Kodak EXR vs Vision Sharpness

[Crysist]

I had seen David Mullen cited as saying "...the contrast has gone down with each generation (EXR, Vision, Vision-2)" in a video essay, so I went to check these stocks out. I had heard elsewhere that they had a very coveted look due to their rendition in various ways, as well as having super fine grain. But looking at the EXR technical data, I found something I didn't expect. The MTF values for EXR are REALLY high compared to later stocks. Why is that? They seem to be nearly twice as good, they're at least touching 200 cycles/mm when they stopped plotting the charts.

Does this relate to having better latitude, did the Vision line trade sharpness for latitude? Is it because of the reduced contrast? Are the charts even right? I've been confused by this for 2 weeks.

 

[Lachlan Young]

I had seen David Mullen cited as saying "...the contrast has gone down with each generation (EXR, Vision, Vision-2)" in a video essay, so I went to check these stocks out. I had heard elsewhere that they had a very coveted look due to their rendition in various ways, as well as having super fine grain. But looking at the EXR technical data, I found something I didn't expect. The MTF values for EXR are REALLY high compared to later stocks. Why is that? They seem to be nearly twice as good, they're at least touching 200 cycles/mm when they stopped plotting the charts.

Does this relate to having better latitude, did the Vision line trade sharpness for latitude? Is it because of the reduced contrast? Are the charts even right? I've been confused by this for 2 weeks.

Think about how too much large-object edge sharpness will affect visual cadence in motion. Stills and MP don't necessarily directly translate.

 

[dokko]

I shot some short films on EXR stocks before the Vision stocks became available. In a lot of ways I preferred the EXR, specially loved the 5248 (100T) and 5298 (500T).

The EXR were grittier, but I liked the grain texture and colors a lot.
The Vision stocks focused on smaller grain and more latitude, which as a trade-off loses some resolution. With the Vision2 series Kodak started to optimize for scanning as well, which meant smoother grain structure but again the trade-off is the loss of some sharpness/resolution.

In a similar development, this is the reason the latest generation of Portra is a bit lower in resolution than the older emulsions were.

 

[Crysist]

Think about how too much large-object edge sharpness will affect visual cadence in motion. Stills and MP don't necessarily directly translate.

I agree that in motion a difference at those levels won't really matter, I'm just wondering about why the difference is this drastic at all. And "backwards" on the metric when it'd worked before. I don't mean to only look at it from a perspective of shooting stills. But, motion picture films look like they're fun to shoot in still photography!

I shot some short films on EXR stocks before the Vision stocks became available. In a lot of ways I preferred the EXR, specially loved the 5248 (100T) and 5298 (500T).

The EXR were grittier, but I liked the grain texture and colors a lot.
The Vision stocks focused on smaller grain and more latitude, which as a trade-off loses some resolution. With the Vision2 series Kodak started to optimize for scanning as well, which meant smoother grain structure but again the trade-off is the loss of some sharpness/resolution.

In a similar development, this is the reason the latest generation of Portra is a bit lower in resolution than the older emulsions were.

Ah, interesting. So some of those aspects do trade resolution. And there were multiple things they were optimizing. I did see that about Portra with the prior NC/VC vs the current lineup.

What specifically about the color on EXR did you like?

 

[dokko]

What specifically about the color on EXR did you like?

hmm, difficult to describe colors, maybe the best way to describe it that they felt a bit more like Provia rather than Portra (only grittier as mentioned).

They also responded quite strongly to uneven spectrum light sources like fluorescent, which was sometimes tricky but gave the scenes a lot of atmosphere.

PS: this was before high quality scanning and digital projection, so my impression are mainly based from what I remember from analog optical prints.

 

[koraks]

The MTF values for EXR are REALLY high compared to later stocks.

I never looked into this, but can you point me to where the difference is? I've done a quick & dirty overlay of the Vision3 250D and the EXR 200T MTF charts and they don't look all that different:

The MTF for the blue channel on Vision3 is a little higher; that's pretty much all. And yes, they plot more data for Vision3, but this might just as well be due to a change in measurement routines where they eek out a little more data than they did before. A lack of data doesn't mean that the performance at that point doesn't exist, of course. It's debatable whether there's any practical utility to plotting MTF data beyond 100lpm or so.

Is it possible you were mistaken due to changes in how Kodak set the vertical scale of the plots?

 

[dokko]

I never looked into this, but can you point me to where the difference is? I've done a quick & dirty overlay of the Vision3 250D and the EXR 200T MTF charts and they don't look all that different:

I think the OP was referring to the 50D, for which the MTF of the EXR seems to be quite a bit higher.

PS: I also just looked at the 500T and the same thing is true there. Grain on the other hand is much smaller on the (500T) Vision3 emulsion (as is to be expected).

50D:

https://125px.com/docs/motionpicture/kodak/5245.pdf

https://www.kodak.com/content/products-brochures/Film/VISION3_50D_5203_TI-2657_Technical-Data_EN.pdf

500T:

https://125px.com/docs/motionpicture/kodak/5298-ti2082.pdf

https://www.kodak.com/content/products-brochures/Film/VISION3_5219_7219_Technical-data.pdf

 

[koraks]

I think the OP was referring to the 50D, for which the MTF of the EXR seems to be quite a bit higher.

Oh yeah, that's an odd one:

Green is Vision3 50D, red is EXR 50D.
Looks like something was changed that affected acutance; maybe DIR-related.

 

[Lachlan Young]

Oh yeah, that's an odd one:
View attachment 363724
Green is Vision3 50D, red is EXR 50D.
Looks like something was changed that affected acutance; maybe DIR-related.

Or the market demand was for extremely smooth and clean rather than sharp and gritty - we have to remember that this is based around an approx 12x enlargement of what is effectively very close to 1/2 frame.

As a side note, one of the bigger problems today is that the codecs for many streaming services really struggle with film grain - they want a super de-noised image so they can get away with compressing the hell out of it. And then they pay to have it re-fake-noised after encoding...

 

[dokko]

one point also is that in times when EXR was designed, pretty much all films were printed optically, which mean we often had a fourth or even sixth generation copy in the theaters. ie:

original camera negative (OCN) → inter positive (2. Gen) → inter negative (3. Gen) → release print (4. Gen)

on blockbuster films, another interpositive and internegative generation was used to get the sheer volumes of prints done (or so I've been told, never been involved in a big production like that). so the release print was a 6. generation copy.

since each copy lost some sharpness, it seems natural that film stocks have been designed to have a rather high acutance.

with scanning and digital distribution this was greatly reduced:
original camera negative (OCN) → digital scan and mastering → digital projection

 

[dokko]

Or the market demand was for extremely smooth and clean rather than sharp and gritty - we have to remember that this is based around an approx 12x enlargement of what is effectively very close to 1/2 frame.

not sure where the 12x enlargement number comes from...

35mm movie cameras usually have a 22mm gate (academy) and projectors mask some of that to about 20mm width.

larger cinemas easily have a screen size of 20meters. so that's roughly an enlargement of 1000x.

 

[Lachlan Young]

not sure where the 12x enlargement number comes from...

35mm movie cameras usually have a 22mm gate (academy) and projectors mask some of that to about 20mm width.

larger cinemas easily have a screen size of 20meters. so that's roughly an enlargement of 1000x.

You might want to do some more careful research about this. It's not about screen size, it's about how big the screen is relative to your distance to it. Hold an 8x10/ 25cm wide print out in front of you while sat in an average seat in the cinema until it covers the width of the screen, that will tell you more about what the actual resolution/ Circle of Confusion requirements are. I used a relatively demanding (0.017mm) CoC to derive an equivalent enlargement size from - much of the cinema world uses looser specifications (often 0.025mm), but my point stands - think about how much visible granularity you'll see in an 8-12x enlargement, and the circumstances in which 50D would be being used (if your subject is already high contrast, you don't necessarily want the film to add a chunk more edge contrast). The other aspect that is often ignored is that MTF sums across each printing/ duplication step, and if this isn't respected you can get some very visually unpleasant effects.

 

[dokko]

You might want to do some more careful research about this. It's not about screen size, it's about how big the screen is relative to your distance to it.

totally agree that the resolution needed for a pleasing image depends a lot on distance, and obviously nobody would ask for the same resolution on a 30meter screen as on a 8x10" print.
the same is true if we print a photo negative once at 8x10" and once at 80x100", the later doesn't need to be as detailed if viewed at a distance. still the former would be a 8x enlargement (from a 35mm negative) and the later a 80x enlargement.

so by that definition, the movie projection is still a 1000x enlargement rather than 12x (at least the way i understand the term "enlargement" as a non-native english speaker : )

There's an interesting phenomena with film projection vs photo prints:
We can project a well mastered film from a DVD, (which is 720pixels horizontal resolution), and it looks very decent on a 2meter screen from say 3 meters viewing distance. When we try to print the exact same image on the same size as a photo that doesn't work for me.

similarly, projecting a 35mm first generation print on a 3meter wide screen looks just gorgeous, while trying to make a 3meter wide print from a 20mm wide crop of a negative usually doesn't.

 

[Lachlan Young]

so by that definition, the movie projection is still a 1000x enlargement rather than 12x (at least the way i understand the term "enlargement" as a non-native english speaker : )

There's an interesting phenomena with film projection vs photo prints:
We can project a well mastered film from a DVD, (which is 720pixels horizontal resolution), and it looks very decent on a 2meter screen from say 3 meters viewing distance. When we try to print the exact same image on the same size as a photo that doesn't work for me.

In projection, the actual enlargement doesn't matter as much as the relative enlargment - ie a movie might be enlarged 100x or 1000x, but if where you view it from is correctly positioned, you will always perceive it to be 10-12x - i.e. a relatively fixed constant compared to the variable of projection throw. It's part of why 4k+ has been a major case of diminishing returns, as you'd have to sit so close as to only ever view part of the screen.

 

[Crysist]

...

Is it possible you were mistaken due to changes in how Kodak set the vertical scale of the plots?

I think the OP was referring to the 50D, for which the MTF of the EXR seems to be quite a bit higher.

...

Oh yeah, that's an odd one:
View attachment 363724
Green is Vision3 50D, red is EXR 50D.
Looks like something was changed that affected acutance; maybe DIR-related.

To Korak's original point, that is half right. I hadn't actually overlaid them to see whether they were close, I was mainly focusing on how far the graphs go and taking a few of the plot values at lower frequencies to compare the different stocks. So it looked a lot more drastic to me.

But also, as dokko states, the comparisons between each 50 speed stock make what I was seeing more clear and show it is rather substantial. The best layer in the Vision stock is as good as the worst layer in EXR. While the increase in latitude is good, I wonder if it sacrifices the resolution maybe a bit much. It seems to have done really well by those figures, I don't see any other color films achieving that.

Regardless, I am curious: I see many people tune their development process and choice of film in B&W by a lot. Achieving certain grain, acutance, mixing up special "recipes" to improve various aspects of the negative. All in development, not printing. Color doesn't seem to have a fraction of that flexibility. There's no fancy developers for each use case with color. So while aspects like the acutance might be formulated into the stock, is there no way to improve acutance during development?

Or the market demand was for extremely smooth and clean rather than sharp and gritty - we have to remember that this is based around an approx 12x enlargement of what is effectively very close to 1/2 frame.

As a side note, one of the bigger problems today is that the codecs for many streaming services really struggle with film grain - they want a super de-noised image so they can get away with compressing the hell out of it. And then they pay to have it re-fake-noised after encoding...

Shhh, don't let James Cameron hear you!

The Abyss and T2 have such great Super35 photography, I cry thinking about what we could have if they were properly remastered...

one point also is that in times when EXR was designed, pretty much all films were printed optically, which mean we often had a fourth or even sixth generation copy in the theaters. ie:

original camera negative (OCN) → inter positive (2. Gen) → inter negative (3. Gen) → release print (4. Gen)

on blockbuster films, another interpositive and internegative generation was used to get the sheer volumes of prints done (or so I've been told, never been involved in a big production like that). so the release print was a 6. generation copy.

since each copy lost some sharpness, it seems natural that film stocks have been designed to have a rather high acutance.

with scanning and digital distribution this was greatly reduced:
original camera negative (OCN) → digital scan and mastering → digital projection

The first generation of Vision even had this big difference, but I don't know how big of a difference it was after all the successive copies. Maybe The film prints from the 90s were a better than those from the 2000s then.

Apparently the smoother look is particularly helpful for higher speed film because it can diminish grain.

 

[dokko]

The first generation of Vision even had this big difference, but I don't know how big of a difference it was after all the successive copies. Maybe The film prints from the 90s were a better than those from the 2000s then.

Apparently the smoother look is particularly helpful for higher speed film because it can diminish grain.

yes, as you mention, in the first generation, the main purpose was to get less apparent grain in the higher speed films. In fact, as far as I can remember there wasn't a first generation 50D and 100T Vision emulsion and they skipped that until Vision2 was introduced.

the data sheet suggests that they did manage to improve MTF slightly from first to third generation while lowering grain even further:

Vision 500T 5263:

https://125px.com/docs/motionpicture/kodak/5263.pdf

Vision3 500T 5219:

https://www.kodak.com/content/products-brochures/Film/VISION3_5219_7219_Technical-data.pdf

 

[dokko]

Regardless, I am curious: I see many people tune their development process and choice of film in B&W by a lot. Achieving certain grain, acutance, mixing up special "recipes" to improve various aspects of the negative. All in development, not printing. Color doesn't seem to have a fraction of that flexibility. There's no fancy developers for each use case with color. So while aspects like the acutance might be formulated into the stock, is there no way to improve acutance during development?

I'm no expert on this, but I suspect it would be possible to change the chemistry to affect acutance somewhat.

The main problem is likely that it would probably also affect color rendering, which is usually a bigger concern than minor changes in sharpness and grain texture.

Also, color is mainly done in commercial labs, where keeping at standarised procedures is more important than tinkering. rather few labs offer custom tailored b/w processing either.

 

[Crysist]

yes, as you mention, in the first generation, the main purpose was to get less apparent grain in the higher speed films. In fact, as far as I can remember there wasn't a first generation 50D and 100T Vision emulsion and they skipped that until Vision2 was introduced.

the data sheet suggests that they did manage to improve MTF slightly from first to third generation while lowering grain even further:

Vision 500T 5263:

https://125px.com/docs/motionpicture/kodak/5263.pdf

Vision3 500T 5219:

https://www.kodak.com/content/products-brochures/Film/VISION3_5219_7219_Technical-data.pdf

Oddly, with the layers in the reverse order. Everything went up but red and blue swapped.

I'm no expert on this, but I suspect it would be possible to change the chemistry to affect acutance somewhat.

The main problem is likely that it would probably also affect color rendering, which is usually a bigger concern than minor changes in sharpness and grain texture.

Also, color is mainly done in commercial labs, where keeping at standarised procedures is more important than tinkering. rather few labs offer custom tailored b/w processing either.

Right, like it's complex but it's for stability of conditions, not changing them up.

I was wondering if the dye "activation" part could be separated from the development of the silver. Say, has anyone tried B&W development tricks with slide film developing? Because doesn't that use separate B&W and color developers? Could you use your "special sauce" in the first development step?

 

[koraks]

I was wondering if the dye "activation" part could be separated from the development of the silver.

Yes, that's theoretically possible, but the question is what purpose it would serve. The silver (halide) captures the light, so if you were to form the dyes by locally creating the same functional group that turns them from colorless into colorful dyes, this chemical process would not necessarily be responsive to light. For instance, I once did this:

This is a small snippet of RA4 paper (similar dyes to ECN2 and E6 film) which was blixed out and washed. Then afterwards I added some developer and silver chloride (made in-situ by combining silver nitrate and sodium chloride), which developed out instantly and the oxidized developer coupled with the dye couplers. While I used silver chloride here to get things going, you could (at least theoretically) also use something else.

So all considered, yes, you could form (colorful) dyes from the colorless dye couplers in a CN material through different routes than the normal development process. However, if you want the film/paper to work more or less intended, i.e. as a material that responds to light (and different colors of it, at that), I don't really see why you'd want to skip over the silver halide already embedded in the emulsion and sensitized specifically to work where it is.

Because doesn't that use separate B&W and color developers?

Yes, it does.

 

[Lachlan Young]

Shhh, don't let James Cameron hear you!

Nothing to do with that - it's more that it would make very clear how much the streaming services are getting away with in terms of terrible compression. They're under quite a lot of pressure both to allow people to originate on film & are also desperate to access archives of IMax etc.

 

[Crysist]

Yes, that's theoretically possible, but the question is what purpose it would serve. The silver (halide) captures the light, so if you were to form the dyes by locally creating the same functional group that turns them from colorless into colorful dyes, this chemical process would not necessarily be responsive to light. For instance, I once did this:

This is a small snippet of RA4 paper (similar dyes to ECN2 and E6 film) which was blixed out and washed. Then afterwards I added some developer and silver chloride (made in-situ by combining silver nitrate and sodium chloride), which developed out instantly and the oxidized developer coupled with the dye couplers. While I used silver chloride here to get things going, you could (at least theoretically) also use something else.

So all considered, yes, you could form (colorful) dyes from the colorless dye couplers in a CN material through different routes than the normal development process. However, if you want the film/paper to work more or less intended, i.e. as a material that responds to light (and different colors of it, at that), I don't really see why you'd want to skip over the silver halide already embedded in the emulsion and sensitized specifically to work where it is.



Yes, it does.

The reason I was considering was that you could then use the same development alterations you'd use with B&W, affecting grain, acutance, contrast, and other factors with a B&W development step, then separately activate the dye couplers afterwards.

The fact that they are separate for E6 development made me think you already could do that for slide film. Is that true? Can you, in an E6 process, swap in a D76 or HC110 or Rodinal or any other B&W developer before the remaining steps?

Nothing to do with that - it's more that it would make very clear how much the streaming services are getting away with in terms of terrible compression.

Right, I was mainly being cheeky on that point. The difference between streaming services and UHD bluray is night and day.

In your prior comment about "adding grain after encoding" did you mean something like the Lowry process, or do you mean recent video codecs which have separate stages to handle film grain?

They're under quite a lot of pressure both to allow people to originate on film & are also desperate to access archives of IMax etc.

Who is? And to what extent? For IMAX, iirc, it affects the branding "Shot on IMAX" vs "Shot for IMAX", where "on" is only allowed when you shoot 15/70.

 

[koraks]

The reason I was considering was that you could then use the same development alterations you'd use with B&W, affecting grain, acutance, contrast, and other factors with a B&W development step, then separately activate the dye couplers afterwards.

The fact that they are separate for E6 development made me think you already could do that for slide film. Is that true? Can you, in an E6 process, swap in a D76 or HC110 or Rodinal or any other B&W developer before the remaining steps?

Gotcha, I see what you mean. As to question #2: yes, you can, and it will certainly affect the final outcome. And yes, in a color negative film, you can do pretty much the same thing: use a B&W developer first to develop just the silver halide image, fix out the non-image silver halide, then bleach it back with a rehalogenating bleach and redevelop the silver halide image with a color developer to form the color dyes. Then bleach and fix again to get rid of the silver image, although you could also leave it in place of course (effectively bleach bypass).

The choice of the first B&W developer will alter the image characteristics in many ways. You may or may not lose certain measures the film manufacturer has taken to control e.g. interlayer effects. If this is a problem, I don't know; it's a case of "experiment and see if you like what you get". It's a fun project for a rainy Sunday if you have a 400ft roll of Vision3 lingering around in the fridge.

 

[Crysist]

Gotcha, I see what you mean. As to question #2: yes, you can, and it will certainly affect the final outcome. And yes, in a color negative film, you can do pretty much the same thing: use a B&W developer first to develop just the silver halide image, fix out the non-image silver halide, then bleach it back with a rehalogenating bleach and redevelop the silver halide image with a color developer to form the color dyes. Then bleach and fix again to get rid of the silver image, although you could also leave it in place of course (effectively bleach bypass).

The choice of the first B&W developer will alter the image characteristics in many ways. You may or may not lose certain measures the film manufacturer has taken to control e.g. interlayer effects. If this is a problem, I don't know; it's a case of "experiment and see if you like what you get". It's a fun project for a rainy Sunday if you have a 400ft roll of Vision3 lingering around in the fridge.

Oh fascinating! So color developers activate the dyes in tandem with the reduction of silver, that's why they're in the same spot as the corresponding silver particles, right? That's why you can't have a step that does it on top of already reduced silver. You need to rehalogenate it to get it back into the undeveloped state and then start color development as normal from there, right?

In summary, just so I have this right:

After exposure: [ Latent silver ] + [ Unexposed silver halide ]

After B&W develop: [ Developed silver ] + [ Unexposed silver halide ]

After Fix: [ Developed silver ]

After Rehalogenating bleach: [ Latent silver (again) ]

Then color development acts as normal on the latent silver halide image.

I never know what to call what the latent image is made of, because it's in a strange state where the later development has been seeded in parts by the exposure, but development will reduce additional silver to form the image. That's why overdevelopment will eventually cause the entire image to become silver, right? Even when barely or not even hit by light? Some explanations of development don't make it clear what chemical state the silver is that will form the image is in.

Also, why is it called rehalogenating bleach? Doesn't bleach remove silver? Is it considering that strictly true because it removes silver... by turning it back to silver halide?

I wonder what results this would produce, that'd be fun to see! B&W development seemed so uniquely fascinating because people were doing all these neat tricks -- for example I saw this guy make some additions which somehow made his images sharper. I suppose there's no guarantee that the color won't also be affected, but seeing what this extra flexibility affords would be cool.

Ah yes, the spare 400ft roll of Vision3 lol. That makes me wonder, since I did want to try different development hacks and change various variables and see what happens, wouldn't the quickest way to test things be to use sheet film? Then you can set up a test scene, shoot one, develop, see how it turns out, adjust, etc.

Thanks a bunch for your responses!!

 

[koraks]

Oh fascinating! So color developers activate the dyes in tandem with the reduction of silver, that's why they're in the same spot as the corresponding silver particles, right? That's why you can't have a step that does it on top of already reduced silver. You need to rehalogenate it to get it back into the undeveloped state and then start color development as normal from there, right?

You got it!
What happens, simply put, is that the developer is oxidized as it reduces exposed silver halide into metallic silver. This oxidized developer molecule then binds itself to the dye couplers that are nearby. The dye couplers themselves are colorless, but once combined with the oxidized developer molecule, they form a colorful dye.

I never know what to call what the latent image is made of, because it's in a strange state where the later development has been seeded in parts by the exposure, but development will reduce additional silver to form the image. That's why overdevelopment will eventually cause the entire image to become silver, right? Even when barely or not even hit by light?

Yes, that's how I understand it as well. The exposed silver halide grains/crystals are still for the most part silver halide (AgBr, AgCl, AgI).

Also, why is it called rehalogenating bleach? Doesn't bleach remove silver?

A rehalogenating bleach links the metallic silver back with a halogen (chloride, bromide, iodide; usually bromide in a color negative process), converting metallic silver into a silver salt. A non-rehalogenating bleach also forms silver salts, but these are easily soluble in water and are thus carried away from the emulsion. Silver halides are insoluble for the most part, so they stay put.

wouldn't the quickest way to test things be to use sheet film?

That's also possible, but budgetary decidedly less attractive for most mortals - including me!
In the end, a bulk roll of 35mm also allows you to do quick iterations on small snippets of film. Just load a small section into a cassette and expose it in-camera, or tape it inside a sheet film holder for experiments with an enlarger etc. etc.

 

[Crysist]

I really appreciate your response, koraks! I know I've strayed away from the original topic quite a bit, but I might be spamming the forum if I post all my questions at once! At least it's more relevant here! Any further questions about the "why" regarding the changes from EXR to Vision might be better sent to Kodak at this point than me guessing.

You got it!
What happens, simply put, is that the developer is oxidized as it reduces exposed silver halide into metallic silver. This oxidized developer molecule then binds itself to the dye couplers that are nearby. The dye couplers themselves are colorless, but once combined with the oxidized developer molecule, they form a colorful dye.

Awesome, thanks!

Yes, that's how I understand it as well. The exposed silver halide grains/crystals are still for the most part silver halide (AgBr, AgCl, AgI).

Ok, cool! That was a bit confusing because before I would see examples explaining it like there was an invisible shadow of all the activated silver where there will be image density upon development. It hides how density differences are captured.

On that note, I wonder what would be a good metaphor or picture to use? Effectively, the relative densities in different parts of the final image make the most sense, since that's what it produces in the end, but that "information" is contained in the latent image by different, far weaker relative densities of silver among mostly unreduced silver halide.

Y'know, I'm really starting to think photography was a miracle in the first place....

A rehalogenating bleach links the metallic silver back with a halogen (chloride, bromide, iodide; usually bromide in a color negative process), converting metallic silver into a silver salt.

"Back"? But the rehalogenating bleach brings those halogens with it, right? Because the fix washed the existing, undeveloped silver halides away.

A non-rehalogenating bleach also forms silver salts, but these are easily soluble in water and are thus carried away from the emulsion. Silver halides are insoluble for the most part, so they stay put.

That makes sense, the silver halides precipitate and get stuck in suspension the collodion/gelatin/etc.

Hm, that's got me thinking, in emulsion-making if the byproducts of forming the silver halide salts are very soluble nitrate salts; do emulsions need to be cut (as in the traditional noodle + wash steps) or is that merely increasing surface area and it would work in other ways? I understand washing liquid gelatin or an uncut block of gelatin directly with water might not make sense, it'd probably take a while to get through a block and I'm not sure if the gelatin would just capture it or get diluted by it, but what about after coating? I guess it's just a bit of convenience that you wash the emulsion once and then the plates you coat are less... salty. But won't that get washed out immediately upon development?

Bleach is a bit of a confusing name, unless it means the strong oxidant aspect, cause the dye couplers seem to survive that step...

Sorry for the cavalcade of follow-up questions!

That's also possible, but budgetary decidedly less attractive for most mortals - including me!
In the end, a bulk roll of 35mm also allows you to do quick iterations on small snippets of film. Just load a small section into a cassette and expose it in-camera, or tape it inside a sheet film holder for experiments with an enlarger etc. etc.

Hm, are you not personally concerned with doing any experiments with motion picture negative but developing in C41? Or do you have ECN2 chemicals? From what I've heard that's the only way to get the colors actually correct and reap the benefits of the higher latitude because it's a lower-contrast developer.

Thank you again for your patience with all my questions!