On the quality differences across formats
Compiled by Q.-Tuan Luong for
the Large Format Page
Richard Knoppow (dickburk@ix.netcom.com)
Where the same emulsion is used for both roll and sheet films the
resolution is the same. Coating thickness is about the same on both
types, its the support which is a lot thicker on sheet film. This is
of no consequence for scattering or loss of resolution.
Some films do have thinner emulsions, Technical Pan, for example,
but again, the emulsion thickness is the same regardless of the
format.
Film has gotten good enough so there isn't a big jump in quality
between 6cm x 6cm and 4x5 as there was in the past. There is still a
big jump between 35mm and 6x6 but there used to be another just about
as great when going to 4x5, no longer, at least for reasonable size
prints. You can still see the difference for prints larger than about
11x14 its just not a stunning as it was many years ago.
From: bg174@FreeNet.Carleton.CA (Michael Gudzinowicz)
A couple years ago, I compared a few lenses for the 35 mm and
4x5 formats, and the Mamiya 50, 65, 100 and 150 lens (newer
models with most recent "M" backs). All negatives were shot on
TMX and were of the same scene, and enlarged to 5x7, 16x20 and
for 120
The results were somewhat predictable. All were sharp at low
magnifications, with granularity detectable at 5X
magnification.
When the negatives were enlarged to 16x20, most of the 35 mm
shots were sharp, though grain detracted from even tone grey
areas. All of the Mamiya prints were very sharp, and
granularity was just detectable, if you know where to look and
had prints from ed Tessars - perhaps
unexpected, but present). The visible difference between
modern 4x5 lenses and the Mamiyas was minimal. If the Mamiya
prints had been enlarged with a diffusion head rather than
condenser, they would have been nearly indistinguishable.
At 30x40, the modern 4x5 lenses, Mamiya 50 f/6.3 and 100 f/3.5
(last model; all black using 55mm filters) were still sharp
when closely examined - granularity was apparent in the MF
prints, but wouldn't be noticed at anything resembling a
normal viewing diportion of typical handheld 35 mm
enlarged to 8 to 11X.
The apertures used were f/8 for 35mm; f/11-16 for MF and f/22
for LF. I hadn't cleaned the inner surfaces of the Mamiyas for
some time, and performance should be better after removing oil
deposited from the shutter. Almost invariably, I've found the
50, 10ilm, and by shooting with less wind (it was
a problem).
After all of that, the MF and LF results are very similar up
to 8x10 or 11x14; at 16x20 modern LF lenses have a slight edge
due to reduced granularity and high contrast; above that,
modern 4x5 wins due to granularity, though the other lenses
are still usable (similar to 35 mm at 16X). I'm accustomed to
8x10 and 11x14 contact prints, and although I don't use those
formats any longer, I rarely print larger than 8x10 or 11x14
form 4x5, so "sharpness" is a non-issue for me, despite the
emphasis of the post.
I'd suggest that you run similar tests on your lenses, and I'd
suggest cleaning the surfaces of the cells near the shutter
before testing since evaporated oil accumulates over the
years, and isn't visible looking through the lenses. Also, use
a 10x loupe on the ground glass of the Mamiya or view cameras
for focusing.
The black Super 23 I used resembles the last of the
Universals, but it has 15 degree swings and tilts. The 100
f/3.5 retracts for full movements at infinity focus, and the
150 and 250 have more than sufficient depth of field at their
hyperfocal distances tor instance, the 50 on
6x9 is equivalent to a 50 mm shift lens on 6x6 moved 15 mm.
Also, the 50 is sharp enough to print a 1 x 3.25" "panorama"
offset from center to correct perspective (equivalent to a 20
mm lens field of view in the 35 mm format). All of the Mamiya
lenses have minimal distortion (if any) - the 50 is a Biogon
design and 65 is a Topogon - both aerial mapping lenses.
--------------------------------------------------------------------
From: bjs@oes.amdahl.com (Barry Sherman)
As empirical confirmation of what Michael says, I'll offer my own
observations. A friend uses a Pentax 67 for landscape work and I've
been consistently impressed with the sharpness and overall image
quality that he can achieve with that camera. At 16x20 I usually
think that my 4x5 images do, on average, have an edge, but it's not
huge and can be easily offset if my taking technique is off at all.
And I've seen work done with the monster Fuji 6x8 SLR which is even
more impressive. I've seen 20x24's from those 6x8 transparencies
which I do think come very close to what I could do with 4x5. It's
difficult, of course, to compare different formats unless one has
done the kind of side-by-side comparison that Michael did and I
trust his results.
---------------------------------------------------------------------
From: sparks@col.hp.com (John Sparks)
I think you will see a bigger difference with format for B&W prints than
with color prints. I work in B&W almost exclusively and have worked
with medium format, 4x5 and 8x10 negatives. In almost all cases, I can
see a difference between all three formats in a B&W 16x20 print. This
difference is significant enough where I think I can generally pick
which format made which print (assuming I can study the print up close
in good light and the prints are more or less traditional landscapes
with everything sharp and from fine grained film and no unsharp masking
was used). In an 11x14 print, 4x5 and 8x10 become harder to
distinguish, but large format can generally be distinguished from medium
format at this size due to greater granularity from medium format. With
8x10 prints, differences are almost impossible to distinguish. Some
subjects make the differences less clear, the more fine details in the
subject, the easier it is to distinguish the different formats.
When I look at large color prints from medium format or even 35mm, I'm
often greatly surprised at the quality possible from fairly small
transparencies and to a lesser extent negatives. I think that a
combination of lower maximum resolution in the color print materials,
smoother "grain" resulting from transparent dye clouds rather than
opaque silver grains, greater impact from color with less reliance on
fine details, and wider use of unsharp masking are all responsible for
making negative/transparency size less important for color than B&W.
I don't know where Peter came up with 3.5 for the [maximum enlargment]
factor, but it agrees pretty well with my observations (at least for
prints from B&W, prints from color allow larger enlargements before I
notice differences between formats). I'm not sure if it's exactly a
constant. I think the max enlargment factor without ANY loss of
quality is smaller for larger negatives (maybe 3X for 8x10, 3.5X for
4x5, maybe 4.5x for 2 1/4" and 5.5X for 35mm). I think this is
because lens quality improves slightly with smaller negatives and film
plane accuraccy improves more substantially so that things don't
improve quite as much as the linear film dimension increase would
predict.
I find that 16x20 prints from 4x5 (~4X prints) look noticably different
from 16x20 prints from 8x10 in most cases. It's definately not grain
since I can't see grain in either, at least from fine grain films like
Tmax. I doubt it's resolution either (at least not resolution measured
by line pairs/mm) since I don't think my eyes can resolve more what you
get from a 4X enlargement.
I think there is something beyond what the resolution number tell you.
Your eyes can detect some of the high-frequency stuff beyond the lp/mm
limit. The numbers I've seen for resolution limits in a print that can
be seen with the naked eye are about 8 or 10 lp/mm. This seems to match
what I can see when looking at a resolution chart with my eyes.
However, I think you can tell the difference between a sharp square wave
10 lp/mm and a fuzzy, noisy 10 lp/mm that you get at the film/lens
resolution limit. To capture this difference, you need a lot more than
10 lp/mm.
From the difference I can see in 16x20 prints (a 2X print from 8x10 or a
4X print from 4x5) or even in 11x14 prints, (a 1.4X or 3X) though this
is a much more subtle difference, I'd guess that you need somewhere
around 30 lp/mm in the print to get the maximum quality that you can see
with a naked eye. I know I want to put my eye right up next to a print
(call this 10"-20"), reguardless of print size. Unless there is a
barrier that prevents close observation of large prints, I find most
people look at prints from about the same distance no matter what size
the print is. If you keep your eye about the print diagonal or more
from the print (which explains why billboards made from 35mm can look
sharp when you drive by them many yards away), I don't see any problems
with any size prints from 4x5, but I don't (and I don't see others)
looking at prints that way.
---------------------------------------------------------------------
Kerry L. Thalmann
So, we keep hearing the argument that modern films and lenses are so
good that there is no longer a need to shoot large format (of course, I
don't buy into that argument, and it is always presented by those who
shoot in smaller formats), but it indeed looks like we need even finer
grained films to get a higher degree of enlargement. Modern lenses (and
for that matter even many 50 year old large format lenses) are capable
of achieving very high on-film resolutions. In the end, it looks like
the grain structure of the film may be the limiting factor in
determining the maximum enlargement. Of course, this grain structure is
a function of the emulsion and will be the same for any emulsion
regardless of the format.
Finally, the normal viewing distance increases as the size of the print
increases. Grain that is objectionable in an 8"x10" print at 1/2 meter
may be unnoticed in a 40"x50" viewed from across the room. This is the
one relationship I have yet to quantify (I still need a formula for the
effects of visible grain as a function of viewing distance). I have
made prints up to 30"x40" from 4x5 with acceptable (for me) results, and
I regularly make 24"x30" prints. This relationship between visible
grain and viewing distance is one more factor that favors the larger
formats. Assuming the same emulsions and a constant enlargement factor,
the final print size, and thus the normal viewing distance increases as
direct function of the film size. So, even though the enlargement
factor is similar, at normal viewing distances, a 20"x24" print from a
4x5 original would appear less grainy than an 11"x14" print from 6x7cm.
This also explains why, in a gallery setting, I can always tell from
across the room, which 20"x24" prints were made from medium format and
which from large format. It is not the lack of sharpness of the film or
lenses, it is simply the effects of the film grain. The grain is just
much more obvious, to my eye, in the prints made from medium format
originals.