Mablo
Well-known
When I look at Tim's excellent test series my first thought was "this film is EI 200 with some very nice latitude". I don't always understand why so many insist on destroying their images by underexposing AND overdeveloping. It's a double whammy you cannot win. Use the latitude and underexpose one or maybe two stops if you must and open up your fast and expensive lens.
anerjee
Well-known
I am learning -- I use Silverfast to give me a linear 48bit negative scan, and then I invert it in Colorperfect. This always gives me better colors than Silverfast negafix. I import the inverted picture into Lightroom, where I do any other modifications.
However, even slight under-exposure introduces lots of color noise when I try and bring the exposure back in Lightroom. I do use Topaz de-noise sometimes, but I'd like to learn if there are better ways to expose and scan.
I do not want to push process -- it is expensive and the results not good. I use ISO 1600 only for contingencies -- where I'd rather have a noisy shot than no shot.
Any advise?
However, even slight under-exposure introduces lots of color noise when I try and bring the exposure back in Lightroom. I do use Topaz de-noise sometimes, but I'd like to learn if there are better ways to expose and scan.
I do not want to push process -- it is expensive and the results not good. I use ISO 1600 only for contingencies -- where I'd rather have a noisy shot than no shot.
Any advise?
Jamie123
Veteran
I don't know about underexposure but Portra 400 seems to take overexposure really well. A while ago I accidentally shot a frame at f5.6 instead of f45 on my 4x5 because I forgot to stop down the lens and it came out surprisingly well. The image was lacking contrast but that could've easily been taken care of in post.
Athiril
Established
Portra 400 is the best versatile film, it is exceedingly sharp and fine grained, and can go in both directions in exposure.
Portra 800 is faster, it is more so than the simple difference in their film speeds when you look at the curves, so Portra 800 should take pushing better than 400. The curves for pushing to 1600 and 3200 with 800 are published too.
In regards to the OP, I have pushed Portra 400 to 25600 out of curiosity.
The resulting image was exceedingly thin, meaning my calculated developing time of 7 minutes and 30 seconds was actually far undeveloped (+15% per stop falls apart after first 2 it appears), while I don't expect a normal image with any shadow detail at all, there should be better colour and a lack of scanner noise (from flatbed scanning), I'd estimate the time is probably double what I used.
Here is one of the best from that roll.
New Portra 400 @ 25600 by athiril, on Flickr
Pushing to 1600 looks great, I always try to incident meter in the main light of the subject when I can, which I point towards the light, not the camera, as I want a reading of the light, not an amalgamation of average of the light and shade, or another light.
OCAU Melb Photowalk Week 2 #6 by athiril, on Flickr
When pushing I would recommend exposing Portra 1/3rd of a stop slower than what you're pushing to, ie: shoot 1250 instead of 1600, 2500 instead of 3200. While it's only a third of a stop, it makes a lot more difference in pushing and it should help keep the margin for metering error on the positive side rather than possibly on the negative.
I scan in Epson Scan. Sometimes I colour balance in Epson Scan, other times I expand the levels to make sure nothing is clipped so I do not have to scan again later if I change my mind on how it looks. Then I colour balance through levels and then finely through colour balance tool in Photoshop, it takes about 10-15 seconds for an accurate balance.
Portra 800 is faster, it is more so than the simple difference in their film speeds when you look at the curves, so Portra 800 should take pushing better than 400. The curves for pushing to 1600 and 3200 with 800 are published too.
In regards to the OP, I have pushed Portra 400 to 25600 out of curiosity.
The resulting image was exceedingly thin, meaning my calculated developing time of 7 minutes and 30 seconds was actually far undeveloped (+15% per stop falls apart after first 2 it appears), while I don't expect a normal image with any shadow detail at all, there should be better colour and a lack of scanner noise (from flatbed scanning), I'd estimate the time is probably double what I used.
Here is one of the best from that roll.
New Portra 400 @ 25600 by athiril, on Flickr
Pushing to 1600 looks great, I always try to incident meter in the main light of the subject when I can, which I point towards the light, not the camera, as I want a reading of the light, not an amalgamation of average of the light and shade, or another light.
OCAU Melb Photowalk Week 2 #6 by athiril, on Flickr
When pushing I would recommend exposing Portra 1/3rd of a stop slower than what you're pushing to, ie: shoot 1250 instead of 1600, 2500 instead of 3200. While it's only a third of a stop, it makes a lot more difference in pushing and it should help keep the margin for metering error on the positive side rather than possibly on the negative.
I scan in Epson Scan. Sometimes I colour balance in Epson Scan, other times I expand the levels to make sure nothing is clipped so I do not have to scan again later if I change my mind on how it looks. Then I colour balance through levels and then finely through colour balance tool in Photoshop, it takes about 10-15 seconds for an accurate balance.
Last edited:
Athiril
Established
Portra 800 is actually a P1600 and P3200 film. It is the only one with curves published for pushing.
The curve for Portra 800 appears .4 or .5 better before hitting the toe of Portra 400 (this is @ 800 vs 400 too), that's 1 and 1/3rd stops to 1 and 2/3rds stops better underexposure latitude and assumably pushability, on top of that it's also a +1 stop faster.
Juan Valdenebro
Truth is beauty
Close to that, but 3200 is too much for it no matter the development: it requires more light than 3200 incident for great tone and grain, because color film is not as simple as B&W and the real problems caused by underexposure are related to color balance, shadows and grain, not to contrast or loss of shadow detail...
Kodak sold for many years (maybe until 7 years ago) a real P1600 color film: it was slide film. Ektachrome P1600 was a wonderful film with big beautiful grain. At 1600 incident it was beautiful after pushing. Amazing skins for that speed in slide film...
Portra800 is great at 1600 with a Push1 and at 2000 with a Push2, and beyond that it suffers. For low light shooting Portra800 pushed is better than Portra400 twice as pushed... With Portra800 at 2000 you can work beautifully at f/1.4 in real low light, and with natural light indoors and a warming filter, skins are a dream...
Cheers,
Juan
Athiril
Established
I disagree. Your problems stem from other people doing/machines your colour balance. Minilabs while have fantastic CCDs in their scanners.. they are completely unrepresentative in terms of colour to what actually is on the film, not because of the CCD, but because of how the output gets manipulated. There is a thread about 160S and 400H on APUG, saying how great the film looks, the 160S looks more contrasty and saturated than Velvia due to the minilab scan. Don't judge off minilab results.
I do not agree with P1600. To paraphrase someone who created stuff for Kodak, C-41 is made for high speed iodide emulsions and pushing, E-6 is not. P1600 didn't have good grain imho, in fact it was terribly coarse, and had both a lack of highlight detail and a lack of shadow detail when pushed, and a lack of sharpness (as in a lack of resolved detail - resolution, not acutance).
I've heard people say the same thing about Provia 400X, but it is god damn awful pushed to 1600 - the results came from the same people saying how great it is. It exhibits the same awful lack of dynamic range.
I do not agree with P1600. To paraphrase someone who created stuff for Kodak, C-41 is made for high speed iodide emulsions and pushing, E-6 is not. P1600 didn't have good grain imho, in fact it was terribly coarse, and had both a lack of highlight detail and a lack of shadow detail when pushed, and a lack of sharpness (as in a lack of resolved detail - resolution, not acutance).
I've heard people say the same thing about Provia 400X, but it is god damn awful pushed to 1600 - the results came from the same people saying how great it is. It exhibits the same awful lack of dynamic range.
Last edited:
Juan Valdenebro
Truth is beauty
My problems?
I wasn't talking about Portra after minilab scans, but about the beautiful tones it gives when wet printed: that's where you can see a film's tone clearly and how it was exposed... Who looks for small grain on 1600 slide film? Who looks for highlight or shadow detail there after a push on direct positive media? Who cares about sharpness and resolved detail on the biggest grained palette available? You.
Who misses the most beautiful big color grain with amazing skins in positive film? Me.
What a shame you don't have some rolls of P1600 I could buy for peanuts!
Cheers,
Juan
Athiril
Established
Well the problem with wet printing is we no longer have different contrasts of RA-4 paper available iirc, so contrast is an issue there, and with the increased contrast, depending on your scene, you may have to clip some of the range, your only option there would be to get creative with potassium ferricyanide.
You still have colour balance controls on an enlarger.
The vast majority of people getting wet prints will be getting wet prints from minilab scans, as that is how wet printing is done, and has been done for a long time now.
IE; Everybody else's wet prints come from a scanner, as RA-4 printing now is exposed with lasers which is read from a digital file, the curve a minilab may or may not slam onto the film would still apply.
So it doesn't really apply to anyone else. I have my Eurogon Dichroic, but I do not have any paper for it yet.
You can see a film's tone clearly and how it was exposed with a scan too. Backlit media looks different to reflective media, just like backlit ilfochromes/cibachromes to reflective prints.
Highlight and shadow detail is important in portraiture. Giving both up is a big compromise for many needs for high speed.
P1600 would be no good by now to use even at 400 I think. Pushing and pulling slides colour casts the image. If you want to correct it and use it as a final product you need to stick it in a projector or lightbox that has colour balance dials on it, or at least stick filters in, eg; I've used my Eurogon as a colour correcting projector for slides since it can project in any direction. Otherwise you have to correct a scan or enjoy the colour cast.
Even pushing 1 stop does. I recently pushed some EBX100 by 1 stop for someone who sent me the film, there is a magenta cast there. It corrects fine if the person wishes to correct their scans of it.
C-41 has colour balancing in all of it's workflows though, during printing or scanning. There isn't an issue as it's supposed to be colour balanced.
You still have colour balance controls on an enlarger.
The vast majority of people getting wet prints will be getting wet prints from minilab scans, as that is how wet printing is done, and has been done for a long time now.
IE; Everybody else's wet prints come from a scanner, as RA-4 printing now is exposed with lasers which is read from a digital file, the curve a minilab may or may not slam onto the film would still apply.
So it doesn't really apply to anyone else. I have my Eurogon Dichroic, but I do not have any paper for it yet.
You can see a film's tone clearly and how it was exposed with a scan too. Backlit media looks different to reflective media, just like backlit ilfochromes/cibachromes to reflective prints.
Highlight and shadow detail is important in portraiture. Giving both up is a big compromise for many needs for high speed.
P1600 would be no good by now to use even at 400 I think. Pushing and pulling slides colour casts the image. If you want to correct it and use it as a final product you need to stick it in a projector or lightbox that has colour balance dials on it, or at least stick filters in, eg; I've used my Eurogon as a colour correcting projector for slides since it can project in any direction. Otherwise you have to correct a scan or enjoy the colour cast.
Even pushing 1 stop does. I recently pushed some EBX100 by 1 stop for someone who sent me the film, there is a magenta cast there. It corrects fine if the person wishes to correct their scans of it.
C-41 has colour balancing in all of it's workflows though, during printing or scanning. There isn't an issue as it's supposed to be colour balanced.
Juan Valdenebro
Truth is beauty
"You can see a film's tone clearly and how it was exposed with a scan too."
What you see is a digital autoexposure photograph (scan) of a color negative... A moment ago, on your previous post you just said the opposite: "Minilabs while have fantastic CCDs in their scanners.. they are completely unrepresentative in terms of colour to what actually is on the film..."
"I recently pushed some EBX100 by 1 stop for someone who sent me the film, there is a magenta cast there."
I've pushed slide film many many times: there's no magenta cast in my rolls, but I believe you got it on yours... I just don't know why... Good night!
Cheers,
Juan
What you see is a digital autoexposure photograph (scan) of a color negative... A moment ago, on your previous post you just said the opposite: "Minilabs while have fantastic CCDs in their scanners.. they are completely unrepresentative in terms of colour to what actually is on the film..."
"I recently pushed some EBX100 by 1 stop for someone who sent me the film, there is a magenta cast there."
I've pushed slide film many many times: there's no magenta cast in my rolls, but I believe you got it on yours... I just don't know why... Good night!
Cheers,
Juan
Athiril
Established
You do not have control of the "exposure" from a minilab, you do if you scan or get a flat scan that encompasses the entire density range of the neg, then you can "print" it as you see fit.
Pushing and pulling causes colour shifts on E-6. It is normal to experience it.
"Compensating for underexposure will result in higher film
contrast and lower D-max, and a color-balance shift is
possible. Time and temperature adjustments give
comparable results.
Compensating for overexposure will result in lower
contrast and a possible shift in color balance, especially in
highlights. Time adjustments tend to produce less
degradation of highlights than temperature adjustments at all
levels of overexposure"
Direct quote from Kodak.
Many old pictures from several years back have simply vanished due to where they were hosted, I will shoot a roll of Portra 800 at 3200, push it, and post the images in this thread in a few weeks time.
Pushing and pulling causes colour shifts on E-6. It is normal to experience it.
"Compensating for underexposure will result in higher film
contrast and lower D-max, and a color-balance shift is
possible. Time and temperature adjustments give
comparable results.
Compensating for overexposure will result in lower
contrast and a possible shift in color balance, especially in
highlights. Time adjustments tend to produce less
degradation of highlights than temperature adjustments at all
levels of overexposure"
Direct quote from Kodak.
Many old pictures from several years back have simply vanished due to where they were hosted, I will shoot a roll of Portra 800 at 3200, push it, and post the images in this thread in a few weeks time.
Pickett Wilson
Veteran
"I will shoot a roll of Portra 800 at 3200, push it, and post the images in this thread in a few weeks time."
Take it out to a high school football field at night and shoot the game with available light, film pushed to 3200. Then post the images. You will not like the results. I'm always suspicious of these push tests. They often look like the person doing the metering was misinterpreting their meter readings.
Take it out to a high school football field at night and shoot the game with available light, film pushed to 3200. Then post the images. You will not like the results. I'm always suspicious of these push tests. They often look like the person doing the metering was misinterpreting their meter readings.
Roger Hicks
Veteran
Though in all fairness, it's a legitimate question what 'meter reading' means in this context. Highlight reading, i.e. incident? At which point, any concept of ISO negative speed goes out of the window.
To return to the OP´s question, you can certainly underexpose Portra 400 more than any other colour neg film I've ever tried, and still get usable pics IF you do your own scanning and post-production, but the relationship of that to speed and metering must always remain disputable.
Cheers,
R.
Last edited:
noeyedear
Established
Wow that will confuse a newcomer. You implication is that a different metering method allows for push processing. A correct exposure however it is achieved will not alter what can be achieved with pushing or pulling in processing. An incident reading is done to give the correct exposure, just the same as a reflective reading, multi spot or the Sunny 16 rule they are all capable of giving the correct exposure and have no bearing on the amount a film can be pushed or pulled.
Roger Hicks
Veteran
A correct exposure however it is achieved will not alter what can be achieved with pushing or pulling in processing. An incident reading is done to give the correct exposure, just the same as a reflective reading, multi spot or the Sunny 16 rule they are all capable of giving the correct exposure and have no bearing on the amount a film can be pushed or pulled.
Dear Kevin,
Not exactly. Or, alternatively, exactly, but it depends on how you interpret your readings, and upon what you mean by a 'correct' exposure. For the latter, I'd argue that there is no such thing -- there is only 'a pleasing exposure', or 'the exposure that you want' -- and the former is most easily explored through extremes.
A broad-area reflected-light or multi-spot reading of a snowy landscape will almost invariably recommend under-exposure, so you need to give a stop or even two stops more than the meter recommends. With an incident light reading of the same subject you may well choose to give half a stop or even a stop less exposure than the meter indicates, in order to get some detail and texture in the snow. With a back cat in a coal cellar the corrections are in the opposite direction.
Now, a floodlit football game consists (or can consist) of black sky and more or less brightly lit players, so what do you meter? And even if you use an incident meter, how worried are you about shadow detail? That's what I meant when I queried 'misinterpreting the meter reading' in an earlier post.
Cheers,
R.
noeyedear
Established
Dear Roger,
I know you to be a man of Knowledge in matters photographic, so I find it a strange place to be disagreeing so much with the message this is sending out.
Although I can see what is being alluded to here, I have to disagree with what is being implied.
Which ever method you use to get the exposure you want, it will not alter the amount of pushing or pulling a film can take. If you use a method of metering that consistently over exposes it does not mean that a film can be pushed more, it means you are not reading the scene correctly. Getting the exposure right is the name of the game, deciding which end to clip the detail is one of the factors we have to take into account, sometimes we get it badly wrong. Using an incident or any other method has zero effect on the films reaction to pushing or pulling.
I have just done a test against Portra and my 1DsmkIII outside. I had grey cards in the scene, the reading from a spotmeter of the grey card, the gossen with an invercone and the Canon with multiaverage all gave readings too close to call a difference, even the meter on the Rolleiflex agreed, I doubt there was half a stop between the lot. So how would an invercone reading give me any advantage in under exposing and pushing the development over a reflective reading in this scene?
Relating metering method to the amount a film can be pushed is very misleading and will confuse anyone new to photography.
Cheers,
Athiril
Established
I'm not a soccer mum so I won't be going to any high school games, I don't think we even that many here, just external u-18 leagues not associated with schools.
In that case, Portra 800 takes it better, it has a longer range before the toe hits, and a better toe.
If you are shooting box speed, and spot meter your darkest reflective detail you want to record and expose at that reading minus x stops at place it at the bottom of the curve, depending on other factors that may well be underexposure at box speed and require a push to achieve normal density for mid tones.
"Now, a floodlit football game consists (or can consist) of black sky and more or less brightly lit players, so what do you meter? And even if you use an incident meter, how worried are you about shadow detail? That's what I meant when I queried 'misinterpreting the meter reading' in an earlier post."
Incident metering the light your subject in will give correct density for mid tones. IE; mid tones will be placed at mid tones.
When you scene is composed mostly of highlights or mostly of shadows, reflective metering will fail you, unless you make a guestimate compensation.
Incident metering gives the same reading no matter the subject if the light source is the same. I would stick to incident meter main subject light instead of metering reflectivity most of the time.
And I completely disagree about the nature of a 'correct' exposure with Roger, there is such a thing as a correct exposure. Correct isn't what gives you the most pleasing image, deviation to give you the most pleasing image is creative choice. Portra 400 is the same sensitivity for everyone, just like 5DII and x ISO is the same sensitivity. Deviations are just personal preference.
My pushing (my set of 1600) wasn't a test, it was a need for more speed, as I do not like simply underexposing (Even Portra 400) and resulting with thinner than normal negs I was hand held shooting with an RB67 in the city. The 25600 is metered for incident, so it is metering the main light of the scene and placing midtones in that light @ 25600, as I wanted to explore the limits, the 7m 30s developing time (38c) is very heavily underdeveloped, at the time I was only doing +15% consecutively for each stop, and that falls apart after first 2 on C-41 which I did not consider. I haven't re-visited it, as it's not worth revisiting. But given it still records midtones (and a bit under) at 6 stops under, you can see it has a long scale to record under mid tones, given how thin it was at the push time though, simple underexposure would have recorded less, so I think if you have :
1/x shutter, f/y aperture ratio, and 400 ISO as your chosen exposure, we'll call it EV Z, and you spot meter the darkest detail and it's EV Z-6, that area would achieve similar density at the 7m 30s push time as that shot which was very thin, is recordable detail, you'd think that with less than half the amount of processing time it'd not show up since the same exposure level as EV Z-6 at 1/x shutter and f/y aperture came out very thin, but the rate at which density builds drops rapidly.
In any case, the Portra 400 v 500T test on twin lens life, with no pushing, while who knows what the metering was, each comparison shot received the same exposure and Portra 400 breaks down earlier than 500T.
Now what I would like to add to this is that scene contrast does not change your film speed. If you have a lit room with a very short range, placing midtones at midtones density is correct (though you may have a different idea of what looks best), and this is what manufacturer sheets would also agree with at box speed for the speed of their film, if you suddenly put up a cutter/flag on the light and shaded half the room from light, or half the subject/etc and that half now drops below what will be recorded on the film. The sensitivity (speed) of your film has not changed, it is not any different to what it was before. Increasing exposure 2 stops may record that shadow detail, you are then placing midtones in the highlight density region, now that is overexposure, even though you have a massive amount of room on C-41 for highlights to extend.
As contrast contraction and expansion isn't a part of the C-41 workflow, nor generally do you need to contract from overexposure anyway with C-41. You can do it if your lab agrees, or you process yourself. But that's a different story.
In that case, Portra 800 takes it better, it has a longer range before the toe hits, and a better toe.
If you are shooting box speed, and spot meter your darkest reflective detail you want to record and expose at that reading minus x stops at place it at the bottom of the curve, depending on other factors that may well be underexposure at box speed and require a push to achieve normal density for mid tones.
"Now, a floodlit football game consists (or can consist) of black sky and more or less brightly lit players, so what do you meter? And even if you use an incident meter, how worried are you about shadow detail? That's what I meant when I queried 'misinterpreting the meter reading' in an earlier post."
Incident metering the light your subject in will give correct density for mid tones. IE; mid tones will be placed at mid tones.
When you scene is composed mostly of highlights or mostly of shadows, reflective metering will fail you, unless you make a guestimate compensation.
Incident metering gives the same reading no matter the subject if the light source is the same. I would stick to incident meter main subject light instead of metering reflectivity most of the time.
And I completely disagree about the nature of a 'correct' exposure with Roger, there is such a thing as a correct exposure. Correct isn't what gives you the most pleasing image, deviation to give you the most pleasing image is creative choice. Portra 400 is the same sensitivity for everyone, just like 5DII and x ISO is the same sensitivity. Deviations are just personal preference.
My pushing (my set of 1600) wasn't a test, it was a need for more speed, as I do not like simply underexposing (Even Portra 400) and resulting with thinner than normal negs I was hand held shooting with an RB67 in the city. The 25600 is metered for incident, so it is metering the main light of the scene and placing midtones in that light @ 25600, as I wanted to explore the limits, the 7m 30s developing time (38c) is very heavily underdeveloped, at the time I was only doing +15% consecutively for each stop, and that falls apart after first 2 on C-41 which I did not consider. I haven't re-visited it, as it's not worth revisiting. But given it still records midtones (and a bit under) at 6 stops under, you can see it has a long scale to record under mid tones, given how thin it was at the push time though, simple underexposure would have recorded less, so I think if you have :
1/x shutter, f/y aperture ratio, and 400 ISO as your chosen exposure, we'll call it EV Z, and you spot meter the darkest detail and it's EV Z-6, that area would achieve similar density at the 7m 30s push time as that shot which was very thin, is recordable detail, you'd think that with less than half the amount of processing time it'd not show up since the same exposure level as EV Z-6 at 1/x shutter and f/y aperture came out very thin, but the rate at which density builds drops rapidly.
In any case, the Portra 400 v 500T test on twin lens life, with no pushing, while who knows what the metering was, each comparison shot received the same exposure and Portra 400 breaks down earlier than 500T.
Now what I would like to add to this is that scene contrast does not change your film speed. If you have a lit room with a very short range, placing midtones at midtones density is correct (though you may have a different idea of what looks best), and this is what manufacturer sheets would also agree with at box speed for the speed of their film, if you suddenly put up a cutter/flag on the light and shaded half the room from light, or half the subject/etc and that half now drops below what will be recorded on the film. The sensitivity (speed) of your film has not changed, it is not any different to what it was before. Increasing exposure 2 stops may record that shadow detail, you are then placing midtones in the highlight density region, now that is overexposure, even though you have a massive amount of room on C-41 for highlights to extend.
As contrast contraction and expansion isn't a part of the C-41 workflow, nor generally do you need to contract from overexposure anyway with C-41. You can do it if your lab agrees, or you process yourself. But that's a different story.
Last edited: