Sharpness diff between 645 and 4x5

This follows on some issues that came up on an earlier post of mine "Help on

Lens Choice" where the issue of sharpness in 8ftx10ft enlargements was

being discussed.

 

I wish to exhibit enlargements. I am planning on acquiring a 4x5 for a project

so that, theoretically, the enlargements will be better than from my Contax

645. But then that is what I think.

 

But is this true?

 

If I photograph the same scene on the Conatx 645 (vacuum back, and the

super 120 Makro) and on a 4x5 (Planning on the Toyo CF) with a good lens,

and make enlargements the same size, how will the results compare?

 

Since I plan on doing landscapes primarily, will the 4x5 make an impact

(compared to the 645)? Thus the 'movement' advantage is minimal?

 

What got me worried is this answer from a gent at Mamiya on the Mac User

Forum. The question is: "Will I see a noticeable improvement in image quality

with a 5x4 over 66 or 67 format"

 

Response from: Paul D'Ambrosio - 11:13am Jul 16, 2001 EST (1.)

Mamiya America Corporation

 

"Quentin, You should not really notice a difference between an image shot

with a 6x7cm camera as opposed to an image shot with a 4x5" camera. We

have seen images taken with a 6.4x5 cameras that have been enlarged to

5'x4' and they look extremely sharp. The main benefits are the movements.

(swings, tilts ect) that you get with a 4x5 view camera. Every photographer

has his/her idea of what camera would be best for him or her. Fortunately we

have many models for you to choose from. I suggest that you visit our site at

www.toyoview.com. There you will find all the specs and information that you

need to help you make the right choice for yourself."

I don't know what that guy is smoking but you should notice a big difference if you are planning to enlarge to 8 feet x 10 feet. Especially from a 645 neg.

Not even close. Wonder what the guy's response would be if Mamiya made LF cameras ...

I did a comparison a while back and a 6x7 with a super sharp lens carries close the detail as a super sharp 4x5 slide, but the 4x5 slide enlargement will be much cleaner and smoother.

 

IMO 645 to 4x5 is no comparison if everything is perfect.

 

Personally my rule of thumb for a super clean enlargement size should be around 10-12x max.

 

Pushing it up with a super sharp lens and a lot of work I have gone up to 16X. You can always go higher still but at the loss of fine detail in print.

 

IE a good print rez is 4 lp/mm in print for a 20" view distance. Most camera systems, no matter how sharp are going to top out at 40-50 lp/mm in average contrast, due to the film, technique etc. 40 lp/mm at the film plane works out to a 10X enlargement in print at 4 lp/mm.

 

So for 645 a 10x enlargement would be 16 x 22. Pushing to a 16x enlargement would yield a 26 x 36 print.

 

But look at a 4x5 foot print as an example.

 

That would be a 29x enlargement for 645

 

You can do it of course but your print rez will be down. Even if you were lucky enough to squeak out 60 lp/mm at the film plane you would still be only printing at 2 lp/mm. If you image was at 45 lp/mm at the film plane you would only get 1.5 lp/mm in print. 8 lp/mm is considered critically sharp in print.

 

So for a 4x5 foot huge print

 

645 = 29x enlargement

670 = 21x enlargement (still to much for me)

4x5 = 12.8x enlargement (perfect if you have a super sharp lens but not much room for cropping) 5x7 would probably be better.

 

8x10 = 6x

 

I saw a Crewdson print a while back that was amazing. It was huge. I think it was something like 5' x 7'. I believe he uses an 8x10 camera of some sort.

 

If you do buy a 4x5 camera buy the sharpest lenses you can find. An average lens and a not so stiff camera might put you back at 6x7 territory. Also a cheap scanner will do the same. Its like dropping a format compared to a drum scan.

"You should not really notice a difference between an image shot with a 6x7cm camera as opposed to an image shot with a 4x5" camera."

 

As a general statement this is a bunch of rubbish, especially when we're talking 4x5' prints! I own a Mamiya 7, which arguably has the sharpest MF lenses available, and I've found that 4x5 starts to outperform it at print sizes around 16x20" (when drum scanned and digitally printed). When I tested the Contax 645 (admittedly I did not use the vacuum back) some time back, 16x20" prints were even softer than those from my Mamiya.

 

Of course, if you are shooting non-resolution intensive subjects (portraits, etc.) then format sensitivity is lessened. But this is the first time I've heard anyone say 645 produces equivalent poster-sized prints to 4x5. Check out this comparison between 6x7 and 4x5 (compares two drum-scanned chromes):

 

http://www.westcoastimaging.com/wci/images/formatcompare.jpg

Mamiya America Corporation (MAC) -- the company Paul D'Ambrosio works for --

markets Toyo large format cameras, LPA Design Pocketwizards, Profoto lighting and

someother lines as well as Mamiya cameras. Paul knows what he is talking about, but

 

Will you see a difference between 4x5 and 6x7cm images printed quite large? yes, but

depending on how the prints are made as well as the photographer's technique the

differences could be significant or minimal. Iexpect you easily see a difference between

4x5 and 6x4.5 format images of the same subject shot on the same film type, shot in the

same light when both are printed to the same very large print size. I'vedone a couple of

24x30 printed portfolios where the origianl images are a mix of 4x5 and 6x7cm. The

"quality" differences in the prints ofthe same type of subjects are pretty much subliminal

butthen again excellent technique was used (cameras on tripod, cable releases, mirror

locked up o nthe 6x7cm camera). if you are ever in Houston ,etexas go Hermann-

Memorial Hospital in the Texas Medical Center and check out the prints for yourself -they

are throughout the first couple of floors of the hospital and cafeteria.

 

The real question is: which type of camera is most appropriate for the photographs Milan

will be making? there are types of work where the more deliberate pace of workign with a

4x5 is not compatible with the nature of the subject matter, and vice-versa.

 

Never discount the ability to use view camera movements to make a large difference in the

aesthetic quality of an image, even with landscapes. a 4x5 camera with movements gives

the ability to shape the image exactly the way you want it to look without compromise .

The trade off for this precision is speed and economics of operation.

 

My thinking is that he'll be better off with a 6x7 camera, maybe a Pentax 67II than a 6x4.5

format camera. From the outside it looks like the best match of tools to project: large

negative, high quality lenses, very portable, lower film and processing costs.

well, 4x5 is about 10cmx12cm, for 120 cm^2 surface area. A 6cmx7cm is 42 cm^2. So, assuming equal sharpness of image, you would record about three times more detail from the 4x5.

 

One of my local camera stores has a variety of vertical banners, the sources of which range from a D70 to 4x5. If you're standing twenty feet away, you can tell the digitals from the medium-and large-formats right away. If you're standing ten feet away, you can tell the difference between the medium- and large-formats right away. If you stand just one or two feet away, you can see that the images from 4x5 aren't as sharp as you thought they were from five or ten feet away.

 

steve

I should have mentioned in the last message, the banners are roughly 10 feet tall.

<I>One of my local camera stores has a variety of vertical banners, the sources of which

range from a D70 to 4x5. If you're standing twenty feet away, you can tell the digitals from

the medium-and large-formats right away. </I><P>given that the d70 is a 6mp camera it

would be interesting to see how an image from a d2x or EOS 1ds mk.2 woud compare.

"Quentin, You should not really notice a difference between an image shot with a 6x7cm camera as opposed to an image shot with a 4x5" camera."

 

If one assumes that all other things are basically equal except the size of the negative, and that the print in question is larger than 11x14, this guy is wrong. I've enlarged hundreds of 4x5 negatives and hundreds of 6x7 negatives over the years. Same paper, same enlarger, same film, same enlarger and brand of enlarger lens, same everything except of course the equipment (Pentax 67 system, various 4x5 cameras and lenses). I've looked closely at many of the prints from each system. Most of the time I can see no meaningful difference between 11x14 and smaller prints from the two types of systems. I've put 11x14 prints from both systems side by side and asked others if they can tell which was made with which system and they can't consisently do so. But with 16x20 prints(the largest I've ever made) there's a noticeable difference in tonal gradation and often in detail between the two. The difference is sometimes subtle and might not be noticed if you weren't looking for it but it's always there even with 16x20 prints. The difference would only get greater as the prints get larger. Clyde Butcher doesn't use an 8x10 camera for his huge prints because he enjoys carrying the 8x10 around in the Florida swamps.

You won't notice any difference. Enlarging either to 8x10 FEET will give such a fuzzy (and grainy) print that it can't be viewed close enough to see much detail, anyhow. Here is an example of a viewable print made from a 22MP back on a 645 camera.<div></div>

Detail.<div></div>

It might help if you come at the problem from a different direction. Think of it as degree of enlargement.

 

Just about any film will make a resonably sharp print at enlargements up to about 10x. For 35mm, that's (24mm)/(25.4 mm/in)(10)= 9.5in, or a 9.5 x 14 inch print. Slower films can take a somewhat greater degree of enlargement before "falling apart" by showing grain and loosing sharpness. Often B&W negatives can handle more enlargement than color, because of the increased number of layers in a color emulsion. I've found, for example, that 12x is about the limit for 160PortraVC. Beyond that the print begins to have that "colored confetti" look in areas of smooth tone like skies.

 

If you really want an 8x10 foot image to withstand close examination and be sharp (nose sharp), clearly you'll need a fairly big negative:

 

(8 feet)(12 in/ft)/(10x enlargement) = 9.6 inches, or a 9.6 x 12 in film. You can stretch that a bit and go for 8x10 film which would be a 12x enlargement obviously.

 

So, if 8x10 ft prints were my goal, I'd skip the smaller formats and go directly to 8x10. If you go to 4x5 instead, you'll find that you are making a 24x enlargement, which will definitely show grain and the breakdown of sharpness no matter what film you use. If you can control viewing distance, or you don't really care if the print is "nose sharp" than this should be fine. The compromise point of a 5x7 negative might be a good spot for you also.

 

Just something to think about. Unfortunately, I don't have anything to sell or any company connections to take advantage of. The only thing I'm pushing is logic. And maybe not much of that ;-)

Milan, I have seen the 11x14 prints of Brian Ellis. He is an expert printer and no, you cannot tell which camera he used. He even has a 35mm that from 4 feet you don't know which is which. At that size your 645 would be fine. When you are talking huge prints given the same film, 4x5 would be way better. Just look at the enlargement factors involved here. For me 16x20 and up would require a 4x5.

A lot of fire here! thanks for your response guys... very insightful and

educative.

 

As Ellis Vener points out, Mamiya makes the Toyo brand of LF cameras, and

there is zero bias in their employees answer.

 

I must add... I plan to shoot TP, scan, and get digital prints for the said

exhibition.

 

It seems that differences will be evident only in 'large' sizes.

 

Also, the viewing distance will play a critical role in judging the difference. (let

us say, the viewing distance is = print diagonal).

 

I accept that camera movements can greatly enhance the aesthetic quality of

landscapes taken on a 4x5, but in real terms, all else remaining the same, is

there a perceptible difference?

 

Thanks

 

Milan

If this was lumber.net would one think a eight foot #2 pine board 1x4 was as stiff as a 2x8 board? <BR><BR>The cross sectional area of a 645 frame is way less than a 4x5 negative. <BR><BR>If 645 is viewed as being the same as 4x5; then why not say it is the same as 35mm, APS, 110 too? <BR><BR>Using vacuum cleaner marketing lingo why not say that a minox is just as good as a 4x5 camera. :)<BR><BR>One of our workers once said his Nikon was "just as good as out process camera" in sharpness for shooting E sized drawings. The process camera can shoot a 24x36 inch negative; about 864 times the area!

I think the key item to remember here is Milan will be shooting primarily landscapes. This is traditionally one of the more resolution-sensitive lines of photography, where increasing format size very quickly becomes a big deal as print sizes increase.

 

People photography, photojournalism, and other subject material are far less sensitive to resolution, and even benefit from a grainy, slightly diffuse look. I have seen some wonderful poster-sized C-prints (non-digital) made from 6x6 and 6x7 film cameras, which appear quite sharp to the typical viewer. But landscape photography is just a different animal. All else being equal, high-resolution detail visibly enhances the image and helps draw the viewer into the photograph.

 

With typical landscape subject matter, using impeccable technique in either case, 4x5 will clearly produce a visibly sharper, higher resolution poster-sized print versus any 645 film camera. Period.

 

I agree that even a 4x5 negative enlarged to this extent will still yield a fairly fuzzy print. I would shoot 8x10 if it were up to me.

>Also, the viewing distance will play a critical role in judging the difference. (let us say, the viewing distance is = print diagonal).

 

I would take this approach.

 

In print

 

4 lp/mm - 20" view distance (204 dpi on a lightjet)

6 lp/mm - +- 15" VD (304 dpi on a lightjet)

8 lp/mm - 8-10" or too damn close, and plus there are not a whole lot of printers that can print at that level. I think I read somewhere the old lightjets could go up to 400 dpi (8 lp/mm) Maybe Lambda prints.

 

For 2 lp/mm I used to have a formula/info for the distance but I cant find it now. If I can find the info I will drop it in.

 

If you dont want to guess, its probably best to start with

 

1. what size you want to print and

2. what view distance you want.

 

That will eliminate all the guess work of what format you should use.

 

So as above you can figure it out from there. Another calc similar to above, 60 lp/mm at the film plane (contax) and printing at 6 lp/mm would be a 10x enlargement

 

Personally I want to print as big as a wall (crewdson size) and be able to look at it like a 8x10 enlargement from fairly close so that locks me into 4x5 and 8x10 and a sharp lens. It would also help to have a vacuum back.

 

> I accept that camera movements can greatly enhance the aesthetic quality of landscapes taken on a 4x5, but in real terms, all else remaining the same, is there a perceptible difference?

 

Yes, but your technique has to be good, and it takes a super sharp LF lens. If you have an average lens, film not so flat etc, you could loose the advantage of 4x5 over 6x7.

 

To me your question boils down to enlargement size vs. view distance.

Troy,

Can we expand on this a little? This statement really hits home with me after my first test with LF

 

>>Yes, but your technique has to be good, and it takes a super sharp LF lens. If you have an average lens, film not so flat etc, you could loose the advantage of 4x5 over 6x7.

 

By technique I assume you mean 1) ability to focus correctly, 2) mastery of the cameras movements to aid to this end, 3) setting widest acceptable aperture to avoid difraction, 4) locking down all the movements, 5) weighting down the tripod, 6) using a sturdy tripod to begin with, 7) using ultra fine grain film, and what else might I be missing in this list ... oh sharp lenses, any comments on a Nikkor 135mm W f5.6?

 

Thanks in advance

Joe

My two cents worth.

 

I shoot 35mm to 8 x 10 and in 8x from 35mm with the same subject and 8 x 10, the "sharpness" is little different, but the TEXTURE and tonality are immediately obvious.

 

It is not so much the grain but what is between the grain, or clumps that make the difference. With a larger neg, there are more clumps and more information per square unit of print area, thus the texture and tonality become the obvious differences.

 

Like what the old european orchestra conductor said: "It not zuh nuts vat counts, but vats betveen zuh nuts".

 

cheers

This is long and are just my opinions so...

 

> Troy, Can we expand on this a little?

 

> you could loose the advantage of 4x5 over 6x7.

 

> By technique I assume you mean

 

I really mean the entire ball of wax.

 

1) ability to focus correctly,

 

Yes, and possibly dialing in your camera, shimming GG, etc.

 

2) mastery of the cameras movements to aid to this end,

 

Less so just for total resolution. For absolute resolution, its probably more important to have a perfectly aligned squared camera and perfectly flat film. This guy has the idea. I think he designed some sort of micro measurement system to make sure the standards were square.

 

http://www.cliffordross.com/R1/R1-project.html

 

Personally I think tilt is slightly overrated, but I do use it when I can, although I constantly find myself in situations where I really cant use it, like where there are trees in the foreground. Also with a tight lens using extensive movements you might be moving into the edge of the image circle that might not be as sharp.

 

3) setting widest acceptable aperture to avoid difraction,

 

That is true up to a point, but all lenses have a sweet spot. With that said after a run at 8x10 I finally realized that you have to shoot a very small aperture a lot of time, like even 1/64 just because of the lens lengths involved.

 

It can be a real PIA.

 

4) locking down all the movements,

 

Stiffer the better.

 

5) weighting down the tripod,

 

That helps especially in the wind.

 

6) using a sturdy tripod to begin with,

 

That probably one of the most important things IMO, but there is a tradeoff when hiking, like afterwards you are too tired to even set up one more shot. I went on a trip last year in the eastern mountains and burned out several times with a 35# pack. I bought a lighter 4x5 setup. Not as stiff though.

 

7) using ultra fine grain film,

 

Yes, that is what i prefer. E100G scans very clean.

 

--------

 

IMO, All in all, after doing a lot of research it comes down to a system approach.

 

The entire camera system, lens, body, back, film, scanner, enlarger all together will determine your end results. IE I imagine a lot of people think that with a 60 lp/mm film and a 60 lp/mm lens your end results will be 60 lp/mm on film, if everything is perfect, but its not really so. Its based on a rule of system resolution and in that case it works out to around 42 lp/mm at the film plane.

 

This formula below was based on a kodak theory that you need a lens with 3x the rez to take advantage of the film rez. Dont know how accurate it is but it seems close.

 

(1/R)2 = (1/R)2 + (1/R)2

 

That works out to about 71% if the film rez matches the lens rez.

 

For 80 and 80 you would end up with around 56 lp/mm, but there are just not that many 80 lp/mm LF lenses around. An example like that might be a high contrast situation, shooting velvia, techpan or E100G, stiff camera, massive tripod with a vacuum or sticky Sinar film holder, and a new MC mega $ Scheider lens.

 

So that would get you to around 56 lp/mm, but then add in the scanner. Scan on an Epson and you can drop that to 30-35 lp/mm so you almost lost half your rez. That is really like dropping a format compared to a drum scan IMO. Same if your LF lens is 30-40 compared to a Mamiya 7 type 6x7 lens.

 

Back to the 56 rez on film. On a drum scanner you would need to over scan it a bit. 60 lp/mm works out to about 3000 dpi if every row of pixels represents a line, but really you need to overscan to see it so you would want to scan it at 4000 dpi. Scanning a 4x5 at 4000 dpi and 48 bit tiff creates an almost impossible file size to work on unless you have a super computer. Seems like its around 1.3 GB.

 

Also at 4000 dpi you are resolving a lot more grain. At 2000 dpi E100G is so smooth on a wet drum scan it is almost identical to digital. Very close and very nice and its pretty much ready to go straight off the scanner.

 

In that case, if I were shooting a specific subject and I knew I wanted to print it huge, I would probably rather shoot it with 8x10 and scan it at 2000 dpi and end up with a cleaner file. That is assuming that you could even set up an 8x10 shot and get the DOF you need. 4x5 is a lot easier to shoot.

 

All that said above is very A.R., but it works fairly well for judging what you will end up with in print.

 

All that said after shooting a few rail cameras, and humping some very heavy Eq, I decided to back off from trying to get the utmost Rez, (IE heavy) and am happy getting slightly less rez, but having equipment that I can actually carry somewhat comfortably.

 

------

 

and what else might I be missing in this list ... oh sharp lenses, any comments on a Nikkor 135mm W f5.6?

 

No I have never used that lens, but I think the concensus is most Nikkor LF lenses are great. I am a big fan of G-Claron lenses myself. The 210 and 240 are some of the shapest lenses I have used. On the cheap a Kodak 135mm Ektar is supposed to be sharp.

 

I usually go here first when looking at lenses and then do a search for lp/mm and then may come here and ask a question or two, contrast etc. Its not a complete list, but its a start.

 

http://www.hevanet.com/cperez/testing.html

 

I, too, am profoundly interested in the comparison of drum-scanned color transparency resolution and print sizes. Near my home there is a superb gallery (Scanlan�s gallery in Park City), featuring landscape and travel color photography, printed largely in the 16x20 to 30x40 inch sizes. I have never taken a roll of film with a medium format camera, though I have used 6x7, 6x9, and 6x12 backs on the various 4x5 cameras that I have used for a number of years for landscape work. Granted, the print sizes above are not as large as those desired by Milan, but my comparisons between the Scanlan�s prints (almost exclusively taken with a 6x7 camera) and my own 4x5 images are that even at 24x30 inch sizes there is virtually no distinguishable difference. My images are drum-scanned at the same lab this gallery uses, and printed on the same paper. Using a crisp loupe, my images made with 110 Schneider, 150 Sironar, and 240 Fujinon-A (easily as sharp as the others�either mine is an unusually sharp sample or this is one of the most overlooked lens out there) on Velvia 50 paper are very sharp. Having said that, and obviously without walking into the gallery with an enlargement of my own to compare side-by-side, I can distinguish no difference, up to 24�x30� between their 6x7 images and my 4x5 images (or other people�s that I have frequently viewed in galleries). I am sure that in larger print sizes the difference would become more pronounced. Forgive this long post, but my query is whether others have observed similar characteristics in the 26x20, 20x24, and 24x30 print sizes between professional scans of 6x7 and 4x5.

I've made wet B&W prints from both 645 and 4x5 negatives. An 11x14" from a TriX 645 negative shows grain. The 4x5 doesn't. Grain doesn't care how sharp the lens is.

I think its easy to have 67 shot that is close to 4x5 although 4x5 is 3x the surface area. All it takes is a super sharp mamiya lens against a not quite so sharp, film not quite so flat or shakey, slightly OOF 4x5 shot and it can get very close.

 

Here is a Mamiya 7 shot against a Sinar 4x5 shot. The Sinar shot was drum scanned at 2000 dpi. The Mamiya 7 was scanned at 4000 dpi, mostly because I was comparing huge enlargements like 40" x 50" at the time.

 

Feel free to down load them or whatever.

 

Also this is not my sharpest LF lens, but it is a good one. Its fairly close and all the detail is there, but it is just cleaner and sharper with 4x5. I think at 16x20 it would be a wash. Both of these are E100G.

 

These have not been sharpened or worked in any way, just resized and mostly straight off the scanner.

 

The Sinar crop is a 2000 dpi drum scan that is roughly 1/3" x 1/4" high on film. The mamiya crop was scanned at 4000 dpi and downsized to match the Sinar crop.

 

http://www.pbase.com/tammons/gallery/mf_vs_4x5

 

Mamaiya 7 resized to 24H x _L 204 dpi (4 lp/mm) dpi to be printed on a lightjet. (I just resized the vert to 24" since the 67 format is more square.

 

http://www.pbase.com/tammons/image/53024463/original

 

Sinar and Rodenstock Sironar 150mm resized to 24H x _L 204 (4 lp/mm) dpi to be printed on a lightjet.

 

http://www.pbase.com/tammons/image/53024464/original

I'm going back to a part of the original post that has not been addressed.

 

Ellis wrote "Never discount the ability to use view camera movements to make a large

difference in the aesthetic quality of an image, even with landscapes. a 4x5 camera with

movements gives the ability to shape the image exactly the way you want it to look

without compromise . The trade off for this precision is speed and economics of

operation."

 

The above is absolutely correct and what has been ignored from the original post is the

choice of 4x5, the Toyo CF. You need to think carefully about the price/performance

tradeoffs with this cameara. It has very limited movements ... no back movements at all

for example. Might defeat your purpose, might not be the best tool for the job. You can

get other cameras in the same price range that give you far more flexability ... the

Tachihara or the Shen Hao. If you like the Toyo approach and design then see if you can

find a good used A, AX or AII .... any of the three are solid performers with a reasonable

collection of movements, unlike the CF.