This topic is available to share feedback or report issues on the feedback survey published at https://survey.blender.org/feedback/2024
Eevee Next is mentioned multiple times in the official summaries as a pain point or an area receiving widespread negative feedback. It’s also mentioned that users are more happy with Eevee Legacy and have performance issues with Eevee Next. How does this official acknowledgement based on community feedback of the problems with Eevee Next affect the development plans and roadmap for Eevee Next?
2 Likes
The EEVEE team has been working hard on performance and reliability in the past months. The ongoing Vulkan work and related improvements will also likely improve performance (it’s often faster in my tests anyway).
18 Likes
It is worth mentioning that the survey is from October 2024, and a lot of work has gone into EEVEE/Vulkan since then. We try to provide some context for the data from the survey, but can’t address every single point.
10 Likes
Some of the Eevee Next feedback was performance related, that’s true, and the current performance improvement work is very much appreciated 
However, I’m not seeing in the data that the negative feedback was exclusively regarding the performance of Eevee Next. What about all the regressions from Eevee Legacy that were reported on the bug tracker but closed as not being in scope for the project? If the official answer is “our plan to address this feedback is to focus only on performance”, that’s fine, I’d just appreciate some clarity on if I’m understanding the response correctly
2 Likes
I did write “and reliability”. Hundreds of bugs have been fixed since the survey, and there have been recent improvements to shadows, light leaking, depth artifacts, etc. It was not just performance, and will not just be performance in the future.
13 Likes
Wow, there were a survey…
In 6.2.1 Projects to focus on, what is the project “Complexity” referring to? It’s fairly high up but I can’t find info about it.
Handling very large scenes basically, lots of objects, dense meshes, lots of materials, instances, etc.
7 Likes
Wow, that is a very good summary of the survey, lots of work must have gone into it ! Thx a lot.
Love all the “Feedback Response” you have include.
It is nice that the online store will be revive.
I think making “Plushies (Suzanne, default cube, open movie characters)” is a good idea.
I own the 2 figurines of Ellie and Rex but I think those (as much as they are very well made, cool and Ton really like them :D, are a bit “niche”, or maybe it is just a trend thing)
It seem to me plush are more popular, as we can see by the recent success of the Godot Plush campaign.
Now good luck shipping thousand of them, probable why they outsource it to Makeship 
3 Likes
I just learned of this survey from the Blender Today post on youtube. I think its odd that, as a donor to both the blender dev fund and blender studio, there were no emails or communication regarding a survey. I use blender professionally every day and had no clue lmao.
I think in the future it could be wise to specifically alert donors/subscribers to the existence of such a survey instead of it just being a blog post.
8 Likes
woah, this was a really interesting insight, although i wish the sample size was a bit bigger maybe
other than that it was a fun read, sad that i missed this survey when it was going on
speaking of Increased Visibility & Marketing, one point mentioned was
- Collaborating with influential content creators (YouTubers, streamers) for fundraising events or mentions.
and one thing that came to mind was a fairly recent podcast with markiplier (a prominent content creator and filmmaker) i recently watched in which he was talking about his usage of 3d software like houdini and blender in his current film project and how he’s learning both of them especially when it comes to procedural stuff/nodes
he mentioned that he’s open to collaborations and that his DMs are open so i feel liek perhaps that’s an avenue when it comes to increasing visibility imo
anyways, thanks for your work on this survey^^
from personal experience they definitely are, unlike t-shirts/cups/figurines people usually buy them more frequently and the cuteness factor definitely makes it a more desirable piece of merch
especially if you have a few characters, say, from sprite fright, the “plush collectors” among us would be more likely to buy a whole set imo
1 Like
Was a freeform text field available for the multiple choice questions? I noticed that for Contribution to Blender the third highest category was “Other” and was mildly curious if we had any further data there.
While I certainly appreciate the openness of the dataset and recognize a certain balance between openness and censorship will have to be struck, i feel like some additional filtering could have been done, Juicy black ball all over the n***** screen is certainly a comment i could have done without (this one is uncensored in the dataset)
I don’t know, almost 7000 responses I think is a pretty darn good result for a survey, pretty much any survey. Combined with the fact it seems many that would have taken part, didn’t even know about it.
Still, as interesting a read as it all was, what will really be interesting is after a couple more and one can look at the trends, etc.
1 Like
I updated the downloadable zip and included a README with info about the license, and a note about the raw data containing potentially offensive language. Thanks for the feedback!
2 Likes
I think the sample size was… okay. bigger is better… we have ten million users or so as a not-quite conservative number I think, 7,000 results is probably a sample size of about 0.07% of the userbase. That’s not quite the best.
however…the issue was in sample method, not sample size.
basically, the survey, if I recall, used survivorship sampling, self-selection and convenience sampling, all of which tend to be very biased, as people who respond tend to be already interested in blender, and probably don’t represent the bulk of users simply by the fact that they knew there was a blender survey! Remember, the only people who took it, had a blender account, were made aware of the survey through browsing official blender channels, and understood english well.
—This means people with strong opinions, users who use the website, and those active in the community were overrepresented, while more casual users, users who get blender through other platforms, and those with different languages, and those with privacy concerns, were probably very underrepresented.
I’m not sure how to reconcile this either. Removing all three of these biases basically requires removing user choice and security options. I agree with the blender account thing, it stops repeat responses, even if many people probably turned away when they learned they had to give yet another entity some personal identifying information.
I think the surveys should be extended, and last a decently long time- months at least, and during that time, outreach needs to happen on a massive scale. At the very least, an email out to every blender account, with a very polite request couldn’t hurt… though most blender users would probably opt out of it.
maybe add a permalink to the first launch setup splash page, saying something like “click here take the blender survey, and tell us how we’re doing!” with a colorful little banner image to catch the eye. As well, steam users get newsposts beside the application when they click on it in their steam library. Having a reminder post a week after any update wouldn’t be obtrusive there.
I think the goal is to nag people in the least obtrusive ways possible.
as for the actual survey, I have some criticism about the questions.
most were pretty good, but the ones that separate field of work or art were definitely arbitrary. Film/Animation field of work is such a massive field, it seems disingenuous to lump it all together, as it does actually contain like half of blender’s modules. It contains VFX, 3D animation, 2D animation, film/animation editing, sound design, motion graphics… and not to mention that you could select VFX AND film/animation, despite VFX being for either that or game development.
This is normally tackled 3 ways-
—say you limit it to 2 choices- then you should say which industries do you work in, and each industry should generate a sub-question, asking what fields in that industry you primarily use blender for. this can be somewhat restrictive and confusing, as many industries share fields (archvis and 3d animation and VFX all having environment artists and compositors, etc)
—you can simply add all fields from all industries, separated by field if applicable. VFX(film), VFX(games), film compositing, film editing, 2D animation, 3D animation, etc… and then give more options for choices, and even allow respondents to order them, from most common to least common, to give a weighted result, too. This gives a massive list, which is hard to parse, but the fairest.
—you can separate the most general answers into their own questions and motivated questions
“if you primarily use blender as an artist, what media of art is it?”
(animation, VFX, archvis, graphic design, none of the above, I primarily use blender for [enter text here])
The takeaway is that very general answers like “artist” or film/animation aren’t very useful in a software primarily geared toward art and animation, and especially when other answers of the same question are a subset of another, more general answer.
the last point I want to make is that the sections on how to make blender better are a bit anemic. If this is by design, that’s fine, in some cases, you want to know use current cases without being complained at… but regardless, it’s important to know what could be. to that end, some whimsical, subjective, and unitless questions work wonders!
“what style of art would you want to make if blender had better tools for it, or is there a style you wanted to try, but found there wasn’t enough support for?”
(realistic 3D, painterly, lo-fi/retro, stop-motion, faux hand-drawn, stylized 3D, cyberpunk, )
“what part of blender wasn’t useful enough to continue working with or had a lack of functionality which stopped you from creating what you wanted, if any?”
(baking, rendering, simulations, sound design, non-linear editing, tracking, texturing, rigging/constraints, posing, modeling, non-destructive workflow, UV editing, import/export…)
sorry for making such a long post, but I really want to collect excellent data in an ethical manner, and the first step is getting an unbiased, large sample size, and the second is asking more useful questions!
4 Likes
Yes, there was a “freeform answer” but I did not process it due to time constraints. Feel free to check the data and let us know if you find something interesting!