Negative film - there's more resolution there than you think

[tyler_norman]

The general consensus is that slide film has better quality and image resolution than

negative (print) film - this is mostly true, however, your color prints are only showing you

half of the image resolution that's in the actual negative.

 

I use a Minolta Dimage Scan Dual III film scanner, which can scan 35mm film into 3,000-

pixel-wide images, pulling out every bit of detail from the film. When I scanned negative

film with it, I was amazed. Kodak Gold 100, for instance, gave me an image comparable in

quality to Ektachrome 100. Kodak Gold 200 and Fuji SuperHQ 200 are of almost equally

excellent quality, with only slighty more grain than their 100-speed counterparts. I've even

scanned from Kodak Max 400 and have gotten fantastic detail and surprisingly fine grain

from it.

 

Bottom line: If you have a good film scanner, don't hesitate to use negative film. You'll be

surprised at how much detail is actually in those negatives.

 

(of course, if you want REALLY great quality, you'll use Velvia 50 :-)<div></div>

[emre]

When I use print film, it is for the dynamic range.

[chip]

Tyler,

 

You haven't even come close to what film can resolve with your 3,000ppi film scanner. Good PMT drum scans can find more resolution in your film up to almost 8,000ppi or so. Your test is limited by you scanning system so your tests are invalid.

 

HTH

[jack paradise]

And there are better slide films than Velvia 50.

 

And as professional photographer Alain Briot has noted, "Velvia increases both the color saturation and the contrast of the original scene, it is important to select scenes that have a low contrast range and low color saturation. In turn, this means you will have to avoid certain specific natural situations which, if photographed with Velvia, will result in blown highlights and overexaggerated colors".

 

Get yourself some Provia 100F.

[scott_eaton]

<I>The general consensus is that slide film has better quality and image resolution than negative (print) film </i><P>Yeah...right....this "concensus" is the same bunch of winners that buy $1200 35mm film scanners, then have to use AutoLevels in Photoshop because they can't understand the software. I've seen more accurate color management with a $20 polaroid camera than this crowd.<P>Attached is a the full frame from a 6k x 4k scan of some old RG-25 I had, and a 1:1 crop. No arguement from me that a good, clean, 4000dpi scan from slow speed print films will compete with anything from slide film. Good scan you have there as an example as well. Better than most I'm seeing uploaded from slide film in terms of clarity and contrast - good job.<P>I suppose if there's all this magical detail in 35mm film at 8,000 dpi I no longer have to use my RB67 chromes/negs at 2000dpi because 35mm film at 8000dpi is just as good, right?

[alexdi]

Scott:

 

To hell with film grain, I'd like to see a high-res copy of that picture.

 

DI

[johann_fuller]

If you look at the technical data of film (fuji publishes it) you will see that the resolution of neg is greater than that of tranny at the same speed. If you look for the actual resolving power of a film scanner rather than the amount of pixels it will output you will come up with zero info from any manufacturer. Theoreticaly there is no more actual information available above about 2800dpi from film - that is a scanner that can actualy resolve 2800dpi - even so called 5400dpi scanners are unlikley to do this across the whole frame. If you have access to a drum scanner and can actualy go beyond 2800dpi in actual resolution all you acheive is better resolution of grain/film structure - you don't get any more image information from the scene you took a picture of.

[gary_ferguson1]

"Theoreticaly there is no more actual information available above about 2800dpi from film"

 

Just out of interest, how did you arrive at that conclusion?

[aaron_w.]

Tyler: Not being a 'film-geek' (no offense fellas), but as a fellow SD3 owner who has had difficulties scanning negs, I am impressed and I'd like to hear about your scanning workflow... Also, do you use the SD3 Scan Utility and which settings for negs?

[johann_fuller]

"Just out of interest, how did you arrive at that conclusion?" - film resolution figures ( line pairs per mm at a selection of contrast ratio's) are (as I said earlier) available from manufacturers - it's not rocket science to workout the resolution in dpi. This is theoretical - when you add lens aberations, focus error, film bulge in the film gate, camera shake, unoptimal dev and exposure etc you have to take 2800dpi as an upper limit.

[mark__5]

"I'd like to see a high-res copy of that picture"

 

why??

[ilkka_nissila]

Interesting number this 2880. Looking at the high-contrast resolving power from the Velvia data sheet, we see that it's 160 line pairs per mm. Now, to render a line pair in a digital image correctly, you need four pixels per line pair. Thus, theoretically we need 25.4*4*160 = 16256 ppi in the scanner to get all information that the film contains. Now, I'm not even going start about Technical Pan.

 

Note that this is indeed theory, in practice 4000 dpi scans are fine as long as they're in focus.

 

However, my experience is that current negative films with the exception of Reala do not scan well compared with iso 100 slide films. At least on the LS-4000, I get much better results from slides than negs, this is due to grain. It may be that with a more diffuse light source, negs give better results, and surely RG25 is an exception (but it isn't made any more so what's the point?). -- Grain in one thing, resolution another, and I think it's possible that some neg films (Reala?) have higher resolution than similar slide films. But I'd much rather have the finer grain for easy scanning.

[emre]

<blockquote>Now, to render a line pair in a digital image correctly, you need four pixels per line pair.</blockquote>

Should you not be using the resolution of the system; i.e., take the lens into consideration? If the scanner has a 100 lp/mm lens, the overall resolution would be 6200 ppi.

[johann_fuller]

160 lppmm is total fantasy for the actual resolution you can acheive in the real world on film. You are doing very very well if you can get 1/4 of this under controled conditions with the best equipment available.

[ilkka_nissila]

Norman Koren says that the 10% MTF point with Velvia and a real lens (apparently the EF 28-70L) is about 67 lp/mm (50% MTF point 37 lp/mm).

 

I've got some 4000 ppi scans of Elite 100 in which some transitions between flower and background are 1 pixel wide, they were taken with the 50/1.8 Nikkor. If a transition is 1 pixel wide, I wouldn't want to scan it at a lower resolution. Of course, these weren't done with a home scanner.

[scott_eaton]

Who seriously gives a flip scanning 35mm film beyond 2000-4000dpi anyway??? I take it most of the people in this thread talking about 4000-8000dpi scans are 35mm shooters who think the world revolves around that format.

 

I hate to tell you this, but a lower rez scan from larger format film looks better. A 1000dpi scan from 4x5 looks MUCH better than 4000dpi from 35mm, and 2000dpi from my 6x7 negs/slides looks much better than 4000 dpi from 35mm. If the posters here were truly interested in improving photographic results vs arguing meaningless film statistics, you'd move to a bigger format, or be shooting digital vs buying better film scanners.

 

I respect the desire to get as much information out of that tiny 24x36mm piece of emulsion via better film scanners and technique, but lets be honest - you have no other choice.

 

In regards to absolute film resolution, those stastistics are based on anechoic room tests done with monochrome test targets. No one bothers to test the resolution of Velvia when it comes to split pairs of alternating red-pink bars in a high key scene. That kind of test would of course show you how limited film is in terms of resolution and how published numbers are vastly exagerated. If your subject matter consists of only high contrast, monochrome detail, then your numbers for film are correct.

[johann_fuller]

Ease up a little! - the actual resolution of film either in a lab or in the field is very important IMO. It doesn't matter if you shoot 35mm or 5x4 - the resolving power of the film is basicaly the same and it's usefull to know this when scanning as it sets a limit to your maximum enlargement and enables you to get an optimal file size. There is no point scanning 5x4 film at 2000dpi if you are only goin to to print A4. Sure a bigger film format will always resolve more as the magnifiction of the original is less - not everyone can afford a good medium format film scanner and I don't see why people who have access to the very good and cheap 35mm scanners can't try and get the best out of them without having to invest in a MF outfit.