From Youtube: Drum Scan vs DSLR vs Epson (Via Nick Carver)

[138S]

So, you've had exactly how many high end scans made off 8x10? I think you have a profoundly different understanding of 'perfect' from everyone who is disagreeing with you.

Look, drum scans for 8x10" are usually only offered at 2000 dpi, costing a fortune. For the 8x10" size in particular the Epson offers a powerful result as it also resolves 2000dpi effective on bed. Anyway result is way better than in your X1, as with it you will need scissors to cut the negative with in two halves and later stitching the "crops".

The Epson from 8x10" offers more than 300MPix effective, which is an insane amount of image quality.

The larger the format the more the Epson excels, for 35mm it's not as good.

I think you have a profoundly different understanding of 'perfect' from everyone who is disagreeing with you.

As I'm making contact prints of 8x10 sheets I can tell you what is technical perfection, after inspecting the print with a x8 magnifier.

To me a 3m print is totally perfect if resolving around 3 lp/mm on the print, so flawless when seen at around 60cm distance.

If you resolve 7 lp/mm in a 3m print then please tell me how you do it.

 

[warden]

I believe 16 x 20 to be on small and over 20 x 24 large.. Inches.

That helps, and it was as I suspected: your small is my large and that's important to know. My Epson scanner gives results that please me when used with medium format negatives for my prints that others here might call "small".

Recently I had my printer scan several of my MF negatives (B&W) with their Flextight so it's easy for me to compare my Epson results to their Flextight. Without a doubt the Flextight is sharper and capable of higher resolution than I can achieve for large prints, and if I need to make 24 x 24" prints I'll definitely use the Flextight. But for smaller sizes the Epson is an excellent, inexpensive choice for me, and the difference between the two is small enough that it doesn't bother me.

 

[warden]

It sounds like you've had a lot of experience with the files from various scanners. Can you post pictures instead of words that show how the Epson behaves compared to a higher end scanner? I'm especially interested in your last quote below about choosing the right part of the image to compare. Can you do a comparison to show what you're saying? I'd be interested to see it.

It's not so much the resolution problems as the god-awful MTF performance - and there are high quality tests that show just how bad the performance really is. The only time an Epson might do better than a high end scanner is when the operator is so inept that the idiot-proofing of the Epson saves them.

There is no 'evidence' to show because it can only be fake. I can say this on the basis of having dealt with dozens of randomly selected V700/750/800/850 scanner outputs - none of which, no matter how much sharpening you applied, got even vaguely close to a high end scan once over 800-1000ppi. All that happened was that the noise got worse.

You clearly don't know what you are looking for in terms of optical performance and are choosing an area that hides the poor performance of the Epson in smooth tone low contrast areas in favour of busy (noise hiding) high contrast edges.

 

[Lachlan Young]

It sounds like you've had a lot of experience with the files from various scanners. Can you post pictures instead of words that show how the Epson behaves compared to a higher end scanner? I'm especially interested in your last quote below about choosing the right part of the image to compare. Can you do a comparison to show what you're saying? I'd be interested to see it.

PM me your email address & I'll send a link. I've found that forum software can do all sorts of odd things to fine detail.

 

[138S]

and the difference between the two is small enough that it doesn't bother me.

Yes, for MF the Flextight has an small advantage over the Epson, but for 4x5 and for 5x7 the Epson is superior, this is because the Flextight sensor has 8k pixels only and when scanning 4" it only can scan 2000 samples per inch, ending in effective 1800 dpi horizontal and 1600 vertical, which is a poor yield.

The Flextight shines in 35mm, where it delivers an insanely good result, but as the format increases it gets worse as the lens takes a larger scan width with only 8k pixels.

For 5x7" the 8000 samples of the X1 are spread in 5", this is 1600 samples per inch yielding around 1400 effective dpi. At this size the Epson blows away the Flextight, as the Epson conserves 2900 dpi effective in the horizontal axis and 2300 in the vertical one, still the Epson requires custom holders to do that, as there are no original 5x7" holders, anyway even on bed (with the scan guide) it also blows away the Flextight.

This is about resolving power, of course the Flextight is more capable than the Epson for extreme Velvia/Provia densities, but difference is way lower when the Epson uses Multi-Exposure.

Anyway the cheap Epson is superior to the luxurious Flextight for LF sheets, no doubt about that.

 

[PhilBurton]

PM me your email address & I'll send a link. I've found that forum software can do all sorts of odd things to fine detail.

Lachlan,

Since I own neither an Epson flatbed scanner nor a drum scanner (!!) this thread has been more or less "entertaining." However I have about 500 MF negatives and maybe 50 prints that I will need to scan at some point. (I do own a Nikon 5000 ED scanner for all my 35 mm material.) But right now, it's time for you to put up or shut up. Forum software is no worse on fine detail than any other sort of web venue. This is your opportunity to convince me not to buy or rent an Epson scanner, but instead to use a scanning service with a drum scanner.

The ball is in your court.
Phil

 

[Helge]

Carver is a nice guy who's adding to the enjoyment of using film for everyone. He's a self-learner and doing a pretty good job I think. I've watched a lot of his videos and they're enjoyable and knowledgeable. Some of you guys are awful tough on him though and his beliefs. Give the guy a break.

I have no doubt that he is a nice guy at heart. But so are most people, that's no special qualification.
I have made my feelings clear on him and his style, I won't go further into that.
He does make some basic mistakes and popularises some ideas that are not ultimately conductive to a fruitful and prosperous future for film overall.
That we are wise to counter where we see fit.

 

[138S]

He does make some basic mistakes and popularises some ideas that are not ultimately conductive to a fruitful and prosperous future for film overall.

Well, explaining a personal experience is fair, testing and using what better works for somebody is OK.

... but that video can be missleading. I'd like to mention what points (IMO) are not fine:

1) He says that the Epson is better than the drum for color negatives. This is fairly wrong, straight from the machine an MF scan of C-41 film from the drum is always slightly better than from an Epson, still the Epson file carefully edited will totally match the drum result, so it's not a bad idea scanning MF C-41 film at home with the Epson, but if the Drum scan is worse this is because a poor edition was made by the sdrum operator.

2) He says that with E-6 film colors of the drum scan are better. This is false, what he sees is the image enhancing corrections made by the drum operator. If he projects the slide then he will see that the Epson is totally matching the colors in the projection, as it is an IT8 calibrated machine, he should learn that one thing is scanning and another thing is applying wonderful presets in the edition.

3) He discards DSLR scanning. My personal tests showed me that with a reversed lens and by stitching (a lot) over 9000 effective dpi can be obtained, then colors can be calibrated or edited to have same result than with the Drum or the Epson, so it's ok if his DSLR workflow is not refined enough, but YMMV in that.

... so the good side is that he tested and compared to know what resources are working for him, but a clear disclamer should be added because YMMV.

Anyway those that had been machinegunning the Epson for a decade and a half have to be a bit disturbed when somebody prefers the Epson over a drum

In fact the Epson is a great piece of gear, but it requires a proficient handling/edition that not all home users are able to do.

 

[warden]

However I have about 500 MF negatives and maybe 50 prints that I will need to scan at some point. (I do own a Nikon 5000 ED scanner for all my 35 mm material.)

Phil, here are two images that might help. I use MF mainly, and have tried Epson and Flextight. I didn't try anything better than the Flextight as I don't take advantage of the resolution of even the Flextight for most of my work/play so there is no need.

Here's the negative (scanned by me with my Epson V700).

Note that I've left the scan fairly flat here so you can see darker details. The final file was made punchier by the staff at the printer and the prints of this image turned out to my satisfaction.

Next is the comparison image. Note that the Flextight hasn't been downsampled, I just manually zoomed out until the Flextight image was the same size as the Epson, which turned out to be 38%.

Neither image has been sharpened, but I adjusted curves to equalize the images. The Flextight is better to my eye, and of course has resolution to spare. If I needed to print very large then the Flextight is the way to go for me and I'll use it again.

But the Epson scan is 5,300 pixels per side, which is a detailed and sharp 17" print @ 300dpi, and that's sufficient for me 90% of the time.

Hope that helps with your decision.

 

[138S]

But the Epson scan is 5,300 pixels per side, which is a detailed and sharp 17" print @ 300dpi, and that's sufficient for me 90% of the time.

If you see the Flextight image it has more noise from a digital sharpening (perhaps made automaticly by the machine), in the Epson image I made a mild touch in the curve and a mild sharpening (posted) to end in a neraly perfect match:

The Epson is a great device that in the right hands is able to challenge very expensive machines !!!


The funny situation is in 4x5" and 5x7" format, where the Epson blows away the luxurious Flextight...

The Flextight is much better than the Epson for 35mm, a only bit better for MF, worse for 4x5", and very inferior in 5x7".

Additionally the Epson makes 8x10"

 

[warden]

Yes I didn't mention the sharpening but of course the Epson (or any scan) would need some before printing. I wasn't aware of sharpening in the Flextight, but I wasn't present when they scanned my negative so maybe you're right. Generally this printer applies sharpening as one of the final steps before printing.

 

[138S]

Yes I didn't mention the sharpening but of course the Epson (or any scan) would need some before printing. I wasn't aware of sharpening in the Flextight, but I wasn't present when they scanned my negative so maybe you're right. Generally this printer applies sharpening as one of the final steps before printing.

But think that Pro machines may make a really good optimization internally while the Epson delivers a raw image, for a fair comparison both the Flextight and the Epson images have to be both optimized to their best,

The flextight image cannot be sharpened much without having overshot, while the Epson requires that sharpening.

I made that matching with only two clicks, but by spending a bit more effort you get a totally perfect match. This discredits totally those that had been machinegunning the Epson, one said in this thread that the Epson only allows 2x enlargements, well... for MF it allows the same than the Flextight, if anyone has a doubt.

 

[warden]

This discredits totally those that had been machinegunning the Epson, one said in this thread that the Epson only allows 2x enlargements, well... for MF it allows the same than the Flextight, if anyone has a doubt.

I'm not interested in arguing with people. I'm just sharing my results for those like Phil that haven't had the chance to compare like I did.

And I'd say the Epson offers similar quality to the Flextight for the size I want to print, but the Flextight is much better for larger sizes than the Epson can achieve while using smaller negatives like this. I experimented by printing this image at 24"x24" with the Flextight and the result was simply fantastic. I don't think the Epson could do that as well as the Flextight did at that size.

But at smaller sizes like 15" (7x enlargement) I'd be hard pressed to find the difference on paper between the Flextight and the Epson with a MF negative.

And 4x5 or larger must be pretty amazing scans, but that's far more data than I can use.

 

[138S]

And I'd say the Epson offers similar quality to the Flextight for the size I want to print, but the Flextight is much better for larger sizes than the Epson can achieve while using smaller negatives like this.

Sorry, but for your MF negative the Epson offers the same quality than the Flextight, after optimizing both images...

I concede that for 35mm the Flextight is way superior to the Epson, but for MF I don't see much a difference, even if pixel peeping.

Still the Flextight allows to scan MF in 2 strips that later can be stitched, if doing that a superior grain depiction can be obtained.

But as said, for 4x5" and 5x7" the Flextight losses.

I'm not interested in arguing with people.

Arguing is nasty, but at the same time some missleading information has to be fighted, people having to make a decision should know the reality: Every scanner has strong and weak points, and operator job is the most important, IMO.

 

[warden]

Sorry, but for your MF negative the Epson offers the same quality than the Flextight, after optimizing both images...

I'm not going to argue the point. I did my own tests and examined the results both on paper prints and on monitors, and I'm satisfied with my conclusions. I'll continue with the wet darkroom for "very small" prints, the Epson for medium format digitized "small prints" and Flextight which I see as worth the expense for medium format digitized "large prints".

 

[138S]

I'm not interested in arguing with people. I'm just sharing my results for those like Phil that haven't had the chance to compare like I did.

And I'd say the Epson offers similar quality to the Flextight for the size I want to print, but the Flextight is much better for larger sizes than the Epson can achieve while using smaller negatives like this. I experimented by printing this image at 24"x24" with the Flextight and the result was simply fantastic. I don't think the Epson could do that as well as the Flextight did at that size.

But at smaller sizes like 15" (7x enlargement) I'd be hard pressed to find the difference on paper between the Flextight and the Epson with a MF negative.

And 4x5 or larger must be pretty amazing scans, but that's far more data than I can use.

Ok, this is a wise decision, and you know better that any other what works for you, just I was pointing that the V700 vs Flextight difference in the side by side you showed is way lower after optimizing the V700 image to its best as seen in post #110.

If difference seen in post 110 is important for you or not this is of your concern, of course, but difference to consider is the one in post 110, not the one in post 109.

 

[PhilBurton]

In fact the Epson is a great piece of gear, but it requires a proficient handling/edition that not all home users are able to do.

So, can someone with Epson scanner experience start a thread called, "Best practice for scanning with Epson flatbed scanners." Be sure to discuss the differences in different models.

 

[138S]

So, can someone with Epson scanner experience start a thread called, "Best practice for scanning with Epson flatbed scanners." Be sure to discuss the differences in different models.

See Post #87.

jawargen made a good v700 scan, Post #109, so with a simple 2pix radius 80% sharpening in Post #110 it practically matches what a high end scanner does.

No that difficut, it's about not degradating information.

 

[warden]

Ok, this is a wise decision, and you know better that any other what works for you, just I was pointing that the V700 vs Flextight difference in the side by side you showed is way lower after optimizing the V700 image to its best as seen in post #110.

If difference seen in post 110 is important for you or not this is of your concern, of course, but difference to consider is the one in post 110, not the one in post 109.

Yes, which is why (like I said earlier) for a 15" print I would stick to Epson. There is very little difference to be found at this size (7X enlargement) so there is no reason to spend for the Flextight scan (for me). The printer prefers the Flextight even for smaller images for Dmax and other reasons more relevant to them than to me.

Bear in mind though that the Epson is maxed out at 100% in that image. There is far more detail in the Flextight, which is why it's worth it to me to use the Flextight on large prints. If I'm paying professionals to make a large print the extra $30 for a 15X enlargement is definitely worth it for me.

But I hardly ever print that big anyway.

 

[138S]

The printer prefers the Flextight even for smaller images for Dmax and other reasons more relevant to them than to me.

Densities usually found in BW or Color negative film are never a challenge for the Epson. At 3.0D the Epson has no problem while those densities would be really hard to print optically with an enlarger.

With Velvia/Provia it's different, if having to recover very deep shadows then the Flextight has an advantage, the Epson with Multi-Exposure makes an amazing job, but the Flextight is superior in that situation, anyway if one usually places shadows at 3.5D then better he buys a good lightmeter.

There is far more detail in the Flextight

Ok, but's not the case of the samples you posted.

In fact the Flextight X1 resolves 2500 dpi effective for 6x6 and 6x7. For 6x4.5 with the frame cut and placed in portrait orientation the X1 resolves 3150 effective pixels. The Epson resolves 2900 Horizontal and 2300 vertical, so in average it's not that far from what the X1 does in 6x6.

My experience is that the 6300 dpi effective of the X1 (or the 6900 of the X5) in 35mm makes a huge difference over anything else, it has to be a very expensive drum scan to compete with that, but really, for 6x6 personally I don't see the difference you say, or better said, I see just a difference like in the Post #110.

The X5 is slightly better than the X1, for 6x6 it resolves 2750 effective.

Please also check post #77 https://www.photrio.com/forum/threa...on-via-nick-carver.172770/page-4#post-2253104 this test is serious and honest, also MF.

 

[Les Sarile]

[QUOTE="PhilBurton, post: 2253569, member: 88328"
Since I own neither an Epson flatbed scanner nor a drum scanner (!!) this thread has been more or less "entertaining." However I have about 500 MF negatives and maybe 50 prints that I will need to scan at some point. (I do own a Nikon 5000 ED scanner for all my 35 mm material.) . . . This is your opportunity to convince me not to buy or rent an Epson scanner, but instead to use a scanning service with a drum scanner.
Phil[/QUOTE]

I started with the 5000 when it first came out and since I only shot a few frames of 6X9 MF film, I also thought I would just get the V700 to scan those. Very big disappointment - scan times are much longer especially when ICE is turned on, Epson ICE is not as effective, the V700 cannot attain "equivalent" resolution of the Coolscan and worse is the sometimes completely wacky Epson color/contrast results compared to Nikonscan+Coolscan in full auto mode across all film type. I ended up purchasing the 9000 shortly before Nikon dropped the line.

In the absence of direct comparisons, one can be very content with working the results from the Epsons, just don't compare it to the Coolscan. In fact, you would be much better served by not making direct comparisons between the Coolscan+Nikonscan with any other desktop+software, Imacon, minilab. If the results you get from those others pleases you, then it is good enough.

 

[warden]

In fact, you would be much better served by not making direct comparisons between the Coolscan+Nikonscan with any other desktop+software, Imacon, minilab.

I thought Nikonscan was no longer available for OSX? I'll check into it so thanks for the heads up. I've been using Vuescan with my Coolscan and the Epson.

 

[138S]

In the absence of direct comparisons, one can be very content with working the results from the Epsons, just don't compare it to the Coolscan. .

The 9000 has a resolving power that is around 3650dpi effective, well over the Epson, but don't think that a higher resolution is always an advantage, most of the times excessive resolving power has no advantage because the limitation comes from the film.

Say that you scan CMS 20 film shot on tripod with an amazing lens at its optimal aperture and with perfect focus: in that case the 9000 will make a clear difference over the Epson,

...but in real photography those conditions are an exception, say that you shot Portra 160, or handheld or taking advantage of DOF, in that case the Epson will equal the 9000 simply because the Epson is outresolving yet what film has.

If you review Post #77 (https://www.photrio.com/forum/threa...on-via-nick-carver.172770/page-4#post-2253104) the Epson matches two Creos and an Scanmate 11000 drum, while it is a good shot the Portra 160 film simply is not sharp enough to challenge the Epson, result is equal not because the Epson is as good as the 11000 drum, but because then Epson is able to extract most of information on the medium, so a better scanning performance won't make a difference.

 

[Les Sarile]

The 9000 has a resolving power that is around 3650dpi effective, well over the Epson, but don't think that a higher resolution is always an advantage, most of the times excessive resolving power has no advantage because the limitation comes from the film.

Say that you scan CMS 20 film shot on tripod with an amazing lens at its optimal aperture and with perfect focus: in that case the 9000 will make a clear difference over the Epson,

...but in real photography those conditions are an exception, say that you shot Portra 160, or handheld or taking advantage of DOF, in that case the Epson will equal the 9000 simply because the Epson is outresolving yet what film has.

If you review Post #77 (https://www.photrio.com/forum/threa...on-via-nick-carver.172770/page-4#post-2253104) the Epson matches two Creos and an Scanmate 11000 drum, while it is a good shot the Portra 160 film simply is not sharp enough to challenge the Epson, result is equal not because the Epson is as good as the 11000 drum, but because then Epson is able to extract most of information on the medium, so a better scanning performance won't make a difference.

I know exactly what you are saying by actual results below.

Controlled test using Fuji Velvia and scanned by Epson V700 and Coolscan 5000.

Full res version of V700 res test -> http://www.fototime.com/11F59FA46FF9497/orig.jpg

Full res version of Coolscan 5000 res test -> http://www.fototime.com/02BB797801DCA89/orig.jpg

"Real world" test using Fuji 100 color negative and scanned by Epson V700 and Coolscan 5000.

Full res version of Fuji 100 scanned by Epson V700 and Coolscan 5000 http://www.fototime.com/535A4899F477014/orig.jpg

Of course you cannot hope to conduct a duplicatable "real world" test as no one can ever know what the lowest common denominator is - shaky hands, bad eyesight, etc.

Probably more important to me is good workflow and color/contrast results in full auto mode. I've scanned tens of thousands of various frames of all types of films with the Coolscan+Nikonscan and only hundreds with the various other scanners. I've never had a bad result from the Coolscan but have encountered many just awful from the others such as this one from Kodak Ektar 100 below.

 

[Les Sarile]

I thought Nikonscan was no longer available for OSX? I'll check into it so thanks for the heads up. I've been using Vuescan with my Coolscan and the Epson.

I keep two Windows Vista machines to run Nikonscan.

I only tried Vuescan early on and conducted this test with Kodak 160VC that it has a specific film profile for. I also used the various modifiers from Vuescan compared to a fully automatic scan from Coolscan+Nikonscan.