priority ranking of applying filters....

[parintele]

Among a long list of things that "could" be better/smarter in LR is one thing that would really improve the workflow in Lr and could solve amny related issues...

as far as i know' LR is not sensitive about picture processing in Develope Module.
that mean user have no control on which modifications affect the picture before others...
many times that is not so important....goind back and forward with exposure, brightness, recovery or fill light, even WB .. the end result will be the same no matter which is the order of applying these modification and that is a good thing...usually.

still, there are situations often met in post processing that require a carefull approach regarding which operations are a priority in order to obtain a better result.


this is the case of Noise Rduction and sharpening filters.


In Lr u do not control when this filters are applied.

for example high ISO images with lot of noise, both luminance and croma. most of the times things are really nasty as High ISO pictures are often associate with extreme exposure "shifts". many times wedding shooters or sport PJ's underexpose '.5 - 1 stop or more in order to be able to use a proper shutterspeed, everything being corrected by increasing exposure back to normal values in PP...
of there are times when subjects and overall frame have a wide DR while still shooting at very high ISO...
for this reason, on top of exposure correction in PP, use of curves, adjustment brush or Fill Light and recovery filters is needed.


resulting noise become somtimes really ugly.

reducing Noise BEFORE some of these operations are done become very important.

i wonder if such options regarding priority of applying some of the filters before others could become possible.
also, i suppose duplicating a filter in various stages of PP could be very usefull...for example a slight NR filter at the beggining of processing, tweaking most of the parameters u need and then apply another NR filter just before sharpening in the end, or something like this.


i know this is possible in PS and generally there are better tools than others for a specific task..still, such options for advanced processing would help a lot of people and would help us acieve potentially better results in some circumstances.

 

[Denis de Gannes]

Applying development adjustments in LR is a non-destructive process and completely different from working with a pixel editing program such as Photoshop. The adjustments you make are stored as instructions for LR to display the image on the screen. You also need to bear in ming that LR's primary function is the processing of raw files with the ability to work with jpeg, tiff, psd etc.
No changes to the original file are made and the creation of a new rendered copy are done through the export process or by sending it to a pixel editor like Photoshop for further editing.
AFAIK in the export process LR make the changes in a fixed order by design so it does not matter at what stage you apply adjustments. However I believe its usually better to get the exposure part sorted out first so the when you make other adjustments you can better view the effects of other adjustments to white balance, color correction, tone etc.

 

[Kiwigeoff]

As well you can place a feature request.

 

[Brad Snyder]

AFAIK in the export process LR make the changes in a fixed order by design so it does not matter at what stage you apply adjustments.

I'll confirm what Denis says that, at present, the 'rendering pipeline' in Lr executes in a fixed order as determined by the engineers as a design constraint.

Should you be able to modify that for technique or artistic vision? I'll leave that to better photog's than I. Seconding Geoff's suggestion, a feature request is probably in order, the more popular ideas receive the most attention.

 

[parintele]

Applying development adjustments in LR is a non-destructive process and completely different from working with a pixel editing program such as Photoshop. The adjustments you make are stored as instructions for LR to display the image on the screen. You also need to bear in ming that LR's primary function is the processing of raw files with the ability to work with jpeg, tiff, psd etc.
No changes to the original file are made and the creation of a new rendered copy are done through the export process or by sending it to a pixel editor like Photoshop for further editing.
AFAIK in the export process LR make the changes in a fixed order by design so it does not matter at what stage you apply adjustments. However I believe its usually better to get the exposure part sorted out first so the when you make other adjustments you can better view the effects of other adjustments to white balance, color correction, tone etc.

as far as i can imagine, LR develope is just a function where u define the variables for each adjustment and also depends by a variable that is fixed , determined by each picture, the raw...

it just calculates the end result for each pixel, in the end is basically also a pixel level thing.

also, while exposure for example is linear function, that mean u add a certain value to all pixels, adjustment brush is not, is much more complicate, coordonates of the pixels affected are also defined by the user...

the same is for noise just u do not define the coordonates but is still non linear function.

well, how exactly works ? each time u modify a parameter/adjustment, LR calculates another visual file format using the same RAW information/"fixed" variable and the rest of variables u modified, using ' or 1 values for variables u did not modified...

basically is also a pixel editing thing using functions amd raw info and adjustment amount numbers as parameters...



i figured out that the order in which those functions are applied is burned into developement engine.
infact that developement engine, as i said, is a big complicated function that put together lots of subfunctions representing each adjustment.


what i ask is would be possible that the engine would allow more flexibility also in DEFINING partially that big developing function my altering the order of it's subfunction components? or maybe not only the order but duplicating a subfunction once at the beggining of the chain, let;s say after WB and EXPOSURE, and once close to final just before sharpening???

that is the question....would be possible thet the developing engine itself to me customizable in some amounts???

and i thought of that startng from a worlkflow many people use...one of the first operations when is a lot of noise involved, is NR...


maybe the actual current engine does just that, would be logic, maybe not, would be nice to know exactly how it works and if possible to customize

 

[Mark Sirota]

I understand what you're asking for, but not why you're asking for it. What problem would this solve?

 

[parintele]

Mark, the problem is this:
shooting in bad lightning, extreme ISO...or advanced processing IN lightroom.

wedding shootings, events, sport related...

in all these areas there is a predispozition to produce high noise levels, cometimes color shifts on some textures (black that turn into greenish or blueish tints, various clothing textures that suffer from the same oversaturated wrong apparent exposure by hipersaturated color, etc..)
also usually u stretch the exposure, often underexposing alittle and recovering some areas in PP...

Noise reduction filters become very important, not only because of luminance noise grain effect but also croma noise which while processing can increase very ugly...

well, if noise reduction is not among the first filters applied, it just increase too much during various stages of PP... applying NR at the end of workflow just do not produce the same results compared to doing that at the beggining of the workflow.

this is thereason why many photograpehrs use it just after setting the basic exposure, before "highlight/shaddows" or curve, contrast, etcetc...and working on certain color or tone areas...


ofcourse, critical pictures can allways be processed in PS where u can control everything but here are 2 main issues:
1. many of these shooters preffer lightroom for 95% of the work, such shooting sessions often involve large volume of pics, LR offers a very good and erghonomic interface and lots of other advantages.

2. LR developing module still works with RAW info not the result after developing the raw, processing raw info totally thru developing function...
RAW keeps lot of usefull information, sometimes more than Tiffs, also allow to go back and forward using the same generous info captured...


for these reasons i consider filtering the noise at the right time WHILE developing and altering images using RAW information would be very usefull...

and the fact is many photographers would benefit from that because LR is more than suited for the photographers that are interested in this aspect..as i said, large volume in one session,extreme ISO. need for speeding up the PP, ergonomy, ranking, organize the materials, easy viewing and collection use, etc...


one who just whoot 5 frames at iso 64'' does not need that as processing 5 in PS with various adjustments is not big deal...try 1''' frames at a wedding or a football game...

the reason i opened this topic is not only expression of a personal thought regarding "wat i consider would be usefull" but rather hear other oppinions, ideas, alternatives, etc.

 

[Mark Sirota]

So why are you so sure that noise reduction isn't already one of the first steps applied?

 

[parintele]

i'm not, i just aked eralier, infact seems nobody really appear to know exactly how things works in this regard...at least i did not find ay refference until now.
it should be public documentation in order to allow users to maximize their results.
Lr and generally adobe products are pro apps.usually pros want ergonomy, ease of use but also as much control as possible in order to acieve maximum end result out of the raw materials...

control in this regard is not yet offered, maybe Adobe programmers already thought about everything and things are indded as they should be...why not make it public, why not make another big plus out of this tweak???

is either a problem of informing the users about a great mechanism or, if not already done, an issue to be approached in future versions.

anyway, user custom control would be nice regarding the subject...

 

[Sean McCormack]

If I were to make an educated (ha!) guess. I would say that Lightroom applies in the order that appears in the XMP data. At least that would seem to be very likely IMHO.

I'm open to being out of the ball park on this, but now that I've said it, does it make sense?

 

[Brad Snyder]

Well, after inspecting a few XMP files, that's as good a guess as any, I suppose. I'm thinking Adobe is playing this along the lines of a 'trade-secret', so I don't expect we'll be able to get Thomas Knoll or Zalman Stern to comment on this.

Liviu, do you participate at the Adobe User to User forums? Have you asked the same question there?

 

[parintele]

hey guys, i just posted an official request form and i also opened anew topic inside adobe forums, LR, requests cathegory..let's see what adobe officials have to say about all this

http://www.adobeforums.com/webx/.59b7beca

 

[parintele]

seems like some people on Adobe forums are not that "kind"
as usuall, less facts but lot of subjective thoughts :x

[COLOR=#''66cc]Lee Jay[/COLOR] - 1:'8pm Jan 31, '9 PST[/B] (#[COLOR=#''66cc]1[/COLOR] of 2)

that mean noise is indeed among the last filters applied, correct me if i am wrong.​

You are wrong.

The pipeline order is carefully thought out and optimized. It is not adjustable as you stated, and I can see no reason it ever should be.

that made me look into a recently created XMP file and here is what i found:

here is a quote from a XMP file, i PRESUME the order if operation inside that file is the one processes are piped, am i wrong again?

<crs:Version>5.3</crs:Version>
<crs:WhiteBalance>As Shot</crs:WhiteBalance>
<crs:Temperature>49''</crs:Temperature>
<crs:Tint>-2</crs:Tint>
<crs:Exposure>-'.57</crs:Exposure>
<crs:Shadows>15</crs:Shadows>
<crs:Brightness>+79</crs:Brightness>
<crs:Contrast>-7</crs:Contrast>
<crs:Saturation>+33</crs:Saturation>
<crs:Sharpness>25</crs:Sharpness>
<crs:LuminanceSmoothing>39</crs:LuminanceSmoothing>
<crs:ColorNoiseReduction>18</crs:ColorNoiseReduction>

<crs:ChromaticAberrationR>'</crs:ChromaticAberrationR>
<crs:ChromaticAberrationB>'</crs:ChromaticAberrationB>
<crs:VignetteAmount>'</crs:VignetteAmount>
<crs:ShadowTint>'</crs:ShadowTint>
<crs:RedHue>'</crs:RedHue>
<crs:RedSaturation>'</crs:RedSaturation>
<crs:GreenHue>'</crs:GreenHue>
<crs:GreenSaturation>'</crs:GreenSaturation>
<crs:BlueHue>'</crs:BlueHue>
<crs:BlueSaturation>'</crs:BlueSaturation>
<crs:FillLight>17</crs:FillLight>
<crs:Vibrance>-21</crs:Vibrance>
<crs:HighlightRecovery>29</crs:HighlightRecovery>
<crs:Clarity>+1'</crs:Clarity>
<crsefringe>'</crsefringe>
<crs:HueAdjustmentRed>'</crs:HueAdjustmentRed>
<crs:HueAdjustmentOrange>'</crs:HueAdjustmentOrange>
<crs:HueAdjustmentYellow>'</crs:HueAdjustmentYellow>
<crs:HueAdjustmentGreen>'</crs:HueAdjustmentGreen>
<crs:HueAdjustmentAqua>'</crs:HueAdjustmentAqua>
<crs:HueAdjustmentBlue>'</crs:HueAdjustmentBlue>
<crs:HueAdjustmentPurple>'</crs:HueAdjustmentPurple>
<crs:HueAdjustmentMagenta>'</crs:HueAdjustmentMagenta>
<crs:SaturationAdjustmentRed>'</crs:SaturationAdjustmentRed>
<crs:SaturationAdjustmentOrange>'</crs:SaturationAdjustmentOrange>
<crs:SaturationAdjustmentYellow>'</crs:SaturationAdjustmentYellow>
<crs:SaturationAdjustmentGreen>'</crs:SaturationAdjustmentGreen>
<crs:SaturationAdjustmentAqua>'</crs:SaturationAdjustmentAqua>
<crs:SaturationAdjustmentBlue>'</crs:SaturationAdjustmentBlue>
<crs:SaturationAdjustmentPurple>'</crs:SaturationAdjustmentPurple>
<crs:SaturationAdjustmentMagenta>'</crs:SaturationAdjustmentMagenta>
<crs:LuminanceAdjustmentRed>+6'</crs:LuminanceAdjustmentRed>
<crs:LuminanceAdjustmentOrange>'</crs:LuminanceAdjustmentOrange>
<crs:LuminanceAdjustmentYellow>-24</crs:LuminanceAdjustmentYellow>
<crs:LuminanceAdjustmentGreen>'</crs:LuminanceAdjustmentGreen>
<crs:LuminanceAdjustmentAqua>+19</crs:LuminanceAdjustmentAqua>
<crs:LuminanceAdjustmentBlue>'</crs:LuminanceAdjustmentBlue>
<crs:LuminanceAdjustmentPurple>-24</crs:LuminanceAdjustmentPurple>
<crs:LuminanceAdjustmentMagenta>'</crs:LuminanceAdjustmentMagenta>
<crs:SplitToningShadowHue>'</crs:SplitToningShadowHue>
<crs:SplitToningShadowSaturation>'</crs:SplitToningShadowSaturation>
<crs:SplitToningHighlightHue>'</crs:SplitToningHighlightHue>
<crs:SplitToningHighlightSaturation>'</crs:SplitToningHighlightSaturation>
<crs:SplitToningBalance>'</crs:SplitToningBalance>
<crsarametricShadows>'</crsarametricShadows>
<crsarametricDarks>'</crsarametricDarks>
<crsarametricLights>'</crsarametricLights>
<crsarametricHighlights>'</crsarametricHighlights>
<crsarametricShadowSplit>25</crsarametricShadowSplit>
<crsarametricMidtoneSplit>5'</crsarametricMidtoneSplit>
<crsarametricHighlightSplit>75</crsarametricHighlightSplit>
<crs:SharpenRadius>+1.'</crs:SharpenRadius>
<crs:SharpenDetail>25</crs:SharpenDetail>
<crs:SharpenEdgeMasking>'</crs:SharpenEdgeMasking>
<crsostCropVignetteAmount>'</crsostCropVignetteAmount>
<crs:ConvertToGrayscale>False</crs:ConvertToGrayscale>
<crs:ToneCurveName>Medium Contrast</crs:ToneCurveName>
<crs:ToneCurve>
<rdf:Seq>
<rdf:li>', '</rdf:li>
<rdf:li>32, 22</rdf:li>
<rdf:li>64, 56</rdf:li>
<rdf:li>128, 128</rdf:li>
<rdf:li>192, 196</rdf:li>
<rdf:li>255, 255</rdf:li>
</rdf:Seq>
</crs:ToneCurve>
<crs:CameraProfile>ACR 3.3</crs:CameraProfile>
<crs:CameraProfileDigest>8E17227'6FEF96E97D62F657E5'DF272</crs:CameraProfileDigest>
<crs:HasSettings>True</crs:HasSettings>
<crs:AlreadyApplied>False</crs:AlreadyApplied>





SO, as u can see here, WB, exposure, contrast, SHARPNESS, brightness, SHADDOWS are ALL listed BEFORE noise reduction...

later on the end of the list there is ANOTHER sharpening listed, so double fiter IS ALREADY applied, thus possible, right???

again, ALL fiters applied BEFORE NR influences noise levels, all increse noise, more or less....
WB influences exposure, try to go 5'''' and see where hystogram goes to the right...
exposure also increase noise in shaddow areas.
contrast increase noise.
brightness also influence noise.
Shaddows are working exactly on most noisiest areas, the shades...

so, it could be something to improve after all...

also, there are 2 sharpness instrucntions there, one at the beggining and one close to the end... so double filtering using the same filter is already there


let's see how things goes further

 

[Denis de Gannes]

This is totally speculative. The xmp appears to show the adjustments in the order they were applied. I recall in earlier posts from persons that are in the know that LR processes adjustments in a specific order and is not done in the order that they are applied.
e.g if you applied an adjustment to exposure first + 55 then later on in the pipeline -85 resulting in a final adjustment of -3', I expect the export will process the net of -3' and not the two adjustments. What is being proposed here make no logic.
As far as I understand LR is using a non destructive process and no adjustments are made to the original file. If you are working with a raw file in the develop module each adjustment is applied to the raw data and displays on the screen. If you return the following day and continue to make further adjustments the original raw data is used again and the accumulated adjustments applied. If you then export to tiff, jpg or psd the resultant file should match what you see on the screen. Do a test apply the nr first in your workflow and export, then apply the nr last and do a second export. Compare the results and I think they will be the same.

 

[parintele]

as far as i understand things should work like that.
raw data is a constant parameter used inside a mathematicall function.


let;s say the function looks like that...
[raw] * [exposure compensation] * [contrast] * [NR] * [brightness] * [saturation] * [ sharpening]
this function takes [raw] parameter, each for everypicture, that is constant value for each picture.
using that number it start producing a image file by "multiplying" that [raw] number with each otehr parameters...

first [exposure compensation], that could be +'.5 stops value or any other numeric value u can set inside LR....
then the result multiply with [contrast], let's say 35 value and so on.... the end result is the image file.

well, when u do not modify one of the parameters it just multiply with 1....that mean the same as previous number, not altered...


BUT, the developing function is not that simple like above...some operations are not like multiplying, unsensitive to order...

if u have a function with + or - instead of only * the order becomes important...

a+b*c is not the same with b+a*c...that is simple math


i say this is the case of applying those filters, one filter applied before others influence differently the final result...

in our case NR...

so, is not important if u raise exposure +1 stop and then decrease -'.5 value so u end with +'.5
not that matters because every time u change one of the parameters, LR calculates ANOTHER image...using all the other parameters, contrast, brightness, etc... IN THE SAME PREDEFINED ORDER but using different numeric values for each parameter...

order the filters are applied is the same, numeric values of each filter is different, the whole calculation is make from scrap, using the raw data and all teh other parameters, even those left as they were before...



all i am saying is order inside the function need to be different, it have nothing to do with how many changes u make to one parameter...


i am not sure i made myself clear, please feel free to ask if you did not get it...

thanks

 

[Mark Sirota]

I'm still perfectly clear on what you're saying. I'm still completely unclear about why you care.

What do you want to do with your pictures in LR that you can't currently do, and why?

 

[parintele]

i am not sure math adn defining mathematical fucntions/computer programming functions use the same addnotation inenglish as it does in romania but i will give it a try.

f(x) - is the first applied filter, let's say WB

then we have g(f(x)) - g is second filter, let's say exposure.

then we have h(g(f(x))) being contrast

then i(h(g(f(x)))) for saturation

j is the display on the screen fucntion, develope raw in order to display image on the screen , ok.

so, raw is a given numeric value.

j(x) = display ( i(h(g(f(x)))) )

so, when the variable take the value raw, we have:

j(raw) = display (i(h(g(f(raw)))) )

that mean operation order is like this.
first u calculate f(raw)...the function f can be any mathematical function, the result could be any number, let's say [a]
then u calculate g([a]), as now a=f(raw)
g([a]) is let's say number

it follows h()...which is [c]
i([c]) = [d]

so, in the end, developed raw , that mean j(raw)= display [d]

that is what u see or what u export, [d]


all the fucntions f, g,h,i can be very complicated mathematical functions inside of which u set another variable parameter, that parameter describe amount of alteration u want to apply...
f function vary depending on the WB alteration u specify
g function vary depending on the exposure alteration u specify
h function vary depending on the contrast alteration u specify
....and so on



so, i am not concerned about that alteration inside the f, g, h, i functions but the order these functions take inside the developement fucntion, that is J function



j(raw)= display (i(h(g(f(raw)))) ) is not the same if you define
k(raw) = display f(h(g(i(raw))))

raw value is the same, parameters inside each fucntion f, g h,i can be the same, the result of the 2 develope fucntions j(raw) and k(raw) can be different.


i hope i did not bother u too much and i hope this mathematicall related example can be better understood...

 

[parintele]

I'm still perfectly clear on what you're saying. I'm still completely unclear about why you care.

What do you want to do with your pictures in LR that you can't currently do, and why?

Mark, is simple and i said it before.
my usual workflow in PS involves NR among the first filters applied.
get rid of as much noise as possible BEFORE i further process the picture because working on exposure increase noise in certain conditions, contrast increase noise, sharpening increase noise, and so on...

if you do not reduce noise u just increase noise with every filter like that and at the end of the process u need to use a very agressive noise reducing filter. if u apply NR first, u also increase the left noise but much less...

aplying NR 1'% at the beggining of the workflow can produce cleaner results compared to applying 5'% NR at the end, not to mention u save tons of details compared to reducing noise at the end in such great amount...


that is all, in PS u can do that, in LR u have no control, as it seems NR is applied after WB, exposure, contrast, shaddows are fixed....even after a primary sharpening process which is idiocy for me, i never sharpen first and after reduce noise...

is a simple thing...

is not extremelly critical, as many other things we all can survive without it... but IF is possible, and considering this section is about wishfull thinking, why not be able to do things better in some circumstances???

this is all about, isn;t it, try to do things little better in order to improve the final result...
most of the times does not matter but again, try to shoot at iso 32'' or 128'' with latest cameras...noise is a real concern.
u can go to PS to take perfect control but why use LR then??? is faster, have great advantages, why not impelment this control element in order to save some time and improve quality...u can not go to PS for each of 1''' pictures...
u can create an action but what is the advantage of shooting raw then if u open a tiff inside PS???

i prefer to do most of the work inside LR on raw's, to extract the maximum quality, especially when quality is not great like iso 2'' shots...

that is all. wishfull thinking, would be nice if ...i can live without it like all others, i just want more and more..

 

[Mark Sirota]

So again -- do you need control over the order, or do you just want to confirm that noise reduction happens first?

Why are you confident that noise reduction doesn't already happen first?

 

[parintele]

so again, indeed:
1. i want to exactly KNOW how things happen, which is the order, if some happens in 2-3 stages , etc...
knowing is the first step.
if we know for sure it works badly and we need the best we can get out of raw material, we might choose to go to PS where we can have total control...

2. reading the XMP info it SEEMS that those variable related to each filter are feeded strangely...
as u can see in above quote, N is among the first but is after :

<crs:Version>5.3</crs:Version>
<crs:WhiteBalance>As Shot</crs:WhiteBalance>
<crs:Temperature>49''</crs:Temperature>
<crs:Tint>-2</crs:Tint>
<crs:Exposure>-'.57</crs:Exposure>
<crs:Shadows>15</crs:Shadows>
<crs:Brightness>+79</crs:Brightness>
<crs:Contrast>-7</crs:Contrast>
<crs:Saturation>+33</crs:Saturation>
<crs:Sharpness>25</crs:Sharpness>
<crs:LuminanceSmoothing>39</crs:LuminanceSmoothing>
<crs:ColorNoiseReduction>18</crs:ColorNoiseReduction>

i gues i haveenough reasons to presume NR is applied after all blue ones, right?


[COLOR=#''''''] also, there is another sharpness applied at the end, so there is a filter that is applied twice, just as i requested for 2 stage NR filter....[/COLOR]

[COLOR=#'''''']3. more control would be nice regarding NR, at least a basic form of it, like a checkbox which make the filter a prority, and moves up to priotity list, a second checkbox for doubling that filter, one being up and one at the end, just before last adjustments like final sharpening...[/COLOR]


[COLOR=#'''''']that is all, we do not know exacly, that is the first problem, seems not perfectly placed, if is "thought" liek it seems woudl be nice to be able to increase the priority even by a simple checkbox, qould be nice toapply it twice as no matter if is among the first, following tweaks also may increase noise, things like saturation and especially luminance on each color can increase noise in some areas. [/COLOR]

[COLOR=#''''''] overall i consider there is enough room for improvement for better results when working in some circumstances (extreme high iso raw's OR extreme processing on various tone intervals / colors)


[/COLOR]

 

[Victoria Bampton]

The only way you're going to get the answers is direct from Thomas Knoll himself, and he's unlikely to hand out that information. Both these forums and the Adobe user-to-user forums will not be able to provide the concrete proof you want, so this is going to go in circles. I'll let it run a little longer, but then I will be closing this thread.

1. The order that the adjustments are applied will be intellectual property. They are unlikely to ever disclose that information. These guys have been doing this for years. They know what will give the best results from their processing engine, so they will use the order of application that will give the best results. It would be illogical for them to do anything else. If you feel you can do a better job, you could build your own raw processor. This is not like processing image data in Photoshop.

2. Ignore the XMP order. It has nothing to do with anything. FWIW, if you look down that list, that appears to be pretty much the order the sliders were gradually to the ACR engine over the course of many versions.

3. There is always room for improvement, but that will not come from changing the order that the sliders are applied.

 

[Mark Sirota]

Thanks for clarifying that you're looking for knowledge, rather than the ability to control the order (at least until after you learn the current order).

However, like Victoria says, I'm sure nobody who actually knows would be willing or able to disclose the answer.

 

[parintele]

yeah, seems nobody really care about such minor aspect, nobody really ask himself how and why is white and not black...so this will be my last post on this one, i will give it a try by mailing to the right/directly involved guys...who knows

so, Victoria:

These guys have been doing this for years. They know what will give the best results from their processing engine, so they will use the order of application that will give the best results. It would be illogical for them to do anything else.

meybe yes, maybe not. u never know, isn' it? if we would not ask for better ways/alternatives would be no progress, right?
think of resizing algorithms, sharpening, noise reduction (people do not use NN or neat plugins because Adobe have perfectly well thought alternatives, isn't it?)

so, nobody is perfect, not even adobe guys...

If you feel you can do a better job, you could build your own raw processor.

in this situation i would not do a request, would i? now what, if anyone ask himself something or rewquest something he shoudl build his own apps?
that is kind of radical approach, i think things are little different...
users propose/request some "crazy" aparentthings , producers decideif and which of those are partially or totally valuable and implement some of it..maybe not this one, true, but the "make your own app" is kind of "don't ask, be happy or die"...

This is not like processing image data in Photoshop.

infact it is pretty similar with applying several adjustment filters one on top of each other...some are order sensitive, some are not, there are also differences but basically it prety similar as a concept. just instead of pixel numeric values from image file u use info from raw file that need extra processing in order to become image information wrote in a specific image file form.
is same pure math, fucntions, variables,etc..

FWIW, if you look down that list, that appears to be pretty much the order the sliders were gradually to the ACR engine over the course of many versions.

i agree XMP can be irrelevant regarding the subject, u can read from a file as u wish, u can start backward reading if u want...

still, i disagree with quoted statement, i qouted froma XMP file on previous pages and there order is seriously different compared to slider order from LR and CR...that being a proof that indeed, app can read and write inside a file as u please.
we can indeed say as it writes and read it also can use theinfo ina different order



just some thoughts, again, is my last post on this regard, is no pint continuig this debate as it just come to a dead end...
could be an idiotic ideal, an already sorted and indeed well thought thing, could be something good in all this, we'll never really know...unless thoseinvolved would let us know somehow, if i will get a valid answer from someone like that i'll keep u updated (maybe is not that top secret, u never know)

cheers

 

[Bruce J]

I've just read through this thread again, and I don't get it. Not trying to further stir up the hornet's nest, but why are you assuming that the processing is done in the order of the tags in the XMP file? I've done a fair amount of XML design and in all cases, the resulting files have nothing to do w/ processing order. They're just tags and they probably occur in the file in some order that makes the file easier for the programmers to read or, possibly, in the order that they occur in the user interface, or in the order that the designer thought of them.
When you process an XML file, it's not read sequentially, it's parsed into a table from which the values are accessed as needed. Think of it as a database of name-value pairs. When you want to process sharpness, you just access the value of the sharpness tag. Unless you have information direct from the Adobe engineers about processing order, I think you're making a huge, unwaranted assumption that the XMP order has anything to do with processing order. Cheers,

 

[parintele]

asumptions comes from lack of info...what can we do except asking how it works or presume??
TRUE, i'e said id 3 times by now, order of sliders inside LR and order inside the XMP might be totally irrelevant.

They're just tags and they probably occur in the file in some order that makes the file easier for the programmers to read or, possibly, in the order that they occur in the user interface, or in the order that the designer thought of them.

yes, there can be 1''' reasons, even thestrongest one: confusing those like me who would try to "guess" what i "assumed" in this thread...maybe they are protectig the secret of engine itself, their "well thought" order...
sill, is not the order in user interface, easyer for programmers to read have nothing to do as not they need to read but the appm and that is exactly what i assumed , order that designer thought of them i really do not think so... i have a programming background myself (studied IT in a special highschool )..when u develope such serious app u just do not write it at first guess, in the end u anyway use a logicall order in everything, keep things organized...again, only if u intention to confuse others u make something without a logic...

bottom line, this thread should really stop.

I started by asking how things work and requesting CONTROL, a simple checkbox for an alternative regarding priority of some filters... might be stupid, might be something...
we already begun assuming and debating XML content and stuff like that, no point wasting time on assumptions

I posted an official request, i also mailes several Adobe partners and developers, shall wait for an oficial response and maybe, if there is something good in this small idea, an update in future versions...

well see, if i get something usefull from all this i will let you guys know.

thank you for spending some time participating in this discussion, at least i found that nobody knows details, nobody really feel need for it.

cheers, let's stop it.