Pages: 1 ... 7 8 [9] 10 11 ... 14   Go Down

Author Topic: New profile making software: Lumariver Profile Designer  (Read 120148 times)

hk1020

  • Newbie
  • *
  • Offline Offline
  • Posts: 19
Re: New profile making software: Lumariver Profile Designer
« Reply #160 on: June 26, 2017, 08:19:52 am »

It's hardly got much publicity yet so you're selling to the CM enthusiasts who've followed your other high end ventures. Sell at a high price and users will demand a lot of support.

CO doesn't have to be expensive. People like myself who bought CO early on can still follow a relatively cheap upgrade path to keep current. It's cost me less than LR overall to stay on the latest version.

I am in a similar situation. I got CO initially for €30 as CO for Sony and payed another €50 in the meantime. My only interest here is to get a replacement for the broken CO profiles. Should I get them I am done and out of here. CO colors are definitely wrong, the OOC jpgs are much closer to reality. Unfortunately, PhaseOne doesn't think so, I tried.

A long time ago I made some experiments with dcamprof even with Anders' help but got mixed results. Colors were great and showed me what could be achieved but the profiles had problems in certain images so eventually I couldn't use them. dcamprof has certainly evolved since then and I might try it again.

So I'd wish I could get a GUI version to make the process easier but twice to thrice the price of the software I want to fix doesn't quite work. And I don't even know yet if it will give me the results I want. Sorry to say that and I understand very well how much work it is to write this software.

Incidentally, can dcamprof produce the same profiles as Lumariver Profile Designer?
Logged

scyth

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 584
Re: New profile making software: Lumariver Profile Designer
« Reply #161 on: June 26, 2017, 09:58:59 am »

Incidentally, can dcamprof produce the same profiles as Lumariver Profile Designer?

author made some new changes that are not ported back to dcamprof code yet... but one might assume that those were not earth shattering ones
Logged

torger

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 3267
Re: New profile making software: Lumariver Profile Designer
« Reply #162 on: June 26, 2017, 10:04:45 am »

I am in a similar situation. I got CO initially for €30 as CO for Sony and payed another €50 in the meantime. My only interest here is to get a replacement for the broken CO profiles. Should I get them I am done and out of here. CO colors are definitely wrong, the OOC jpgs are much closer to reality. Unfortunately, PhaseOne doesn't think so, I tried.

A long time ago I made some experiments with dcamprof even with Anders' help but got mixed results. Colors were great and showed me what could be achieved but the profiles had problems in certain images so eventually I couldn't use them. dcamprof has certainly evolved since then and I might try it again.

So I'd wish I could get a GUI version to make the process easier but twice to thrice the price of the software I want to fix doesn't quite work. And I don't even know yet if it will give me the results I want. Sorry to say that and I understand very well how much work it is to write this software.

Incidentally, can dcamprof produce the same profiles as Lumariver Profile Designer?

If you just need to make one profile and have the time but not the money using DCamProf suits you better. It's free and for the most part it can make the same profiles as LRPD. There are a bunch of new features in LRPD which hasn't been backported to DCamProf (will probably do it eventually, but will be some time) but the features used for the default profile is there. LRPD is speciality software while the big name raw converters are large volume products, pricing isn't really comparable. If everything would be priced the same as large volume products there would be no niche products, or simply free or none. I do understand though that €100 is hefty if you just want to make one profile for a €300 compact camera, and I may look into catering that market later on, but for now I think DCamProf is the way to go for those users.

Regarding problems in certain images I'm always interested in having a look on those. I have a bit of a backlog currently so I'm slow to reply on complex issues though, but I'm always interested in looking at issues that can improve the software. Note that sometimes when it comes down to extreme conditions like lasers or other sorts of narrow band lights one have to make a tradeoff. Bundled profiles generally always choose robustness and let overall accuracy/quality suffer, while LRPD/DCamProf is more geared to best performance in good (=broadband) light. When shooting nightscapes and similar specially light situations when accuracy is unachievable/meaningless anyway it may be worthwhile using a maximally robust profile. You can probably do those with LRPD and DCamProf as well by tuning parameters but I haven't really focused on that so far.
« Last Edit: June 26, 2017, 10:22:27 am by torger »
Logged

hk1020

  • Newbie
  • *
  • Offline Offline
  • Posts: 19
Re: New profile making software: Lumariver Profile Designer
« Reply #163 on: June 26, 2017, 11:14:15 am »

If you just need to make one profile and have the time but not the money using DCamProf suits you better. It's free and for the most part it can make the same profiles as LRPD. There are a bunch of new features in LRPD which hasn't been backported to DCamProf (will probably do it eventually, but will be some time) but the features used for the default profile is there. LRPD is speciality software while the big name raw converters are large volume products, pricing isn't really comparable. If everything would be priced the same as large volume products there would be no niche products, or simply free or none. I do understand though that €100 is hefty if you just want to make one profile for a €300 compact camera, and I may look into catering that market later on, but for now I think DCamProf is the way to go for those users.

If someone has profiles to share for a Sony NEX-6 and a A6500 for CO I wouldn't mind to get a copy... 

And it is not the camera price but the relation of CO vs LRPD. Actually, I find software pricing really difficult. Buy a certain piece of hardware and you get the software essentially for free and if it is the wrong brand the same software costs you an arm and a leg. You just feel ripped off if you payed.

Regarding problems in certain images I'm always interested in having a look on those. I have a bit of a backlog currently so I'm slow to reply on complex issues though, but I'm always interested in looking at issues that can improve the software. Note that sometimes when it comes down to extreme conditions like lasers or other sorts of narrow band lights one have to make a tradeoff. Bundled profiles generally always choose robustness and let overall accuracy/quality suffer, while LRPD/DCamProf is more geared to best performance in good (=broadband) light. When shooting nightscapes and similar specially light situations when accuracy is unachievable/meaningless anyway it may be worthwhile using a maximally robust profile. You can probably do those with LRPD and DCamProf as well by tuning parameters but I haven't really focused on that so far.

Thanks, nothing current. My experiments date back two or three years. I did send you the problematic images and you used them to further improve dcamprof. I want something robust which can be used for every image. I don't want to have to tinker with different profiles for different images. It just has to work  ;) But it has to be better than CO. I can spot this look through rose-tinted glasses immediately. We are not in the analog film era anymore, you can do much better these days.
Logged

scyth

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 584
Re: New profile making software: Lumariver Profile Designer
« Reply #164 on: June 26, 2017, 11:29:15 am »

When shooting nightscapes and similar specially light situations when accuracy is unachievable/meaningless anyway it may be worthwhile using a maximally robust profile. You can probably do those with LRPD and DCamProf as well by tuning parameters but I haven't really focused on that so far.

time for a LrPD NightScape Edition or LrPD BlueLED Edition  ;D
Logged

alain

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 465
Re: New profile making software: Lumariver Profile Designer
« Reply #165 on: June 26, 2017, 06:18:01 pm »

Regarding pricing, nearly all C1 users I've had contact with are pro users or very high level non-professionals, and C1 is not a cheap raw converter so the volumes are lower and pricing higher, and there's also more user support as users this segment more often have an interest in making advanced tuning. Lots of medium format users or very high end 135 cameras. The DCP space is more volume, more casual and less professional, and thus I made a specific product for that. It's the Basic edition that's intended to be the odd one out, not the others. I see it as a big jump down rather than a big jump up :-). Sure I'd love to sell tens of thousands of this and just need to charge €10, but profiling is niche software which quite few use, it's a connoisseur thing, and the relative low volume leads to steeper pricing (LRPD isn't the most expensive profiler out there...). Otherwise you can just as well give it away for free like DCamProf or Adobe DNG Profile Editor.

That said I might still consider a Basic/ICC version later on if the interest is proven to be large but I haven't seen much of that so far.

Ok.

It was not for the price itself that I asked it, just for the price difference.

BTW. I just bought the repro version.  Now finding time to test it ;-)
Logged

torger

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 3267
Re: New profile making software: Lumariver Profile Designer
« Reply #166 on: June 27, 2017, 01:39:38 am »

time for a LrPD NightScape Edition or LrPD BlueLED Edition  ;D

Heh, I don't think there's a need for a separate edition, but there may be a need for some rather different tradeoffs. This area is one of the aspects that I'm interested in doing further R&D. Desaturate and adding green (eg the central bandwidth) to the profile seems to be good ways to increase robustness in these situations, but it of course affects accuracy and look.

The key problem to work with is that you need a really wide transition zone to not get a disconnect between your normal range of colors and the extreme handling, and so far I haven't seen a way that makes it possible to keep a natural and accurate normal range and a super-robust extreme range at the same time, at least for certain cameras, as the extreme handling transition zone must stretch way into normal range colors. I'm speculating a bit here, but I've got the feeling that cameras that try hard with high ISO performance makes their color filters over-sensitive (as filtering away photons means losing signal) which makes it harder to make a sane robust profile, as they're more concerned with showing "a color" than "an accurate color".

I often say that color filters in cameras don't matter as much any longer, and for the normal range colors I think it's true the camera profile is pretty much in almost total control over the look, but when it comes to extreme color handling I think it's a factor. A camera that registers a strong signal in color ranges where the human eye doesn't makes it much harder to work with.
Logged

Ethan Hansen

  • Full Member
  • ***
  • Offline Offline
  • Posts: 148
    • Dry Creek Photo
Re: New profile making software: Lumariver Profile Designer
« Reply #167 on: June 28, 2017, 06:18:17 pm »

With all this discussion about licensing, why does Lumariver not respect the licenses of software it uses? A cursory examination shows the code contains LibTIFF, ExPat, and LittleCMS libraries yet none of the required license information is present. Depending on what other codebases are linked in, there are also signs that GPL code could be present as well.

markanini

  • Jr. Member
  • **
  • Offline Offline
  • Posts: 58
Re: New profile making software: Lumariver Profile Designer
« Reply #168 on: June 28, 2017, 07:03:21 pm »

Can I import a colorimetric DCP and apply tone curves and looks on top of it?
Logged

torger

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 3267
Re: New profile making software: Lumariver Profile Designer
« Reply #169 on: June 29, 2017, 02:59:27 am »

Can I import a colorimetric DCP and apply tone curves and looks on top of it?

There's no support for importing a raw profile at this time. You can import and inspect in text format, but not use it to add stuff on top. The software is designed to make profiles, not modify existing. It wouldn't be an impossible feature to add, but a quite big one so there must be a strong interest in it...
« Last Edit: June 29, 2017, 04:00:37 am by torger »
Logged

torger

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 3267
Re: New profile making software: Lumariver Profile Designer
« Reply #170 on: June 29, 2017, 03:07:44 am »

With all this discussion about licensing, why does Lumariver not respect the licenses of software it uses? A cursory examination shows the code contains LibTIFF, ExPat, and LittleCMS libraries yet none of the required license information is present. Depending on what other codebases are linked in, there are also signs that GPL code could be present as well.

Actually we do, or "I" in this case as this has been a one-man project (except for my colleague that helped me getting started with QT).

I've worked with many open and closed source licenses for more than 15 years, both commercially and non-commercially, both using and making. I've shared a lot of my work for free, been involved for a long time and is still involved in the open-source community. This is both my passion and profession. I respect software developers and I'm not in any case trying to misuse third-party libs. If there's some mistake I'll correct it, but I go through the licenses of the third party components to make sure I can use them commercially. Here's that license section: http://www.lumariver.com/lrpd-manual/#acknowledgements maybe I should cite licenses of the standard libs like ExPat tiff lcms but it's often not required. Not all libs require that license information is repeated in the docs, but I shall check it again. Some don't want to be cited in the docs other want it, it varies even with similar licenses so it's quite easy to make a mistake. The only GPL (a copyleft license that cannot be used in closed source software) code in there is from DCamProf which I am the copyright owner and have licensed to Xarepo which owns the Lumariver code.

Citing licenses in the docs should of course be done if required, but what's the really important thing is to make sure commercial use is ok at all, and that is what I check first before even using a library, while I may have been a bit sloppy in the citing roundup -- I did go through them but "legalese" (license language) can be a bit tricky to decode (and therefore there are many misconceptions on what's required and not). I hope I haven't missed anything, and I'm better at reading licenses than most software developers and laymen are, but now when you complain about it I became a bit uncertain...

Anyway taking LittleCMS as you mention as an example, it uses the MIT license https://opensource.org/licenses/mit-license.php,

"Permission is hereby granted, free of charge, to any person obtaining a copy of this software and associated documentation files (the "Software"), to deal in the Software without restriction, including without limitation the rights to use, copy, modify, merge, publish, distribute, sublicense, and/or sell copies of the Software, and to permit persons to whom the Software is furnished to do so, subject to the following conditions: The above copyright notice and this permission notice shall be included in all copies or substantial portions of the Software."

So what does it mean that the copyright notice should be included in the Software? It means that if you integrate the code into your software (rather than use a pre-packaged lib like I do) the source code files must include that statement, so it doesn't suddenly change copyright or license so the company using the code suddenly think they wrote it themselves. It's for the software development. It doesn't mean that the license needs to be cited in the user docs or the lib must be mentioned. LittleCMS is used by pretty much every cross-platform software that has active color management by the way.

Funny you would mention expat (an XML parsing lib, also uses MIT license like lcms), and the MacOS version of LRPD actually uses the lib distributed with MacOS itself so it's not in the application package. The lib is there because Adobe's DNG SDK links to it, LRPD itself doesn't actually use it directly. Again it doesn't require citing in the docs. If it would tons and tons of software would need to mention it as this lib is very commonly used.

Looking at the libtiff BSD license it uses an unclear wording regarding citing (most BSD licenses use a clearer wording), so there maybe I should actually cite the license in the user documentation, most seem not to though. But you know what -- now when the software is finished and distributed I shall go through all the libs once again and mention them in that section and if uncertain like in the libtiff case I'll provide the license, it's better to mention one too many times than one too few, as there's always someone that assume I don't care about which libs I use and how I use them, but I can assure you I do care.

That said I do know it's quite common that some developers that don't really know much about open-source and licensing just use it as they please without giving it any thought on what's allowed, so I understand your suspicion especially if you don't know who I am and my history in software development. As an author and contributor to many open-source projects I'm very much aware of that situation.

Note that cross-platform commercial software like LRPD often use lots of open-source libs, as the cross-platform stuff usually is from the unix world which has a long tradition of permissive open-source licenses. Being license advisor to my former employee I know this scares business people, and there's been quite some FUD thrown around making the case that it's not safe to make commercial software other than using say Microsoft .NET or MacOS XCode as the alternatives would be riddled with dangerous licenses that force your code to be open-source (GPL is the big bad monster to those...). However this is not the case, you do need to check the licenses, and indeed GPL does put that type of limits, but nearly all common libraries has a permissive license and nearly always you just need to use a pre-packaged lib and don't need to modify the lib itself which makes it even less complicated. And the citing requirement thing is rarer than most seem to believe.

If you have more questions about licensing and libs we use in LRPD you can send me a private message. I'm myself quite interested in the subject, but I guess most readers are more interested in what LRPD actually does ;D
« Last Edit: June 29, 2017, 10:03:50 am by torger »
Logged

torger

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 3267
Re: New profile making software: Lumariver Profile Designer
« Reply #171 on: June 29, 2017, 12:11:57 pm »

To avoid future misconceptions and really be in the clear I've now beefed up the acknowledgements section including license citing from libtiff, and libarchive that I actually had missed has the type of BSD that requires citing (as it uses a bunch of old code that has that type of license, rarely used these days): http://www.lumariver.com/lrpd-manual/#acknowledgements
Logged

Ethan Hansen

  • Full Member
  • ***
  • Offline Offline
  • Posts: 148
    • Dry Creek Photo
Re: New profile making software: Lumariver Profile Designer
« Reply #172 on: June 29, 2017, 01:21:04 pm »

To avoid future misconceptions and really be in the clear I've now beefed up the acknowledgements section including license citing from libtiff, and libarchive that I actually had missed has the type of BSD that requires citing (as it uses a bunch of old code that has that type of license, rarely used these days): http://www.lumariver.com/lrpd-manual/#acknowledgements

Thanks for clearing matters up! I'm glad to see you are definitely aware of licensing issues and went through the effort to ensure commercial usage was allowed.

I now need to make time to put LRPD through its paces. We still use an ancient product that requires running on a Windows XP virtual machine with Photoshop 7 (circa 2003). It remains the best product for creating camera profiles for C1. It is useless for Lightroom or other DNG applications and I certainly would like to get away from our own internal, hacked together approach.

scyth

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 584
Re: New profile making software: Lumariver Profile Designer
« Reply #173 on: June 29, 2017, 02:57:10 pm »

We still use an ancient product that requires running on a Windows XP virtual machine with Photoshop 7 (circa 2003). It remains the best product for creating camera profiles for C1.

what is it ? the name !!!
Logged

Ethan Hansen

  • Full Member
  • ***
  • Offline Offline
  • Posts: 148
    • Dry Creek Photo
Re: New profile making software: Lumariver Profile Designer
« Reply #174 on: June 29, 2017, 03:11:57 pm »

what is it ? the name !!!

ColorEyes 20/20. Off the market for over a decade.

jrp

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 322
Re: New profile making software: Lumariver Profile Designer
« Reply #175 on: July 02, 2017, 08:34:36 am »

Many thanks for this tool, which is exactly what was needed.

I'll buy this later in the summer when I really have time to play.

A couple of cosmetic suggestions:

* buttons / settings not available in the edition that you are using should be hidden, to reduce clutter
* it would be helpful to include tool tips or other means of self-documentation to obviate the need to jump backwards and forwards between the program and your (excellent) documentation
* the tool could give some overall indication of the quality of the target and the quality of the profile.  (Perhaps that is only available in the more advanced editions than the basic version I tried.)  Quite a few of tweaks that I tried made little difference to the preview (when rendered).

Thanks again.
Logged

torger

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 3267
Re: New profile making software: Lumariver Profile Designer
« Reply #176 on: July 03, 2017, 01:49:09 am »

Many thanks for this tool, which is exactly what was needed.

I'll buy this later in the summer when I really have time to play.

A couple of cosmetic suggestions:

* buttons / settings not available in the edition that you are using should be hidden, to reduce clutter
* it would be helpful to include tool tips or other means of self-documentation to obviate the need to jump backwards and forwards between the program and your (excellent) documentation
* the tool could give some overall indication of the quality of the target and the quality of the profile.  (Perhaps that is only available in the more advanced editions than the basic version I tried.)  Quite a few of tweaks that I tried made little difference to the preview (when rendered).

Thanks again.

Thanks for the suggestions, they're all good. It comes down to time required to implement them and other priorities. I keep a list though, adding these on there!

Many tweaks does indeed have a small impact, especially if you only view the target image. It's almost impossible to view changes in skintone on a flat skintone patch on a target for example, what you need is real pictures and do A/B swapping between new and old (profile comparison tab), then changing tone operator (for example) becomes much more visible. Changing gamut compression is invisible unless you have an image with high saturation colors etc.

Regarding quality of the target there's no artificial intelligence there to identify if it's good or bad, but you can with some experience see it. Poor lightness matching of the grayscales usually mean uneven light (open up the tune matrix dialog to see matching), light dark patches means glare, oversaturated profile means glare, the glare number beside the (experimental) glare compensation checkbox means either uneven light or glare. However, due to that matrix matching is really robust and LUT correction is per default conservative a bad target shot doesn't necessarily mean a bad general-purpose profile, there's many factors involved, which does mean that such an artificial intelligence to make a good evaluation of the quality is quite tricky to implement.

However, just automatically identifying if the lighting is uneven or there's much glare would be quite easy to implement. That would be triggered by 90% of the casual CC24 shots though and since those shots generally still make good general-purpose profiles I'm not sure how valuable such a feature would be.
« Last Edit: July 03, 2017, 02:01:07 am by torger »
Logged

scyth

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 584
Re: New profile making software: Lumariver Profile Designer
« Reply #177 on: July 03, 2017, 08:47:11 am »

However, just automatically identifying if the lighting is uneven
typically (one might hope that readers always do this actually ) if you have a decently flat-fielding surface (I use Novoflex XL grey card) behind the target then FRV (fastrawviewer) will nicely illustrate visually how uneven is the light using its simulation of overexposure option ... I have a steel sheet painted with matte black paint on which with rare earth (painted black) magnets I mount that grey card and on top of it - a target... looking @ form of the shadows from those tall & thin cylindrical-shaped rare earth magnets help to make the light (I use one source) ~orthogonal to the target surface (I shoot with lens at angle and with light orthogonal - I rather have as even illumination  as possible with the light source I use - Solux halogen bulb)
Logged

kwalsh

  • Full Member
  • ***
  • Offline Offline
  • Posts: 101
Re: New profile making software: Lumariver Profile Designer
« Reply #178 on: July 05, 2017, 03:23:55 pm »

Dcamprof saved my tail last year with LR/ACRs totally effed up profiles for the E-M5II.  Just noticed this new GUI based tool.  Looks amazing and awesome.  Will have to purchase and start using later in the summer when I actually have some time again.  Honestly the new manual alone could be sold for the asking price - more useful information on building generic profiles than I've seen anywhere.  Amazing work.  Glad I'll have an excuse to contribute some money to the effort.

One question before I forget...

What do you now think about dual illuminant profiles?  It seems StdA is drastically different from D50 in many ways and perhaps the interpolation is actually undesirable unless you really are right in the middle and can't/don't do a specific profile shot?

I guess my question is that if I try to creatively "warm" a D50 scene using LR/ACR WB sliders would I be better using a single illuminant D50/D65 or similar profile when making that WB adjustment rather than a dual illuminant profile that will start interpolating in a simplistic and perhaps "wrong" way towards the vastly different StdA profile?  Or is that interpolation exactly what I want in that kind of situation?

I understand with a LookTable you almost certainly want to use single illuminant profiles since there is only one LookTable and it is unlikely you'll use the same table for StdA and D50.  But I'm more wondering about the rest of the profile under "small" changes in WB - better to stick with one set of tables for the closest illuminant and leave interpolation to "large" changes.

Related question - are the "base look" options and tone operators tuned differently for StdA and D50?  Meaning if I just leave all "defaults" to create a D50 and StdA profile will those profiles in fact likely have different LookTables?  If so how does the software decide which to use when doing a dual illuminant profile?

Hope my questions are at least understandable if not downright stupid...
Logged

torger

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 3267
Re: New profile making software: Lumariver Profile Designer
« Reply #179 on: July 06, 2017, 01:56:01 am »

Many/most cameras really struggle when you get down to StdA (2700K), they're not really optimized for that, and you often gain from making manual tradeoffs in the optimization because the matching is so bad that the automatic optimization may not match with the tradeoffs you would prefer. The advanced tutorial has an example: http://www.lumariver.com/lrpd-manual/#tut_step4

That was not what you asked about though, but I just wanted that out of the way :). In any case a 2700K profile is generally much more of a compromise than 5000K, and on top of that tungsten profiles is often used in sub-optimal mixed light, as if you want good colors you'd strive for good light, which would be higher temperature. I haven't really investigated where the "breakpoint" is, probably different between cameras, but I assume that already at 3500K it's much better, and above 4000K I think the matching would be as good as at 5000K.

Dual-illuminant is about convenience, not about precision. You don't need to manually swap between different profiles. The interpolation as defined by the DNG spec is indeed simplistic, but I think it's okay regarding the often suboptimal situations a StdA profile is used in anyway, and good light is almost always 5000K and above when the high temp profile will take over. I have only briefly made some experiments with interpolation vs real profile for targeted temperature and from what I remember my conclusion was that it's okay, but a real profile is of course always better. When creeping down towards StdA it's certainly better to use a dual-illuminant profile with StdA than just a D50 profile with WB adjustment. Maybe at 4000K the interpolation do more harm than good, but the difference is not large.

I use dual-illuminant myself, for convenience. But then my shooting style is such that for all shots that matter the light is ~D50 or above (meaning that the daylight profile is then used alone), and the StdA is for my needs just a fallback for say documentary style indoors usually in bad light so I'm simply not as picky there. I rarely come across "real" StdA these days, it's generally a mix of artifical lights, often with bad CRI. If one is picky, shooting a color checker at the scene and make a custom profile for the venue, often with manual tuning, would be the way to go for that tricky lighting.

Regarding if it's better to stick with fixed profile than interpolation for small changes, I don't think it matters that much as the interpolation will have very weak effect when close to a main illuminant.

It is an interesting subject, would a 3000K-5000K be better than 2700K-6500K profile for example? Maybe, I don't know as I haven't done enough experiments. To test the simulation capabilites of DCamProf using SSF would be a place to start. I personally use the combo StdA-D50 or D55, I think D65 is unnecessarily high. It happens quite a lot from 3000K down to 2700K (the temp scale is not linear when it comes to color) so I'm thinking maybe one should step up a bit on the low, but I haven't made any serious testing there. It would be interesting to hear if someone else has some experience or ideas to share.

The "base look" is tuned for daylight. As it's subtle it won't cause any carnage at StdA, and maybe it even works good there, I just haven't tuned it for that specific light.
« Last Edit: July 06, 2017, 02:06:46 am by torger »
Logged
Pages: 1 ... 7 8 [9] 10 11 ... 14   Go Up