Skip to main content

CLO Help Center

How can we help you?


  • silvialis

    I have the same doubt, How can I create a thick fabric like a knitted fabric? I am looking for a thicker jumper...and my final renders looks so thin..

  • silvialis

    And another issue, when I add a chest embroidery in my sweater and I add Displacement Map, the logo disappear.. check my pictures. Do you know anyway to add Displacement and and not to lose the logo? 

    Thanks a lot!

  • ottoline


    When you add the embroidery logo you have to ensure the distance and hence grey scale on the image map is appropriate for the combined texture you have underneath, along with it's depth offset of the base fabric texture (rib knit).  



    _ _ _


    For knitwear you need great textures and matching frequencies as a start point, these often are not the domain of Adobe Substance as they do really poor weave structure (2D) as apposed to 3D in general as the light off the yarn doesn't have the levels of frequency needed for the complexity of yarn twisting at the many light angles you would find on knotted media (light transport within yarn is critical). They are kind of lifeless and flat when created, so my suggestion is always to toss out Adobe Substance as a start point it's a bit of a red herring for knitwear or going to make matters complicated. Shift to a true yarn based weave simulator (even blender can do this basic deconstruction) as a start point, then make sure you generate the right complimentary digitized texture maps.


    See below true knit weaves are made up of simple geometric data

    It's quite easy to generate these 

    Then add in the noise frequencies and presto bake down all the correct maps for a great 2D knotted texture simulation.



    You can make it do almost anything ... in knots. This level of custom knot frequency (at speed) is impossible in Adobe Substance as it also only deals with one side of the fabric texture, so right tools for the job will make the textural map generation for the knotted knit so much better, as the transmission level in any weave contributes to the noise frequency for an organic woven textile structure.

    _ _ _ _

    And where you use scan digitization as a start point you want the virtual weaving to match the same level of detail so if you use a combination you cannot tell one start point from the other.

    Below weave digitization and then A.I. deconstruction and digital reconstruction so the organic small sample seamlessly becomes the large sample. 

    These images use an entirely new digitization process that we developed to capture weave and detail, unlike traditional scan digitization for fabric texture lighting where you need many multiple images to reconstruct a texture which can be slow to process, the approach  I developed extracts the full detail of the weave complexity in a single image (one unconventional digital image) with infinite number of light directions present in 1/1000th of a second, and with zero camera lens distortion as there is no lens !  A breakthrough in fabric texture process and capture. Fast, blindingly accurate, amazing light frequency and low cost hardware - under $300 verse over $20,000.  So it's not the hardware it's how you think through frequency, and approach at the get-go.

    My suggestion is if you want a lively knit wear texture for render you need to place the correct 'knit' 'knot' frequency and noise associated with that textile.

    You also want perfect masking of the knots as this allows you to push the depth/displacement to get the organic look all structured weaves have. 

    A great normal map also makes a huge difference.

    At all times I keep maximum frequency in.


    So 'wooliness' of a fabric is directly proportional to the 'data' on the yarn strand that you place in, rather than chop out, and that combines with the frequency of the randomness spread as you reconstruct a digitized textile, all the stuff that gets tossed out when resolution is not set properly for the fabric type. The actual depth map can then be relatively simple.

    As the knot complexity (open nature of weave) goes up, then the need to hold masking to the weave pitch becomes significant, as does curvature on the secondary (rib) frequencies that many fabrics exhibit at distance viewpoints.

    It's this play of frequencies in deconstruction and then the many maps for reconstruction that allow you to introduce factors like 'silky shimmer' and 'softness' (light transport via varying weave optics in the warp weft), organic randomness within a loose yet defined WIF weave pattern 'structure'.  eg: how much the warp weft deviate across a fabric off the loom.

Please sign in to leave a comment.