fredmiranda.com Login
This topic is in the 2011 archive. You can read it and like its posts here. It is read only.

Archive 2011Raw Photo Processor

  
 
philip_pj
Offline
Details
p.3 #1Raw Photo Processor
What's clear is that ACR6/LR3 is certainly not outclassed if at all, which is very important for minority camera users, as Adobe must take care of Canon primarily. And it has a very good toolset

denoir, it is a balancing act, indeed - how good can one get the image in a short timeframe, colour, detail, contrast, noise handling, tonal gradation. ACR is definitely good enough for me because its toolset is very good. It may not be enough for users looking for the last 1-2% of detail who are prepared to work with more apps in workflow.

Bif, strong opinions are fine, we are adults and want to get the knowledge base advanced! Cough up your RC favourites, man..

Jae, the camera makers and RC devs give themselves a leg up with 'behind the scenes' algos - here is a short quote from my favour paper source (c't digital photography):

"It is certainly possible that camera makers use deconvolution techniques in their RAW-compatible models, but it is a good as impossible to get any information on the subject. The same is true of the internal sharpening algorithms that manufacturers often build into their equipment..."

So it starts with the camera and goes on from there down the workflow. It seems it is as much 'their' RAW image as ours...you can see their quandary, they need to produce a 'very good' starting image if at all possible, that makes it easier for new and advanced users alike to get where they want to go. So purity is a lower priority for them.
Nov 04, 2011 at 08:18 PM
philip_pj
Offline
Details
p.3 #2Raw Photo Processor
"..like all demosaicing algorithms includes a sharpening process."

Ken, not to question the veracity of this statement, but I saw this statement recently, again in c't:

"Sharpening is most often performed after demosaicing...The Raw Therapee documentation confirms this assumption and Sony confirmed it when asked. However, Canon, Nikon and Adobe refused to give us any information on the subject."

I had been interested in a really good RC based on the (seemingly fallacious) assumption that edits done pre-demosaicing might enhance image quality. It seems to me that the only advantage is that you may gain something from work done on the 'instruction' file in the RC before it becomes an output file as a jpeg, tiff etc. Sometimes, the more you know, the less you know...
Nov 04, 2011 at 08:52 PM
Bifurcator
Offline
Details
p.3 #3Raw Photo Processor
FlyPenFly wrote:
On my monitor, from your samples denior, the ACR conversion seems to discard shadow detail... That could be just the higher default contrast profile.


If you like that look simply save your ACR presets with the shadow slider increased. It's better than having to contend with RPP's reduced detail (global and micro contrast) across the rest of the image. I mean if you already have ACR. Obviously if a person is broke and needs a free utility then RPP can suffice.




RustyBug wrote:
Bif ... you mentioned ACR in top 3 ... what do you consider the other 2, maybe even round out your TOP 5.

I also have a preset in ACR that sets everything to zero, so I can go straight to PS with the RAW mostly in tact. Otherwise, the OEM DPP gets used at times as well.

BTW ... I dissent at you being the "Nay-Sayer".
You turn into the "Yeah-Sayer" whenever I'm being a "Nay-Sayer" ... which in turn sometimes makes you the "Nay-Nay-Sayer".


Yeah, that's true, I always want to consider the other side of things. It's not an attempt to be arbitrarily contrary tho. When someone says left I always wanna examine what's to the right and consider up, down, forward, and backward. I mean that other guy already went left so he can tell us what's there.





philip_pj wrote:
Bif, strong opinions are fine, we are adults and want to get the knowledge base advanced! Cough up your RC favourites, man..


We are adults? Hey, speak for yourself man! Hehehehe... But I'll post some stuff after the image display here gets sorted out and I have some time to go over my reasoning, etc.
Nov 05, 2011 at 12:49 AM
douglasf13
Offline
Details
p.3 #4Raw Photo Processor
Where does it state that RPP uses deconvolution sharpening during demoisacing??
Nov 05, 2011 at 03:47 AM
kosmoskatten
Offline
Details
p.3 #5Raw Photo Processor
I am not sure though, my RPP files still look way better than ACR, though I have been cheapskating and I haven't the latest ACR. The RPP user interface is a drab though, I'd be happy with something that had better flow to it.

Considering the ubiquitously myriadic plethora of RAW converters I haven't had the time, nor interest in buying and trying. Bif/Denoir: if you can provide some more thoughts on the matter it would surely help me out as I have plenty of images I need to work on throughout the winter and if there is ONE converter that stands out from the crowd I am willing to invest. I haven't the patience to dabble with several converters. Oh, I work on a Mac.

/ Henrik
Nov 05, 2011 at 04:00 AM
Mitch Alland
Offline
Details
p.3 #6Raw Photo Processor
The tests above by denoir and Bifurcator don't match at all what I find with the GXR M-Module (a camera without an AA filter), which is that the resolution of RPP is much better than that of Aperture and enough better than that of ACR to be significant. Actually, I have to place Bifurcator's "tests" in quotation marks because they seem to be by someone who hasn't bothered to make an effort to figure out how to use RPP, which makes them worthless. The other thing is that people seem to be dealing here only with resolution, but as important is the superior performance of RPP with regard to color rendition, which has a clarity and transparency that is better than that of either Aperture and Lightroom. One example of that is the first picture in my post above (p.1 post 19), which has some small paintings lying on a table: some of the pairings have a red head on the background if a violet field — the RPP rendering of the relations of the hues, intensities and saturations of these two colors is much more lifelike and accurate than that of Aperture. In general, RPP renders reds very well, which I can't say for the other raw developers that I've tried.

The other thing that I like about RPP are the film presets, particularly K64. It's not that I set out to simulate film when I process digital files, it's just that when i achieve a look a like it is usually a ilm look.

—Mitch/Chiang Mai
Days and Nights in the Forest
Nov 05, 2011 at 04:38 AM
Tariq Gibran
Offline
Details
p.3 #7Raw Photo Processor
As I have mentioned before, my experience with many different raw converter/ camera combos is that some combinations work better than others. That's the case with Fuji X100 files plus RPP vs ACR and may also be the case with the Ricoh GXR-M module as well. ie. Testing other camera raw files from different cameras - Panasonic, Leica, etc. - with RPP, for instance, is not going to offer accurate info on how RPP is going to work with the x100/ Ricoh GXR-M files.
Nov 05, 2011 at 07:07 AM
kwalsh
Offline
Details
p.3 #8Raw Photo Processor
philip_pj wrote:
"..like all demosaicing algorithms includes a sharpening process."

Ken, not to question the veracity of this statement, but I saw this statement recently, again in c't:

"Sharpening is most often performed after demosaicing...The Raw Therapee documentation confirms this assumption and Sony confirmed it when asked. However, Canon, Nikon and Adobe refused to give us any information on the subject."


Hi Philip,

All I meant, but didn't say specifically, is that any demosaicing algorithm includes something akin to sharpening in its basic demosaicing algorithm. Some software will do a separate step of what we'd traditionally consider "sharpening" as well that you may or may not be able to turn off or adjust. Even without this additional step the various demosaicing algorithms affect high spatial frequencies differently, they have a "native" amount of sharpening within the algorithm's kernels even if it isn't immediately obvious as a separate sharpening step.

Demosaicing is at its heart a form of interpolation, all interpolation involves filtering kernels and sharpening is just a high pass filtering kernel. If different demosaicing kernels have different high pass characteristics this is very much like a difference in sharpening between the two kernels. Convolve a sharpening or bluring kernel with your "native" demosaicing kernel from one algorithm and it will look a lot like the kernel from a different demosaicing algorithm.

Many of the algorithms with weaker "native" sharpening look suprisingly comparable to algorithms with higher "native" sharpening once sharpened to match. This is not to say all algorithms are the same by any stretch, rather that sometimes the differences are masked by simple sharpening effects that are easily equalized across different demosaicing algorithms. This is clear if you do a lot of comparisons between ACR/LR and other converters. You can achieve a wide range of sharpening effects in ACR and even vary the sharpening algorithm with the detail slider (100% is pure deconvolution, 0% is almost pure USM). Once you've tweaked ACR sharpening much of the detail differences with other RCs disappear and you are left with more interesting residual differences that are a bit more subtle.

As others pointed out detail and sharpness is only one part of what makes a RC "good" but sharpening has a very strong affect on perceived detail. Get that out of the way and you can focus on color processing, noise character, highlight and shadow recovery, etc. etc. etc.

You are definitely right about "the more you know the less you know", I feel that way the more I look at and try to understand practical image processing! In general every simple assumption I try to make eventually proves to be insufficient.

Ken
Nov 05, 2011 at 09:00 AM
RustyBug
Offline
Details
p.3 #9Raw Photo Processor
Hmmm ... let's see. Every OEM has their own magic/secret/proprietary version of how a RAW file is generated. NOBODY but them knows what the "magic formula" is in their "secret sauce". In that regard, I think Tariq's point @ combinations has merit.

I know that while I do use PS from ACR set at zero as my usual workflow, I am usually well served when I break out Canon's or Kodak's OEM software first ... extra hassle I avoid a lot of the times, but seeing the difference it can make suggests two things to me ...

1. OEM "secret sauce" encoding is well mated with OEM "secret sauce" decoding
2. No one single RAW converter will be best for all.

I think (in my limited, non-technical perspectives) that ACR may have the broadest acceptable scope as a third party RAW converter, but that broad scope may not necessarily make it the best for all, especially those with their own ideas of what constitutes discriminating style, taste and technical excellence with minimal penalties en route.

Somewhere on the spectrum is good enough, good, very good, great, excellent, awesome and perfect (etc.). None will ever be perfect, and somewhat like optical design, there will always be trade-offs inherent to its design & function.

I'm looking forward to hearing Bif's Top 5 (or Top 3) and rationale. Not that I'll go "Oooh-Ahhh" that's the best one ... but it'll give me some perspective on how their characteristics "fit" into my approach & philosophies. I'm hoping to find one I like better than ACR enough to invest in it.

Even though I demo'd several more than a year ago and didn't find anything I really liked ... I've probably changed a bit since then. Will be interesting to see what I think this year.

Edited on Nov 05, 2011 at 09:12 AM · View previous versions
Nov 05, 2011 at 09:10 AM
AhamB
Offline
Details
p.3 #10Raw Photo Processor
Mitch Alland wrote:
In general, RPP renders reds very well, which I can't say for the other raw developers that I've tried.


That's interesting because the reds look deficient in denoir's, FlyPenFly's and bifurcator's test samples. What is it you need to do to get decent colors with RPP? I presume the film presets change the colors quite a lot.
Nov 05, 2011 at 09:11 AM
kwalsh
Offline
Details
p.3 #11Raw Photo Processor
douglasf13 wrote:
Where does it state that RPP uses deconvolution sharpening during demoisacing??


I don't think it specifically states that. That said, RPP uses an algorithm like this one:

http://www.eie.polyu.edu.hk/~enyhchan/J-TIP-CDemosaicking_using_VarCD.pdf

Long into that paper you can see that they, and most others, pair the basic demosacing algorithm with a "refinement scheme" that is typically separate algorithm (reference 14 in that paper). For the above method it is and adaptive filtering method that is very much like deconvolution sharpening (there are in fact both blind and adaptive deconvolution sharpening methods, this one appears to be adaptive). I don't know exactly what RPP is doing since they don't say but I'd assume it is what most of the practitioners in academia are doing and that appears to be using adaptive spatial filters post demosaic.

Ken
Nov 05, 2011 at 09:26 AM
Mitch Alland
Offline
Details
p.3 #12Raw Photo Processor
AhamB wrote:
Mitch Alland wroteIn general, RPP renders reds very well, which I can't say for the other raw developers that I've tried.
That's interesting because the reds look deficient in denoir's, FlyPenFly's and bifurcator's test samples. What is it you need to do to get decent colors with RPP? I presume the film presets change the colors quite a lot.


While my experience with RPP is still limited, I haven't had any problem getting good colors. If there is a camera profile for your camera, the colors will be accurate — at least they are for the GXR-M — and will depend on your settings for (compressed) Exposure, Black Point and Saturation and Cold/Warm (which I haven't used). It seems that if there is no camera profile for your camera, the Saturation will have to be set quite high (30-50), while if there is, fairly low settings or zero are best.The film presets do affect the colors, although KR64 is fairly "safe". Actually, once you get the hang of it, RPP is easy to use.

One thing that I've learned is that it's usually best to set Local Contrast in the 10-30 range. I've been using 0-5 — and that's not enough. But I'd like to know what other RPP users think about this setting.

—Mitch/Chiang Mai
Ricoh GXR M-Module
Nov 05, 2011 at 09:28 AM
Tariq Gibran
Offline
Details
p.3 #13Raw Photo Processor
Mitch Alland wrote:
AhamB wrote
Mitch Alland wroteIn general, RPP renders reds very well, which I can't say for the other raw developers that I've tried.
That's interesting because the reds look deficient in denoir's, FlyPenFly's and bifurcator's test samples. What is it you need to do to get decent colors with RPP? I presume the film presets change the colors quite a lot.
While my experience with RPP is still limited, I haven't had any problem getting good colors. If there is a camera profile for your camera, the colors will be accurate — at least they are for the GXR-M — and will depend on your settings for (compressed) Exposure, Black Point and Saturation and Cold/Warm (which I haven't used). It seems that if there is no camera profile for your camera, the Saturation will have to be set quite high (30-50), while if there is, fairly low settings or zero are best.The film presets do affect the colors, although KR64
...Show more →

I also use the 10-30 range for local contrast in RPP. One thing about the film presets is that they are not neutral in color rendering and that, after all, is why they look more film like. You will see color casts in the shadows for instance with k64. Reds, greens and blues will take on a different hue depending on the film emulation type. What I find nice about RPP is exactly this aspect. ACR can be made to mimic it with a lot of tweaking but this tweaking may look different with different scenes/ subjects. RPP seems to do a great job of this with different scenes if you prefer that look. Thus, what I can do in 2 minutes with RPP might take 10-15 minutes of work to give a similar look.
Nov 05, 2011 at 09:51 AM
Mitch Alland
Offline
Details
p.3 #14Raw Photo Processor
Tariq Gibran wrote:
...One thing about the film presets is that they are not neutral in color rendering and that, after all, is why they look more film like. You will see color casts in the shadows for instance with k64. Reds, greens and blues will take on a different hue depending on the film emulation type. What I find nice about RPP is exactly this aspect. ACR can be made to mimic it with a lot of tweaking but this tweaking may look different with different scenes/ subjects. RPP seems to do a great job of this with different scenes if
...Show more →

Tariq, you've described very well the the important thing about the film presets in RPP: how good they are and how much more difficult it is to achieve such a film look with other raw developers. I've been using mainly the K64 one, and it does remind me of Kodachrome, which is the reason that I named a thread on RPP that I starred in the Leica Users Forum "You Can Stop Missing Kodachrome Now".

—Mitch/Chiang Mai
Paris au rythme de Basquiat
Nov 05, 2011 at 10:03 AM
Tariq Gibran
Offline
Details
p.3 #15Raw Photo Processor
Mitch Alland wrote:
Tariq Gibran wrote...One thing about the film presets is that they are not neutral in color rendering and that, after all, is why they look more film like. You will see color casts in the shadows for instance with k64. Reds, greens and blues will take on a different hue depending on the film emulation type. What I find nice about RPP is exactly this aspect. ACR can be made to mimic it with a lot of tweaking but this tweaking may look different with different scenes/ subjects. RPP seems to do a great job of this with different scenes if you prefer that look. Thus, what I can do in 2 minutes with RPP might take 10-15 minutes of work to give a similar look.
Tariq, you've described very well the the important thing about the film presets in RPP: how good they are and how much more difficult it is to achieve such a film look with other raw developers. I've been using mainly the K64 one, and it does remind me of Kodachrome, which is the reason that I named a thread on RPP that I starred in the Leica Users Forum "You Can Stop Missing Kodachrome Now".

—Mitch/Chiang Mai
Paris au rythme de Basquiat


Thanks. Like both you and now Henrik, I come from a strong film background, including processing, scanning, printing, etc. RPP appears to be pretty convincing in its emulation - I think even better than some of the dedicated photoshop plug-ins.
Nov 05, 2011 at 10:14 AM
Mitch Alland
Offline
Details
p.3 #16Raw Photo Processor
Tariq Gibran wrote:
Thanks. Like both you and now Henrik, I come from a strong film background, including processing, scanning, printing, etc. RPP appears to be pretty convincing in its emulation - I think even better than some of the dedicated photoshop plug-ins.


For color photography RPP certainly produces a more film-like look than any of the Color Efex modules. However, I do a lot more B&W, for which, so far, I've been using the RPP color TIFF files for conversion to B&W with Silver Efex running under Aperture. What I find is that I need fewer adjustments in Silver Efex this way, except for any burning and dodging that I may want. Is anyone here doing the B&W conversion in RPP? If so, how do you find the results and do you have any workflow suggestions for this?

—Mitch/Chiang Mai
Days and Nights in the Forest
Nov 05, 2011 at 07:51 PM
RustyBug
Offline
Details
p.3 #17Raw Photo Processor
Bump for Bif's picks
Dec 02, 2011 at 02:35 AM
FM ForumsLeica & Alternative GearJoin Upload & Sell
  
 

You are not logged inLog inRegister