Houdini 20: The Little Things

[MJV]

12 hours ago, eikonoklastes said:

 

A camera switcher can be set up very easily already by using the same prim name for all your cameras and running them through a Switch LOP.

 

Unfortunately, no focus target option yet.

Thanks  It's not about there being an available workaround for me. It's more that it makes no sense whatsoever to me that they bothered to make a light mixer because Solaris is supposed to be where you stage stuff, but didn't think to make a camera switcher for the same reason. It doesn't make logical sense. Also, I know that there is a workaround for there not being a camera target yet, but that is also a rediculous oversight. These things do no instill confidence in me. Also, not being able to change the camera focul length in the editor? These are basic things that have been in 3d software for 30 years already.

[Mike A]

I think we have to consider the background here... Sidefx's traditional market, and priority, is high end VFX - as you well know.  The production pipeline for that is totally different to the daily work of a solo Indie artist. Modellers model, lighters light, animators animate etc. Yes, that means there are frustrations for the Indie artist with our do-it-all requirements. I think Sidefx do realise that they need to cater for these other workflows - Christin Bargiel noted the widening of the Sidefx customer base at the London presentation - so hopefully they will increasingly take that into account. But, I think if you use Houdini, you need to accept that as an Indie you're on the periphery of the current customer base and that will mean some frustrations.

 

My USD know-how is close to zero, but I think I read about some issues with USD and camera switching, so that may explain the current lack of that feature. 

 

I do think SideFX could do with focussing some more on quality of life / ease of use aspects - but, like many other developers I suspect they have a considerably longer list of things to do than resources to deliver them.

Edited November 12, 2023 by Mike A (see edit history)

[MJV]

5 hours ago, Mike A said:

I think we have to consider the background here... Sidefx's traditional market, and priority, is high end VFX - as you well know.  The production pipeline for that is totally different to the daily work of a solo Indie artist. Modellers model, lighters light, animators animate etc. Yes, that means there are frustrations for the Indie artist with our do-it-all requirements. I think Sidefx do realise that they need to cater for these other workflows - Christin Bargiel noted the widening of the Sidefx customer base at the London presentation - so hopefully they will increasingly take that into account. But, I think if you use Houdini, you need to accept that as an Indie you're on the periphery of the current customer base and with that will mean some frustrations.

 

My USD know-how is close to zero, but I think I read about some issues with USD and camera switching, so that may explain the current lack of that feature. 

 

I do think SideFX could do with focussing some more on quality of life / ease of use aspects - but, like many other developers I suspect they have a considerably longer list of things to do than resources to deliver them.

I completely agree, and you're talking to someone who thinks Houdini is a great and wonderous piece of software, I honestly love it, and at the same time it is everything you've said, which is a lot, but that's it. Talk of widening the customer base is a year old and R20 didn't budge even an inch in that direction. That's a little surprising to me but also perfectly understandable from their perspective, given their niche as you said, and I think it's not a matter of dealing with some frustrations or not but simply being clear eyed about what Houdini is and is not. It's a specialized tool for high end particle simulations, and related systems, covering a wide variety of special effects areas for studios and film production companies. Despite its power it was simple enough even for me to get up to speed with in under a year, with the help of people here, to build the pretty complex node based, user input driven model I had in mind for Houdini when I started.  A model that afaik would not even today be possible in Cinema because of a few specific limitations in Cinema's nodes system that were never developed or further addressed after the first technology preview, as far as I know. Unfortunately at the same time the most mundane take-it for-granted tasks that one might do in Cinema without even much thought suddenly require complex networks and multiple work space hierarchies in Houdini, leaving the user to educate themselves to understand why it's all necessary. You can't really do anything in Houdini slapdash or just good enough to get-the-job-done. Anything one might do has to be done well enough so that it works and follows the rules of the larger eco system and can be handed off to other departments as a complete asset. 

 

Unfortunately nothing I was hoping for ended up in Houdini 20, as far as I can tell.  I had a lot of hopes here and there but none of my wishes came true. 

[HappyPolygon]

20 minutes ago, MJV said:

I had a lot of hopes here and there but none of my wishes came true. 

 

Now I'm curious 

[Mike A]

9 hours ago, MJV said:

Unfortunately nothing I was hoping for ended up in Houdini 20

 

I empathise - it's frustrating when that happens. I was also hoping for lots of things that didn't arrive, but hey, that's how it goes for Maya / Max / C4D / Blender / Houdini... or any other. : )

 

The funny thing is I've just read a thread with various VFX artists who were 'disappointed with H20' because they felt their needs for even more powerful VFX features were being ignored : )  It must be tough being a developer : )

 

You've just got to go with the 'least bad' option - or a combination.  

 

Edited November 12, 2023 by Mike A (see edit history)

[Mike A]

Another welcome improvement:  A much improved HDA creation dialogue with better branch and version control.

Edited November 14, 2023 by Mike A (see edit history)

[eikonoklastes]

On 11/11/2023 at 9:47 PM, MJV said:

Thanks  It's not about there being an available workaround for me. It's more that it makes no sense whatsoever to me that they bothered to make a light mixer because Solaris is supposed to be where you stage stuff, but didn't think to make a camera switcher for the same reason.

 

It's the officially recommended way to do that in Solaris, so I don't believe that they are even considering putting in a camera switcher (and honestly, I agree with them on this).

 

On 11/11/2023 at 9:47 PM, MJV said:

Also, I know that there is a workaround for there not being a camera target yet, but that is also a rediculous oversight.

Fully agree with you on this, though. I would love to have an actual focus target object out-of-the-box in there myself.

 

On 11/11/2023 at 9:47 PM, MJV said:

These things do no instill confidence in me.

Like @Mike Asaid, I get the frustration on not getting things that you've been looking forward to. It does not detract from the sheer amount of things that were added, though, and those things instill an enormous sense of confidence in me. I'm not talking about the high-end creature FX stuff like feathers and APEX rigging, but the sheer volume of user-friendly additions and tweaks that were made to Solaris and Karma and the Graph Editor, in addition to the many, many little things they improved all over the place - I'm still discovering them as I continue to explore.

[MJV]

I was searching for something online, anything that might suggest R20 is worth a look and came across this: 

 

Possibly the greatest example of why one should not use Houdini ever made. What a train wreck. Poor guy.

 

And here is the only other review I found of R20 yet online: https://youtu.be/eyTVegGsEZE?si=RdSWUif_AbguseWY

[MJV]

12 hours ago, eikonoklastes said:

 

It's the officially recommended way to do that in Solaris,

 

I love the qualifier here, "in Solaris". So when does someone set up a camera "in Solaris",  and why? I found it difficult to figure out. I think I finally settled on doing it in Object level and ported the result through to Solaris which I had to use because of Karma.

 

(I edited this post to be more concrete and intelligible)

[Mike A]

That video may well be the greatest example of why you shouldn't believe everything you find on YouTube : )

It's a guy fumbling around not knowing what he's doing and switching on stuff that was clearly labelled as early experimental (why do you think the 'on switch' is buried as code in a text file? )

 

Just for the record:  from what I've used and seen over the last few days, I think it's probably the most impressive upgrade I've ever seen for any 3D software.

 

But if it's not to your taste, that's fine. C4D is great, use that.

 

 

Edited November 16, 2023 by Mike A (see edit history)

[eikonoklastes]

12 hours ago, MJV said:

 

I love the qualifier here, "in Solaris". So when does someone set up a camera "in Solaris",  and why? I found it difficult to figure out. I think I finally settled on doing it in Object level and ported the result through to Solaris which I had to use because of Karma.

 

Object level has a camera switcher built-in (it's called Switcher). Object level, however, has no concept of prims or prim names, whereas USD/Solaris does, and can leverage that to enable a camera switcher by using the same prim name for all the cameras that you want to switch between.

[eikonoklastes]

Steering this back on course, Karma CPU and XPU can both read point cloud data directly to do attribute transfers at render time. This is a massively useful feature that Mantra has, and I'm glad to see it finally come to Karma. This works independent of the mesh resolution, so it can allow for extremely efficient setups.

 

Here is a video describing it: