Dynamic range of slide film?

[srichards]

In the latest ap mag they say slide film has a large dynamic range and that is an issue for scanners. Everywhere else on t'internet everyone says the opposite apart from the odd lone voice that agrees with ap.

So which is true and are there any actual proper scientific tests including the known ev range? Also I wonder whether the nay sayers are confusing dynamic range with exposure latitude hence saying it is low on dynamic range because they haven't exposed correctly? But if exposure is so critical then wouldn't it follow more logically that slide film has a narrow dynamic range as exposure errors manifest more easily?

Both sides can't be right . . .

 

[Steve Smith]

I think you are confusing dynamic range of light levels it can record with it's physical light transmitting range.

A slide will have quite a large difference between the lightest, least dense parts and the darkest, most dense parts.

However, it is not capable of recording such a large range of light levels as negative film.

Slide film has a very low exposure latitude - so low that I would say it is zero. In my opinion, the exposure latitude of a film is the amount you can get it wrong and recover from it in printing. As a slide is the final product, it has no latitude for correction.


Steve.

 

[s162216]

When your scanning on most consumer scanners then slide film certainly has a small dynamic range simply because the scanner itself does! A lot of people have posted about being disappointed with their scans of slides that they've posted here, the simple reason being that the scanner they've used lacks the dynamic range to pull the detail out of the shadows unlike projection where you can in effect blast through the shadows as you've got a powerful light source!

Having a dedicated 35mm scanner helps as I can use multi-exposure scanning which really increases the already fairly impressive dynamic range, but even then it sometimes struggles when there are really deep shadows in the slide.

 

[srichards]

I see.

I've been scanning in some old slides and I didn't really notice an issue. There were only odd shots with no shadow detail where exposure was also totally off. Scanner still seemed to drag out quite a bit of detail even on slides I could barely see through!

Sounds like a great excuse to get a slide projector too

 

[s162216]

It does of course depend on the scanner that you were using, what do you have?

 

[abbandon]

I seem to remember reading some articles by Tim Parkin on On Landscape regarding film dynamic range and his testing there of but I can't find the right link at the moment.

Most "disscusions" suggest that most slide films have about 4 stops range with Provia perhaps having five stops while negative films have about sixteen stops.

Another general rule of thumb for colour is that you can get four stops in a print.

Film density is another matter density for negative materails is generally accepted to be 2 while its 4 for slide film.

The majority of manafactuer scanner specifications are umm overly optomistic / a distiontion of the truth / based on theoretical values rather than measured ones.

It is only really drum scanners that have density ranges approaching 4.

 

[srichards]

I seem to remember reading some articles by Tim Parkin on On Landscape regarding film dynamic range and his testing there of but I can't find the right link at the moment.

Most "disscusions" suggest that most slide films have about 4 stops range with Provia perhaps having five stops while negative films have about sixteen stops.

Another general rule of thumb for colour is that you can get four stops in a print.

Film density is another matter density for negative materails is generally accepted to be 2 while its 4 for slide film.

The majority of manafactuer scanner specifications are umm overly optomistic / a distiontion of the truth / based on theoretical values rather than measured ones.

It is only really drum scanners that have density ranges approaching 4.

My scanner is minolta scan elite 5400 ii. Bought it new in about 2005 ish. Did find some German review site which tested scanners and the Nikon and minoltas were both close to spec. The epson flatbeds they tested were way off. Resolutions were way under what was quoted.

 

[s162216]

My scanner is minolta scan elite 5400 ii. Bought it new in about 2005 ish. Did find some German review site which tested scanners and the Nikon and minoltas were both close to spec. The epson flatbeds they tested were way off. Resolutions were way under what was quoted.

Thats the reason why you had no problems with shadows then! The dynamic ranges of those Minolta dedicated scanners are fantastic, I sometimes still wish that I had waited for one of those to come up cheap on ebay and get one rather than getting my Reflecta Proscan 7200 but for my purposes the Reflecta is 95% of the time good enough anyway and when reviewed by Scandig (the German website you mention) it did come close to what was advertised (3250 dpi compared to the 3600 advertised - way better than the Plustek models, similar values despite them advertising 7200 dpi!) in both density and resolution , much better than any of the flatbeds or Plustek models did. I keep multi-exposure on all the time when scanning slides simply because of the ease it makes in resolving details in deep shadows and the multi passes not being a problem like on a flatbed which have trouble lining up the scan heads for every pass.

 

[ChrisR]

Having a dedicated 35mm scanner helps as I can use multi-exposure scanning which really increases the already fairly impressive dynamic range, but even then it sometimes struggles when there are really deep shadows in the slide.

Samuel, I think you perhaps have a Plustek; I have a 7500i and use Silverfast 6.6. I've tended to use a 4-pass multiscan "to minimise noise and grain", rather than the 2-pass multi-exposure. This is partly because I couldn't work out the difference, and partly because I couldn't see the difference (and 4 is better than 2, right?). It sounds like you have done some tests and satisfied yourself that multi-exposure really works, so I'm trying it now. Not a proper comparison test, just trying it anyway!

 

[s162216]

Samuel, I think you perhaps have a Plustek; I have a 7500i and use Silverfast 6.6. I've tended to use a 4-pass multiscan "to minimise noise and grain", rather than the 2-pass multi-exposure. This is partly because I couldn't work out the difference, and partly because I couldn't see the difference (and 4 is better than 2, right?). It sounds like you have done some tests and satisfied yourself that multi-exposure really works, so I'm trying it now. Not a proper comparison test, just trying it anyway!

As a said above, I have a Reflecta Proscan 7200. The difference between the multi-exposure and multi-pass is simple: the multi-exposure takes 2 scans one at normal exposure and the other at a much brighter exposure to get into the deep shadows that were likely not picked up during the initial normal exposure scan, these 2 scans are then combined. It is important to note however that mutli-exposure only shows benefits on slides, it has no use on negatives (as obviously the 'shadows' on the negative are actually highlights).

With multi-pass however the exposure remains the same but allows the scanner to reduce noise as it combines the scans and can reduce the noise accordingly (as obviously noise will not be in the same place for every scan).

I don't use Silverfast, I use Vuescan but the multi-exposure on slides really shows how much more detail you can pull out of the shadows in comparison to not using it and by itself it will actually reduce shadow noise anyway. Plus that in Vuescan you can actually get it to both! For slides I usually get it to do 4 passes and one multi-exposure pass. I'll find some time this weekend to do a comparison with and without multi-exposure.

Sam

 

[srichards]

The scary thing is that the 5400 is still £500 secondhand. That's what I paid for it new!

I think I'm going to have to find some seriously back lit and contrasty scenes that I'd need HDR for with the sony and try some slide and neg film to see what happens. I'll only be satisfied with scientific experiment that I've seen with my own eyeballs. Gloomy churches with stained glass windows and buildings against brilliant blue skies are probably the kind of thing that would test both.

Slide film was often used in fashion and commercial photography so I still don't get why they'd use anything which had less light level differentiating ability. If colour neg was better than you'd use colour neg unless they chose slide because it had the same gamut as print?

 

[s162216]

Slide film was often used in fashion and commercial photography so I still don't get why they'd use anything which had less light level differentiating ability. If colour neg was better than you'd use colour neg unless they chose slide because it had the same gamut as print?

Probably because it was easier to put on the printing presses, like why newspaper photographers mostly shot slide as well, in pre-computer days for negative it was necessary to make a 'dupe' of the negative on slide film which would obviously lower the final quality; slide films were also seen as higher quality sharpness etc wise so you can probably see why it was used.

 

[timparkin]

I seem to remember reading some articles by Tim Parkin on On Landscape regarding film dynamic range and his testing there of but I can't find the right link at the moment.

Most "disscusions" suggest that most slide films have about 4 stops range with Provia perhaps having five stops while negative films have about sixteen stops.

Another general rule of thumb for colour is that you can get four stops in a print.

Film density is another matter density for negative materails is generally accepted to be 2 while its 4 for slide film.

The majority of manafactuer scanner specifications are umm overly optomistic / a distiontion of the truth / based on theoretical values rather than measured ones.

It is only really drum scanners that have density ranges approaching 4.

I've bracketed a few different films and the dynamic range is a lot more than you'd think. About 8 stops for Velvia, Provia, Astia, etc. Plus or minus a couple of thirds.

The thing with Velvia is that it gets very, very dark. You'll lose a couple of stops with a typical flat bed or dedicated scanner.

Velvia has the same dynamic range as Provia but provia's shadows are a lot brighter and hence it scans better.

When you're metering though, even with a spot meter you only average an area - albeit a smaller area. Most 1 degree areas contain variations in brightness, especially shadows, and hence it's good practise to expose from -3.5 to plus 2 (-4 at a push).

Interestingly, blues 'clip' at lower exposures than red/yellows. Hence if you're shooting a sunset, you'll get yellow tones all the way up to +3.

If you want to check a transparency, punch a small hole out of some thick card or plastic and hold it over the film and then hold the whole lot over a bright bulb. You'll see into shadows a lot better this way - your eyes suffer from flare just like a camera lens and the surrounding brightness will close your iris - you'll be surprised at how much more you can see.

 

[Andysnap]

Interesting, I'd not thought about eye-flare (new word just invented) I shall give this a go.

Another thing that has got me pondering is Samuel's mention of multi-exposure. Can I asked how you achieve this please.

Ta

Andy

 

[s162216]

Another thing that has got me pondering is Samuel's mention of multi-exposure. Can I asked how you achieve this please.

Ta

Andy

Its an option in Vuescan and certain versions of Silverfast, as I explained above it basically combines two scans with different exposures to increase the dynamic range of the scans. Unfortunately its use is limited on flatbeds as they have difficulty in aligning the scan head for each pass so there will be a likely reduction in sharpness.

 

[Andysnap]

Thanks Sam, I'm still using the Epson software at the moment but I might upgrade later and give it a try.

Andy

 

[abbandon]

Hi Tim, I was not doubting the greater range of slide film just quoting what the generally accepted values are.

My scanner is minolta scan elite 5400 ii. Bought it new in about 2005 ish. Did find some German review site which tested scanners and the Nikon and minoltas were both close to spec. The epson flatbeds they tested were way off. Resolutions were way under what was quoted.

Minolta 5400

A resolution test results for Konica Minolta DiMAGE Scan Elite 5400 II an effective resolution of 4200 dpi.

I tested the Konica Minolta DiMAGE Scan Elite 5400II so shows in horizontal and vertical direction, a difference of about 1000 dpi resolution. If one forms from the two measured values ​​an approximate average, one arrives at an effective resolution of 4200 dpi. Measured at the nominal resolution of 5400 dpi are the only 78%, that is more of a poor result. Compared with other film scanners at the present time (April 2005), this value means the highest resolution I have ever measured with a film scanner.

A small flaw, the measurement of the resolution with my Konica-Minolta scanners: A first resolution test I have done after several hours of continuous operation of the film scanner, showed only a very modest resolution of 2900 dpi. This value was reproduced by several rescan of the USAF test target. The high resolution of 4,000 dpi values ​​I scored again until a reboot after several hours of rest of the scanner. This effect, I can not explain it, because the film scanner works even with LEDs that do not heat up and not wear out.

In other journals lower resolutions were measured down to 3600 dpi for this device for some of them. This is obviously a disappointing value, especially since there are 4000 dpi scanner from Nikon, almost reach the nominal resolution.

Perhaps I didn't have the patience but I did not find anythiing aboout its D range.

Aztek has an interesting site hidden away

Results for the Nikon



The Hutchcolor scanning guide has a lot of usefull info as well as a practical way to measure a scanners D range sadly the require items for this are not cheap.

Some when I will have a go at comparing the different scanners I own after fixing up the Howtek and haveing the funds to get hold of the required targets etc.

 

[srichards]

Ah. It was in German so I probably didn't understand it fully or erased it from my memory

4000 ish dpi isn't that bad. I think some of the epsons tested were showing 2/3rds quoted resolution from what I remember. 5400 ii quoted dmax is 4.8 but I guess if they were exaggering then it is more likely to be closer to 4 in reality.

I wonder if the edited out one in the list is a nikon or a minolta?

 

[s162216]

I wonder if the edited out one in the list is a nikon or a minolta?

Seeing as the Coolscan 8000 is on the list I would assume its not a Nikon, it could also be a number of other scanners apart from a Minolta, Canon did manufacture a similar level film scanner around that time for instance.

TBH flatbeds like the Epson's are never going to really live up to their marketing claims really, the Canon 9000F claims 9600 dpi for instance (which the sensor can resolve!) but only resolved up to about 1700 dpi when tested and even the Epson V700 only reached about 2300 dpi.

For most films (exlcuding specialist high resolution films like Technical Pan), 4000 dpi will be enough anyway to resolve the full detail, I do remember finding a study once where they actually tested the resolution characteristics of various films and the results ranged from about 2400 dpi for very fast B&W through to about 4000 dpi for slow slide films. I'll try and find that at some point.