OB1.7
- Janusz Biela
- Posts: 3265
- Joined: Mon Mar 13, 2006 10:39 am
- Location: Finland
- Contact:
OB1.7
test renders from 1.7
- Attachments
-
- room_bath_A.jpg (630.01 KiB) Viewed 9146 times
-
- room_bath_B.jpg (365.92 KiB) Viewed 9064 times
-
- room_bath_C.jpg (287.86 KiB) Viewed 9094 times
-
- room_bath_D.jpg (359.55 KiB) Viewed 9069 times
-
- room_bath_E.jpg (375.82 KiB) Viewed 9203 times
- Janusz Biela
- Posts: 3265
- Joined: Mon Mar 13, 2006 10:39 am
- Location: Finland
- Contact:
- Janusz Biela
- Posts: 3265
- Joined: Mon Mar 13, 2006 10:39 am
- Location: Finland
- Contact:
Wow - truly amazing! You are at the stage where your renders are so good, the quality of some of the small details lets them down. This isn't a criticism, because for most rendered images, you know its CGI, so don't look too carefully at some of the details. These are so good, that little inconsistencies draw your eye because everything else is so perfect.
For example there is a really nice touch of irregularity in the horizontal lines in the blinds, which is excellent, but at the same time there are no handles on the windows, although they are modelled like opening casements.
Also the texturing on these images isn't quite correct, particularly the grain of the timber texture, which should usually run lengthways e.g shelf under the window - narrow face. In reality, the timber grain on the bed might be different too.
There are real technical reasons for this. Veneer is peeled along the length of the logs, and it comes in longer, narrower strips. Because matching veneer can be tricky, carpenters tend to make best use of it by keeping the grain in the long direction. Standard veneered panels tend to be around 1.2x2.4m, but planks can be longer, but maybe 250mm wide and a maximum of 20mm thick. Anything other than standard sizes (even in individually-designed furniture) is special order, much more expensive and has a long order time. I guess you don't need to know the detail, but it is just background to why the wood grain almost always runs parallel to the long direction.
Also the edges of the books and wicker texture on the chaise longue and wardrobes isn't up to the standard of the rest of the image.
However, to be honest, I'll never approach this standard, it's really stunning, but because you are so good, anything that doesn't immediately look right detracts.
I really feel uncomfortable pointing these things out, because it's so much better than I could produce, even if I spent months trying, but I thought that as you seem to be such a perfectionist, anything that improves the images even more is probably a good thing.
For example there is a really nice touch of irregularity in the horizontal lines in the blinds, which is excellent, but at the same time there are no handles on the windows, although they are modelled like opening casements.
Also the texturing on these images isn't quite correct, particularly the grain of the timber texture, which should usually run lengthways e.g shelf under the window - narrow face. In reality, the timber grain on the bed might be different too.
There are real technical reasons for this. Veneer is peeled along the length of the logs, and it comes in longer, narrower strips. Because matching veneer can be tricky, carpenters tend to make best use of it by keeping the grain in the long direction. Standard veneered panels tend to be around 1.2x2.4m, but planks can be longer, but maybe 250mm wide and a maximum of 20mm thick. Anything other than standard sizes (even in individually-designed furniture) is special order, much more expensive and has a long order time. I guess you don't need to know the detail, but it is just background to why the wood grain almost always runs parallel to the long direction.
Also the edges of the books and wicker texture on the chaise longue and wardrobes isn't up to the standard of the rest of the image.
However, to be honest, I'll never approach this standard, it's really stunning, but because you are so good, anything that doesn't immediately look right detracts.
I really feel uncomfortable pointing these things out, because it's so much better than I could produce, even if I spent months trying, but I thought that as you seem to be such a perfectionist, anything that improves the images even more is probably a good thing.

- Janusz Biela
- Posts: 3265
- Joined: Mon Mar 13, 2006 10:39 am
- Location: Finland
- Contact:
bigstick wrote:Wow - truly amazing! You are at the stage where your renders are so good, the quality of some of the small details lets them down. This isn't a criticism, because for most rendered images, you know its CGI, so don't look too carefully at some of the details. These are so good, that little inconsistencies draw your eye because everything else is so perfect.
For example there is a really nice touch of irregularity in the horizontal lines in the blinds, which is excellent, but at the same time there are no handles on the windows, although they are modelled like opening casements.
Also the texturing on these images isn't quite correct, particularly the grain of the timber texture, which should usually run lengthways e.g shelf under the window - narrow face. In reality, the timber grain on the bed might be different too.
There are real technical reasons for this. Veneer is peeled along the length of the logs, and it comes in longer, narrower strips. Because matching veneer can be tricky, carpenters tend to make best use of it by keeping the grain in the long direction. Standard veneered panels tend to be around 1.2x2.4m, but planks can be longer, but maybe 250mm wide and a maximum of 20mm thick. Anything other than standard sizes (even in individually-designed furniture) is special order, much more expensive and has a long order time. I guess you don't need to know the detail, but it is just background to why the wood grain almost always runs parallel to the long direction.
Also the edges of the books and wicker texture on the chaise longue and wardrobes isn't up to the standard of the rest of the image.
However, to be honest, I'll never approach this standard, it's really stunning, but because you are so good, anything that doesn't immediately look right detracts.
I really feel uncomfortable pointing these things out, because it's so much better than I could produce, even if I spent months trying, but I thought that as you seem to be such a perfectionist, anything that improves the images even more is probably a good thing.

Very thanks for some critics (this is best for me).
This render is TEST render (I was test some behaviour Kray)
I concentrate on ENGINE KRAY. I know most problems in this render (many obiects need UVmapping). This scene much stress my computer , need around 2.4 G of RAM for render.... I have 2 Giga....
I tested: speed of translucency, save GI, spot lights near lamps and luminosity (small luminosity inside lamps over bed) background and small lumi surfaces with new shader KrayPhotonNumberMultiplier, amount bounce lights (5-20 bounce), HDR exposure in Image Editor LW (I was render to HDRi), multi textures and gradients on flowers, effect reflection on walls, DoF in LW+difference FSAA in Kray, new setup oversampling, etc.
I back to this scenes when delete some obiects (need free RAM).
thx
Jeez Johnny, my only crit is that you don't leave enough room for crits. 

It is interesting how once you get past a certain level of photorealism in lighting and surfacing that the mind starts to look for finer/smaller elements and details to chase down and find it as CG.
Great stuff none the less.


It is interesting how once you get past a certain level of photorealism in lighting and surfacing that the mind starts to look for finer/smaller elements and details to chase down and find it as CG.
Great stuff none the less.
http://www.digitaldomain.com
AKA - KC
AKA - KC
- Janusz Biela
- Posts: 3265
- Joined: Mon Mar 13, 2006 10:39 am
- Location: Finland
- Contact:
update.
SG_CCFilter, 6 bounce, sunsky+SUN on lumi sphere, DoF from real lens camera FSAA G10 BOX filter 0.5, no oversample - standart setup without tailers cmds (only recurse 6; limitdr -1;)
After rendring I used unlinear SG_CCFilter, HDR Exposure 0.8/50%, Corona filter: all in Image Editor in Layout.
Time 1h - 1:30h.
SG_CCFilter, 6 bounce, sunsky+SUN on lumi sphere, DoF from real lens camera FSAA G10 BOX filter 0.5, no oversample - standart setup without tailers cmds (only recurse 6; limitdr -1;)
After rendring I used unlinear SG_CCFilter, HDR Exposure 0.8/50%, Corona filter: all in Image Editor in Layout.
Time 1h - 1:30h.
- Attachments
-
- room_bath_c.jpg (170.28 KiB) Viewed 8988 times
-
- room_bath_b.jpg (306.49 KiB) Viewed 8957 times
-
- room_bath_a.jpg (432.82 KiB) Viewed 8951 times
Last edited by Janusz Biela on Sat May 31, 2008 9:54 pm, edited 1 time in total.
- Janusz Biela
- Posts: 3265
- Joined: Mon Mar 13, 2006 10:39 am
- Location: Finland
- Contact: