OK it’s not perfect but why do we need retopo? To use a model easier (+ zRemesh), decrease quality for animation or games (+ zRemesh),
easier unwrap it (+ zRemesh), maybe easier to texture paint (+ zRemesh), easier multires - I don’t get this one. If you already came from a top level sculpting software to retopo your model why do you want to sculpt it again? 
But (+ zRemesh). What else? This is not perfect, but it saves hours and hours for game developers, sculpters (yes you Michalis love your sculpts by the way) Of course manual workflow will give you more freedom - that’s how nature works… well technically. Anything I haven’t named?
@ArasTM when talking about a good mesh for animation you talk about a mesh that deforms the way you want it to. It’s not simply a matter of “decreasing the quality a little”, if you try to animate a zRemeshed mesh you might try to raise an arm only to discover that your entire shoulder flattens out because of bad geometry. Also when talking about production-quality animations for games or film there’s no settling for lower quality because the artist got lazy. They want perfect results regardless what workflow you think is the best or the fastest, if you can’t satisfy their needs they’ll hire someone else to do your job.
Not to sculpt the whole thing again. Shrinkwrapiing (zb project) will adopt all these already sculpted details.
Pores, wrinkles etc etc will be added though. You need 4 M or more for such high freq details. For alpha based brushes - projections…
An evenly distributed quad mesh, only retopology+multires can provide it.
So, you press a button called z remesher and you have the topology you need for fine sculpting.
This is how I understand it.
Another retopology is needed for low poly assets, animation movies ask for this, game engines for that one, normal map baking for something else… what should we expect from an auto retopology application? Let’s be reasonable.
Z remesher is a fine tool. Pixologic always provides such fine tools. Always close to an artist’s needs. Without asking more and more money.
Sorry ninthjake, same time posting.
@NinthJake, @Michalis, imagine small (2 - 15 people) indie dev team. They are broke students which need budget, kickstarter doesn’t work because it’s not USA (oh the mighty (not so mighty (I should use less of these) nation)) and they want the budget fast. So what they could do is realease a small alpha/beta of the game, but that requires some time, the faster they get it the better. Now out of those 2 - 15 people there are 1 - 4 3d artists which makes ~5 assets per hour and they could make a character in 5 - 10 hours without retopo. Now let’s say the game artists want it in one month (crazy) and they need a lot of assets plus characters, let’s say it’s simillar to skyrim so it has a lot of characters. Either it’s 7 hours per character with zRemesher or it is 12 hours per character. If they want to make 100 characters (not that much) it’d take 1200 hours which is ridiculously a lot (50 days), or it could be 700 which is 29 days - exactly on time but no sleep://. Not all game devs require perfect models or assets, in fact most (percentage) companies want more assets that looks fine but not perfect or manually retopologed. In movies it could varry, but still there are low budget made animations and there’s titanic. A lot of text I’ve put here to illustrate it for no clear reason, well whatever.
You’re talking like the mis-aligned spiral fests zremesher sometimes spits out are still useable, just not completely optimal. That’s not the case. Sometimes they’re just downright unusable. The lost time retopoing gets made up later when you can just smoothly work on UVs and skinning instead of wasting hours tearing out your hair because the topo is still all wrong.
Like I said, a separate guide where you mark two loops, for instance one at the shoulder and one at the wrist, and the algorithm MUST keep a series of separate loops in between, would solve 99% of the gripes about ZRemesher. Unless you’re doing a very specialized character, there are very few places on the body where topology is a huge dealbreaker. Fingers, eyes, mouth, elbows/knees, and shoulders/hips. As long as everything else is relatively well distributed and has a decent surface flow, there shouldn’t be any huge issues. I feel like the above implementation would be fairly simple for Pixologic to implement.
This first time I ever used Zremesher I had spirals all over the shop. Pretty cool but also useless (and ugly).
@coljwood
There is some workaround on this issue. Group cuts etc.
The problem is: It takes some more time.
The question is: How much time are you ready to spend on a so and so topology? More than 10 mins?
If it turns to be more than 10 mins we should better go for manual retopology.