I want to add support for PBR materials extension to OBJ, but the way
current I/O code syntax handling was done made it quite cumbersome
to extend the number of MTL textures/parameters.
Simplify all that by removing FormatHandler template on "syntax"
that gets routed through keyword enums, and instead just have
simple `write_obj_*` and `write_mtl_*` functions.
Simplify MTLMaterial to not contain a map of textures (that is always
fully filled with all possible textures), instead now there's
a simple array. Rename `tex_map_XX` to `MTLTexMap`.
All this does not affect behavior or performance, but it does result
in 170 fewer lines of code, and saves a couple kilobytes of executable
size.
When allocating new `CustomData` layers, often we do redundant
initialization of arrays. For example, it's common that values are
allocated, set to their default value, and then set to some other
value. This is wasteful, and it negates the benefits of optimizations
to the allocator like D15082. There are two reasons for this. The
first is array-of-structs storage that makes it annoying to initialize
values manually, and the second is confusing options in the Custom Data
API. This patch addresses the latter.
The `CustomData` "alloc type" options are rearranged. Now, besides
the options that use existing layers, there are two remaining:
* `CD_SET_DEFAULT` sets the default value.
* Usually zeroes, but for colors this is white (how it was before).
* Should be used when you add the layer but don't set all values.
* `CD_CONSTRUCT` refers to the "default construct" C++ term.
* Only necessary or defined for non-trivial types like vertex groups.
* Doesn't do anything for trivial types like `int` or `float3`.
* Should be used every other time, when all values will be set.
The attribute API's `AttributeInit` types are updated as well.
To update code, replace `CD_CALLOC` with `CD_SET_DEFAULT` and
`CD_DEFAULT` with `CD_CONSTRUCT`. This doesn't cause any functional
changes yet. Follow-up commits will change to avoid initializing
new layers where the correctness is clear.
Differential Revision: https://developer.blender.org/D15617
With the ultimate goal of simplifying drawing and evaluation,
this patch makes the following changes and removes code:
- Use `Mesh` instead of `DispList` for evaluated basis metaballs.
- Remove all `DispList` drawing code, which is now unused.
- Simplify code that converts evaluated metaballs to meshes.
- Store the evaluated mesh in the evaluated geometry set.
This has the following indirect benefits:
- Evaluated meshes from metaball objects can be used in geometry nodes.
- Renderers can ignore evaluated metaball objects completely
- Cycles rendering no longer has to convert to mesh from `DispList`.
- We get closer to removing `DispList` completely.
- Optimizations to mesh rendering will also apply to metaball objects.
The vertex normals on the evaluated mesh are technically invalid;
the regular calculation wouldn't reproduce them. Metaball objects
don't support modifiers though, so it shouldn't be a problem.
Eventually we can support per-vertex custom normals (T93551).
Differential Revision: https://developer.blender.org/D14593
While fixing T100302 (rBd76583cb4a1) I did not realize that the
change in imported vertex order would actually matter. Turns out, it
does for morph targets / mesh shape keys. So redo the fix in a way
that does not change the vertex order. Fixes T100421.
Report T98781 and part of T97642: the MTLMaterial info only captures
image nodes and the default socket values. When the image information
is present, do not emit the socket defaults - the .MTL spec states
they are multiplied together, but the default value is not used
in blender when the socket is connected.
Also contains svn tests repository update to extend the test coverage,
and update test expectation outputs.
As part of the previous fix (D15410), the importer got code to track
min & max vertex indices used as part of the mesh faces. However, if
faces refer to a "sparse" (i.e. non-contiguous) subset of all vertices,
then the imported mesh would contain all the vertices between min & max
range.
Replace that with proper tracking of actually used vertex indices
for each imported mesh. Fixes T100302.
This does affect import performance a tiny bit, e.g. importing Blender
3.0 splash scene goes 21.7s -> 22.1s, and importing rungholt.obj
goes 2.37s -> 2.48s.
Importer related tests have a bunch of vertex changes in them, since
now vertices are added in the order that the faces are referring
to them. Which incidentally matches the order that the Python based
importer was creating them too.
The Python based exporter was replacing spaces with underscores
in object/group names, mostly to handle cases where names could begin
or end with spaces. The new exporter was not doing that. Note: spaces
in material names were already handled by the new exporter.
Fixes T97769. Updated test coverage expectations; one of the test
files has an object with a space in the name.
Python based OBJ importer, as well as glTF2 importer, are creating
"placeholder" images for texture images that can't be found. These
are empty textures (displayed as magenta), but with their file paths
set so that File > External Data > Report Missing Files can report
them as missing.
Make the new C++ OBJ importer do the same as well. Fixes T99502.
The Python based importer had logic to immediately turn image paths
into relative-to-blender-file paths, if user preference for relative
paths is used (which is on by default). The new importer code did not
have that. Fixes T100076.
The new OBJ/MTL importer was creating a new image for any referenced
texture, even if another material (or another property of the same
material) already referenced the same texture. Make it use
BKE_image_load_exists function just like Collada or USD importers do.
Fixes T100075. Extended test coverage to count imported images;
without the fix import_cubes_with_textures_rel would have incorrectly
created 5 images instead of 4.
The Python based importer had a special case handling of "no faces in
the whole file at all", where it ended up treating the whole file
as essentially a point-cloud-like object (just loose vertices, no
faces or edges). The new importer code was missing this special case.
Fixes T100017. Added gtest coverage that was failing without the fix.
Previously, when creating "very large" (tens-hundreds of thousands)
amounts of objects, the Blender code that was ensuring name
uniqueness was the bottleneck. That got recently addressed (D14162),
however now sorting of IDs by their names is the remaining bottleneck.
Name sorting code in Blender is optimized for the pattern where names
are inserted in already sorted order (i.e. objects expect to get added
near the end of the list). By doing this pre-sorting of objects
intended to get created by an importer (USD and OBJ, in this patch),
this sorting bottleneck can be largely removed, especially with very
high object counts.
Windows, Ryzen 5950X, import times:
- OBJ, splash screen scene (26k objects): 22.0s -> 20.7s
- USD, Disney Moana scene (250k objects): 585s -> 82.2s (10 minutes -> 1.5 minutes)
Reviewed By: Michael Kowalski, Howard Trickey
Differential Revision: https://developer.blender.org/D15506
The importer code was written under incorrect assumption that vertex
data (v, vn, vt commands etc.) are grouped by object, i.e. follow the
o command, and that each object has its own vertex data commands. This
is not the case -- all the vertex data in the whole OBJ file is
"global", with no relation to any objects/groups; it's just that the
faces belong to the object, and then they pull in any vertices they
like.
This patch fixes this incorrect assumption in the importer:
- Vertex data is now properly global; no need to track some sort of
"offsets" per object like it was doing before.
- For each object, face definitions track the minimum & maximum vertex
indices referenced by the object, and then all that vertex range is
created in the final Blender object. Note: it might be (unusual, but
possible) that an object does not reference a sequential range of
vertices, e.g. just a single face with vertex indices 1, 10, 100 --
the resulting Blender mesh will have all the 100 vertices (some
"loose" without belonging to a face). It should be possible to track
the used vertices exactly (e.g. with a vector set), but I haven't
done that for performance reasons.
Reviewed By: Howard Trickey
Differential Revision: https://developer.blender.org/D15410
Address the issue by re-working line continuation handling: stop
trying to parse sequences like "backslash, newline" (which is the
bug: it should also handle "backslash, possible whitespace, newline")
during parsing. Instead, fixup line continuations after reading chunks
of input file data - turn backslash and the following newline into
spaces. The rest of parsing code does not have to be aware of them
at all then.
Makes the file attached to T99536 load correctly now. Also will extend
one of the test files in subversion tests repo to contain backslashes
followed by newlines.
As pointed out in a comment on T89421, if a MTL file contained
something like: `map_Ka -o 1 2.png` then it was parsed as having
offset `1 2` and the texture filename just a `.png`. Make it so that
mtl option numbers are parsed in a way where the number is only
accepted only if it's followed by whitespace.
Differential Revision: https://developer.blender.org/D15385
The new OBJ/MTL importer was already handling case T89421
correctly, but there was no test coverage to prove it. Extend
the tests to parse various forms of "-o" and "-s" (one, two, three
numbers).
The old Python OBJ importer had a (somewhat confusingly named) "Keep
Vertex Order -> Poly Groups" option, that imported OBJ groups as
"vertex groups" on the resulting mesh. All vertices of any face were
assigned the vertex group, with a 1.0 weight.
The new C++ importer did not have this option. It was trying to do
something with vertex groups, but failing to actually achieve
anything :) -- the vertex groups were created on the wrong object
(later on overwritten by "nomain mesh to main mesh" operation);
vertex weights were set to 1.0/vertex_count, and each vertex was only
set to be in one group, even when it belongs to multiple faces from
different groups. End result was that to the user, vertex groups were
not visible/present at all (see T98874).
This patch adds the import option (named "Vertex Groups"), which is
off by default, and fixes the import code logic to actually do the
right thing. Tested on file from T98874; vertex groups are imported
just like with the Python importer.
Reviewed By: Howard Trickey
Differential Revision: https://developer.blender.org/D15200
OBJ vertex color related tests were not producing identical results
across various platforms, primarily due to sRGB<->Linear color space
conversions.
While D15193 has just made the color space conversion accuracy match
much closer between platforms, it's still not 100% the same.
This change reduces the amount of decimal places used for exporting
vertex colors, to 4 digits (down from 6). Vertex normals were
already always printed with 4 digits, and colors are conceptually
similar (usually 0..1 range etc.).
This makes the vertex color tests pass again, so re-enable them
after adjusting to 4 decimals expectations.
Adds support for vertex colors to OBJ I/O.
Importer:
- Supports both "xyzrgb" and "MRGB" vertex color formats.
- Whenever vertex color is present in the file for a model, it is
imported and a Color attribute is created (per-vertex, full float
color data type). Color coming from the file is assumed to be sRGB,
and is converted to linear upon import.
Exporter:
- Option to export the vertex colors. Defaults to "off", since not
all 3rd party software supports vertex colors.
- When the option is "on", if a mesh has a color attribute layer,
the active one is exported in "xyzrgb" form. If the mesh has
per-face-corner colors, they are averaged on the vertices.
Colors are converted from linear to sRGB upon export.
Reviewed By: Howard Trickey
Differential Revision: https://developer.blender.org/D15159
To match the existing Python .obj importer, and to make it easier for
the user to determine which object is which, use the filename for the
default object name instead of "New object".
Differential Revision: https://developer.blender.org/D15133
A new experimentatl STL importer, written in C++. Roughly 7-9x faster than the
Python based one.
Reviewed By: Aras Pranckevicius, Hans Goudey.
Differential Revision: https://developer.blender.org/D14941
The OBJ parser was primarily using StringRef for convenience, with
functions like "skip whitespace" or "parse a number" taking an input
stringref, representing an input line, and returning a new stringref,
representing the remainder of the line. This is convenient, but does
more work than strictly needed -- while parsing, only the "beginning"
of the line ever changes by moving forward; the end of the line
always stays the same. We can change the code to take a pair of
pointers (begin of line, end of line) as input, and make the
functions return the new begin of line pointer. This makes the return
value neatly fit into a processor register, which StringRef did not.
On Windows, this does result in non-trivial speedups in the actual
OBJ file parsing part, due to Windows calling convention where return
values larger than 64 bits are returned via memory. Does not
measurably affect performance on Mac/Linux, because the calling
convention there uses a pair of 64-bit registers to return a
StringRef.
End-to-end times of importing several test files, on Windows
(VS2022 build, Ryzen 5950X):
- Monkey subdivided to level 6, no normals (220MB file): 1.25s -> 0.85s
- Rungholt minecraft level (270MB file): 7.0s -> 5.8s
- Blender 3 splash scene (2.4GB file): 49.1s -> 45.5s
The full import process has a lot of other overhead besides actual
OBJ file parsing (mostly creating actual blender objects out of
parsed data). In pure parsing, in the monkey test scene above, the
parsing part goes 1.0s -> 0.6s.
Reviewed By: Howard Trickey
Differential Revision: https://developer.blender.org/D14936
New OBJ exporter is missing "Path Mode" setting for exporting .mtl
files. The options that used to be available were: Auto, Absolute,
Relative, Match, Strip Path, Copy. All of them are important. The new
behavior (without any UI option to control it) curiously does not match
any of the previous setting. New behavior is like "Relative, but to the
source blender file, and not the destination export file".
Most of the previous logic was only present in Python based code
(bpy_extras.io_utils.path_reference and friends). The bulk of this
commit is porting that to C++.
Reviewed By: Howard Trickey
Differential Revision: https://developer.blender.org/D14906
Fixes T97794 (which is a reintroduction of an older issue T67266 that
has been fixed in the python importer, but the fix was not in the C++
one). Some software produces OBJ files with mtllib statements like
mtllib "file name in quotes.mtl", and the new importer was not stripping
the quotes away.
While at it, I noticed that MTLParser constructor was taking a StringRef
and treating it as a zero-terminated string, which is not necessarily
the case. Fixed that by explicitly using a StringRefNull type.
Reviewed By: Howard Trickey
Differential Revision: https://developer.blender.org/D14838
Even if available OBJ/MTL format documentations don't explicitly specify
which characters can possibly separate keywords & arguments, turns out
some files out there in the wild use TAB character after the line
keywords. Which is something the new 3.2 importer was not quite
expecting (T97417).
Fix this by factoring out a utility function that checks if line starts
with a keyword followed by any whitespace, and using that across the
importer. Also fix some other "possible whitespace around name-like
parts" of obj/mtl parser as pointed out by the repro files in T97417.
Reviewed By: Howard Trickey
Differential Revision: https://developer.blender.org/D14782
Continued improvements to the new C++ based OBJ importer.
Performance: about 2x faster.
- Rungholt.obj (several meshes, 263MB file): Windows 12.7s -> 5.9s, Mac 7.7s -> 3.1s.
- Blender 3.0 splash (24k meshes, 2.4GB file): Windows 97.3s -> 53.6s, Mac 137.3s -> 80.0s.
- "Windows" is VS2022, AMD Ryzen 5950X (32 threads), "Mac" is Xcode/clang 13, M1Max (10 threads).
- Slightly reduced memory usage during import as well.
The performance gains are a combination of several things:
- Replacing `std::stof` / `std::stoi` with C++17 `from_chars`.
- Stop reading input file char-by-char using `std::getline`, and instead read in 64kb chunks, and parse from there (taking care of possibly handling lines split mid-way due to chunk boundaries).
- Removing abstractions for splitting a line by some char,
- Avoid tiny memory allocations: instead of storing a vector of polygon corners in each face, store all the corners in one big array, and per-face only store indices "where do corners start, and how many". Likewise, don't store full string names of material/group names for each face; only store indices into overall material/group names arrays.
- Stop always doing mesh validation, which is slow. Do it just like the Alembic importer does: only do validation if found some invalid faces during import, or if requested by the user via an import setting checkbox (which defaults to off).
- Stop doing "collection sync" for each object being added; instead do the collection sync right after creating all the objects.
Cleanup / Robustness:
This reworking of parser (see "removing abstractions" point above) means that all the functions that were in `parser_string_utils` file are gone, and replaced with different set of functions. However they are not OBJ specific, so as pointed out during review of the previous differential, they are now in `source/blender/io/common` library.
Added gtest coverage for said functions as well; something that was only indirectly covered by obj tests previously.
Rework of some bits of parsing made the parser actually better able to deal with invalid syntax. E.g. previously, if a face corner were a `/123` string, it would have incorrectly treated that as a vertex index (since it would get "hey that's one number" after splitting a string by a slash), instead of properly marking it as invalid syntax.
Added gtest coverage for .mtl parsing; something that was not covered by any tests at all previously.
Reviewed By: Howard Trickey
Differential Revision: https://developer.blender.org/D14586
- Was not exporting "Poly" curves at all,
- Had a crash when a single object contains multiple curves of different types -- it had a check for "is this nurbs compatible?" only for the first curve, and then proceeded to treat the other curves as nurbs as well, without checking for validity.
Fixed both issues by doing the same logic as in the old python exporter:
- Poly curves are supported,
- Treat object as "nurbs compatible" only if all the curves within it are nurbs compatible.
Added test coverage in the gtest suite. While at it, made "all_curves" test use the "golden obj file template" style test, instead of a manually coded test that checks intermediate objects but does not check the final exported result.
Reviewed By: Howard Trickey
Differential Revision: https://developer.blender.org/D14611
The new 3.1 OBJ exporter code had incorrect code to determine which vertex group a polygon belongs to -- for each vertex, it was only looking at the first vertex group it has, and not using the group weight either.
This 99% fixes T96824, but not 100% on the user's submitted mesh -- exactly two faces from that mesh get assigned a different group compared to the old exporter. Either choice is "correct" given that on these two faces there are two vertex groups with equal contribution. The old Python exporter was picking the group based on internal python group name map order, whereas the new C++ exporter is picking the group with the lowest index, in case of ties. I'm not sure if it's possible to fix this TBH, will have to wait until the importer is also C++.
While at it, the new vertex group calculation code was doing a lot of redundant work for each and every face (traversing group lists several times, allocating & freeing memory), so I fixed that. Exporting a 6-level subdivided Monkey mesh with 30 vertex groups was taking 810ms, now takes 330ms.
Reviewed By: Howard Trickey
Differential Revision: https://developer.blender.org/D14500
The all_objects.blend test scene (in subversion tests repo) contained an
object with a subdivision surface. Which changes vertex positions
slightly, depending on used OpenSubDiv version and the compile flags. It
seems that the intent of the test was "test export of meshes that use
modifiers", so I changed that object to be a cube with a simple "taper"
modifier instead.
While at it, changed OBJ exporter test code to always print the
"expected and what we got" text difference details, when a test fails.
Much easier to see than just "the files are different" output. The code
to print that was behind an off by default flag for some reason.
This diff should get comitted together with updated all_objects templates
in subversion tests repo.
Reviewed By: Sebastian Parborg
Differential Revision: https://developer.blender.org/D14597
This commit furthers some of the changes that were started in
rBb9febb54a492 and subsequent commits by changing the way surface
objects are presented to render engines and other users of evaluated
objects in the same way. Instead of presenting evaluated surface objects
as an `OB_SURF` object with an evaluated mesh, `OB_SURF` objects
can now have an evaluated geometry set, which uses the same system
as other object types to deal with multi-type evaluated data.
This clarification makes it more obvious that lots of code that dealt
with the `DispList` type isn't used. It wasn't before either, now it's
just *by design*. Over 1100 lines can be removed. The legacy curve
draw cache code is much simpler now too. The idea behind the further
removal of `DispList` is that it's better to focus optimization efforts
on a single mesh data structure.
One expected functional change is that the evaluated mesh from surface
objects can now be used in geometry nodes with the object info node.
Cycles and the OBJ IO tests had to be tweaked to avoid using evaluated
surface objects instead of the newly exposed mesh objects.
Differential Revision: https://developer.blender.org/D14550
This takes state of soc-2020-io-performance branch as it was at
e9bbfd0c8c7 (2021 Oct 31), merges latest master (2022 Apr 4),
adds a bunch of tests, and fixes a bunch of stuff found by said
tests. The fixes are detailed in the differential.
Timings on my machine (Windows, VS2022 release build, AMD Ryzen
5950X 32 threads):
- Rungholt minecraft level (269MB file, 1 mesh): 54.2s -> 14.2s
(memory usage: 7.0GB -> 1.9GB).
- Blender 3.0 splash scene: "I waited for 90 minutes and gave up"
-> 109s. Now, this time is not great, but at least 20% of the
time is spent assigning unique names for the imported objects
(the scene has 24 thousand objects). This is not specific to obj
importer, but rather a general issue across blender overall.
Test suite file updates done in Subversion tests repository.
Reviewed By: @howardt, @sybren
Differential Revision: https://developer.blender.org/D13958
Original report (T96763) only reported the issue of double-space before the texture path, but while adding test coverage I found some other issues that I fixed while at it:
- Incorrectly emits two spaces between `map_Xx` keyword and the texture path, leading to some 3rd party software not finding the textures,
- Emissive texture map (`map_Ke`) was not exported,
- When Mapping node is used on the texture UVs, the "Location" and "Scale" values were mixed up (location written as "scale", scale written as "location).
Added gtest coverage.
Reviewed By: Howard Trickey
Differential Revision: https://developer.blender.org/D14519
On Windows/MSVC this gives a minor (~20%) speedup presumably due to a faster float/int formatter. On macOS (Xcode13), this gives a massive speedup, since snprintf that is in system libraries ends up spending almost all the time inside some locale-related mutex lock.
The actual exporter code becomes quite a bit smaller too, since it does not have to do any juggling to support std::string arguments, and the buffer handling code is smaller as well.
Windows (VS2022 release build, Ryzen 5950X 32 threads) timings:
- Blender 3.0 splash scene (2.4GB obj): 4.57s -> 3.86s
- Monkey subdivided level 6 (330MB obj): 1.10s -> 0.99s
macOS (Xcode 13 release build, Apple M1Max) timings:
- Blender 3.0 splash scene (2.4GB obj): 21.03s -> 5.52s
- Monkey subdivided level 6 (330MB obj): 3.28s -> 1.20s
Linux (ThreadRipper 3960X 48 threads) timings:
- Blender 3.0 splash scene (2.4GB obj): 10.10s -> 4.40s
- Monkey subdivided level 6 (330MB obj): 2.16s -> 1.37s
The produced obj/mtl files are identical to before.
Reviewed By: Howard Trickey, Dalai Felinto
Differential Revision: https://developer.blender.org/D13998
This is patch D14349 from Aras Pranckevicius.
The logic in the code was _completely different_ from the documentation
and what the python exporter in 3.0 did. The new code assumed that
"export material groups" meant "append material name to the object name",
and was only ever kicking in when the "export object groups" option was
also checked. But the proper behavior (as in 3.0 exporter & the online docs),
is to emit g objectname_materialname before each usemtl line. Which is something entirely else.
This is patch D14347 from Aras Pranckevicius.
Instead of scaling "the scene" (i.e. transform vertices by object matrix,
then multiply by scale factor), it was instead first applying the scale
factor in local space, and then transforming by the object matrix.
This applies patch D14343 from Aras Pranckevicius, with a description:
The new 3.1+ OBJ exporter did not have correct logic when faced with
non-uniform & mirrored (negative on odd number of axes) object scale:
- Normals were not transformed correctly (should use inverse transpose of the matrix),
and were not normalized,
- Face order was not "flipped" when transform has negative scale on odd number of axes
(visible when using "face orientation" viewport overlay).
Use a shorter/simpler license convention, stops the header taking so
much space.
Follow the SPDX license specification: https://spdx.org/licenses
- C/C++/objc/objc++
- Python
- Shell Scripts
- CMake, GNUmakefile
While most of the source tree has been included
- `./extern/` was left out.
- `./intern/cycles` & `./intern/atomic` are also excluded because they
use different header conventions.
doc/license/SPDX-license-identifiers.txt has been added to list SPDX all
used identifiers.
See P2788 for the script that automated these edits.
Reviewed By: brecht, mont29, sergey
Ref D14069
Fixes T95384. New exporter was missing a fix for T94516 that recently got applied to the python exporter.
Also changed the obj export tests code so that when save_failing_test_output is requested and MTL result is different from the golden expectation, it is saved as well, similar to how it's done for the OBJ file result.
Also fixed conflicts due to the change in file writing in the new obj exporter
in master, and fixed one of the tests that was added in master but not 3.1.
