160VC old vs 160VC new vs 100UC vs Reala vs Rebel XTi

Well I did my test I said I was going to do. I got the last roll of old 160VC

from the store along with a roll of Reala. I got my other roll I had of 100UC

ready along with the new Portra 160VC. My idea was to see how the resolution,

grain, and color would compare against the old 160VC and the new 160VC. But I

also wanted to see how the grain of 100UC would be against 160VC. And since

Reala is a favorite- I through that in the test. But my main desire was to see

how 35mm film would compare in terms of resolution and detail against my Rebel

XTI 10 megapixel SLR camera. So I had a day off from the lab and took all 5

cameras out for an afternoon shoot. I later got the rolls developed with no

prints. I scanned all those rolls here at home on my 3200dpi Minolta Scan Dual

4. I did some color changes on each roll since each film had a severe color

cast with some shots. But you will notice they are not matched to each other.

I was just trying to get things in the ballpark. I wasn't too concerned about

color quality and contrast differences on most shots. I left contrast as per

the scan and did no sharpenning on any of these shots- including the digital

files. The easiest and most accurate film to scan with little changes was the

new Portra 160VC. 100UC seemed to go magenta, while Reala was very cyan green.

The Reala roll was overexposed compared to the other rolls, as I think my one

camera is out. I shot everything at f8 on all cameras. I left shutter speeds

up to the camera itself. The cameras in use were a Elan 7, 1N-HS, AE-1, and FT-

B. I used a 50mm lens on all 35mm cameras. The Rebel XTi I resorted to using

my 17-40L zoom to get the camera in the 31mm range. I very much doubt the 17-

40L is much sharper then my prime 50mm lenses. The Elan 7 and 1N-HS shared the

same 50mm 1.8 lens, while the FT-b used the FD 50mm 1.8 and the AE-1 used the

FD 50mm 1.4 lens. I had my tripod with me for the shots, but only 3 cameras

had a base plate to use with the tripod. The other 2 cameras I just rested on

the head. This shows up with a shot or two with movement. The day was very

overcast. I just couldn't get any sun to do this, but being the films they

were (other then Reala), they are good with flat lighting. The film that had

the most color was not 100UC, but the new Portra 160VC. 100UC was very sharp.

SO here are my samples and take a look for yourself. After seeing these

results, Im convinced my 10 megapixels SLR had more resolution then 35mm film.

Plus the images are cleaner with no grain. I'll still use film, but after all

this I will be using my Rebel XTi more often.

 

Here's the first pic full frame.

And here are the 100% crops from the full rez images. I uprezzed the Rebel XTi images to be the same as the scanned 35mm film. This is the one example that the digital image was out of focus compared to the film shots.

Here's the second full image.

And the 100% crops from this shot.

Third full image.

And 100% crops from 3rd image.

Fourth image. Full frame.

Fourth frame 100% crops.

Fifth image full frame.

And 100% crops.

Since the weather was drab out with poor lighting, I took only one real shot for color. I did this between the old and new 160VC. The new film shows way more saturation, but in this case contrast seems higher as well. Since I still have the remaining part of those rolls, I might do a second test in the future with more color and maybe some people shots. I like the new 160VC, if it is a bit over the top in color. I dont know if I'd switch from Reala though, but I might for 120 film.

One last comment- now if only the color quality of digital would be as good as the film scans. I would suspect the digital images would need quite a bit of corrections to get it close to the film scans. The best color was from 160VC new with 100UC or old 160VC being 2nd best. The Reala in this test was out to lunch in terms of color. This was not my standard results with Reala.

Coorection: The second batch of 100% crops shows the digital image out of focus, not the first set of images. Also pardon the spelling mistakes and gramar

Scot,

 

Doesn't a digital camera perform sharpening algorithms? Could this be why the pictures from the Rebel look sharper than from the scans of the negative.

 

And of course, there is an extra element involved regarding the film...the scanner. What would the comparison look like if a better scan was made, for example, a high-end drum scan? That would be a more interesting comparison.

 

Anyway, thanks for the pictures. It looks like the new portra 160VC is a really great film and is an improvement over the old version

I used the XTi in RAW mode, so I dont believe there was any sharpenning done at all. All were processed with Photoshop Elements 5 with the Raw converter.

 

Also a drum scan may be better, but to me it seems the grain is getting in the way as is in terms of resolving detail any further.

 

I will still use my 120 film cameras, but 35mm might be left to B&W and E-6 from now on.

But Scott...the digital camera must be doing some kind of processing, even if the results are outputted as RAW. After all, there is a Bayer pattern, or something similar, on the sensor. Interpolation algorithms are employed to fill in the missing colors. I've often wondered if these interpolation algorithms also contribute to edge enhancement.

RAW takes the info direct from the chip. All processing including Bayer algorythms are done in post. But I checked the program to be safe and I just noticed the sharpenning was turned to 25 in the Raw converter, so darn it the results are slightly skewed. I checked the same pic of one without sharpenning and its a bit softer. But the problem with using USM with film scans is that it enhances grain, which can be a problem. I suppose doing USM and then NeatImage noise reduction would be an improvement. I'll do the one pic and post a couple updates.

Here is the 3rd set image from the XTi without any sharpenning- I turned it to zero. You can see its softer, but still holds up well.

Here is the same pic 3 steps. First is Raw image from XTi with no USM in raw, but USM done in Photoshop. Second is new 160VC sharpenned same amount as digital Raw file using USM in Photoshop. Third is the second image then processed using Neat Image, then a second set of USM passes done on that file. I did two passes of USM in photoshop for each image, plus a third and fourth pass on Neat Image file.

Thanks, Scott. Its amazing how good the digital is. I for one still like film cameras because: 1) I like the bigger screen/image size in the viewfinder than an APS-sized sensor digital camera; 2) I like the dynamic range of film; and 3) I can shoot slides. But hey....the results of the Rebel look great. I just ordered an F6 as a birthday present for myself, so I'm looking forward to more fun with film. But someday, I might just join the bandwagon and include digital in my camerabag. My main concern, though, is how long do the sensors last before pixels start to die? Any experience with that?

rebel XTi is the only camera that doesnt show blinds on the right side of the crops (sample2full.jpg)

Pls, don't forget that rebel's image is about 10x14inches (300dpi) while scanned image is at least 15x 21 at 300dpi

Scott, thanks for the test.

 

I did something similar with my 20d, only comparing it to sensia 100 using my Scan Dual 4. In short, there was little if any appreciable gain from scans from the Scan Dual 4 when compared to the 20d. That might be different if I were using a 4000dpi Nikon film scanner (like Les uses) or a drum scan. But I wasn't. What I have is a Scan dual 4, so that's what's relevant in my case.

 

tTwo thoughts: First, as others have noted you're not really comparing the resolution of film (at least not directly ) with your XTi, but the capabilities of the Scan Dual 4. Having said that, second, I still find it very helpful. I have a Scan Dual 4 which I've used for a couple of years now. I've always been thrilled with it's results. I could almost certainly get more out of a scan if I were using a Nikon, not to mention a drum scan, but that's all hypothetical in my case. What I have is a Scan Dual 4. So I find tests like this helpful given my circumstances. I'm not going to spend big bucks on a film scanner with higher resolution. Seems to me the expense doens't outweight the relatively mild gain in resolution. I"d rather save for a 5d or some such.

 

So in sum, these are helpful. But let's be clear that you're comparing a DSLR to a film scanner--a very helpful thing to do, but not quite the same as comparing directly to film.

That rebel xti looks quite good, seems to have all of the resolution without the grain which is nice. Of course then film has more dynamic range and you can shoot film when it gives you a useful advantage and shoot digital when that has the advantage.

As for the sample 2 pics- I mentioned the digital pic was out of focus. So in that batch the results are skewed.

 

Yes I should say Im comparing scanned film to digital pics. This is the workflow I have available, so this is what will work for me. I end up getting better results from the XTi in my case. Cleaning up grain in the image can be difficult, especially when trying to get detail out of that grain. I still like films colors better, but digital produces clean images- so its a trade off.

<p>Thanks for taking the time to shoot these, Scott. I've been anxious to get my hands on the new Portra for some time, but still haven't found any, so your comparisons with old Portra, 100UC and Reala are of considerable interest to me.</p>

 

<p>A couple of comments about your conclusions relative to your Rebel, however...</p>

 

<p>First, in terms of drum scans and bringing out more grain, keep in mind that drum scans are fluid mounted and tend to do a very good job of bringing out the best in detail while helping to tame scratches and grain.</p>

 

<p>I'm not trying to suggest that you should drum scan your 35mm film or that the comparison is somehow "unfair". It's completely "fair" for your purposes as it uses the tools you would actually use - end of story. I'm simply responding to your comment relative to drum scanning and pointing out that a good drum scan might well bring out more detail without adding grain.</p>

 

<p>And secondly, the all important issue of sharpening. I would never, ever, ever globally sharpen film scans due to exactly the problem you see with grain accentuation. There're a number of techniques, packages and plugins out there which address this issue.</p>

 

<p>If you're not familiar with this, I'd highly recommend you check out <a href ="http://www.creativepro.com/story/feature/20357.html">this article</a> by the late Bruce Fraser at the Creative Pro website. Better yet, check out his book "Real World Image Sharpening". I developed actions based on Fraser's approach years ago, and later moved to the Photokit Sharpener Photoshop package that his company developed. The differences, especially with film, are not subtle!</p>

 

<p>Not trying to incite a film vs digital debate here - anything but! I shoot both and find value in both. I would argue, however, that getting the most out of film in a digital workflow does require more time and different skills, knowledge and patience. These issues often impact the conclusions reached in any film/digital comparison.</p>

 

<p>Scott</p>