To push one stop or fix in DI

[Ray Lavers]

So I was shooting all my none sync stuff on a Motorize Bolex with a lens adapter for Nikon Lenses I had forgotten about the the compensation needed for the viewfinder 1/3 of a stop. The next Day my AC pulls out an 85 filter in the filter holder. Which is 2/3's of a stop.

 

Now I'm underexposed for all my Bolex stuff for 1 stop

My stock is 500t Vision 3 7219

 

I have some free time with a colorist from Technicolor, I was wondering should I get them to fix it or push it all one stop in development.

 

 

Thanks,

Ray

[David Mullen ASC]

I would just process it normally and correct in the D.I. Pushing doesn't really "save" underexposed shots anyway, it just adds density to the areas that got exposed already, increasing overall contrast & grain as a result.

 

And a 2/3-stop underexposure isn't that bad. The fact that the image is more orange now might be a bigger issue because your blues are underexposed more than the reds.

[Ray Lavers]

I would just process it normally and correct in the D.I. Pushing doesn't really "save" underexposed shots anyway, it just adds density to the areas that got exposed already, increasing overall contrast & grain as a result.

 

And a 2/3-stop underexposure isn't that bad. The fact that the image is more orange now might be a bigger issue because your blues are underexposed more than the reds.

 

 

yeah your right.. It's almost like I'll have to add color to my shadows. or change the look completely. It's a Low key boxing scene. All backlit and top lit with many Diva 400's hanging from a grid... I guess we will see what happens.

 

Thanks David for the heads up.

 

PS are you looking for an Intern?

[Richard Tuohy]

So I was shooting all my none sync stuff on a Motorize Bolex with a lens adapter for Nikon Lenses I had forgotten about the the compensation needed for the viewfinder 1/3 of a stop. The next Day my AC pulls out an 85 filter in the filter holder. Which is 2/3's of a stop.

 

If its a Rex or SB(M) series bolex (or older, but you said it had a viewfinder), you also have to factor in 1/3rd of a stop for the 150 degree shutter opening (unless you set your meter to that in some way). If the meter is assuming a 180 degree shutter (such that 24 fps = 1/48th of a second) then the total compensation required for shutter opening and viewfinder prism is 2/3rds stops. With 85 filter you thus would have 1 and 1/3rd stops under exposure.

Just pointing that out for clarity to others who might read this post in future.

[Brian Pritchard]

I believe that it is worth pushing up to 1 stop. You do get an actual increase in speed (probably not a complete stop). If you process normally so that the film is effectively underexposed no amount of DI work will retrieve information that is not there in the film. You will be able to increase the contrast and density with DI but that bit of extra processing might just retrieve some vital information in the shadows. The extra processing will increase the contrast and the grain but that might be acceptable for the extra information you obtain.

Brian

[K Borowski]

Don't have time to go into it in detail right now, but I disagree with David.

 

Pushing a stop may be worth it, esp. considering it is a low-ley scene.

 

Shadow detail can't be gotten back from a push, though, just highlight and contrast can be pulled up.

 

 

With or without a push, an improperly exposed (especially under) shot will be compromised in the end-product and will only match with properly-exposed material to a certain point.

 

Will post a more thorough explanation when I get back from work today.

[Brian Pritchard]

Shadow detail can't be gotten back from a push, though, just highlight and contrast can be pulled up.

 

Not correct I am afraid, the whole point of pushing is to increase the shadow detail. In fact it is possible to lose detail in the highlights if you push process but normally if you are pushing an under-exposed scene the highlights will just increase in density.

Brian

[Ray Lavers]

Not correct I am afraid, the whole point of pushing is to increase the shadow detail. In fact it is possible to lose detail in the highlights if you push process but normally if you are pushing an under-exposed scene the highlights will just increase in density.

Brian

 

Hey guys I appreciate all the answers, thankfully I was told that in the bolex I was using, it has 2 setting for the filter. I assumed that the 85 was in but actually it wasn't, It was in the standby position . So really I just screwed up on the bolex viewfinder compensation . Nothing a little DI won't fix.

[David Rakoczy]

Just to continue this discussion, I would Push as it is easier to bring things down than up from an abyss. The Tests I shot for a Fotokem DP dinner revealed that a one stop push did no perceptible deterioration of the image with respect to Grain and Contrast. 1 1/2 and more... yes. We shot both 35mm and S16mm. It was at that point I abandoned anything faster than 200t when shooting S16 unless we really really want Grain. For Night Ext I prefer pushing 200t than shooting 500t for example. of course, there are times when you absolutely (have) to use (and push) 500 if need be......

 

It is true, on the other hand, it can be corrected (to a degree and differently) in DI.

[David Mullen ASC]

Pushing can't add information that wasn't originally captured, it just increases density on the negative to what did get captured, while increasing contrast, the base fog level, grain, and to some degree, there is a bit of a color shift to the blacks.

 

With only a 2/3-stop underexposure, it won't hurt to push it one stop for a D.I. but it won't help much either -- either way, you have the same picture information to scan off of the negative. So it's more a difference of look, whether a scan of a thinner negative that is brightened electronically looks better than a scan of a denser negative that doesn't need electronic brightening. Black levels in either case can be manipulated in the D.I., but the flatter look of the unpushed negative may be easier to correct. I think the main difference you are going to see, besides the increase in contrast from pushing, is noise from brightening a thin negative versus grain from transferring a pushed negative.

 

So it's half-dozen one way or the other.

 

Now for printing, there are advantages to a pushed negative since it will print at higher printer light numbers than a thin negative and thus the blacks will be stronger.

[David Rakoczy]

Now for printing, there are advantages to a pushed negative since it will print at higher printer light numbers than a thin negative and thus the blacks will be stronger.

 

Absolutely, and the test I shot was for Print, and from that, you could discern what you'd have at the DI phase... so I have no problem pushing one stop.

[Stuart Brereton]

Nothing a little DI won't fix.

Are you actually going through a digital intermediate stage, or are you referring to color correction in telecine? They are different things. It might be nit-picking, but using incorrect terminology leads to confusion.

[David Mullen ASC]

Absolutely, and the test I shot was for Print, and from that, you could discern what you'd have at the DI phase... so I have no problem pushing one stop.

 

You just have to be aware that some telecines and scanners prefer negatives that are not too dense so just because something is optimal for printing doesn't necessarily mean it is optimal for scanning.

 

This is where comparative testing comes in handy.

[Ray Lavers]

Are you actually going through a digital intermediate stage, or are you referring to color correction in telecine? They are different things. It might be nit-picking, but using incorrect terminology leads to confusion.

 

 

Yep I'm going through a DI , However I'm referring to the end of the process when I conform my DPX and do my final grade with a colorist at Technicolor.

[Stuart Brereton]

Yep I'm going through a DI

 

I wasn't picking on you in particular. Unfortunately, there are many people who think that the terms DI and telecine/color correction are interchangeable.

[Brian Pritchard]

Pushing can't add information that wasn't originally captured, it just increases density on the negative to what did get captured, while increasing contrast, the base fog level, grain, and to some degree, there is a bit of a color shift to the blacks.

 

This is just not true. it would only be true if you did not get an actual increase in speed with push processing. I measure sensitometric strips on a regular basis and you can see the speed increase when you push process.

 

If you push process a 50 ASA film 1 stop, then it is, as far as exposure concerned, equivalent to using a 100 ASA film, so you will have more information than if you used the film at 50 ASA. If the push processing is giving an actual increase in speed then you will have more information than if you did not push. If you under-expose by one stop then iDI cannot bring up information that is not there.

 

Push processing will increase contrast, base fog level and grain but if you desperately need detail in the shadows you need to push process if the film is under exposed.

 

Brian

[K Borowski]

Pushing can't add information that wasn't originally captured, it just increases density on the negative to what did get captured, while increasing contrast, the base fog level, grain, and to some degree, there is a bit of a color shift to the blacks.

 

With only a 2/3-stop underexposure, it won't hurt to push it one stop for a D.I. but it won't help much either -- either way, you have the same picture information to scan off of the negative.

 

Right, pushing nowadays almost always is done in relation to what that will do to the scanned image, not the printed image.

 

 

While increased contrast is something that doesn't help a scanner so much, increasing highlight density on the negative can help, depending on the shooting situation.

 

 

So, it'd really depend on the contrast in the original scene, degree of underexposure, and the final viewing format what would produce optimal results.

 

 

 

It's amazing how many people on here (not calling you out Dave M., but you Brian) don't understand pushing, when I remember reading and comprehending it as a sophomore in H.S. The act of pushing increases highlight development, and therefore contrast, but doensn't generally result in any true speed gain, at least with standard development techniques. It increases grain.

 

While I was reading about Double-X back then, the same principles apply as long as we are using silver-halide based films.

 

 

 

The advice in the late '70s, when neg. stock was, what, 100T was to underexpose two stops. Push one stop, and make up the other stop in printing.

 

While things have changed a great deal since then, I think the notion to make up half in processing and half in printing (the DI suite) is still the best way to go about it.

 

 

 

Either types of noise (grain or pixelation) become unacceptable in non-abstract situations. So it is best to observe how they interact with one another and to optimize the processing to render the best possible negative for the scanner.

 

My own personal experience has been that scanners tend to deal better with pulling image information out of dense areas than with thin areas. So, as far as highlight detail, which is the only area that can be recovered from an underexposed shot without loss of actual image over noise, I think scanners would generally produce better, less noisy images with more density than less.

 

 

But that is a good point as well, that it depends on the type of scanner. It also depends on whether the lab is willing to adjust the analog gain on the scanner with your problem roll. A lot of labs can, but don't want to do this; it takes some convincing.

 

 

Ray: Is there any possibility that you can do a clip test or shoot a separate test so that you can evaluate the process throughout your entire workflow?

 

 

Otherwise, I'd need to know your contrast ratio, printing chain on both the normally and underexposed footage to give you definitive advice. You probably have gotten more information here than you WANT to.

 

 

Just keep this in mind:

 

PUSHING BRINGS BACK HIGHLIGHT DETAIL AND *PERCEIVED*

IMAGE SPEED BY INCREASING THE CONTRAST, BUT NOT SHADOW SPEED (SHADOWS FALL TO BLACK).

 

SCANNING UNDEREXPOSED (THIN) AREAS ADDS NOISE.

 

PUSHING ADDS GRAIN (NOISE).

 

SCANNERS ALSO HAVE A HARD TIME WITH DENSE AREAS (OVEREXPOSED/PUSHED HIGHLIGHTS).

 

SCANNERS CAN ADD PIXELATION WITH EXCESSIVE GRAIN, ESPECIALLY WITH 16MM.

 

DIGITAL IMAGING SOFTWARE TENDS TO RENDER A LESS NOISY JOB DECREASING CONTRAST THAN INCREASING IT.

 

 

So you have to weigh each of these factors before you make a final decision what to do. That last point hasn't been brought up in this thread that much, but is especially important.

[Brian Pritchard]

It's amazing how many people on here (not calling you out Dave M., but you Brian) don't understand pushing.

 

Wow, I have worked For Kodak Research, been Technical Director of three labs including one of the biggest in London, I have been in the industry for 49 years and I still don't understand pushing. Time I gave up.

 

Brian

[David Mullen ASC]

Wow, I have worked For Kodak Research, been Technical Director of three labs including one of the biggest in London, I have been in the industry for 49 years and I still don't understand pushing. Time I gave up.

 

Brian

 

Brian, I don't understand how you can believe that forced processing will increase the sensitivity rather than just increase the density of the negative. The sensitivity of the stock is based on the size of the grains and their efficiency in capturing light. Forced processing comes after exposure so it can't go back in time and make those grains better at capturing photons. It can only take what was captured and increase the percentage of silver halide grains that get converted to silver. That's not an increase in sensitivity but it does compensate for underexposure by taking what did get captured and increasing its density back to normal levels. So it can allow one to rate a stock faster but it is not a true increase in the sensitivity of the stock.

[K Borowski]

Brian, I am sorry if I came off as personally attacking you.

 

I tried hard not to do so, but it appears as if you still took it as such.

 

 

You agree that there is a difference between real and perceived speed gains, right? I'd say that, to a point (less than a whole F/stop's speed) you can achieve some small speed increase, although film stocks still tended to be rated slower than they actually are if you read them under a densitometer.

 

 

The only system that was designed for true speed increase (at the cost of atrocious grain buildup) was B&W negative. Color negative, followed by B&W reversal, and color reversal, were a dismal 2nd 3rd, and 4th in comparison, respectively.

 

 

 

 

If you've been at this as long as you say, you'll remember the time, then, when Kodak officially said that they didn't sanction push-processing at all. I *think* they've softened their stance since. . .

 

 

Anyway, I am on your side here. I think there are some practical, not perceived benefits, to push processing in the era of film scanners and DIs, just a benefit that isn't as great as it was in the days of film finishes.

[Brian Pritchard]

Brian, I don't understand how you can believe that forced processing will increase the sensitivity rather than just increase the density of the negative. The sensitivity of the stock is based on the size of the grains and their efficiency in capturing light. Forced processing comes after exposure so it can't go back in time and make those grains better at capturing photons. It can only take what was captured and increase the percentage of silver halide grains that get converted to silver. That's not an increase in sensitivity but it does compensate for underexposure by taking what did get captured and increasing its density back to normal levels. So it can allow one to rate a stock faster but it is not a true increase in the sensitivity of the stock.

Push processing does give a real increase in speed. The information is captured by the emulsion, the amount of information that appears on the film will depend on the amount of development. If you under-develop (pull process) you will lose some of the recorded information and likewise push processing will bring up information that you would not get if you didn't push. It is not a question of making grains better at capturing photons they have already been captured and require the extra development for them to be revealed.

Brian

[Brian Pritchard]

I also meant to say you only have to compare a Kodak control strip normally processed with one that has been push processed to see that extra steps are revealed indicating that there has been an increase in speed.

Brian

[K Borowski]

Right, but isn't some of this due to an increase in base fog, and the limitations of the human eye?

 

 

Even at a normal processing time, the D-min+ areas of a control strip (that are almost or completely indiscernible from one another visually) will read differently. They're hard to find, by eye or by automatic densitometer, but they are still reading three to five points differently from one another.

 

 

With a push process, these same squares on an ECN strip still read differently but their differences are DECREASED, due to underexposure. It isn't as bad as with normal processing on underexposed film, but you still lose contrast between individual steps.

 

Where the increase in contrast comes is between the shadows and the highlights. The areas that have received adequate exposures gain density. The shadow areas don't gain nearly as much, and, as David M. said, some if not all of that is due to fog.

 

 

Ever shoot a basketball game on B&W neg, pushed a couple of stops, where the players are "hidden" when they enter the shadow areas? Unfortunately, I have.

 

D. Mullen has spoken about this numerous times on here, so I don't want to take this as my own advice, but he recounts how Kubrick would give a whole stop's pushing for every 2/3 stop of increase in exposure (maybe starting at 250 EI on a 500 stock too, I forget).

 

 

So even if there is a slight speed increase, which I am willing to concede, you don't get an extra stop's speed with a one stop push, even if you start out at an EI of 250 with a 500T stock.

[Brian Pritchard]

So even if there is a slight speed increase, which I am willing to concede, you don't get an extra stop's speed with a one stop push, even if you start out at an EI of 250 with a 500T stock.

 

Karl, if you read my posts carefully you will see that I did not state that when you push 1 stop you get 1 stop increase in speed.

 

If you want to see what Kodak have to say about push processing then this is the link.

http://motion.kodak.com/US/en/motion/Suppo...mation/push.htm

 

This is what they say about speed increase:

Speed:

ER speed increases with Push processing may be significantly less than needed to offset speed loss from under exposure. With the given conditions above, Push 1 processing produces ER speed increases of less than a 1/3 of a stop and Push 2 produces ER speed increases of around 1/2 of a stop. Push 2 taken in combination with two stops of under exposure represents a stop and a half loss in real speed. The end result is smoky shadows in prints.

 

Brian

[David Rakoczy]

Besides the decrease in grain perception (esp. in S16), this is another good reason to rate stocks 2/3s slower... Most underexposures are between 2/3 and 1 stop under so, if you forget to compensate for an 85 for example, you are still in 'normal' range.

 

I have to agree with Mr. Pritchard and anyone who has ever pushed has heard the lab express to them that they will not yield a full 1 stop on a 1 stop push. I also think there is a bit of semantics going on in that pushing does not increase speed (on the low end) because as David M. stated.. it occurs after the fact and we can not bring to life something that is not there in the first place. At the same time, pushing can bring to life things that are under between the mid and lows.

 

On the Fotokem Test I saw no increase in Contrast or Grain in a 1 stop push (which is really 2/3s) in either 35mm or S16mm.