DarkDreamsVR
patreon
Blender, and the possibility of a new SFMVRSTITCH
🕑 Added 2019-04-10 13:07:10 +0000 UTCAbove is a teaser pic of a possible SFMVRSTITCH program that removes the issues in our current model, such as stretching and pixelization. I had someone (dd209) ask a question on the front page, and I thought this would be a good time to address it to everyone as well as bring up the potentially exciting news. Here is the question:
"Hey DD, I can't remember if I've asked this question before (bad memory), but is there an inherent benefit to VR scenes that were originally made in SFM over those made in Blender? I ask because it seems like Blender VR scenes are much "fuzzier" than SFM scenes, making them look a bit grainy. I think there are some fantastic Blender scenes that look almost photorealistic, but they don't seem as good in VR."
Here is my response:
Hey dd209. I am not sure if you asked this before either, but you're not the first. There are plenty of reasons to go with SFM over Blender at the moment, at least for VR animations.
The most basicand common reason I give is the time needed for a render. SFM's rendering process is not as complicated as Blender's, and is extremely fast in comparison. If I were to attempt Blender animations instead of SFM, you would see our weekly animations go from 4+ a week to 1 at most. The length of Blender animations would also pale in comparison. I set a goal at the beginning of this year to have our animations maintain 15 seconds without looping, as well as 30 seconds of sound without looping. This means for every animation we put out you are getting 15 fresh seconds of video (at 60 fps in 4k+) and 30 fresh seconds of audio IN VR. There aren't any other artists or studios out there that can boast that 4 times a week (let alone once a week including exclusives, remaster, and studio release that are minutes in length). If we switched to Blender, you would see an extremely drastic change in pure seconds released each week. I have continued to raise the bar each month for Dark Dreams, and that isn't going to stop anytime soon. But as I see it, switching to Blender would not "raise the bar". Especially when you compare our animations to the Blender ones that are out there.
This isn't the only reason, however, that we continue to use SFM. SFM has an ever-growing database of models, maps, sound, textures (and more!) that we not only use, but add to at SFMLab. Yes, there is a plethora of models out there that aren't SFM, but SFM has a distinct advantage in relation to anything rule 34. SFM has new models of this type every day, that are free and well organized, that Blender does not.
Another reason to stick with SFM is Blender would require our team to all put down SFM and take time to learn Blender itself. Yes, this will happen someday, as I have set a goal of us learning Blender. I feel that VR will grow and as it does, so will it's quality and demand. I don't think this will be anytime soon, but I do understand it is a necessity. It's one of the reasons our next goal for the Patreon is to get a secondary render computer. Not only would we have an easier time rendering SFM animations, we would also have the ability to set one computer to render for days if needed (and with length of Blender animations that is often the case).
Simply put, switching to Blender is not something we can do AND successfully maintain our current growth. We are slow growing, as is VR, so halting our progress by switching to Blender is not beneficial to us or our patrons. Blender has the capability to put out amazing animations that are photorealistic, but only if they are done correctly. Often times the animations I see from Blender are lit incorrectly, or in a manner that isn't aesthetically pleasing... Or they are fuzzy, as you put it... Or the camera work is wrong... These are all aspects I have been working on every day for years, and I feel I have done a good job of growing in those areas (there is always room for improvement, though). But even if I have these parts down, I would give up so much pure time in actual video releases just to raise the quality a little bit. It's just not worth it.
Besides, we may or may not have a new version of SFMVRSTITCH in the works as we speak. I can't make any promises, but if it does happen we would see better looking SFM animations in VR. Here's a teaser. Notice the lack of stretching and pixelization that is typical of our older animations?
https://sfmvr-stitch.neocities.org/imgs/thenewversion.jpg
Generally speaking I have figured out what exact settings cause the stretching and pixelization. But with a new iteration of SFMVRSTITCH, I wouldn't need to worry about these settings anymore and my hands wouldn't be constrained with the way we render or set up our videos. This means that our animations would improve EVEN MORE so in quality. This is all assuming this program happens (I am hoping to have a demo tested by the end of the month). So, keep your fingers crossed, as things are only looking better for Dark Dreams.
As you can see, things are still growing and looking better for Dark Dreams and VR in general. While I think a significant part of the reason our animations look so good is the post work we put into them, this program would take them a step further. For now it's just a matter of time and patience. As for Blender, I am not against it down the road. I just don't see it happening right now. Especially with the prospect of this new stitching program for us to use. Here's to hoping it happens. If it doesn't, that's okay. We'll continue doing what we do best in the vein of raising VR awareness.
-Dark Zalivstok
Comments
Dark Dreams
I am not sure how much of the post you read, but in it I explained that the photo is a teaser pic of a possible update to the SFMVR_STITCH program that will eliminate some of it's previous issues. Assuming all goes to plan, I am hoping to have a beta for it by the end of the month to test.
Moxige
The SFMVR_STITCH version is different from the version I am using. Is this new?
