Curve Merging or Averaging

[mkochsch]

What's the easiest way to add two ACV files, or curve files, together? What I'm currently doing is making a second curve to "fine tune" the first curve and then reapplying the second curve after I've applied the first curve. What I would like to do is merge the two curves together as a new curve file.
~m

 

[Joe Lipka]

I believe that you could just apply the curves to succeeding layers and that would accomplish your goal.

But then again, it would require experimentation to be sure...

 

[jd callow]

I think Joe is correct. You could also try to add/subtract the differrences of your inputs and outputs manualy. I suspect that you are looking for an automated method that will merge the curves and I don't know of one.

 

[amphoto]

Curve merging

Alternatively, what you can do is apply each curve in succession to a standard greyscale step tablet (in photoshop), and then measure the altered values of the step tablet after both these curves have been applied. You can thus build a composite adjustment curve in much the same way that you build a standard adjustment curve.

 

[mkochsch]

I believe that you could just apply the curves to succeeding layers and that would accomplish your goal.

But then again, it would require experimentation to be sure...

I'm already doing that. I think were I'm wanting to go is a way of visually comparing curves. Ideally, if I could open two or more curves in a program window and compare how changing certain darkroom variables is affect the slope of the curve, the toe, the shoulder, the density. A tweaker utility.

 

[Joe Lipka]

You just solved your problem. Copy your photo so you have two versions, one with each curve. Open both at one time and experiment with the curves as you like.

That's one approach.

 

[mkochsch]

You just solved your problem. Copy your photo so you have two versions, one with each curve. Open both at one time and experiment with the curves as you like.

That's one approach.

Let me clarify....After running a curve calibration, if one applies the resulting curve to another calibration target further adjustments and improvements to the original curve can be made. It's possible even a third iteration would yield an even better curve. Liam Lawless proposed such a method in Post Factory #9 ["Tweaking the Digital Negative", p.36]. The problem is doing this by hand is painfully slow.
Changes in humidity, coating method, temperature, time etc. all may affect the emulsion on the calibration targets when printed. One person I know will actually print out a half dozen or more and averages the results. If I could take the six curves from that person's calibration images and average them into one curve it would simplify things. No dealing with multiple layers of curves, having to remember to turn on or off certain layers. Produce just one really good curve.
A system like this would also have the advantage of letting us compare how curves are changing over time -- easily. Kind of like running the calibration on your monitor every two or three weeks. Also, one could compare curves from different coloured emulsions in the same family or different families altogether.

I've taken the first step toward writing a program to do this. I'm not a programmer by trade so it might a little time.
Guillermo Luijk the author of Tone Hacker (if you haven't checked it out yet http://perso.wanadoo.es/j.luijk) has already given me some help and tips on reading the binary acv file. I now know how it is formatted and can read it. So now it's time to lock myself in woodshed and see where it leads me...a pui tarde.
~m

 

[Colin Graham]

Probably I'm looking for something completely different, but I've been playing with multiple curves a little today and the only luck I've has so far is applying the secondary curve first, before all the drama of the main curve is added, otherwise the tones I want to tweak are impossible to locate. But next I'm going to try applying each curve on a seperate layer and merging before inverting and filling, seems like this may be the most neutral and 'WYSIWYG' route. It would be very nice to be able to overlay 2 or more curves, to be able to compare them visually.

edit- Just tried the 2 seperate curve layers and it added a surprising amount of density compared to the sequential curves on the background layer...So I probably need to experiment with the blend options.

 

[mkochsch]

Pierre-Oliver is working on an idea like that here:
http://www.pierreoliviertavernier.com/Joomla
The software is still in it's infancy but it's coming along. Works using Paint Shop Pro (free to 30 day demo) and not PhotoShop however.
~m

 

[sanking]

The problem is doing this by hand is painfully slow.
Changes in humidity, coating method, temperature, time etc. all may affect the emulsion on the calibration targets when printed. One person I know will actually print out a half dozen or more and averages the results. If I could take the six curves from that person's calibration images and average them into one curve it would simplify things. No dealing with multiple layers of curves, having to remember to turn on or off certain layers. Produce just one really good curve.
\

If the objective is to combine several curves to produce a curve that compensates for the inconsistencies of process it would seem to me that the average curve is destined to produce less than ideal results most of the time.

The best practice IMO is to learn what is necessary so that you can always get repeatable results with the process, then you would have only one curve that is pretty much perfect. This means temperature and humidity control in the workroom, and consistent procedures of coating and drying.

Sandy

 

[mkochsch]

If the objective is to combine several curves to produce a curve that compensates for the inconsistencies of process it would seem to me that the average curve is destined to produce less than ideal results most of the time.

The best practice IMO is to learn what is necessary so that you can always get repeatable results with the process, then you would have only one curve that is pretty much perfect. This means temperature and humidity control in the workroom, and consistent procedures of coating and drying.

Sandy

Wow did I really write that a year ago?! How time flies. Sandy I think I was still "roughing" in my colour choice for blocking at this point. That is, choosing the first white square in the colour step wedge and then curving twice. A few months later I started doing the "refinement" procedure (sampling the first white square on the 101) to get the "perfect" blocking colour rather than apply a second curve to fix the excess density of the first.
BUT, I still think that the curve averaging idea is worthwhile pursuit, for at least one good reason. Coating. Hand coating materials is always going be a bit of a question mark and probably have some variation. Bless those whose every calibration test turns out like the last. I'll put my money on the law of averages. Rest assured I haven't forgotten the idea.
~m

 

[sanking]

Wow did I really write that a year ago?! How time flies. Sandy I think I was still "roughing" in my colour choice for blocking at this point. That is, choosing the first white square in the colour step wedge and then curving twice. A few months later I started doing the "refinement" procedure (sampling the first white square on the 101) to get the "perfect" blocking colour rather than apply a second curve to fix the excess density of the first.
BUT, I still think that the curve averaging idea is worthwhile pursuit, for at least one good reason. Coating. Hand coating materials is always going be a bit of a question mark and probably have some variation. Bless those whose every calibration test turns out like the last. I'll put my money on the law of averages. Rest assured I haven't forgotten the idea.
~m

I did not pay any attention to the dates and thought this was a new thread. It is true that there will always be some variation in hand coating materials. On the other hand, with careful control of humidity, temperature and procedures one should be able to make an edition of 10-12 prints in processes like carbon, vandyke, kallitype and pt./pd that are virtually identical in density and contrast.

Sandy

 

[Kees]

I found the Curve Merging Method!

Hi all,

After long search I found a method in this article by Martin Evening:

http://photoshopnews.com/2007/09/05/how-to-express-blend-modes-as-curves/

Basic idea is a tip from Russell Williams at Adobe about saving curve data in a photoshop raw image file. A straight curve is represented by a 256x1 pixel regular gradient. Apply your curves on this file and resave it as a new curve. Stunning idea, and it works!

This is how I use this method for curve merging.
Let's say you have two curves: curve_01.acv and curve_02.acv where curve_01.acv is your first charthrob curve and curve_02.acv is a second pass charthrob curve for finetuning.

1. open any file in photoshop and open the curves palette. You see a straight line curve.
2. click on the pencil button next to the curve button. You enter manual curve editing mode. Save this straight curve as a .amp file. Name it merge_curves.amp for example. (if you don't click on the pencil you get a .acv instead of a .amp file and that doesn't do the trick)
3. rename this file to merge_curves.raw making it a photoshop raw file
4. open this raw file in photoshop
5. photoshop shows a dialog asking for the image dimensions and some more details
6. in this dialog put width: 256 pixels, height: 1 pixel, channels count: 1, depth: 8 bits, header size: 0.
7. you just created a regular 256x1 pixel grayscale gradient representing all the data in a straight curve. Save (a copy of) this file for later use

8. in comes the magic!
9. tada.....

10. apply curve_01 to this file
11. apply curve_02 to the allready curved file (doing exactly as you would do with your real file)
12. flatten, if needed, and save as: my_merged_curves.amp
13. open the image you want to curve
14. apply my_merged_curves.amp et voila!
15. if you want a regular .acv file click on the curve button next to the pencil button in the curves palette and save the curve again.

Many thanks to Martin Evening and Russel Williams for pointing this out.

It looks somewhat complicated but with an allready saved raw file it is a snap.

Enjoy,

Kees

 

[donbga]

Hi all,

After long search I found a method in this article by Martin Evening:

Thanks Kees! Nice find.

Don Bryant

 

[donbga]

Hi all,

After long search I found a method in this article by Martin Evening:

http://photoshopnews.com/2007/09/05/how-to-express-blend-modes-as-curves/

Hi Kees,

I just tried this and it works beautifully.

Thanks once again,

Don

 

[Katharine Thayer]

The best practice IMO is to learn what is necessary so that you can always get repeatable results with the process, then you would have only one curve that is pretty much perfect. This means temperature and humidity control in the workroom, and consistent procedures of coating and drying.

Sandy, I don't think it's about inconsistency of practice, but simply about fine-tuning the curve. Once the curve is perfected it's the curve, but it can take a couple of iterations to get the curve perfect, and the question is, how to merge the first curve and the second iteration that fine-tunes the curve.

 

[sanking]

Sandy, I don't think it's about inconsistency of practice, but simply about fine-tuning the curve. Once the curve is perfected it's the curve, but it can take a couple of iterations to get the curve perfect, and the question is, how to merge the first curve and the second iteration that fine-tunes the curve.

Katharine,

If you read the thread you will see that the old message from Michael, to which I responded, clearly states that his interest in curve merging is because of process inconsistencies.

There may be other valid reasons for curve merging, about which others may want to comment. However, it is my opinion that merging curves because of process inconsistencies is not good practice.

Sandy King

 

[mkochsch]

Hi all,

After long search I found a method in this article by Martin Evening:

http://photoshopnews.com/2007/09/05/how-to-express-blend-modes-as-curves/

Basic idea is a tip from Russell Williams at Adobe about saving curve data in a photoshop raw image file. A straight curve is represented by a 256x1 pixel regular gradient. Apply your curves on this file and resave it as a new curve. Stunning idea, and it works!

...

It looks somewhat complicated but with an allready saved raw file it is a snap.

Enjoy,

Kees

Sweet. Good get.
~m

 

[Katharine Thayer]

Katharine,

If you read the thread you will see that the old message from Michael, to which I responded, clearly states that his interest in curve merging is because of process inconsistencies.

I was going by the original question:

What's the easiest way to add two ACV files, or curve files, together? What I'm currently doing is making a second curve to "fine tune" the first curve and then reapplying the second curve after I've applied the first curve. What I would like to do is merge the two curves together as a new curve file.
~m

There may be other valid reasons for curve merging, about which others may want to comment. However, it is my opinion that merging curves because of process inconsistencies is not good practice.
[/QUOTE]

I agree.
kt