hey guys,
is it normal that adding layers would cause the ztl file size to double or triple?
i currently have a detailed model that has 6 subdivison levels and at around 40mil. i’ve tested that if i bake all my layers down, the ztl is only around 200mb. but with the layers (4 of them), the file is around 2gb.
just not sure if this is normal. it takes a long time to save the ztl each time and want to know if im doing anyting wrong.
thx in advance!
A Layer is like a duplicate of your model, so yes, they will always add to the file size.
thanks for the quick reply marcus_civis!
i’ve seen artists that use many layers for sculpting details like wrinkles and pores on their characters. so its normal that the ztl will end up over 2gb + ?
the huge file size and slow down in performance just makes me wonder if theres something i’ve done wrong in the workflow.
my model is 1 subtool for head, body, hands and legs. and other subtools are for eyes, nails, teeth, etc.
How are you saving your model?
just saving by Tool > Save as and saving the ztl (2gb file size)
ztool only consist of 3 subtools:
1 - full body (head, body, arms/hand, legs/feet) - 26mil points, 6 subd, 2 layers consisting of wrinkle and details
2 - eyes, mouth interior (teeth, tongue) - 18k points, 1 subd
3 - finger nails and toe nails - 5mil pts, 6 subd
don’t think i can seperate the head, body, arms and legs as different subtools right? since these should be connected. i do have polygroups for them tho, but still feels its heavy and performance is a bit lag when sculpting in subd 6 even when only the head polygroup is visible.
thx again!
You’re not doing anything wrong. Splitting up the body won’t make any difference to file size and although you might see a performance improvement it would make many operations more difficult.
I think the large file size must simply be that the layers are having to store a lot of data. I did a test with the DemoSoldier, increasing the res to 32 million and adding a few layers. The file size increased but not by a huge amount. But then I didn’t really do much on the layers, so obviously ZBrush only stores data for the points that have actually changed. If you’ve got polypaint and/or textures with the model then those will add to the file size.
26 million is quite a lot. Do you actually need all that? If you are generating displacement or texture maps then you’ll have to use 8K maps to have any hope of capturing all that detail.
yea the model is quite heavy with 26mil, but seems to be the only way to have the model dense enough to sculpt fine details like small wrinkles and skin details throughout the body, especially the head. I was provided with a base mesh that was retopologized in maya that starts at 25k poly. I brough that into zbrush and added details through each subdivison level.
Even at current subd6, there are spots on the face that are not dense enough to add finer wrinkle details, but i’ll have to skip those as its already too heavy for sure.
I will be generating displacement maps at 4k to be handed down to surface/texture.
there are technical side of things in the whole production workflow that I’m not familiar with, so thats why im thinking theres something wrong with my work process. Used to use zbrush for just sculpting and fun, but now that I’m working for cg production, requirements and process is quite different
If the face does not have enough density then you may lose detail there when you generate your maps. 4K won’t capture what you’ve got. It’s simple math - a 4K texture map has about 16 million pixels, so that is the maximum available. If each pixel stored the data from a single polygon then you could capture the detail from a 16 million polygon model. In practice the UVs never cover the whole of the texture map. Depending on the unwrapping method perhaps two thirds of the map will be actually available, so you’re really looking at getting around 11 million polys onto a map, so expect to see some of your detail gone.
Topology for a human model will normally have many more polygons for the face and other important areas, and fewer for the back and areas that matter less. This will also be reflected in the UVs which will give a larger area to the face. If your UVs are like that then perhaps the face detail will be mostly OK but you will lose detail elsewhere.
ahh thx! didnt kno about that for uvs. still lots to learn about the whole production of creating characters from zbrush to deliver for animation.
the topology is prepared with denser polygons in the face, hands and feet, so the details should be good.
i’ve figured out why my ztool file size is that large now altho dont really know the logic behind it.
if the layer is left on recording mode when saving… it causes the file to be over 2gb. if its out of recording mode, then the ztool is only 800mb
That’s interesting, thanks for letting us know.