Thanks Anselm,

I had a play with dg_HSLTool. Using it I have got the best results yet and it 
is frankly a very cool little tool. I managed to do an acceptable hue stretch 
after a few attempts. Again, spill of effect is a problem, but the sat 
threshold curve seems to serve most purposes. 

It is far from intuitive, but if a user knows there way around curves and 
understands the principles, then it is ok. Some hue ranges are more difficult 
than others. 

A hue compress seems more difficult and I have not yet got it to work. 

If my coders have time I will ask if they can implement a plugin. Personally I 
would love to see it implemented, but it is difficult to get students to do 
anything if they don't see a publication as the end result. 

Many, many thanks for the feedback. This is very useful. 


Martin 




On 10 Jun, 2014, at 1:24 PM, Anselm Lier <[email protected]> wrote:

> I often use the HueShift node to smoothen skin tones. But I find it difficult 
> to configure it for other hues. And I have to key out the skin first.
> I recall now that there is a gizmo for a Hue-vs-Hue curve approach on 
> Nukepedia: 
> dg_HSLTool v1.0
> 
> After playing with it for a minute I guess that would do most of what I 
> expect in that regard.
> But maybe the tool you are developing is even easier to adjust. I would like 
> to see it in action some day.
> 
> Best,
> 
> Anselm
> 
> 
> 
> --------------------------------------------
> 
> Anselm Lier
> Haslerspitzstr. 5
> 81673 München
> 
> Phone:        +49 (0) 151 165 178 38
> Mail: [email protected]
> 
> Am 10.06.2014 um 06:18 schrieb Martin Constable <[email protected]>:
> 
>> I don't know the HSV tool very well. Always understood it as Photoshop's Hue 
>> Sat adjustment on steroids. 
>> 
>> But... as far as I am aware the HSV tool can:
>> - Replace hue
>> - Cycle the hue values
>> 
>> As with the 'perform contrast adjustment on the H channel' method, it is 
>> difficult to stop the effect spilling over into neighbouring hues. 
>> 
>> Functionally, it seems simply to take one hue value and replace it with 
>> another. The docs are a bit zen on the exact function of the range slider 
>> parameter, but it seems to spread the source color range. 
>> 
>> Our tool does something like this: It takes a user defined contiguous 
>> selection of the hue histogram and stretches (or compresses) it. 
>> 
>> I have compared the two results using the Vector Scope and the effects are 
>> clearly different, though I admit that currently our approach can produce 
>> 'gappy' histograms (I am assured that this is solvable).  
>> 
>> 
>> Thanks for the feedback Diogo! These are very useful comments.   
>> 
>> 
>> 
>> On 10 Jun, 2014, at 11:47 AM, Diogo Girondi <[email protected]> wrote:
>> 
>>> Perhaps I'm missing something, but isn't that what the HSVTool does?
>>> 
>>> 
>>> cheers,
>>> Diogo
>>> 
>>> 
>>> On Mon, Jun 9, 2014 at 12:16 PM, Elias Ericsson Rydberg 
>>> <[email protected]> wrote:
>>> You could do a key on a specified range of hues using the keyer, using that 
>>> as a mask for a grade node. That node manipulates the red(rgb->hsv first) 
>>> would probably give you the results you're looking for?
>>> 
>>> Don't have nuke in front of me but I think that could work.
>>> 
>>> Best regards,
>>> Elias Ericsson Rydberg
>>> 
>>> måndag 9 juni 2014 skrev Martin Constable <[email protected]>:
>>> 
>>> Thats a good idea and actually something I already tried.
>>> 
>>> After some fiddling similar results can be had. And the results themselves 
>>> are sometime comparable.
>>> 
>>> However, it is very difficult to control and the effects is very likely to 
>>> spill over into other color values. Drawing an adjustment in ColorLookup 
>>> can offer more local control but is even more difficult to master.
>>> 
>>> 
>>> 
>>> On 9 Jun, 2014, at 5:22 PM, Adrian Cruceru <[email protected]> wrote:
>>> 
>>>> Try this:
>>>> 
>>>> Use a colorspace node to change the current color space from linear to HSV 
>>>> then use a grade node to color correct just the red channel (which becomes 
>>>> the Hue chanel after the colorspace node) then use a second colorspace to 
>>>> get the colors back into the linear color space
>>>> 
>>>> Any color correction applied to the red channel between the two colorspace 
>>>> nodes will only alter the hue, so you cna compress it and change it as you 
>>>> want
>>>> 
>>>> Hope it helps
>>>> 
>>>> Cheers
>>>> 
>>>> Adrian
>>>> 
>>>> 
>>>> On Mon, Jun 9, 2014 at 10:01 AM, Martin Constable <[email protected]> 
>>>> wrote:
>>>> Unforgivably I accidentally deleted the questions from the source file.
>>>> 
>>>> They are:
>>>> The questions must have got accidentally deleted. They are simple:
>>>> - Can such an effect be easily achieved with existing tools?
>>>> - Would you find such a tool useful?
>>>> 
>>>> Apologies
>>>> 
>>>> 
>>>> On 9 Jun, 2014, at 3:44 PM, Martin Constable <[email protected]> wrote:
>>>> 
>>>>> Dear Nuke users,
>>>>> 
>>>>> 
>>>>> I am writing to you in order to ask for your help on a research project 
>>>>> that I am engaged in with here at Nanyang Technological University. I 
>>>>> hope you won't regard this as spam.
>>>>> 
>>>>> I am working with some engineers on a tool whereby the local hue contrast 
>>>>> of a region may be expanded or compressed. Exactly what we mean by this 
>>>>> will take a bit of explanation which I hope is delivered in the linked 
>>>>> pdf.
>>>>> 
>>>>> We are now hoping to publish and are looking for feedback from 
>>>>> professionals as to the usefulness and novelty of our approach. Ideally, 
>>>>> we would like people who have a special interest in color.
>>>>> 
>>>>> A very short description is here, as is two questions we would like to 
>>>>> ask:
>>>>> https://dl.dropboxusercontent.com/u/3788380/HUE_TOOL_FEEDBACK.pdf
>>>>> 
>>>>> We are working to a tight schedule (deadline = this Friday 13th) and do 
>>>>> not have time for a full user study. However, one or two lines of 
>>>>> feedback will be very helpful, especially from users of your caliber.
>>>>> 
>>>>> I thank you all sincerely in advance
>>>>> 
>>>>> Martin Constable
>>>>> 
>>>>> 
>>>>> 
>>>>> CONSTABLE Martin | School of Art Design and Media | Nanyang Technological 
>>>>> University
>>>>> 3-21, 81 Nanyang Drive, Singapore 637458
>>>>> Tel: (65) 6316-8729 GMT+8h | HP: (65) 9329-4064 | Fax: (65) 6795-3140 | 
>>>>> Email: [email protected]
>>>>> 
>>>>> 
>>>>> 
>>>>> _______________________________________________
>>>>> Nuke-users mailing list
>>>>> [email protected], http://forums.thefoundry.co.uk/
>>>>> http://support.thefoundry.co.uk/cgi-bin/mailman/listinfo/nuke-users
>>>> 
>>>> _______________________________________________
>>>> Nuke-users mailing list
>>>> [email protected], http://forums.thefoundry.co.uk/
>>>> http://support.thefoundry.co.uk/cgi-bin/mailman/listinfo/nuke-users
>>>> 
>>>> 
>>>> 
>>>> --
>>>> Adrian Cruceru
>>>> df(x)
>>>> _______________________________________________
>>>> Nuke-users mailing list
>>>> [email protected], http://forums.thefoundry.co.uk/
>>>> http://support.thefoundry.co.uk/cgi-bin/mailman/listinfo/nuke-users
>>> 
>>> _
>>> 
>>> _______________________________________________
>>> Nuke-users mailing list
>>> [email protected], http://forums.thefoundry.co.uk/
>>> http://support.thefoundry.co.uk/cgi-bin/mailman/listinfo/nuke-users
>>> 
>>> _______________________________________________
>>> Nuke-users mailing list
>>> [email protected], http://forums.thefoundry.co.uk/
>>> http://support.thefoundry.co.uk/cgi-bin/mailman/listinfo/nuke-users
>> 
>> _______________________________________________
>> Nuke-users mailing list
>> [email protected], http://forums.thefoundry.co.uk/
>> http://support.thefoundry.co.uk/cgi-bin/mailman/listinfo/nuke-users
> 
> _______________________________________________
> Nuke-users mailing list
> [email protected], http://forums.thefoundry.co.uk/
> http://support.thefoundry.co.uk/cgi-bin/mailman/listinfo/nuke-users

_______________________________________________
Nuke-users mailing list
[email protected], http://forums.thefoundry.co.uk/
http://support.thefoundry.co.uk/cgi-bin/mailman/listinfo/nuke-users

Reply via email to