KT 2008 Echo Edition Teaser!!
-
@unknownuser said:
by DzineTech on Wed Jan 23, 2008 4:59 am
From the images posted, I really don't see what there is to get excited about. All the images are grainy and aren't anywhere near as good as what VrayforSketchUp (even in its frustratingly buggy first release) can produce.
You may be right, but remember...It's FREE!
-
@dzinetech said:
From the images posted, I really don't see what there is to get excited about. All the images are grainy and aren't anywhere near as good as what VrayforSketchUp (even in its frustratingly buggy first release) can produce.
Read what's under the pictures, please.
-
@unknownuser said:
Read what's under the pictures, please.
Your point being? Of course I've read the annotation below the images. What counts is the quality of the final image produced before post-processing and in my opinion its not that impressive based on the examples provided. Yes the KT team do an admirable job working on a program that is free for all to use yet I personally dont think that it should be used as an arguement for the image quality it produces.
-
I would expect any render engine's to have the capability to reduce or removed completely any noise/grain by amending settings /changing presets hence why I'm surprised that the images showcasing the next edition are so noisy. I may be the only one not impressed by what I see but I'm willing to be truthful in what I see rather than be a SketchUp community 'yes man'.
With all due respect to yourself Kim and JJ, there is a huge amount of noise on the first 2 images, far less on JJ's last image. I'm surprised you can't see it but perhaps seeing that you are exposed to KT on a daily basis and immersed within the KT team and community as a whole, you are familiar with this output. I dont believe there is any merit in myself slagging off anyone's work even via comparrison to what I would deem acceptable but I will say that if we agree these images fall into the category of photo-realism, then for any image to become acceptable I would elimate as much noise as possible.
-
@dzinetech said:
@unknownuser said:
Read what's under the pictures, please.
Your point being? Of course I've read the annotation below the images. What counts is the quality of the final image produced before post-processing and in my opinion its not that impressive based on the examples provided. Yes the KT team do an admirable job working on a program that is free for all to use yet I personally dont think that it should be used as an arguement for the image quality it produces.
Let me walk you through it. "Rendered in MLT(BPT) mode for 5 hours... (single core machine)"
Unbiased render method. Thus grainy result after rendering on a single core machine for only 5 hrs.
I could go on like this, but sometimes you just gotta with Mies and think: less is more.
-
The problem with such unbiased spectral renders is time it takes for the renderer to run long enough and get the image right and converge on the final correct solution and that isn't an option for many people with deadlines..
-
What are you trying to point out here? That KT can't produce decent images? Or that KT can produce decent images, though it might take some time?
-
@dzinetech said:
yet I personally dont think that it should be used as an arguement for the image quality it produces.
I fully agree with you...!
The point is that you can get renders completely free of the grains... It's all a matter of the render preset that have been used and how long time you let it "cook"...
The "teaser" I made was stopped after 15 mins... If i had let it cook for i.e. 30 minutes, you would not be able to see any grains...
When I get home, I will post some new images...
BTW... I don't see much grain in JJ's images... My clients would be very pleased with this quality, but perhaps your clients wouldn't..??
-
I'm not claiming either. If you read my original post I'm saying that the examples shown don't do anything to impress me. What is the point in showing a half cooked image? Would you give that to a client and expect he/she to be all impressed by it. I'd imagine most client have never heard of MLT nor care either. What they care about is the final image they get from you, not how you get it. My comments on this thread are based on images posted here - not the overall capabilities of the software. I'll happily look at examples of KT that Frederik has offered to see how impressive this new build actually is.
-
Ah. Now I understand. When you wrote:
“From the images posted, I really don't see what there is to get excited about. All the images are grainy and aren't anywhere near as good as what VrayforSketchUp (...) can produce.”
you actually meant:
“From the images posted, I really don't see what there is to get excited about. All the images are grainy and aren't anywhere near as good as what VrayforSketchUp (...) can produce. Then again, the annotations say these were done on a pretty low-grade machine, using an unbiased render method. So, in fact, it’s quite feasable that if one were to render these on a decent computer for the same amount of time, the result would be quite different. Let me rephrase that: that’s a certaintity. So ... in the end, it might be quite possible that KT can in fact, given you have the time, produce images that are better than Vray’s, as unbiased rendering, in terms of quality, generally is the better solution.”
It seems we agree. Which, of course, is a good thing.
-
Now you are putting words in my mouth. Sure KT "might" be able to produce a superior image using an unbiased render method (maxwell, indigo, MLT etc..) if it was given enough time but what use is that to us when we more often than not have to get images to our clients in a shorter time than we would like. This fire-fighting analogy makes unbiased rendering not an option for most professionals when biased rendering can be made to look like any unbiased render. I'm not sure how many hours it would take for a KT MLT render to get anywhere near the quality of a vray render on the same machine but I doubt its worth the wait.
I saw the following quote on another forum regarding such a topic which made me laugh:
*"This is all pointless really, because with a little cunning, you can easily reproduce Maxwell's superb quality of rendering from within Vray.Simply follow these steps:-
- remove your current cpu from motherboard and replace with a nice i386 running at 33Mhz
- Take out all 4 of your 1Gb memory sticks and replace them with a couple of 4Mb sticks
- Pour some porridge over your harddrive (we dont want it spinning too fast)
- Book your plane tickets to Hawaii, then hit "Render"
- Go on holiday for 2 weeks
- Hey Presto - by the time you get back your teapot will be beautifully rendered.
There is only one problem since Vray will have made a nice 'clean' render. DONT WORRY ! We can fix this. Open up Photoshop, go to Filters, select "Add Noise" (personally I prefer about 12%, Uniform & Monochromatic for best Maxwell effect).
Thats it ! your done. easy.
"* -
"Now you are putting words in my mouth. "
-
@dzinetech said:
you can easily reproduce Maxwell's superb quality of rendering from within Vray.Simply follow these steps:-
- remove your current cpu from motherboard and replace with a nice i386 running at 33Mhz
- Take out all 4 of your 1Gb memory sticks and replace them with a couple of 4Mb sticks
- Pour some porridge over your harddrive (we dont want it spinning too fast)
- Book your plane tickets to Hawaii, then hit "Render"
- Go on holiday for 2 weeks
- Hey Presto - by the time you get back your teapot will be beautifully rendered.
There is only one problem since Vray will have made a nice 'clean' render. DONT WORRY ! We can fix this. Open up Photoshop, go to Filters, select "Add Noise" (personally I prefer about 12%, Uniform & Monochromatic for best Maxwell effect).
"[/i]You made by day .
Hilarious .....
and in some way true. Unbiased methods can be a pain when you want to be productive.(Not that I am planning to mingle myself in the discussion...I heard this one before)
-
@dzinetech said:
I'm not sure how many hours it would take for a KT MLT render to get anywhere near the quality of a vray render on the same machine but I doubt its worth the wait.
It seem that you're just ignorant here...
How can you state something like that..??
You might be a big fan of Vray (which btw is a very good render...), but claiming that it's not worth waiting for in KT proves that you don't know what you're talking about..!!I suggest that you take a look at the Comparison among Unbiased Render Engines at the KT Forum...
-
wow! this thread has gone pretty well in a wrong direction...
Dzinetech.. i am an ardent VrayForSU User , and i am not the least bit doubtful about its superior render quality and speed!
infact the images which i posted were to merely to hight the new features such as Instancing tool.. where u can make whole lots of realistic grass without having to choke ur machine.. and the new IES lights whereby u can save tons of time without havin to tweak the spots and point lights.
these renders were to merely highlight those functions (i have even made it clear in the annotations provided) and yea the renders do have noise.. but i did post the original renders (with no post pro involved), except for the third exterior render, to avoid any misunderstanding of any sorts.
besides this KT project is solely done by a one man army.. with a few supporters helping him to test and provide feedback, and its free... i am not saying that this is the best render engine out there.. but it comes pretty darn close to one if u take the price to performance factor!
and i just finished the first render using the biased render solution available in KT..
finished in 40mins 17 secs.
and here is the render in full size. -
@frederik said:
@dzinetech said:
I'm not sure how many hours it would take for a KT MLT render to get anywhere near the quality of a vray render on the same machine but I doubt its worth the wait.
It seem that you're just ignorant here...
How can you state something like that..??
You might be a big fan of Vray (which btw is a very good render...), but claiming that it's not worth waiting for in KT proves that you don't know what you're talking about..!!I suggest that you take a look at the Comparison among Unbiased Render Engines at the KT Forum...
Seems like I touched a nerve and the KT users are on the defensive!
Frederik: the point I make is perhaps a generalisation on unbiased renderers yet it is hardly unfounded. Whilst apps such as Maxwell can produce excellent images if using a correct setup, waiting a week for that image to be produced at working resolution is a joke. Unbiased renderers are notorious for the time they take and assuming KT's unbiased setting works on the same basis using the same algorithms etc..., I can't see why it would be any different. I don't think this is a particularly unfair deduction based on the evidence of every other unbiased renderer and professionals who can't use them because of how slow they are. Am I to expect that arch viz firms will now drop vray/mental ray etc.. for KT Echo because it can deliver the same high end images in a time frame that is the fraction of a Maxwell render? I doubt it very much but I do look forward to seeing what users can produce with the new build.
-
@dzinetech said:
Seems like I touched a nerve and the KT users are on the defensive!
I guess you did... But does that come as a surprise to you..??
@dzinetech said:
Am I to expect that arch viz firms will now drop vray/mental ray etc.. for KT Echo because it can deliver the same high end images in a time frame that is the fraction of a Maxwell render?
It's very clear that you have no clue about what render methods are available in Kerkythea...
Please note that KT doesn't solely have unbiased render methods available... KT has both biased and un-biased methods... That's why I - in my first post - wrote that it all depends on which render preset you choose to use...
I don't know any other render engines capable of having both biased and unbiased methods available...Didn't you take a look at the Comparison among Unbiased Render Engines thread...?!?
On page 2 you'll see that for that particular scene, KT beat Vray in speed...@dzinetech said:
I do look forward to seeing what users can produce with the new build.
And I'm sure that you'll get a pleasant surprise...
Just one final comment...
@unknownuser said:
Besides, here's a small lesson about unbiased rendering:
-
Unbiased rendering was first introduced back in 1986. Yes!
It was not introduced by a commercial renderer. Just marketed
nicely (but this is just my opinion). -
Writing an unbiased renderer is simpler than writing a biased
one! Yes! Developing a path tracer or even Kelemen's MLT is easier
than writing photon mapping. -
There are many more unbiased renderers that you can think off!
Virtually all ray tracers can turn to unbiased rendering relatively
easily. Sunflow for example can operate in path tracing mode.
Overall! I see users reproducing the same information found in
commercial leaflets but never digging a bit more in theory or.. history! -
-
Sorry Frederic but these comparison are biased. They actually do not paint the good picture. Look at this render on the Vray forum.
http://asgvis.com/index.php?option=com_smf&Itemid=90&topic=3351.0
Now that render took less than 10 minutes. How do I know is because i know the artist. Now compare that little pot and complexity of that scene on the Vray forum. -
Sepo, I'm not sure you've read what I wrote...
@frederik said:
On page 2 you'll see that for that particular scene, KT beat Vray in speed...
I can only repeat... for that particular scene...
I know Vray very well and also the superior render quality and speed...
And I also acknowledge that Vray will beat KT in many scenes, but again it all depends on the render preset used... -
That is exactly my point Frederik. It all depends on the presets. This guy uses special preset to render exterior images in less than 10min. Normally ordinary user would use standard one and it could take 40 minutes. So we should just stop these comparisons as they are fruitless.
As far as that particular scene again I bet you this guy can render it in seconds.
Advertisement