Redshift Output
Progressive Rendering 漸進式渲染
WARNING 警告
Due to a bug in Softimage (SOFT-8133), progressive rendering may occasionally result in hanging Softimage. To work around this, Redshift will abort progressive rendering if it detects that the main Softimage thread is hung for more than a few seconds.When this happens, Redshift will issue an error 『Redshift detected a hang in the main Softimage thread and has aborted the render to prevent a full application hang』.由於在Softimage（SOFT-8133中存在BUG,漸近式渲染可能會偶爾導致Softimage掛起。為了解決這個問題，一旦它發現主要Softimage進程掛起秒後，Redshift將中止漸進式渲染。當發生這種情況時，Redshift將發出錯誤信息「Redshift detected a hang in the main
Softimage thread and has aborted the render to prevent a full application hang （Redshift在主要Softimage 進程中檢測到掛起，它將終止渲染以阻止應用程序完全掛起）」。If you get this error, you can simply render again.
如果你遇到了這種錯誤，那麼只需要再將渲染。
Progressive rendering refers to Redshift"s interactive rendering mode. It renders the framebuffer in multiple passes which usually start up noisy and get progressively cleaner. It is useful in getting fast feedback when shaders, meshes or lights are edited.
漸進式渲染指的是 Redshift的交互渲染模式，它會分階段緩衝渲染，這些中間階段通常以雜點開始並漸漸地變得清晰。在著色器、多邊形物體或光源被編輯時，能得到快速反饋。
When progressive rendering is enabled, certain renderer features and options have no effect. These are:
當啟用漸進式渲染時，渲染器的某些特點和選項沒有效果，其中包括：
1、All unified sampling settings (including filtering)
所有的Unified Sampling（整體採樣率）設置，包括Filtering（模糊方式設置）。
2、Subsurface scattering. Redshift"s SSS implementation is computed as a separate pass so it"s only available to production (non-progressive) renders
次表面散射。Redshift的SSS渲染是一個獨立的過程，所以它只適用於產級（非漸進式的）渲染。
3、Photon mapping (including caustics). Redshift"s photon mapping implementation is computed as a separate pass so it"s only available to production (non-progressive) renders。 Photon Mapping（包括Caustics）。Redshift的Photon Mapping
計算也是一個單獨的計算過程，所以它只適用說產品級（非漸進式的）渲染。
4、Irradiance cache, irradiance point cloud. These are not needed because progressive rendering computes GI in a brute-force way. Irradiance Cache和Irradiance Point Cloud。
因為漸進式渲染只會以Brute-Force方式來計算全局照明，因此這兩種GI方式不影響漸進式渲染。
5、All parameters that have to do with "number of samples". Examples include the number of samples for Depth-Of-Field, number of samples for glossy reflections or refractions, number of area light samples, etc. 與Number of Samples（採樣數目）相關的所有參數，例如用於景深的採樣數目、用於光澤反射的採樣數目以及面積光採樣數目等 。
Unified Sampling 統一採樣
For information on unified sampling, please refer to the unified sampling page.
要了解Unified Sampling（統一採樣），請參考Unified Sampling統一採樣一節。
Sample Filtering採樣過濾方式
Filter Type And Size 過濾方式和採樣大小
Sample filtering defines how the per-pixel samples will be combined together to produce the final pixel color. There are plenty of options here: Box, Triangle, Gauss, Mitchell and Lanczos. The first filter type (「Box」) is the blurriest of all and the last filter (「Lanczos」) is the sharpest. The most 「neutral」 is the Gaussian filter, which is also our default.
採樣過濾定義了如何將像素採樣到的信息融合到一起，以產生最終的像素顏色，此處有多種選擇：Box（方盒）、Triangle（三角形） 、Gauss（高斯）、Mitchell（米切爾）、Lanczos。第一個過濾方式BOX是所有過濾器中最模糊的，最後一個過濾器Lanczos是最銳利的。最中中立的是Gauss過濾器，它也是默認選擇。
Gauss
Lanczos
You can control the final 『blurriness』 of the filter using the 「Filter Size」 parameter. A thing to remember here is that, the sharper the filter type, the larger the filter size typically needs to be. For example, Gaussian filters work fine with values like 2.0, 2.5 and 3.0 while Mitchell typically works best with values like 3.0, 4.0 or 5.0.
你可以使用Filter Size採樣大小參數來控制過濾器的最終blurriness（模糊強度）。此外需記住，Filter Size越銳利，通常所需要的Filter Size越大。例如Gaussian 過濾器的最佳值是2.0、2.5以及3.0，而Mitchell最佳值通常中3.0、4.0、5.0。
Note注意
A sharpening filter such as Mitchell or Lanczos in conjunction with a small filter size (like 2.0) will accentuate pixel 『jaggies』as well as noise!對於銳度高的過濾器，如Mitchell或Lanczos，假如Filiter Size過小（如2.0），可能會導致渲染結果邊緣出現鋸齒，還可能會有噪點。
Note注意
On some renderers 『filter size』 refers to the filter radius. On Redshift, 『filter size』 refers to the filter』s diameter! For this reason, if you』re porting a scene from a different renderer, you might have to double the filter size value in order to get similar results.對於一些渲染器而言，Filter Size指的是過濾採樣區域的半徑，而在Redshift中，Filter Size指的是過濾器採樣區域的直徑！所以如果你從不同的渲染器移植場景，那麼為了得到相同的平滑結果，你必須將Fiter Size小大加倍。
Choosing a filter type and filter size highly depends on what you are rendering. For example, still images can often get away with sharpening filters such as Mitchell or Lanczos. Animations, on the other hand, usually work best with blurrier filters, like Gauss. The reason is that 『jaggies』 are particularly visible on animations!
選擇Filter Type和Filter Size應該完全取決於你所渲染的對象，例如，靜態圖像通常可以使用銳利的過濾模式，如Mitchell、 Lanczos。另一方面，動畫通常適合運用模糊一些的過濾模式，如Gauss，原因是「邊緣鋸齒」在動畫上尤其明顯。
The images below were all rendered with min 4 samples and max 16 samples.
下圖最小4採樣和最大16採樣來進行渲染。
Gauss, Size 2.0. A bit too sharp around the rings, especially considering the number of samples (up to 16)
Gauss 2.0環的周圍有點太銳利，特別是此時的Filter Size已經很大了（多達16個）。
Gauss, Size 3.0. Ok. 高斯大小3.0可以
Mitchell, Size 2.0. Very sharp features. Almost as if antialiasing is disabled. Mitchell filters require larger sizes!
Mitchell 2.0非常銳利，就像是禁用了反鋸齒，Mitchell過濾模式要求更大Filter Size。
Mitchell, Size 3.0. Still a bit aliased but looking better.
mItchel 3.0邊緣還是有點粗糙，但看起來好了很多。
Mitchell, Size 4.0. Sharp and relatively smooth. The mitchell filter typically works best with sizes like 4.0 and 5.0
Mitchell 4.0，銳利且相對平滑，Mitchell過濾通常適合採用4.0和5.0的大小。
Max Subsample Intensity最大子採樣強度
Usually monitors can show colors ranging from black (0.0, 0.0, 0.0) to white (1.0, 1.0, 1.0). Any color that is brighter than white (like 10.0, for example) will still be rendered as white. But the internal workings of a renderer do care about these 「overbright」 colors.
顯示器顯示的顏色範圍通常是從黑色(0.0, 0.0, 0.0)到白色(1.0, 1.0, 1.0)。比白色明亮的任何顏色（如10.0）仍然將渲染成白色，但是渲染器的內部工作方式確實會考慮到這些「過分明亮」的顏色。
Scenes containing very bright lighting and/or strong emissive materials can be very hard to antialias. The reason has to do with sample filtering: say a pixel was rendered with 64 samples. If most of these samples are mid-gray (0.5, 0.5, 0.5) but there is a single sample that is extremely bright (100.0, 100.0, 100.0), that sample will 『dominate』 the final pixel color.
包含有強光或者自發光材質的場景很難變得平滑，其原因與採樣過濾方式有關：假設一個像素用64次採樣進行渲染，如果大部分的這些採樣處於中間灰度（0.5，0.5，0.5），但是仍然有單個極其明亮的採樣（100，100，100），那麼該採樣將「主導」最終像素顏色。
The issue can be addressed by limiting the intensities of individual samples. For example, ensure that no sample is brighter than 2.0. The 『Max Subsample Intensity』 control allows the user to adjust that limit. The higher the 『Max Subsample Intensity』, the harder for the renderer will be to antialias the image.
通過限制Max subsample intensity（最大子採樣強度）可以解決該問題。例如要確保沒有採樣值比2.0更要亮，可以用Max subsample Intensity來調整該限制，Max Subsample Intensity越高，渲染器就越難去平滑該圖像。
The image below contains a self-illuminating torus with intensity 100.0. The AA settings are min 4 samples per pixel, max 16 samples per pixel.
下圖包含有強度為100.00的自發光環面，AA設置是每像素最小4次採樣，最大16次採樣。
Max Subsample Intensity set to 128. All the rings are smooth except the self-illuminated one. That part of the image looks as if no antialiasing is enabled.
Max Subsample intensity設置為128，所有環都是平滑的，除了自發光環之外，圖像中這個區域分看起來像是未啟用反鋸齒。
Max Subsample Intensity set to 4.0. While there is a little bit of improvement, the self-illuminating ring still shows aliasing artifacts!
Max Subsample Intensity設置為4.0，雖然此處有一點點改善，但是自發光環仍然顯得是粗糙。
Max Subsample Intensity set to 1.0. The self-illuminating ring now renders smoothly.
Max Subsample Intensity設置為1.0，現在自發光平滑地渲染。
There are a couple of important drawbacks to using 『Max Subsample Intensity』:
使用Max Subsample Intensity時，有幾個缺點：
1、If you are generating an HDR image, bright pixels will be clamped to that value. If you were hoping to apply bloom effects as a post-process that could limit your ability to do so.
如果你正在生成HDR圖像，明亮像素會受該值限制。如果你希望使用爆炸特效做後期處理，這可能會受很大影響，以至於無法後期調節
2、Clamping the per-sample values means that 「light energy」 is lost. The effect is especially apparent with high antialiasing settings (i.e. many samples per pixel). The lost energy appears as a darkening effect, especially in a blurry effect like Depth-Of-Field.
限制每個採樣值意味著會損失「光能量」。對高反鋸齒設置（即每個像素有很多採樣），損失的能量表現為偏暗，尤其是在後期使用景深模糊效果一類的調整時，效果尤為明顯。
The only current solution for both cases is raising the 『Max Subsample Intensity』 which, unfortunately, will re-introduce harsh edges on strongly illuminated objects.
這兩種情況目前唯一的解決方案是提高「Max Subsample Intensity」。不幸的是，這樣會給強烈的發光物體重新引入粗糙的邊緣。
The images below demonstrate the 「energy loss」 with an aggressive 『Max Subsample Intensity』 setting.
下圖演示了Max Subsample Intensity設置造成損失的能量的問題出現。
Max Subsample Intensity set to 1.0. Even though some of the specular highlights are strong, the depth-of-field effect is weak on them.
Max Subsample Intensity設置為1.0，雖然一些鏡面高光區域仍然很亮，但經過景深模糊之後，這些區域變得很暗。
Max Subsample Intensity set to 4.0. Notice the stronger blurry specular highlights.
Max Subsample Intensity 設置為4.0，可以看到，強烈模糊過的高光仍然保持了亮度。
Max Secondary Ray Intensity最大次級採樣射線強度
Just like 『max subsample intensity』 limits the intensity of primary rays for antialiasing, depth of field and motion blur, 『max secondary ray intensity』 does the same for glossy and GI rays. This feature is useful for reducing fire-flies that may appear in glossy reflections or refractions, which can occur when a small percentage of rays are unlucky enough to sample extremely hot light sources, when most of the rays do not. The ideal solution would be to use as many rays as it takes to clean up the noise caused by fire-flies, but often this is not practical since many thousand, if not infinite, ray samples would be required. Clamping glossy samples is a simple (though not physically correct) solution, which helps ensure that hot samples are no brighter than a specified max amount.
與Max Subsample intensity限制了一級採樣抗鋸齒、景深和運動模糊的強度一樣，Max secondary Ray Intensity（最大次級射線強度）對Gloss Reflection和GI採樣射線施加同樣的影響。這有助於減少可能會出現在Glossy Reflection或者Refraction上的Fire-Flies（螢火），這是因為少部分射線恰巧採樣到極強的光源，而其他射線卻錯過了光源。理想的解決方法是增加足夠的採樣射線，直到消除這種螢火噪點。但是這種辦法並不太可行，因為往往需要太多（不是無窮多）的採樣射線才到達到的效果。將Gloss Samples（光澤採樣）的亮度降低（儘管不是物理正確）是一個簡單的方法，這使得採樣到的熱點亮度不至於過大。
AOV Processing AOV處理
This option allows you to apply a 『clamp』 (intensity limit) to the AOV color and AO channels. For colors, the clamp will be applied to each color channel (RGB). Clamping occurs prior to any AOV filtering. For examples of how and when these settings should be used, please refer to the AOV tutorial page.
你可以通過該選項對AOV顏色和AOV通道亮度進行限制。對於顏色，「鉗制」工具將會被運用到每個顏色通道中（RGB），鉗制工具影響先於任何AOV過濾。想了解這些設置應該如何被使用以及何時被使用的例子，請參考AOV章節。
Enable Deep Output 使用深度輸出
When this option is checked, Redshift can render deep images. For more information regarding deep rendering and deep EXRs, please refer to this page.
當勾選這個選項時，Redshift渲染Deep-Ouput。想獲取有關Deep渲染和Deep EXR文件的更多信息，請參考關有EXR的內容章節。
Enable Clamping / Max Value使用數值範圍鉗制/最大值
When 『Enable Clamping』 is checked, the maximum value allowed for each color channel or AO is defined by 『Max Value』
當Enable Clapping被勾選時，每個顏色通道和AOV通道的亮度最大值就是Max Value。
Use with caution. Note that this clamping will only be applied to AOV channel buffers, not during the beauty pass, so final composited results may not yield the same results as the beauty pass.
請謹慎的使用，需要注意該種鉗制將僅僅應用於AOV通道緩衝器，而不是最終的顏色（Beauty）渲染中，所以最終通道的混合結果可能與最終的顏色(Beauty)渲染結果不同。
Disable Importance-based Optimizations禁用基於權重的優化
By default, Redshift adjusts the numbers of samples of different effects like glossy reflections and refractions, area lighting and brute-force GI based on the visual importance of the pixel. For example, darker reflections can get away with fewer glossy reflection samples. However, doing so, can produce noisy AOV and negatively affect the final comp.
在默認情況下，Redshift會根據肉眼觀察像素的特點調整了光澤反射和折射，區域光源以及Brute-Force GI採樣的數量。例如，較暗的反射可以使用較少的光澤反射採樣數量，但是，這樣做會產生帶噪點的AOV並對最終合成產生不利影響。
This option disables these optimizations and guarantees that the required number of samples will be used, irrespective of whether the reflection/refraction/lighting is dark for a pixel.
該選項禁用這些優化並確保使用規定數量的採樣，而不管反射、折射、照明像素是否是很暗的。
Sampling Overrides強制採樣
These options allow you to apply global overrides or scales to the number of samples for reflections, refractions, ambient occlusion and lights/shadows ray categories. This is useful for tuning performance or removing noise across the whole scene without having to manually edit materials or lights individually. Overrides are disabled by default.
這些選項允許你為以下採樣類型設定一個全局的強制數值：
Reflections反射
Refractions折射
Ambient occlusion環境遮擋
Light/Shadow 光源/陰影
通過這些設置，就無需手工再一一到相應位置手工設定這些數值。強制選項默認都是關閉的。
Any override will only affect rays of one sample or more. Zero samples will remain as zero samples.
所有強制設置只會影響那些至少設置1的採樣。如果某個採樣設置為0，那麼強制不起作用。
Override Samples強制採樣
Checking this will enable the global samples override for the given ray category. Once checked, you will be presented with the following override options:
通過勾選將為給定的採樣射線類別啟用全局採樣設置，一旦勾選，將看到以下覆蓋選項
1、Replace Samples – replaces the samples for the given ray category
Replace Sample替換採樣次數------用指定的採樣射線數目取代原來採樣次數
。
2、Scale Samples – scales the samples for the given ray category
Scale Samples（縮放採樣次數）------將原來次數縮放這個值。
Samples / Samples Scale 採樣/採樣縮放
The override value for the number of samples.
用設定數值覆蓋原採樣數。
Photometric Units 光度學單位
Certain features of Redshift such as photographic exposure, physical sun/sky and IES light support require knowledge of the 「units to meter」 and 「candela to square meter」 settings. It』s important to set these values correctly, otherwise lighting coming from physical light sources might appear too dim or too bright.
Redshift的某些特性，如Photographic Exposure、Physical Sun/sky以及IES光，需要使用Unit to Mater和Candela to Square Mater的量度。正確地設置這些值是非常重要的，否則IES燈可能會顯得太暗或太亮。
If you』re working with centimeters (i.e. 1 world unit is 1 centimeter), the units to meter scale should be set to 100. That』s because, in this case, 100 world units means 100cm, which means 1 meter. If you』re working with meters, then it should be set to 1, because 1 world unit means 1 meter.
如果你是以厘米做單位（1個標準單位厘米），那麼單位米的縮放應該被設置為100，因為100個標準單位=100厘米=1米，好么它應該被設置為1（1個標準單位=1米）。
Additionally, photographic exposure, IES lights and the physical sky/sun use the 「candela to square meter」 setting. Please make sure to attach a photographic exposure lens shader when using IES lights and physical sun/sky, otherwise your lighting will appear too bright or too dim.
Quantizeto 8-bit and Dither數字轉換為8位和顏色抖動
Most computer displays are able to show up to 256 shades of gray even if you are viewing an HDR image. Depending on the monitor capabilities and calibration, there』s the possibility of 「banding」 artifacts on slow-changing gradients.
大多數計算機顯示器即使你查看的是HDR圖像，也只夠顯示256個灰度。根據顯示器能力和校準能力，很有可能在緩慢的漸變中出現「帶狀」紋理
Dithering can help with that!
色彩抖動可以用於消除這一問題!
In Redshift, dithering only works on 8-bit images, i.e. non-HDR images. So, assuming you don』t care about HDR output, you can enable the 「quantize to 8-bit and dither」 option and get rid of any banding artifacts.
Redshift的色彩抖動只對8位輸出圖像（非DHR圖片）起作用，假設你並不關注HDR輸出，你可以使用Quantize to 8-bit and Dither(數字轉換為8位和顏色抖動)選項，這樣就可以消除任何帶狀紋理。
The left image shows the banding near the bottom. Please note that, depending on your monitor calibration, the effect might not be very noticeable.
左圖的底部會顯示帶狀條紋，根據顯示器的校準能力，該效果可能不是很明顯。
No Dithering 沒有顏色抖動
Quantization and dithering enabled 開啟顏色抖動
Because this option quantizes to 8-bits per color channel, the final image will lose color precision in the dark tones. Additionally, any pixel that』s overbright will have its intensity clamped to a value of 1.0. For these reasons, if you do care about HDR you should keep this option disabled and, instead, apply dithering or other kinds of processing on an external image editing program.
由於該選項將每個色彩通道數字轉換為8位，最終圖像會在暗區損失一些顏色準確性。而且，所有郭亮的區域最終像素都將被限制為1.0。因此，如果你希望用HDR，應該禁用此選項，轉而在其它後期軟體中做類似的顏色抖動調整。
推薦閱讀：