Purple Heart
The woods were donated by Steve
Knight (a very generous supply of exotics) or came from my private stock
(most purchased from Southern Lumber in San Jose); Lee
Valley helped me to obtain a couple of woods I had difficulty acquiring
on my own.
Methodology
All tasks and ratings involved with this comparison were
performed by the investigator alone (i.e., me). All planes were allowed
several weeks to acclimatize to the temperature and humidity of my shop. All
plane soles were checked for flatness with a Bridge City Tools straightedge,
and then lapped flat if necessary using "scary sharp" techniques
on a long piece of glass. All plane blades had both their faces and backs
flattened (as needed) and backs were polished by hand using waterstones;
bevels were made flat and sharpened to 0.5 micron on my Lee Valley Power
Sharpening System. The consistency and quality of the edge was checked each
time with a 5X magnifier.
The wood samples were milled rectangular, usually with a
1.5 to 1.75 inches wide face, so that all planes would see the same width of
surface to be cut, the wood was then planed smooth with a freshly sharpened
Lee Valley #4.5 with standard pitch and bevel angle. Sufficient cuts were
made to insure that adjustment was such as to produce optimal performance
from the plane. This became the benchmark condition with the Lee Valley #4.5
at 45° used to establish the surface condition of each wood sample.
Then one of the comparison planes followed to see if any
difference could be discerned. Sufficient cuts were made to insure the best
adjusted performance was achieved from the plane. If adjustment required
more than six strokes, the blade was pulled and re-honed, then adjustment
proceeded again. No evaluation cuts were made with a blade that had more
than six prior adjustment cuts. Notes were taken on a separate rating sheet
for each wood sample. Two photos of each surface were shot (one of the area
of maximal "damage" and one represented the larger view of the
sample) and coded so that the plane was not readily identifiable.
The Lee Valley #4.5 blade would then be re-honed at 0.5
and used to renew/return the wood surface to the Lee Valley #4.5 level, then
another plane was tried with the same follow up steps. This was repeated
over and over until all the planes had been tried on a given sample of wood.
If planes were close in performance, back to back comparisons were made to
refine awareness of their differences. Then the process was repeated for
another species of wood. This continued until each plane had been compared
on a sample of each wood species. Then the entire process was repeated using
a second sample of each wood. Planes were selected in a quasi-random order
for each wood species (that means I tried to use different orders of planes
for each sample, but random selection was not guaranteed).
Ratings of the planes were based on several criteria,
which for a given wood might include some combination of surface smoothness
(both visual and tactile), depth of tear out, amount of tearout, clarity of
the wood surface, etc. There were no fixed criteria, as different woods
displayed different characteristics and were handled differently by the
planes. For example, some woods displayed a lot of tearout, but the clarity
of the remainder of the wood was consistently very good; while another wood
may have shown no obvious tear out, but there were several differences in
how smooth the wood felt to the hand.
A point was awarded for each discernible improvement in
overall performance. For some woods, there were several discernible levels
of performance (i.e., the surface left by plane B was better than the
surface left by plane A, and the surface left by plane C was better than
that provided by plane B.) Points were awarded to each plane according to
how many discernible steps in performance it represented relative to the
benchmark Lee Valley #4.5. Thus, for a very difficult wood, like cocobolo,
the plane that performed the best was able to achieve a score of 5 on one
sample, representing that five different levels of performance were
discernible between the surface left by this plane and the benchmark plane
while using the full range of planes on this wood. In contrast, on a few
woods, no discernible differences were detectable between any of the blades,
thus all planes were rated as “0” relative to the benchmark performance.
It is important to understand that these ratings
represent, in the statistical sense, an interval scale. Though there was an
arbitrary "0" point (the benchmark plane’s performance), the
"0" did not represent poor performance and the maximum score did
not represent a "perfect" surface. There were some woods on
which an essentially perfect surface was achieved by the benchmark plane
(which isn’t too surprising since it has been very favorably compared to
some of the best of standard angle planes), and even the highest rated plane
may not have achieved flawless performance on other woods. The intervals
were determined by the overall group of planes’ variation in performance,
relative to the benchmark plane, and did not represent any externally
established structure. Thus, the intervals do not represent equal steps (the
steps ultimately being determined by the performance of the planes), but
rather represent rankings, in which ties can occur among two or more planes.
After all the samples had been planed, the scores for a
given plane on both samples of a wood were entered into a statistical
database (SPSS) and averaged. The developed photos for a given wood (both
samples) were then all laid out on a table and again ordered according to
overall discernible differences. Unfortunately, the photos did not allow for
as fine a discrimination as did the direct evaluations which involved the
ability to feel the surfaces and more easily consider a flawed area in the
context of the overall surface. It was decided after starting the
evaluations, but before analyzing the data, that if the rankings differed by
more than one level of placement, between photos and direct observations,
then one half the difference in ranking would be added (which might be a
positive or negative score) to the rankings achieved through direct
observation. Fortunately, these complications were avoided in that the final
blind ranking of the photos was never in conflict with the direct
observations, though sometimes fewer categories of difference were
discernible through sole use of the photographs.
Results
First, results will be listed by overall ratings for
each plane (an obvious order by bedding angle and style becomes apparent).
If you want just the overview, you can stop there. Then planes will be
ranked according to the number of their first and second place finishes
(including ties). Finally, each plane’s rating will be given for each
wood. This lets you see which planes did best on a given wood and also gives
you a good idea of how much variation in plane performance was present for a
specific wood. You can consider planes and bedding angles with reference to
the woods you most often use, or you can gain an idea of what planes tend to
always do a good job over your own personal group of woods.
Again keep in mind that the Lee Valley #4.5 standard
angle smoother is the benchmark and thus always presents with a score of
zero. Bedding angles are as close as I was able to measure on the planes
used, and may or may not agree with the manufacturer’s specifications.
|
Rankings based on total
rating points accrued over all woods
|
|
Plane
|
Angle
|
Score
|
|
Lee Valley Low Angle
|
32°
|
-4.0
|
|
Mujingfang Ebony
|
40°
|
-0.5
|
|
Lee Valley #4.5
|
45°
|
0.0
|
|
Knight Finish
|
47°
|
4.0
|
|
ECE/Primus Imp.
|
50°
|
5.5
|
|
Lie-Nielsen #4.5 HA frog
|
50°
|
6.0
|
|
Clark &Williams Coffin
|
55°
|
6.0
|
|
LV #4.5 back bev.
|
60°
|
8.5
|
|
Mujingfang Rosewood HA
|
62.5°
|
13.5
|
|
Knight Japanese Infill
|
47.5°
|
13.5
|
|
Stephen Thomas Infill
|
47.5°
|
19.5
|
|
Ranking based in order of
number of 1st and/or 2nd place finishes
(These rankings do not include the three woods on which all planes
produced the same level of surface. Multiple ties may have occurred
on a given wood.)
|
|
Plane
|
Angle
|
1st places
|
2nd places
|
| Lee Valley #4.5 |
45° |
0 |
0 |
| Mujingfang Ebony |
40° |
0 |
1 |
| Lee Valley Low Angle |
32° |
0 |
2 |
| ECE/Primus |
50° |
0 |
2 |
| Lee Valley #4.5 back bevel |
60° |
0 |
4 |
| Lie-Nielsen #4.5 HA frog |
50° |
1 |
1 |
| Knight Finish Smoother |
47° |
1 |
2 |
| Clark & Williams Coffin |
55° |
1 |
3 |
| Mujingfang Rosewood HA |
62.5° |
2 |
3 |
| Knight Japanese Infill |
47.5° |
4 |
2 |
| Stephen Thomas Infill |
47.5° |
7 |
2 |
|
The following are the
plane ratings on each of the wood samples. The ratings are based on
the average score from two samples of each wood. Again, each point
represents one discernible level of difference between at least two
planes.
|
| |
Wood Used
for Comparison
|
|
|
Plane
|
Angle
|
1
|
3
|
4
|
5
|
7
|
8
|
9
|
10
|
12
|
13
|
Combined
Score
|
|
Lee Valley Low Angle
|
32°
|
-2.0
|
-1.0
|
-1.0
|
0.5
|
0.5
|
-1.0
|
1.0
|
0.0
|
1.0
|
-2.0
|
-4.00
|
|
Mujingfang Ebony
|
40°
|
-1.0
|
-1.0
|
1.0
|
0.0
|
0.0
|
-1.0
|
2.0
|
0.0
|
0.5
|
-1.0
|
-0.50
|
|
Lee Valley # 4.5
|
45°
|
0.0
|
0.0
|
0.0
|
0.0
|
0.0
|
0.0
|
0.0
|
0.0
|
0.0
|
0.0
|
0.00
|
|
Knight Finish
|
47°
|
-1.0
|
-2.0
|
2.0
|
0.5
|
1.0
|
2.0
|
1.0
|
0.5
|
0.0
|
0.0
|
4.00
|
|
Clark & Williams Coffin
|
55°
|
-1.0
|
-1.0
|
3.5
|
1.0
|
0.0
|
0.0
|
0.0
|
0.0
|
1.0
|
1.5
|
5.00
|
|
ECE/Primus
|
50°
|
-1.0
|
-1.0
|
3.0
|
0.5
|
0.5
|
1.0
|
1.0
|
0.0
|
0.0
|
1.5
|
5.50
|
|
Lie-Nielsen #4.5
|
50°
|
0.0
|
-1.0
|
2.0
|
0.5
|
1.0
|
1.5
|
0.5
|
0.5
|
0.5
|
0.5
|
6.00
|
|
Lee Valley #4.5 back bevel
|
60°
|
2.0
|
-1.0
|
2.0
|
0.5
|
0.5
|
2.0
|
1.0
|
0.5
|
0.5
|
0.5
|
8.50
|
|
Knight J. Infill
|
47.5°
|
1.0
|
0.5
|
2.0
|
1.5
|
0.0
|
2.0
|
3.0
|
0.0
|
1.5
|
2.0
|
13.50
|
|
Mujingfang Rosewood HA
|
62.5°
|
3.0
|
1.0
|
4.5
|
1.0
|
0.0
|
1.0
|
2.0
|
0.5
|
0.5
|
0.0
|
13.50
|
|
Stephen Thomas Infill
|
47.5°
|
1.0
|
2.0
|
4.5
|
1.5
|
1.0
|
4.0
|
2.0
|
1.0
|
0.5
|
2.0
|
19.50
|
Index to woods used for
comparison
| 1. |
Blackwood Acacia |
One of the most difficult woods; no
plane produced a perfect surface. |
| 2. |
Bloodwood |
All planes produced an excellent and
indistinguishable surface. |
| 3. |
Bubinga |
No plane provided a perfect surface,
but little difference between planes. |
| 4. |
Cocobolo |
Another of the most difficult woods;
no plane produced a perfect surface. |
| 5. |
Ebony |
Best planes provided a good surface. |
| 6. |
Ipè |
All planes produced an excellent and
indistinguishable surface. |
| 7. |
Lacewood |
Best planes produced an adequate
surface. |
| 8. |
Lignum Vitae |
Best planes produced an excellent
surface. |
| 9. |
Fiddleback Maple |
All surfaces were good, but best were
exquisite. |
| 10. |
Curly Oak |
All surfaces were good to very good. |
| 11. |
Straight Grain Oak |
All planes produced a very good
surface. |
| 12. |
Padauk |
All surfaces were good to excellent. |
| 13. |
Purpleheart |
Poor to very good surfaces. |
Discussion
I wrote in the introduction, and I think it is good to
reiterate it now:
In this review, there were only two primary
objectives:
To determine if high angle planes, as a group, tended to perform better
than standard angle planes on difficult to plane woods (i.e., should you
obtain a High Angle plane if you work with difficult hardwoods that are
prone to tearout, fuzzing, etc.); and, more specifically, to find which
planes and/or plane configurations yielded the best surface on a range of
particularly difficult to surface woods.
It is important to keep in mind that the results from
this investigation do not necessarily, or even likely, represent the
performance of planes on less dense and commonly used woods such as walnut,
mahogany, cherry and poplar. It is also good to keep in mind that the
findings do not so much represent a comparison of manufacturers, as they are
a comparison of plane styles and configurations; the difference in
performance between the Lee Valley #4.5 with an effective cutting angle of
45 degrees and the same plane with an effective cutting angle of 60 degrees
is a good example.
Plane Characteristics vs. Quality of Surface Achieved
There were only two consistent findings associated with
good performance across the range of woods: that infill style planes
performed better than any other style; and for non infill planes,
performance was consistently better as effective cutting angle (bedding
angle plus back bevel angle, if any) increased. These findings are obvious
upon casual comparison of effective cutting angle and total performance
points, and, with respect to effective cutting angle, is confirmed by
elementary non-parametric correlational analysis (Spearman Correlation
Coefficients). Taken as a total group of all eleven planes, effective
cutting angle correlates at the 0.697 level (a perfect correlation would be
1.0), a strong correlation, and is found to be significant at the 0.017
level (fewer than 2 times in a hundred could such a finding occur by
chance). The finding is extremely strong if we remove the two infill planes
from consideration. For the remaining nine planes effective cutting angle
and total performance points is correlated at the 0.9874 performance level,
with a significance of 0.000 (in words, less than one time in 1000 would
this occur by chance).
Other plane characteristics, such as total weight or
blade thickness failed to achieve a significant correlation, with or without
the inclusion of the infill planes. While the top performing plane was the
heaviest (the Stephen Thomas infill), for those tied for second place in the
category of total performance points, one ranked 5th from the top and the
other was two places below that. Put another way, at 760 grams, the
Mujingfang Rosewood HA plane that tied for second in total points is little
more than one third the weight of the top ranged plane, and only a little
more than half the weight of the plane it tied with for second place. So in
general, on these dense difficult woods, weight doesn’t correlate with
improved performance, but angle does. This does not mean that weight does
not matter, but likely that performance is related to some function of both
angle and weight (as is suggested by the relatively low angle but high
weight infill rankings).
Blade thickness was equally un-associated with plane
performance. Though this goes against some conventional wisdom, it does not
surprise me that much. All of these blades were bedded more fully than some
typical Bailey style planes. With each plane, the blade support extended at
least all the way down to the top of the bevel, and in the case of the Lee
Valley Low Angle plane, to just short of the blade edge. The thinner the
blade, the shorter the bevel width, thus the closer the bedding came to the
blade edge. With a blade bedded near to its edge, I personally doubt that
blade thickness is that much of a factor, once a certain basic thickness is
achieved, which in this investigation would suggest anything around 1/8th
inch is sufficient. More likely, thicker blades are most important on Bailey
style planes where when the frog is advanced and bedding ends well above the
blade edge.
None of these planes had what has traditionally been
considered a thin blade. The thinnest blade comes in at just slightly under
1/8th inch, most were at 1/8th inch, and only one blade was considerably
thicker than the others (the Knight Finish/Smoothing plane at 1/4 inch).
Four planes had blades that were tapered, so the thickness at the top of the
bevel was used as the measuring point for thickness, this resulted in
measuring the thickest blade section for three of the planes (i.e., those
blades thinnest at their top) and one of the thinnest sections for one plane
(the Japanese blade that tapered down from a thick top).
The presence or absence of a chip breaker, by itself,
seemed to have no influence on plane performance with these woods. Half the
planes used one, half the planes didn’t. Of the high rated infills, one
used a chip breaker, one didn’t. Particularly telling, of course, is that
the Lee Valley #4.5 ranked very differently depending upon the effective
cutting angle used, but in both cases, the same chipbreaker was employed.
Perhaps, but this is only unsupported speculation, at these higher angles a
tightly coupled chip breaker (such as is employed on the ECE/Primus which is
coupled to the blade with two screws) is more important for its ability to
reinforce a thinner blade, than for it any chip breaking performance per se.
Half the planes had a wood sole, half the planes had a
metal sole. No association was found between sole material and surface
finish on the woods planed. While two of the three top ranking planes had
metal soles, so did two of the three lowest ranking planes. Sole material
did make a difference with regards to usage characteristics, but not to the
quality of finish obtained. It is worth noting here, that all plane soles,
whether wood or metal were waxed with Renaissance microcrystalline wax.
Blade width was basically eliminated as a characteristic
of the planes’ performance, as all wood samples were made no wider than
the narrowest blade. However, this does not mean that blade width might not
play some significant role on tasks that involve wider boards or panels.
Functional Characteristics vs. Quality of Surface Achieved
Shaving thickness itself was not in a 1:1 relationship
with quality of surface, and not all planes were able to make equally fine
continuous shavings. Some planes took very fine even shavings and provided
an excellent surface (most commonly the Stephen Thomas, Knight Japanese
Blade Infill and the Mujingfang High Angle), but some planes producing fine
shavings left a rather mediocre surface while other planes produced much
thicker shavings but left a better surface.
User variable plane characteristics, such as blade
extension, mouth width (on those planes where this could be adjusted), and
cap iron/wedge tension were not consistent between planes, woods, or even
individual samples of the woods. Each were and needed to be adjusted to
obtain the best performance for the planing task at hand, and no overall
settings appeared to apply for any of the planes.
User qualities
User qualities break down primarily into ease and
refinement of adjustment, the ability to move the plane across the wood, and
general matters of fit and comfort.
Adjustment
This issue is surely biased by my limited experience
with wood planes. I have eight wood planes that do not have mechanical
adjusters (and four that do), and have used this style for many years, but I
am clearly not a master of wood plane adjustment. With that caveat made
clear up front, I’m going to discuss one of the most important
distinctions between planes, as I experienced it, other than the surface
finish they achieved.
The ability to achieve fine and consistent changes in
planing performance as a result of changes in blade extension and alignment
is what I call adjustment control (a stupid name, just the best I came up
with). To me, this is more than just a reflection of the refinement and
precision of adjustment mechanisms, but also includes the ability of the
plane to translate these changes in blade position into a change in cutting
performance. The Stephen Thomas plane was simply a delight to adjust and
repeatedly demonstrated the highest levels of adjustment control. It simply
allowed me to make very fine adjustments in meaningful blade extension that
translated into distinguishable changes in differences in the
characteristics of the shaving made and/or surface revealed. No other plane
came even close. No other plane could I take to a new wood with a freshly
installed blade and so quickly achieve the optimal setting for the shaving
or surface I wished to achieve. Consistent with my definition of adjustment
control, it allowed for easily achievable very fine settings that resulted
in distinguishable and meaningful changes in plane performance. At risk of
repeating myself, this plane was able not only to make fine changes in blade
position, but respond to those changes. It has spoiled me for all other
planes, on this characteristic alone.
Interestingly, the plane that ranked second in this
respect was the Lie-Nielsen #4.5. It was notable superior to the other
planes. Even though the Lee Valley #4.5, in my experience, has the better
adjuster and I find it slightly more precise when comparing the two planes
in their standard angle configuration, it did not show as great a degree of
functional control over the plane’s cutting as did the Lie-Nielsen.
Perhaps the angle and back bevel has something to do with this, but that is
a consideration I have absolutely no foundation for. Regardless, the
Lie-Nielsen #4.5 with its HA frog, was meaningfully (at least to me)
superior to all planes but the Stephen Thomas.
Roughly tieing for third place was the ECE/Primus, the
Lee Valley #4.5 in both its configurations, and the Mujingfang High Angle.
These planes I could be happy with, had I not experienced the Stephen
Thomas.
The remaining planes were acceptable except for one. The
disappointment was the Knight Japanese Infill. I found it a huge hassle to
adjust and it never provided the sense of control that was offered by the
Stephen Thomas, the Lie-Nielsen or even the Mujingfang High Angle plane and
the Primus. When finally set perfectly, the Knight Infill performed
excellently, as the total points figures show, but I just hated having to
remove the iron as I knew it was going to be a struggle to properly adjust
it again. Part of my trouble adjusting this plane is likely explained by a
certain lack of skill adjusting by hammer, but I suspect the combination of
the tapered Japanese blade and the very tight fixed mouth also contributed
to the frustrations I encountered with this blade. I’m sure in normal use,
I’d use this plane sparingly, just so I didn’t have to sharpen and
readjust the blade again, but it sure works well when set correctly.
Movement across the wood
Two factors are involved here, but I shall consider them
concurrently. The first factor is the ability to plane right through a
resistive surface, such as the densest woods or woods with notably varying
density. The second factor relates to the ability to change the direction of
the plane to accommodate shifting grain. In this case two planes stood out,
one for each of the factors.
Again, the Stephen Thomas plane was just wonderful in
its ability to seemingly effortlessly maintain the momentum of the planing
stroke throughout its entire length. Surely this is largely related to it
being the heaviest plane, but likely other factors, such as control over
blade extension, ergonomic factors such as grip shape and placement, and
effective cutting angle all contributed to its ease of movement through the
wood. Stephen's plane is over six pounds. I loved the weight, but I'm not
sure everyone would be very happy with it. I really do plane a lot, and have
the associated task specific musculature. I'm not sure someone else would
handle the Stephen Thomas or some of the other HA planes in the same way or
with the same control as I do, but if the unsupported weight of a plane is
not an issue, the Stephen Thomas makes use of its weight and other
characteristics very successfully when making cuts through the denser woods.
Second in this regards were the two metal planes
followed by the Knight Japanese Ironed Infill. Again weight seemed to be the
primary underlying factor, though the former two planes also use a more “Western”
style grip, as does the Stephen Thomas. Also quite acceptable was the
Mujingfang HA, which is a plane that felt and performed like a plane much
heavier than it really is. I should note that I used this plane without its
optional crossbar that will allow pulling and a somewhat different sort of
pushing grip.
The Primus plane, the C&W and the Knight Finish
plane often seemed to lack the authority to continue with cuts along the
length of a dense wood such as purpleheart or cocobolo. As a result, they
required greater physical effort to move through the wood stroke, and became
more difficult to control as the activity of the larger muscles predominated
over the more precise smaller directional control muscles.
A point worth bringing up here, although it might well
be considered with respect to several other topics, is blade width. Clearly
as angle goes up, resistance increases for a given blade width. Obviously,
the wider the plane blade, the higher the resistance. Ideally, a lighter
plane with a higher angle, would use a narrower blade, while a heavier plane
might more easily allow for a higher, more resistive, angle. I can’t say
that this idea is supported by either the data on quality of surface
achieved, nor by my own subjective experience using the planes. Perhaps that
is because manufacturers have intended that as blade width increases, so
also does the overall plane weight. Still, one wonders about the trend of
Lie-Nielsen to make its widest planes available with high angle frogs, and
what might be the relative performance of a heavy plane but with a narrower
blade used at high angle. A future study might involve back bevels applied
to a Lie-Nielsen Bronze #4 compared to an iron Lie-Nielsen#4.5 with similar
back bevels.
The other factor is one of ability to manipulate the
plane to change directions to account for changing grain and other wood
variations. In this area, as was also revealed in last year’s review of
the Lee Valley #4.5, the ECE/Primus Improved Smoothing Plane stands alone.
There is something about the slippery Lignum Vitae sole, the Continental
handle arrangement, and perhaps some other unidentified characteristic that
allows the Primus to stand unchallenged in this respect. Particularly on
cocobolo, where some synergistic interaction takes place with this plane,
the ECE/Primus just effortlessly skims across the surface, cutting a fine
shaving and instantly responding to the lightest muscular input to change
direction or force. On this particular wood, it is like going from a Stephen
Thomas Porsche Turbo to a Primus Formula One car.
Ergonomics
I’ve touched on this previously, but it is worthy to
return to as a topic of its own. In this investigation, I placed all other
considerations secondary to sheer surface finish for the specialized high
angle "super smoother". Still, all other things being equal, it is
nice to have a plane that is comfortable for extended use. This is of course
a highly personalized consideration, and one that is totally subjective. For
me, two planes stood out, though the two planes that ranked the highest for
me couldn’t be more different: the Stephen Thomas and the ECE/Primus. The
Stephen Thomas took a while to get used to, with its smaller enclosed grip,
but once I adjusted to the enclosed grip, it soon became one of my
favorites.
The ECE/Primus, with its Continental front horn and
rounded rear notch was always a pleasure to hold, and allowed rapid changes
in direction without having to contort one’s hands or wrists. I have
always found the ECE/Primus, with its unusual (in North America) grip, to be
a particularly refreshing break from long planing sessions with a Bailey
style plane.
The Mujingfang High Angle plane also was comfortable.
Its Taiwanese style "roofed" front body and cross bar allow for
several comfortable gripping positions.
The Bailey style grips found on the Lie-Nielsen and Lee
Valley planes have not become the Western standard without reason. I like
the Lee Valley rear tote slightly better, it having a slightly different
angle and length than the more traditional tote found on the Lie-Nielsen,
but both are good for both brief and extended use.
The Clark &Williams was new to me, and I was
impressed with how readily this small, coffin style of plane felt right in
my hand. I wish it had more mass, but I suspect one could comfortably use
this plane for extended periods of time.
The Knight planes fall at the bottom of my list. His
Finish/Smoothing plane is too blocky for comfort and makes no accommodation
for grip. It is acceptable for shorter periods of planing, but became less
comfortable to use the longer a planing session went on.
The Knight Japanese Blade Infill, is one of the most
beautiful planes to view, but one of the least practical to use for extended
periods. The grips are very smooth and fairly comfortable when the plane is
riding on the surface of the wood, but this plane is fairly heavy and there
is no accommodation made for lifting the plane from the wood. The shape of
the Bailey style grips are the best for lifting a heavy plane, allowing it
to be easily gripped both forward and aft. The English infill style is
second, with the enclosed tote offering at least a one handed grip. But the
heavy Knight Japanese Blade Infill is a disaster, requiring a tight pincher
grip to be used to pick up the plane after every stroke. The style that was
generally acceptable for a light wooden plane, becomes a recipe for
repetitive stress injuries when applied to such a heavy plane. The good news
is that Steve is aware of this and is quite actively researching and
modifying his planes to achieve better gripping shapes.
Limitations of this investigation
Try as one might, an investigation such as this is
fraught with error variance (factors unrelated to the focus of study). There
is so much difference in wood, settings and tuning, tuning abilities between
people, style of use, etc., it makes such a investigation almost futile.
For one, it would be simply overwhelming, both in
expense and time, to consider all the many suitable planes and the various
configurations available to the woodworker. For example, it would have been
desirable to have included one or more true English style infills (Norris,
Spiers, Preston, etc) as well as their more modern replicas. It would have
also been useful to examine at least several planes able to support multiple
back bevels, lets say an ECE/Primus, Lee Valley and Lie-Nielsen with back
bevels at 50, 55, 60, and 65 degrees.
There are advantages in maintaining consistency by using
one rater, but it would have been a stronger study if it had been possible
to use three raters using all the planes on all the woods.
There is such difference between samples of a given
wood, not to mention differences between woods, that all rankings must be
considered only crude approximations to what will likely be one’s own
experience. I tried to minimize this problem somewhat by basing ratings
always on at least two samples of a given wood, but this should not be
considered sufficient. To be truly valid, something more like eight samples
of each wood would need to be evaluated. A good example of why this is true
can be found by comparing the performance of planes used in this
investigation with those that were also used in my previous Lee Valley #4.5
review which included comparisons to the Lie-Nielsen #4, Lie-Nielsen #4.5,
Lie-Nielsen Low Angle and ECE/Primus Improved Smoothing planes. Now granted,
I didn’t employ such elaborate methodology in that earlier review, but it
is illustrative to compare the following words from the previous review to
the present findings:
Purpleheart. L-N LA did best here, followed by
L-N4.5 and V4.5. ECE/Primus prone to chatter on this wood. Not a hint of
chatter from the LV4.5 and virtually as good from the L-N4.5.
In contrast, in this investigation, the Lee Valley Low
Angle plane performed quite poorly on Purpleheart while the ECE/Primus
achieved a fairly good surface. I suspect the differences between then and
now is not the difference between the Lee Valley and Lie-Nielsen low angle
planes, but likely is the result of different samples of wood.
Woods, and the individual sample of wood before one, is
the primary arbiter of what plane will perform best at a given time!
Conclusions
Again, I want to reiterate, there were no “bad”
planes in this investigation. Everyone of these planes is well made, and
works very well indeed on some woods and for some purposes. However, not all
of these planes were as suitable to application on the dense, difficult,
sometimes heavily figured woods deliberately selected for this
investigation.
So, was there a "best" plane? In my mind there
certainly was. The Stephen Thomas infill planed at a superior level more
consistently than any other plane. Its ability to do that at a relatively
modest effective cutting angle implies that it will handle a wider variety
of woods than the higher angle planes. It felt good, it was easy to adjust,
it breezed through the densest woods, it made some of the finest and most
even of shavings; simply, it was just wonderful to use and a true asset as a
tool to quickly achieve the finest surface planing can accomplish. Alas,
wonderful comes at a very high price, one that exceeds all the other planes
in the investigation combined, and for more reasons than price alone, it
will be a plane of very limited availability. If you have the money for it,
though, it won’t disappoint and will likely even save a professional hand
tool user money in the long run.
Was there a best value? I think so. A Lee Valley #4.5
with an extra blade or two for back bevels is by far the best value in my
opinion, and is also one of the most versatile of planes (along with a
Lie-Nielsen with a 45 degree frog). For roughly $200 (including an extra
blade), you can have a level of performance I suspect can be brought to only
a little short of the best. Using back bevels, I expect the Lie-Nielsen #4.5
will offer much the same performance, but at about twice the price (again,
including an extra blade for back beveling). Still, not a bad price for such
a versatile plane.
Another plane strongly in the running is the ECE/Primus
Improved Smoothing Plane, with its adjustable mouth, it will likely respond
equally well to back bevels, and at a price comparable to the Lee Valley.
For me, the lessons most safely drawn from this
investigation (and for the reasons I express above, all speculation should
be done very cautiously) is that higher blade angles are generally useful
for denser woods, that back bevels work and are an inexpensive way to
achieve a range of effective cutting angles, and that a plane with an
adjustable mouth is almost essential to exploit the use of a back beveled
blade. It appears also, that there is something about the infill design,
that allows such planes to perform over a wide range of wood types, with but
only a modest increase beyond Standard pitch.
There is a saying in medicine to "treat the
patient, not the machine", (that is, pay attention to whether the
patient is breathing, not just whether the machine is saying the patient is
breathing, or not). I think a similar saying is a good way to end here,
"Use the plane that you find works best for the wood before you, not
the plane any review or investigation has suggested was the best plane for
their woods".
I’m sure I’ve failed to address some important
points, and simply don’t have the knowledge or wisdom to respond to
others. Now, on to some investigations of super high angle planes
(over 60 degrees) and how all this relates to scraper planes. Just don’t
expect the results for a while. :-)
