What's your initial reaction to this graph?

Is the 4k field somehow saturated?

Do you write 4k's yourself?

Do you agree that the period from 2007-2009 was a golden age of 4k's ?

Has Crinkler taken the size aspect out of the equation?

Has Crinkler caused better quality in 4k intros?

Has Crinkler caused more similarities / less uniqueness among 4k intros?

Why do you think 4k's are losing ground to other size-restricted compos?

Are you using Crinkler for compression?

If you do agree - what caused this "golden age" ?

Unfortunately no insights but maybe it was just about time for them :)

No, I think there's still lots of good stuff being released

Not at all, 4096 bytes is a difficult constraint by any measure

I guess coders explore 1k trying to push it to the limit like calvin rotane https://www.youtube.com/watch?v=5yiszCErHs0 ;) and 8k is enough for more complex works. Maybe in 2 years 16k will be the new hip shit on PC. At the moment - looking at the graphs - 4k is between 2 extremes and maybe kinda uninteresting to coders/groups.

better compression (Crinkler)

No, I think there's still lots of good stuff being released

Not at all, 4096 bytes is a difficult constraint by any measure

Yes, because it performs better in terms of compression

Not sure. But maybe they became more popular at that age. Also,. at later ages people outside the demoscene learn about it through 4ks and 64ks, since it's the only technical thing today that can draw attention. I also see the decline. People today are bored of the same 4k raymarchers. Maybe with the advantage of GPUs and shaders, that era was just right for a lot of new stuff (and it's easier to fit 4k raymarchers than polygons, although some of the popular 4ks were still polygon based) and later we started seeing the same and the same boring 4ks. So, maybe the new GPUs were right for new creativity and ideas to arise in that aspect but later it declined because there is not much new stuff to do with the same technology.

Yes (technical reasons prohibit further progress)

Not at all, 4096 bytes is a difficult constraint by any measure

Generally speaking, sometimes tiny intros are more popular to the lonely coder, especially when there is no time or motivation for much bigger demos or 64ks. You have a week and making a full blown demo or filling 64k with lot's of procedural stuff is out of the question.

Then, 1k is the laziest option, you can sometimes even avoid using music, having the excuse there were no bytes left (which makes you ashamed, finding some 1ks that have still good music and wondering how). 4ks is a nice average solution, but I find that with some music not much space is left and you easily reach the limits and have to remove or simplify some of your code. Surprisingly enough, 8ks feel much more. I never thought I would reach 5.5k and still had no time couldn't fill the space. You could have 3-4 times graphical content of the 4k. But anyway, so 8k is appealing, because it's not exactly 64k to need a lot of code and lot's of parts, but also is a nice breath from 4k to do some more stuff without feeling much restricted. Also it's the new thing, so data is not enough to say where it will go. If more parties adopt the 8k, which I'd love to, then we might see a rise and new levels in what can be achieved. Maybe that could be a breath for not evolving 4ks. As long they also do something else than raymarchers (I miss traditional polygons in intros).

Yes, because it performs better in terms of compression

d3dx9 improvements, shader model 3, better compression (Crinkler)

I was unaware of this trend - but yes - this might add up. There's a while since i've been really impressed by a 4k.

4k was in 2005-2006 a quite easy format to do, the expectations was (compared to now) quite low - and therefore easy to make "something". This does not explain the trend of released 4k the latter years, but maybe the popularity.

No, I think there's still lots of good stuff being released

I used to, but don't anymore

Not at all, 4096 bytes is a difficult constraint by any measure

Again, 4k is still a hard category to impress in. 8k's are quite fresh - and opens a lot of possibility towards new thinking - 8k is A LOT more space than 4k given the compression ratio's that you're able to get. This can be spent on creating a new tier of intros. 1k intros is what 4k's used to be - smaller experiments with lesser expectations.

Personally I think this trend is quite nice, it shows the scene finds new and fresh challenges all the time instead of just watching soccer with their pants off.

Peace.

Yes, because it performs better in terms of compression

d3dx9 improvements, shader model 3, better compression (Crinkler), better compatibility (Crinkler), tools for creating content

No, I think there's still lots of good stuff being released

Not at all, 4096 bytes is a difficult constraint by any measure

Mainly because of the challanges, and the inmense high bar

Yes, because it performs better in terms of compression

No, I think there's still lots of good stuff being released

Although a lot of good releases came out during that time, I've not seen many from the period before then.

Not at all, 4096 bytes is a difficult constraint by any measure

1k is gaining popularity among those that find 4k not restrictive enough, while 8k is gaining popularity among those that feel too restricted by 4k, but find 64k not restrictive enough.

i would say 4Ks went from obscure to popular in the 2006-2011 period

No, I think there's still lots of good stuff being released

Partly - but it's still difficult to match 4096 bytes

it made it easier for coders who dont wanna bother with packing issues

no, other than the hausse of danish releases :)

- 1Ks became what 4Ks were before they become "really good", quick show off one one good effect.
- 8Ks are just a novelty that triggers some coders to give it a go. frankly i personally find it a crap category because so far most 8Ks look like 2 superglued 4Ks :)

competition drove interest and quality?

Availability of tools like crinkler and 4klang in combination with powerful enough gpus to implement techniques like raymarching. That combined with a decline of 64k intros.

Yes (the genre has been pushed to its limits)

I used to, but don't anymore

Not at all, 4096 bytes is a difficult constraint by any measure

Saturation effect as it happened to 64k before. Making the next best 4k got increasingly hard. So focus shifts to other categories or leads to new categories.

Yes, because it performs better in terms of compression

shader model 3, better compression (Crinkler), tools for creating content

No, I think there's still lots of good stuff being released

better compression (Crinkler)

First of all, no offense.

Well, don't look only at the pseudo golden age. Look at the longer range and please don't try to predict the future. You might be right, you might be wrong - you will know when it happens.

64k was dead, remember? Golden age? I'm not into declaring things dead.

There is a steady rise before that: Before crinkler existed (people used cab droppers/20to4), before there was d3dx. This is extremely multicausal.

When I started doing 4k's in 2004 it was really easy to compete - others joined, we had a lot of fun and developed skills. There was a lot of fresh blood coming into the scene doing 4k's and I was part of it. We were young.

No, I think there's still lots of good stuff being released

I can't take the last option. Because I don't know whether it's going to happen again or not. Maybe?

I don't agree on predicting the future. So that's a "there was a peak" - fact. Declaring a golden age is like a death declaration for me. So that's a FUCK NO.

Not at all, 4096 bytes is a difficult constraint by any measure

No. What a bullshit question, I want that pepperschnaps.

Oh hi.
Can't answer to some questions properly due to missing free text:
-------------------------------------------------
"Is the 4k field somehow saturated?"
No - but I think the current quality average really sucks. There are still a couple of good releases here and there - but no "killers".
But the bar for a 4k killer release is pretty high, so that also might hinder newcomers to develop proper skills over time, since it's kind of pissing you of when a release from 2007 is way better than what you are doing.
-------------------------------------------------

-------------------------------------------------
Has Crinkler taken the size aspect out of the equation?
Definitely no, there is always room if you go really low level and some guys (maybe not the majority) still do that. What do you think are crinkler compression reports for? If you want the maximum you can squeeze in you have to "team up" with your toolchain not just use.
-------------------------------------------------

-------------------------------------------------
Are you using Crinkler for compression?
Yes, because I love sizecoding but am too lazy to write my own compressor (and I'm way more into graphics coding than compressors), Crinkler does a hell of a job there - doing it better would take to much time and effort. Not interested. It does not take away the sizecoding part at all. See above.
-------------------------------------------------

1k is mostly crap. I hope it dies. Really.

Looking forward to good 4k and 8k releases AND THE PROMISED PEPPERSCHNAPS. Maybe 4k will be back. 8k has certainly some interesting potential: You still can explore new grounds there - that makes it certainly somewhat interesting. Sceners will do whatever the fuck they want, so let's see what happens next. Looking forward to some nice intros from your side!

I hope that helped. Sorry for the large amount of F-words ;)
Not really sorry.

Yes, because it performs better in terms of compression

As you might notice, there is a steady increase till the peak. d3dx9 was around 2005, way before the 4k count started rising. Crinkler is awesome, but that might be a chicken/egg thing?

Yes I have. And it's not just the 4k intros. In that era computer programming blossomed due to the fact that many factors started viewing programming as a means to increase their profit. From individuals to companies, we have a large range of people during that period, the so called 'golden years of programming', that give all their talent to coding. Also, computers aren't the same as they were in the past years. Chips get enhanced, as well, new cpu technologies (dual core, and quad core), new gpu technologies (SLI) and so on. Also, we have new programming IDEs (Visual Studio 2008) and scripting languages. In general, computers are blossoming during that period. After all, isn't debris from farbrausch (2007), lifeforce from asd (2007), elevated from rgba (2009) some of the all time top demos? The above fit and answer the 'Why?' question perfectly. So, no surprise as to why the best 4k intros, tools and demos, were released during that period.

No, I think there's still lots of good stuff being released

Partly - but it's still difficult to match 4096 bytes

What the packer has to do with this? A packer is nothing more than a packer. Seriously get over it.

What the packer has to do with the intro content? No.

People want to test new fields and do new stuff. Why 64ks were prominent in early 00s? People got bored of doing demos and wanted to try something new. Why 4ks rose during the late 00s? People got bored of doing 64ks and wanted to try something new. Why 8k and 1k are the dominant trend these days? People got bored of doing 4ks and wanted to try something new. If you do the same thing over and over again during the years you simply get bored. It's not the packer, the tools, the hardware and so on... simply people want to try something new. I was eating yoghurt with chocolate for years and liked it, got bored eventually with the flavor, tried yoghurt with fruits, stuck with it. period. over and out. :)

All of the above, plus what was mentioned in the first text.

IMO it was a trend. Demos and 64k became too much work and had too many good competitors. 4k was considered a niche where you were able to get a high rating. Later then, the trend went to web based demos/intros.

Yes (the genre has been pushed to its limits)

I used to, but don't anymore

Not at all, 4096 bytes is a difficult constraint by any measure

As already said, yes. To other size-restricted compos and to web-based demos/intros

Yes, because it performs better in terms of compression

That´s when "Raymarching" appeared as rendering technique..
Nowadays everyone uses this technique,.. But at it´s first appearance it was a blast!

No, I think there's still lots of good stuff being released

Not at all, 4096 bytes is a difficult constraint by any measure

Yes, because it performs better in terms of compression

I recall the motto "4k is the new 64k" became popular in the demoscene around 2007.

No, I think there's still lots of good stuff being released

Undecided. There have always been good 4k intros.

Not at all, 4096 bytes is a difficult constraint by any measure

Maybe it is just a trend, just like 64k intros were fashionable in 2000 - 2007. Perhaps in a couple of years 128k intros will become popular to create.

No, I think there's still lots of good stuff being released

I used to, but don't anymore

Not at all, 4096 bytes is a difficult constraint by any measure

Yes, because it performs better in terms of compression

Yes (the genre has been pushed to its limits)

Not at all, 4096 bytes is a difficult constraint by any measure

No, raymarching caused most of the similarities

4k has been pushed to it's limits, time to try using similar techniques in another size limited category. a good example is mercury's timeless

raymarching, IQ/rgba's works

Reasons for spike-up:
- Tools have improved (Crinkler, 4klang)
- CPUs got fast enough to (pre)calc really nice-sounding soundtracks
- Gfx cards got powerful enough to use inefficient but small techniques (raymarching)
- Before NVScene'08, NVidia distributed 8800 Gfx cards like candy. People REALLY targeted the high-end cards in a way they didn't before.
- Combined, lots of "new frontier" space. Coders could do things that were never done before

Reasons for spike-down:
- top-notch 4Ks have become a lot of work: You need as much coding, music AND designing effort as in oldskool 64Ks.
- Raymarching has become cliche, nobody is impressed with it anymore.
- Same with 4klang, people complain if you use f.e. the default 4klang instruments.
- Youtube means people have no clue about how fast an intro really runs. If you spend lots of time and bytes to make it run on average hardware, very few people will notice.
- I'm not sure, but I think the amount of people voting on Pouet has been going down the last years. So that's a downward bias.

No, I think there's still lots of good stuff being released

Didn't notice it, but your graph looks convincing.

Not at all, 4096 bytes is a difficult constraint by any measure

Yes, by simply leaving more room for content

Nope, raymarching and Crinkler are completely unrelated.

- Assembly replaced the 64K compo by 1K.
- Revision introduced an 8K compo.

So if the 2 biggest parties make it an official category, you're bound to see more releases. And compared to 4Ks:
- 1Ks can be coder-only, no need for a designer, or even a musician sometimes.
- 8Ks leave more space for more complex coding techniques or ideas/content that doesn't fir in 4K.

Yes, because it performs better in terms of compression

d3dx9 improvements, shader model 3, better compression (Crinkler), tools for creating content, I'd rather say Better/Faster Gfx Cards, I use OpenGL so D3DX improvements are not relevant to me

No, I think there's still lots of good stuff being released

Partly - but it's still difficult to match 4096 bytes

Yes, because it performs better in terms of compression

quality expectations for demos and 64ks started requiring too much graphic assets. 4ks became the right size to explore fewer quantity with higher quality. crinkler also helped push the trend a few bytes.

No, I think there's still lots of good stuff being released

I used to, but don't anymore

Not at all, 4096 bytes is a difficult constraint by any measure

like 64ks, it's been a bit saturated with good examples of what you can squeeze in there, so in part it lost a bit of the appeal. 1k now allows for sound, which makes for great 1 trick pony effect showcases. 8k is rising due to the novelty factor, people can throw there their 4ks that didn't quite fit without having to completely rape / rewrite them, maybe add a few better sound and graphic assets.

Yes, because it performs better in terms of compression

shader model 3, better compression (Crinkler)

No, I think there's still lots of good stuff being released

Not at all, 4096 bytes is a difficult constraint by any measure

mentor/TBCs 'fingerprints' are all over the half top 10 4ks, mostly concentrated in the period in question.

Further searching shows mentor/TBC and iq/rgba are each involved in 8 out of the top 40 each. (16) which isn't far off half the top 40 4ks ever produced.

To me that indicates a big contribution/activity increase at that time by them and others as the cause for the spike.

Yes, because of compatibility

Where are the axis on that graph ? Is that Windows only 4k intros or ... ?

I think the folks from Loonies & RGBA just had a too much free time on their hands during these years.

Work and life probably caught them up around 2010. My own demoscene production peak was before my kid turned 1yo and I was on parental leave. That's how it goes.

Also around 2007 some tools ( yes, Crinkler, 4klang and co ) and techniques ( mainly SDF ) were getting explored and more mature.

No, I think there's still lots of good stuff being released

Not at all, 4096 bytes is a difficult constraint by any measure

1k is new 4k: where one trick poneys are still OK
8k allow to go wilder.

4k is a very interesting category, with very high standard.

Its kind of surprising. The only reason i can think of is that maybe the 4k music tools like 4klang or else were released in that period of time or became popular and that gave the whole 4k intro scene a major push.

No, I think there's still lots of good stuff being released

Partly - but it's still difficult to match 4096 bytes

because people may like to play around with with other specifications when they feel something new has come up. "old" categories may feel boring to them then.
1ks are on the rise imho because you dont need a team to produce that, its mostly just one effect without any music and that gives the coder a possibility to get fast possitive feedback.

better compression (Crinkler), better compatibility (Crinkler), 4klang

"Oh look, a visual representation of people getting bored with raymarched effects."

No, I think there's still lots of good stuff being released

I've done a few, but it's never been a main focus

It was the golden age of a new rendering technique that fit well with the 4k constraint.

Partly - but it's still difficult to match 4096 bytes

1k is extreme, appeals to the hardcore size coders.
8k allows for a lot more freedom and is thus easier to create something unique in it.

Yes, because it performs better in terms of compression

No, I think there's still lots of good stuff being released

Yes, because of compatibility

better compression (Crinkler)

Better tools/frameworks..? And after a while ppl got tired of making 4k intros because it got "easy"..

No, I think there's still lots of good stuff being released

Partly - but it's still difficult to match 4096 bytes

better compression (Crinkler), tools for creating content

There's not much to see in 4k anymore. 64k and other compos provide more freedom.

Yes (the genre has been pushed to its limits)

Not at all, 4096 bytes is a difficult constraint by any measure

As I said, I don't think there's much to see in 4k anymore. Maybe at some point in the future again, when we have even more powerful shaders and systems. 1k provides even more challenge to the size aspect and 8k gives more freedom to create small but complex demos.

shader model 3, better compression (Crinkler), better compatibility (Crinkler), tools for creating content