ISO.

[Gregory Middleton]

Hello Rob,

If your exposure was accurate at 1.5 under key , then i would expect a little more detail. If its 2 under and maybe a little more then that image does not look too much out of line.

The clouds in the distance have all the detail in them, even in the brightest areas. This also tells me a lot about your exposure. If your colorist saw you were exposing for the highlights as you intended then the image seems reasonable.

Getting exactly the shadow detail you want is much more controllable in a video transfer than straight print. Even at that level of exposure you could probably get a little more out of your shadows, but not too much.

Things that could have contributed to you not getting exactly what you want are

1) daylight exposure being off by 1/2 stop. Clouds thicken and sometimes you dont notice the small change

2) stock. 64D is a little more contrasty than some others, so you will notice that in darker areas. I shot almost 1/2 my first feature on that stock.

3) incident vrs reflected exposure. His pants and jacket are dark. 2-3 under incident on a dark or near black subject will be VERY dark. if he was wearing a light grey t-shirt and faded jeans it would appear quite different.

4) Lab and Transfer differences. Processing can be a little off. Transfer can be interpreted a little one way or another. Trying to save highlights in certain areas can lead to shadows being a little more crushed.

 

The best way to learn is to be present when you transfer, if possible. Perhaps they will let you come back in a look at that particular shot again. they you will know what you are working with.

 

best of luck.

greg

 

 

Perhaps I'm not making myself clear: I did not intend the areas that were 3 stops under to be 'graded back up to proper exposure', or anything of the sort. these areas were simply the shadows of the picture. at roughly three stops under the set stop, my concern was at the fact that there was no detail in these areas. that's all. and my question was whether it was reasonable to believe that the grade may have contributed to that.

 

Rob

[Stephen Williams]

What did your lightmeter say the stop should be? I would have expoed that scene about T5.6-8 (T2.8-4 +ND 0.6 in reality) asuming 24 fps 180 shutter.

[K Borowski]

4) Lab and Transfer differences. Processing can be a little off. Transfer can be interpreted a little one way or another. Trying to save highlights in certain areas can lead to shadows being a little more crushed.

 

Greg, agree with all you are saying except here. Assuming it's a reputable lab, processing tolerances with ECN-2 are extremely tight. OTOH, it's +/-1/6 of a stop (.05) before you take corrective action and +/-1/3 (0.10) of a stop before you stop running film through the machine and still keep your Kodak certification.

 

While ECP has run to sh!7 these days, especially with the big labs, ECN-2 processing is incredibly tight.

 

 

Most of the problems are subjective decisions made in telecine, improperly aligned or calibrated scanners/telecines, or miscommunication. With students there is the added factor of TK guys "knowing better" than what the customer tells them. Sometimes they're right. However, often they are wrong, just don't care, or have bigger fish to deal with who are actually paying a decent amount of money.

 

There's a bunch of politics/BS in this sometimes. But as with all human endeavours, you have to rise about it if you want to succeed in this world.

 

 

I always shoot a grey chart, then MEASURE it on the film with a densitometer. Look at the neg. The neg doesn't lie. It tells you both if the processing was bad and if the TK was bad (assuming you check and calibrate your camera before you shoot anything important, of course).

[Gregory Middleton]

Densitometer is the only way to truly tell what's going on, you are right about that Karl.

I hope the lab is reputable, but I have known labs that do vary the speed of the machines depending on work load, but that would be very rare. The image doesn't really suggest anything funny going on processing wise so certainly don't want to offend anyone there. I do think its important to remember all the variables when analyzing something like this though.

Of course filtration can vary the exposure as well and is often not compensated for correctly. Most diffusion is 1/6th of a stop for eg.

 

Greg, agree with all you are saying except here. Assuming it's a reputable lab, processing tolerances with ECN-2 are extremely tight. OTOH, it's +/-1/6 of a stop (.05) before you take corrective action and +/-1/3 (0.10) of a stop before you stop running film through the machine and still keep your Kodak certification.

 

While ECP has run to sh!7 these days, especially with the big labs, ECN-2 processing is incredibly tight.

 

 

Most of the problems are subjective decisions made in telecine, improperly aligned or calibrated scanners/telecines, or miscommunication. With students there is the added factor of TK guys "knowing better" than what the customer tells them. Sometimes they're right. However, often they are wrong, just don't care, or have bigger fish to deal with who are actually paying a decent amount of money.

 

There's a bunch of politics/BS in this sometimes. But as with all human endeavours, you have to rise about it if you want to succeed in this world.

 

 

I always shoot a grey chart, then MEASURE it on the film with a densitometer. Look at the neg. The neg doesn't lie. It tells you both if the processing was bad and if the TK was bad (assuming you check and calibrate your camera before you shoot anything important, of course).