5x7 vs. 4x5 scanning help

[mikossant]

I am really wanting to get into large format photography and want to know if 5x7 scanned on an epson v750 will display more detail than a 4x5 scanned by a top of the line drum scanner when printed at 40x50.
Thank you

[Don Hutton]

I haven't done a direct comparison with 4x5 and 5x7, but I have done a B&W comparison with 4x5 and 8x10 - drum scan of the 4x5 at 4000DPI on a Howtek 4500 and the 8x10 scanned on an Epson 4990 at 2400DPI (I believe that the V750 is maybe just slightly better). Actual resolution was not that different - however, the drum scan delivers far more microcontrast and overall a much better file - and that's on a 4x5 to 8x10 comparison. I've made 16x20 prints from scans of 8x10 negs on the Epson and then made the same 16x20 prints from a drum scan and there are very visible differences in the final prints - mostly related to microcontrast. In color, I would expect these differences to be even greater, especially from chromes.

Most people I have spoken to about the relative merits of consumer flatbed vs drum scans of large format film presume that the differences will be all about resolution - that's only a small part of it - microcontrast, dynamic range and lack of any noise are probably bigger factors to consider. If you have any doubts, I'd highly recommend that you spring for a drum scan from a reputable operator (like Lenny Eiger) and make up your own mind. I'd suggest it won't take long.

[Ron Marshall]

5x7 is only 25% wider than 4x5, so not much of an advantage in terms of enlargement potential.

It really depends on your personal standard. Some people are happy with a 20x24 from a consumer flatbed scan of 4x5, others will only go as high as 11x14.

The largest I would be happy with from a consumer scan of a 5x7 would be either a 16x20 or a 20x24, depending on the subject, the importance of detail, and viewing distance.

[Ken Lee]

If detail and huge prints is of primary concern, don't fiddle around with 4x5 or 5x7. Consider 8x10, because of some simple arithmetic:

From 8x10, a 40x50 print is only an enlargement of 5x, so with a high-end lens that delivers 60 l/mm, you will end up with 12 lines/mm - which is still critically sharp. Starting out with smaller film, you have to enlarge by 8x or 10x, and you have reached the critical point of 8 lines per mm, where things look mushy - and that's presuming you have a perfect scanner. But you don't, even with a drum scanner. With a consumer flatbed, you can't get anywhere near that. Maybe 5x. If you need to crop your image, then, as they say... forgetaboutit.

With an 8x10 negative, and a consumer flatbed scanner, you can get away with a 5x enlargement. Of course, you will need a rather hefty computer to perform corrections on the file, especially if it's in 48-bit color ;)

That being said, what you may deem acceptable, another may not. It's best to make a test for yourself, with your equipment and your criteria. The money and exasperation you save, will be... your own.

[Lenny Eiger]

5x7 is only 25% wider than 4x5, so not much of an advantage in terms of enlargement potential.

I think this math is a little faulty. I used to think that way as well. However, 4x5=20 and 5x7=35, it's actually almost double the area (1.75 actually). Quite a bit more than 25%.

As I have said too many times, I have been very disappointed with the Epson scanners in comparison. I am sure there are some that can clean it up better than others. However, what about a third possibility - a great drum scan from a 5x7? ;-)

Lenny

[Don Hutton]

From 8x10, a 40x50 print is only an enlargement of 5x, so with a high-end lens that delivers 60 l/mm, you will end up with 12 lines/mm - which is still critically sharp. And if you're working at typical 8x10 apertures to maintain necessary depth of field - i.e. f32 and smaller, you will never even get 50LP/mm....

[Walter Calahan]

Personally, I'd worry more about the content of the image then whether 4x5 or 5x7 has the technical specs. Content will carry an image more than lines per millimeter, etc.

[Don Hutton]

Personally, I'd worry more about the content of the image then whether 4x5 or 5x7 has the technical specs. Content will carry an image more than lines per millimeter, etc.I'd agree 100% with that Walter, but the question is really a technical one, where I presume, there is a assumption that the content is going to make a 40x50inch print worthwhile anyway...

[audioexcels]

I'd agree 100% with that Walter, but the question is really a technical one, where I presume, there is a assumption that the content is going to make a 40x50inch print worthwhile anyway...

Don,

Are you doing wet or dry scans of the larger sheet film on the scanner bed and would you say that your expertise with the scanner is to the extent that you are getting the absolute most you can out of it?

Makes one have to consider why to bother with anything larger than 4X5 other than to print out of camera negs or if the person simply doesn't like 4X5 format (they can shoot 5X7 as an option for the OP to have the different ratio) glass as well as a 5X7/8X10 piece of GG.

In other words, if one cannot get the results even out of a Howtek 4500 which is a nice scanner, but not as nice as some of the better scanners that can be used to to (drum/pro-flatbed) scan the film, but cannot even match them with an 8X10 neg on a consumer flatbed, the 8X10 or larger is strictly for out of camera silver prints only since most will not be enlarging 40X50" prints, let alone anything larger than 20X24 as a maximum size.

[Don Hutton]

Don,

Are you doing wet or dry scans of the larger sheet film on the scanner bed and would you say that your expertise with the scanner is to the extent that you are getting the absolute most you can out of it?
I don't presume that I can get the "absolute most possible out of the scanner". I have about 8 years of experience scanning and have used various flatbeds for scanning LF film for about 6 years. I don't believe my scanning experience with the Epson is a severely limiting factor in the final file. For the tests, I used a dry scan as I have no way of conveniently doing a wet scan of an 8x10 on the 4990 - IMO, the bed size is too limiting and you are looking for a scanner full of scanning fluid. Personally, having done a fair bit of wet scanning of 4x5 negs on the same scanner, I don't believe wet scans on the 4990 are a "dramatic" improvement - whatever advantages they offer, are, IMO, pretty small.


Makes one have to consider why to bother with anything larger than 4X5 other than to print out of camera negs or if the person simply doesn't like 4X5 format (they can shoot 5X7 as an option for the OP to have the different ratio) glass as well as a 5X7/8X10 piece of GG. I wondered about this - so a couple of weeks ago, I shot an identical scene with very similar lenses on both 4x5 (with a Fujinon A 180) and on 8x10 (with a Fujinon A 360); color neg (Fuji Pro 160S) and B&W (Tmax100) and drum scanned the negs and made prints, and crops from "very large" prints. On a 24x30 some differences become visible - subtle, but there. On a 40x50, the advantages of the 8x10 become quite apparent - finer details, tonal transitions etc are just better. Standards of print quality are highly subjective and it's almost pointless trying to ascertain what others perceptions of quality are - you have to define these things for yourself.


In other words, if one cannot get the results even out of a Howtek 4500 which is a nice scanner, but not as nice as some of the better scanners that can be used to to (drum/pro-flatbed) scan the film, but cannot even match them with an 8X10 neg on a consumer flatbed, the 8X10 or larger is strictly for out of camera silver prints only since most will not be enlarging 40X50" prints, let alone anything larger than 20X24 as a maximum size.I'm not sure what you mean here - my results suggested that the 4x5 scanned on the Howtek was way better than an 8x10 scanned on the Epson 4990. And FWIW, I think the Howtek is an excellent scanner - I'd choose it over any flatbed at any price and I'm not sure there are not many drum scanners out there I would swap it for except perhaps an Aztek Premiere. I'd take it over a Tango anytime. I've yet to see a flatbed scan which is as good as the best drum scans I've seen and I appreciate the control of the variable aperture on a drum scanner which is not available on any flatbed.

[audioexcels]

[QUOTE=Don Hutton;364421]

Thanks Don.

Op's question has had some mixed answers, but I know you would have a massive wealth of experience since you have both scanner types sitting side by side. I don't think there would be any improvement on the V750 than with the 4990 other than some little bit I guess. Regardless, and for the OP, it has been said numerous times that while things are subjective, a drum scan and even a pro-flatbed scan of a 4X5 sheet will yield far superior results past the threshold of the consumer (V750) scanner, with the threshold considered to be a 3X enlargement out of the machine. If doing such large prints, I would consider the 5X7 only to get that bit more surface area/resolution from the film to give better looking results at 40X50...as you mentioned, as soon as things get to that 24X30 barrier or so, the 8X10 starts to look more impressive detail wise. Some would maybe even claim that they can see a great difference at even 16X20 size between a drum scanned 45 and 810 piece of film.

Don't want to mention his name, but Ted had done many tests of pro-flatbeds and drum scanners (including the Howtek 8000), and found no differences in terms of the clean output from each of these scanner types. Some would argue with him about drum scans being superior, but he simply let things be and say that the pro-flatbeds were equally good, though I'm fairly certain he said select few were actually superior to even the best drum scanners.

Not to put a drum vs. pro-flatbed war onto the table since it's off-topic, but is to say that he and many others that have been involved with scanning for many years feel similar about particular pro-flatbeds.

[Don Hutton]

Scanners are operator dependant. I've seen some pretty horrible scans from supposedly highly experienced operators. Like I said, one needs to develop one's own idea of what works and what doesn't, and doing one's own testing reveals far more relevant answers to most questions. On the pros and cons of formats, I'd suggest going out and shooting a particular format for a year or two - you'd be amazed how quickly you can come to perfectly logical conclusions and about the pros and cons.

[Ron Marshall]

I think this math is a little faulty. I used to think that way as well. However, 4x5=20 and 5x7=35, it's actually almost double the area (1.75 actually). Quite a bit more than 25%.


Lenny

Lenny, I mentioned width, not area. A 40 inch wide print from a 4x5 neg. is a ten times linear enlargement; from a 5x7 it is an eight times enlargement, ie. 25% greater enlargement. The width is the limiting dimension.

[D. Bryant]

5x7 is only 25% wider than 4x5, so not much of an advantage in terms of enlargement potential.


Ron,

In my experience there is a very noticeable difference between 4x5 and 5x7 when scanned or enlarged optically.

Don Bryant

[Ken Lee]

"On a 40x50, the advantages of the 8x10 become quite apparent - finer details, tonal transitions etc are just better."

This is exactly what we would expect. By this degree of enlargement from 4x5, we have reached the limit for even the best "taking" lens.

A 40x50 print, from a 16x20 negative made with a good lens like a 600mm Fujinon A, should do even better. With only a 2.5 X enlargement, we get to keep even more of the original data given by the taking lens.

Questions ? Ask your friendly neighborhood contact printer. ;)

[Don Hutton]

"On a 40x50, the advantages of the 8x10 become quite apparent - finer details, tonal transitions etc are just better."

This is exactly what we would expect. By this degree of enlargement from 4x5, we have reached the limit for even the best "taking" lens.

A 40x50 print, from a 16x20 negative made with a good lens like a 600mm Fujinon A, should do even better. With only a 2.5 X enlargement, we get to keep even more of the original data given by the taking lens.

Questions ? Ask your friendly neighborhood contact printer. ;)
Ken

There are diminishing returns as the format gets bigger and bigger - basically, you are forced to operate the lens a long way from optimum - essentially in severley diffraction limited territory and you will get to a point where the smaller format at a bigger enlargement will produce a better final print than a bigger negative at less enlargement. While the Fujinon 600 A may be capable of putting 45 LP/MM onto film under test conditions, at f45 or smaller under actual 16x20 shooting conditions, it may only make 20LP/MM etc... I don't know where this point is reached because I haven't needed to find out and it obviously depends on the subject, but I do know, it's not on 8x10! I have shot a fair bit of 11x14 and 12x20 and the step up in quality is not linear because of depth of field/diffraction limitations. A couple of times on 12x20, I simply had no way of making a composition work without areas which were not going to be critically sharp. I've never had that experience on 4x5.

[Brian Ellis]

I haven't done a direct comparison with 4x5 and 5x7, but I have done a B&W comparison with 4x5 and 8x10 - drum scan of the 4x5 at 4000DPI on a Howtek 4500 and the 8x10 scanned on an Epson 4990 at 2400DPI (I believe that the V750 is maybe just slightly better). Actual resolution was not that different - however, the drum scan delivers far more microcontrast and overall a much better file - and that's on a 4x5 to 8x10 comparison. I've made 16x20 prints from scans of 8x10 negs on the Epson and then made the same 16x20 prints from a drum scan and there are very visible differences in the final prints - mostly related to microcontrast. In color, I would expect these differences to be even greater, especially from chromes.

Most people I have spoken to about the relative merits of consumer flatbed vs drum scans of large format film presume that the differences will be all about resolution - that's only a small part of it - microcontrast, dynamic range and lack of any noise are probably bigger factors to consider. If you have any doubts, I'd highly recommend that you spring for a drum scan from a reputable operator (like Lenny Eiger) and make up your own mind. I'd suggest it won't take long.

What's "microcontrast?" Is that the digital age term for local contrast or is it something different?

[Don Hutton]

What's "microcontrast?" Is that the digital age term for local contrast or is it something different?
The same I believe...

[Ron Marshall]

Ron,

In my experience there is a very noticeable difference between 4x5 and 5x7 when scanned or enlarged optically.

Don Bryant

I'm not saying that there is not a noticable difference.

What the OP asked was would the difference between 4x5 and 5x7 compensate for the difference between a consumer flatbed scan and a drumscan. I don't think it would, because in the smallest dimension 5x7 is only 25% larger than 4x5!

Do you think it would be enough to compensate?

[Lenny Eiger]

I'm not saying that there is not a noticable difference.

What the OP asked was would the difference between 4x5 and 5x7 compensate for the difference between a consumer flatbed scan and a drumscan. I don't think it would, because in the smallest dimension 5x7 is only 25% larger than 4x5!

Do you think it would be enough to compensate?

My answer would be no. It wouldn't.

However, what is the OP interested in? What kind of print does he want to make? Is the aesthetic about critical sharpness - or overall sharpness? Tonality? Color? When anyone asks a question like this they ought to post a web site showing the type of images they want to make... Then it would be much easier to speak directly to the point.

Lenny

[Ken Lee]

While the Fujinon 600 A may be capable of putting 45 LP/MM onto film under test conditions, at f45 or smaller under actual 16x20 shooting conditions, it may only make 20LP/MM etc... I don't know where this point is reached because I haven't needed to find out and it obviously depends on the subject, but I do know, it's not on 8x10!

Thanks for that excellent clarification. I forgot that the high numbers we see in lens testing, are seen in the center of the circle of coverage.

A 16x20 contact print will still dazzle, because 20 lp/mm (or 20 cycles/mm as Emmanuel Bigler has called them so wisely) and no grain at all, really delights the eye.

So I stand corrected - and now I don't have to run out and grab that 16x20 camera :cool:

[D. Bryant]

I'm not saying that there is not a noticable difference.

What the OP asked was would the difference between 4x5 and 5x7 compensate for the difference between a consumer flatbed scan and a drumscan. I don't think it would, because in the smallest dimension 5x7 is only 25% larger than 4x5!

Do you think it would be enough to compensate?
Oh in that case no as Lenny said.

Don

[audioexcels]

5X7 is very close to 2X the size of 4X5. There may be a 25% difference in one direction, but given the person would want to have even a 40X50 drum scanned print, the 5X7 would smoke the 4X5 by basic laws of physics/math...well, given same operator, same well done negatives, etc.

I would bypass 4X5 since I don't think it can produce a solid 40X50" print unless someone has some wonderful tricks up their sleeves (i.e. getting the precise exposure with the precise amount of information on the sheet, then drum scanned and fully controlled by the best operator out there, and then printed by the best printing person out there).

40X50" just seems like a huge stretch for "any" person to manage a print that most would be happy with. Viewing from feet away, sure, it can be done just fine, but I hate looking up close and seeing garbage anywhere on the print.

But 5X7, again, has close to 2X the sheet area, and a well executed shot/drum scan/etc. should be 2X easier to get not only a 40X50" print, but a much better looking one.

Just my humble opinion for whatever it is likely not worth...:D