Jump to content

RyderSK

Members
  • Posts

    1760
  • Joined

3 Followers

Personal Information

  • Country
    Slovakia

Recent Profile Visitors

The recent visitors block is disabled and is not being shown to other users.

RyderSK's Achievements

Newbie

Newbie (1/14)

10

Reputation

  1. Hah, I almost shared that post few days ago :- ). That was nice write-up.
  2. I personally render at 8k because my clients do both print and web, but I've yet to see them upload images to web higher than 2k (and even that is quite rare). Honestly, when I watch our website statistics, I am bit depressed :- ). 60perc. is from mobile users and almost 80perc. of those are 320px and less. Yup, 320 px. So much for the detail we do :- D. A good advice is to always render higher than you need unless it costs you a lot more time(and money). It's safe-proof when clients come later requesting more resolution than they thought they need. Personal anecdote: There were few jobs where we worked alongside other studios. When the work was given to printers, we were the only studio whose work was never returned with our resolution (8k) but others only submitted 5k by default (seems to be some odd industry standard, where does it come from ? A3+ 300DPI ?) and were tasked with re-rendering it to higher fidelity. Yes all these problems wouldn't exist if clients knew exact parameters they will need, but they rarely do, so it's better to be pro-active.
  3. The legacy was great. Not interested in the Outlier/Onion/etc.. just the simple old system. If they could kept it across every CAD suite, not sure why Max needed this convoluted mess. In fact, I almost stopped using layers workflow because of it.
  4. You might be surprised at the results :- ) It's still hit&miss sometimes but I've been getting fantastic results upscaling some older textures. I even tried upscaling 2k render and comparing it to my native 8k render and it was impressive. It's definitely good enough for watching from distance. To be honest, you shouldn't need to do any upscaling and sending 3k image is fine, parity in DPI/PPI is beneficial for vector graphics and typography which will render at their sharpest but I understand that when dealing with ignorant print shop or client it's better to be safe and just bake-down the raster image into giant file size for peace of mind. Just note about letsenhance, it has limit on file size mostly, so after you have done your post-production just upload a jpeg. I personally like the "magic" option which hallucinates new detail. You can test it in meantime on any render you have on hand so you don't trust it blindly. If it won't give you sufficient result you may just upscale your image using regular bicubic in Photoshop. (Latest Photoshop CC update provided partly AI-enhanced upscaling too, it has to be turned on in settings and then selected as correct method in upscaling).
  5. The DPI doesn't have to correlate to actual image resolution. For 8 meters away, you may literally upscale 1920px wide image and it will look crisp :- ). But if you do want to keep up with the DPI=PPI logic, just upscale the image in post. https://letsenhance.io/boost is fantastic, it does 4x upsampling, so just downsample using bicubic back to desired resolution afterwards.
  6. Some nice point Derek :- ). Yeah cliches, trends and fads come and go like within any endeavor, but the ever-lasting rant against them was always bit akin to "why don't these people get a real job/'jerb' ?" A slightly mocking response stemming from preconceived notion of what "real" archviz is, or should be. (My other favorite discussion is the one "who's 'more' artist?" , i.e. "my tools/approach are better than yours".) There are some people who made a career by creating the best 'house in the forests' and went on to start industry-leading London studio. As in the saying: "If you build it, they will come". But to not derail thread: My personal pet peeve (of which I am guilty too) is pointless close-up to show interesting shader/model, unconnected in any way to the project. My inner LEGO soul wants to do it, but I try to show restraint when possible :- ).
  7. It's mainly because of 3dsMax + Corona, it's very hungry combination, lot more than other renderers. And there are some factors within 3dsMax than can multiply this (by not properly flushing the cache after heavy subsequent renderings,etc..). That coupled with lot of multi-tasking and always having multiple heavy scenes and Photoshop files open at same time, makes having lot of memory a faster workflow as it avoids any swapping. I was actually quite surprised to run to wall with 64, I wouldn't have thought of that. Then again, I thought the same with 16 few years ago. The memory price levels are nauseating. And it's grotesque some brands keep coming on top of this with some ridiculous "luxury" items like super-fast, RGB-lit modules, bringing exactly none benefits apart from being fool parted with his money on pointless gimmick. Who on earth considers that a valid investment ?
  8. Where did you see that ? So far all I've read are vague predictions about small gradual price decline over the whole year. Right now locally, 128 GB set of DDR4 (8x16) starts at 1700 in market around me. 2 years ago I bought used 128 GB DDR3 ECC for 200 euros : / As Nikolaos said..all I've upgraded recently was nice 32" 4k monitor as well :- ). Everything else is sky-high.
  9. There has been steady decline even in CAD world in capacity to benefit performance-wise from PRO cards like Quadro, simultaneously with growing capacity of mainstream cards. There are some good benchmarks out there (which I am lazy to go look for) that shows it's actually some rather surprising minority of CAD suites that do benefit (Siemens NXT) while most do not that much, or at all. Their benefits are ever-shrinking and are currently tied to precision that is cutoff in mainstream cards (10bit color output, double precision floating point for accuracy where needed (medical/fintech/etc..), ECC memory. Those who do need OpenGL and different floating-point performance modes can actually buy mainstream VEGA card from AMD since that one doesn't restrict the featureset unlike nVidia, not much crippling going on. Just look at the latest humbug around nVidia's drivers EULA explicitly forbidding 'datacenters' from using GeForce cards outside of blockchain. Mainstream cards aren't just good-enough, they're great and identical in performance unless you fall into the specific need of niche feature. Regarding the sad reality of price hike, GPUs aren't even the worst offender, have a look at the cost of DDR4 memory, particularly 128 GB sets (8x16). Well priced Threadripper is hardly a boon when you'll pay twice as much for memory to accompany it.
  10. As James Vella wrote, the issue isn't in hardware, but 3dsMax itself. There is nothing that can speed-up 3dsMax viewport above the point that it's capable of. Most of the time, the slowness of viewport isn't even tied to graphical performance, but the amount of data it shuffles for the objects in it. That's why 3dsMax suffers from high draw-call (too many objects vs few) and when those objects are tied to complex hierarchy, or have too many modifiers. It's the issue of the core. The apex of viewport speed came with 2016 release which vastly improves upon anything before it. 2017/2018 offer some specific speedups with high-poly objects, as long as you have big graphic memory (These two releases are apparently problematic with graphic memory consumption). Since 2012 the default DX11 based Nitrous viewport no longer benefits from dedicated OpenGL driver of Quadro, rendering the performance of Quadro and GTX identical to most workloads.
  11. Thank you so much for honest and in-depth answer Nils, always very valuable to get his sort of insight. The 'one/two sentences' was just to make sure I don't waste your time but still possibly source some info :- ).
  12. Would you be willing to talk ( one or two sentences only really ) about the transition to this general creative/branding agency direction (very clear vibe I got from the web) ? Was it a natural result of studio's evolution and client's demands or did you always wanted to focus away from architectural visualization ? It's general curiosity. I have lot of admiration for bigger studios. But it interests me how they change, some kept clear focus, some have varied interests and some try to position themselves where visualization seems like a tiny partial side-product, almost an afterthought, despite their portfolio heavily dominated by architectural work. Or maybe it's me reading the 'Studio'(About) page wrong given they all start to be bit ambiguous. Perhaps it's branding, positioning,etc.. connected only to architecture/urban planning/real-estate. And not in general terms (for any sort of brand, from Ketchup to Verizon).
  13. https://en.wikipedia.org/wiki/Nudge_theory Helps in every facet of life :- ) Super true advice. For years I was obsessed with the product to point I didn't notice I put my clients so ahead of my own personal quality of life, sacrificing not just weekends but months including Christmas holidays. Half the issue was my own obsession, half the issue was letting be easily manipulated by clients considering everything "important" ( by constantly stressing the point, making up fake deadlines, etc.. ) The fact is, almost never was it that important. It's just the language of today. It pushes you to sacrifice yourself on behalf of others when it absolutely isn't necessary. I wish I had forced down the need for absolute honesty, reclaim my agency, respect for boundaries and setting realistic expectations and acceptance of reasonable compromise earlier than I did. But this career is never-ending work-in-progress, learning of past mistakes. TLDR advice to former self: There is no respect or love to be earned by sacrificing all the personal time to please every request client throws at you. The work will not benefit and result won't be better. With creeping depression it's not like you would be able to do any work at all. Work/Life balance might be cliche and doesn't denote single solution or approach but it's important matter to cater to. What you say is very much true and always had been. Yet I would like to provide slight counter-argument, or rather, possible alternative :- ) With the rapidness of changes in current times, chasing the 'new thing' can be a vicious circle. There are only so many 'new things' and only so many windows of opportunity to be the first. Sometimes it's worth to step away and just provide good old-fashioned quality. People who come up with something first will always gain most attention. But those most sought after will be those who will make it best. With the possibility I will eat my words I would rather illustrate that. Tesla has brought the attention, infrastructure and public availability to electric cars, but it's already becoming clear the company is racing against a wall. The reality just might be, the biggest success will yield someone else (and that your future electric car might rather be Porsche). Tablet and smartphones predates Apple, but it was Apple who improved upon the concept and made it popular. (Second time this year I use them as example, I swear, I don't own single product and I use Lumia and SurfacePro ) Two years ago, the VR and real-time revolution started haunting me. It had click-bait articles around every corner waiting to give you anxiety, it had cult followers at conferences telling you how you'll be forgotten in ashes if you don't start offering VR today. Yesterday was late, it was VR or die. At first I laughed at it, then it crawled back in my mind and gave it some consideration, but in end ultimately resorted to focusing on what I was already good at and do that better. Just single simple images. And I can tell just looking at many of those early adopters what result it brought: attention fracture leading to broader range of mediocre work. It might have worked for select big-sized studios as supplementary service, but for freelancers and small studios, the results look unimpressive. I am by no means advocating to ignore the changes. But chasing it first is not ultimately what everyone needs to do to be successful and competitive. TL : DR advice: Have a clear focus early and stick to it. Whether that's being revolutionary pioneer or simply mastering your niche
×
×
  • Create New...