Modern digital photography is very much associated with image processing programs. Sometimes, just the photo editor creates exactly that photo, which we ultimately consider the work of the photographer.
RAW converter
Modern digital photography is insidious, sometimes it turns into ordinary mathematical tricks, which I already wrote about in articles'Subpixels', 'Gigapixels', 'JPEG", "Tricks with RAW'and'RAW setup'. Any photograph taken in the so-called raw format - RAW, for printing or viewing on a computer must be converted (converted) into a simpler format with an unambiguous representation of the data. This program is usually called a RAW converter ('rav-converter'). There are just a lot of such converters, but due to the specifics of the raw files of each individual camera, the converters sometimes cannot recreate the “correct” photo from this raw data. It is often said that for the best results you need to use only the “native software” that sometimes comes with the camera.
Below is my experiment that everyone can repeat. The essence of the experiment is very simple - you need to convert the same raw file with different converters and see what happens :) The photo from the experiment was taken on Nikon Nikkor-SC Auto 1: 1.4 f = 50mm, I specifically chose a lens without microprocessor contacts to exclude any specific image enhancement by the converter by correcting lens imperfections from the database. The picture was taken in difficult lighting conditions: very cloudy weather, two flashes.
Link to the archive with photos. (The archive takes 96.9 MB to view the packed files, the archive should be unpacked). Attention: the photos are not for the faint of heart, they are especially dangerous for scammers and pixel worms :)
The archive contains the following files:
- DSC_5549.NEF - original RAW file obtained during shooting. Nikon D700, 14bit, lossless compression, control vignetting, ADL, noise reduction at slow shutter speeds, noise reduction at high ISO - were off. All shot in Picture Control mode SD (standard mode): contrast correction - 0, brightness - 0, saturation - 0, hue - 0, sharpness - position 5 out of 10. White balance - "flash". If you turn on all the native "improvements", then third-party software copes with displaying data even worse. EXIF You can look at any available program that you are used to. If there is no such program, then it is easy to do using the service on http://regex.info/exif.cgi.
- DSC_5549.JPG - original JPEG file obtained during RAW + JPEG shooting, settings are described above. JPEG L with quality priority.
- DSC_5549 (acr-original) .jpg - file converted with Adobe Camera Raw 6.0 (ACR 6.0) in default mode.
- DSC_5549 (acr-auto) .jpg - file converted with Adobe Camera Raw 6.0 in auto mode.
- DSC_5549 (capturexn2-original) .jpg - file converted with CaptureNX 2 v 2.4.5 in default mode.
- DSC_5549 (capturexn2-pt) .jpg - file converted with CaptureNX 2 v 2.4.5 with PT selected capture mode.
- DSC_5549 (lr-auto-tone) .jpg - file converted with nude Adobe Photoshop Lightroom 5.2 using the 'Auto tone' command. When exporting to JPEG, the quality was 100%, other settings were default.
- DSC_5549 (lr-original) .jpg - file converted using nude Adobe Photoshop Lightroom 5.2 in default mode with standard profiles installed along with the LR setting. When exporting to JPEG, the quality was 100%, other settings were default.
- DSC_5549 (photoshop-from-original-jpeg-with-auto-levels) .jpg - this is how the processed on-camera JPEG looks like with Photoshop's “auto levels” command.
- DSC_5549 (viewnnx2-after-quick-adjust) .NEF - this is what ViewNX 2 did from the original RAW when installed white balance 'Calculate automaticaly' and PT mode is selected.
- DSC_5549 (viewnx2-auto-pt) .JPG - ViewNX 2 on install white balance 'Calculate automaticaly' and PT mode is selected.
- DSC_5549 (viewnx2-original) .JPG - ViewNX 2 by default
- DSC_5549 (inverted original jpeg) .JPG - original JPEG rotated 90 ° for viewing with built-in utility for Windows OS
- picassa-view-original-nef-file.jpg - this is how the original RAW file looks when viewed by Google Picassa, a screenshot of the program window.
No comment.
Configuring a converter is a very delicate and necessary task.
I would like to share some experience of using converters. I really like the way native ViewNX and CaptureNX display a photo. In order to reproduce the image as accurately as possible - since I saw it on the camera display and in real life, I use the “native” converters. Refinement (retouching, color correction and other boring routine) is done in Photoshop. I like a bit of a sadistic mechanism: first correct the original RAW (NEF) file using ViewNX, export the result to TIFF (16-bit uncompressed), and then finalize it in Photoshop. Thus, I get the parameters of the photo closest to the original. ViewNX has three very strong negative aspects:
- changes the original RAW file, after which the camera will not see it
- no noise reduction function, which greatly limits ViewNX when converting images at high ISO
- requires tremendous computing power
CaptureNX is also a very interesting and very useful converter with many subtle "native" settings that other converters do not have (for example dust removal function) But for some reason, he didn’t take root with me, I use it only for very thoughtful picking in RAW files.
Well, of course, Lighroom comes in handy when you need to deal with large file sizes. Nothing better for batch processing RAW has yet been invented :) I already began to notice that I behave like a gamer - when working in LR, I keep my left hand near the 'asd' keys for quick access to some functions. There is nothing better than turning on LR in the evening, pressing the letter 'F' twice (switch to full screen mode) and starting to fancy a photo.
LR load when processing photos in the Develop tab. Processor I3-3240, 8GB DDR3
Leave your thoughts about converters in the comments.
Thank you for attention. Arkady Shapoval.
Explain who knows. The NEF source is about 12MB in size. Converting to JPEG in ViewNX or Capture compresses them to 9MB. If I open RAW in Photoshop and process it (noise suppression, white balance, exposure compensation, sharpening) with subsequent saving in JPEG (maximum quality - 10), I get files of 2-3 sometimes 4 MB. What is the reason for this? I tried to process files that were previously converted to VNX in PS, but the noise is reduced, and the sharpness is increased by an order of magnitude worse than when working with RAW.
every JPEG save is lossy. more than 1 time it is better not to do this
I am wondering why, after a single compression in PS, the files are much smaller than after the same operation in VNX. The file is much smaller.
check color settings - 8 or 16 bit
No effect. From 12 MB, the file was compressed to 2,8 MB with 8 and 16 bits. I may not have written it correctly. I do the conversion in the built-in add-in PS Camera RAW 6.0., Because it has the best noise reduction results, well, for all open files, you can change the necessary parameters at once.
Different compression algorithms, that's all.
Clearly, the algorithm. The question is: are Adobe programmers so talented that even a shrunken file will print a picture no worse than after another converter (VNX) that does not compress the information so much or is it such a packer that Nikon is not strong? Why else I ask - it always surprises me that my friend on Canon, comparable to my Nikon model, always had photos 2-3 times easier, even when I shot JPEG.
Does VNX have quality settings?
Perhaps they want to maintain the highest possible quality, not taking into account the desire of users.
The same Nikon (for example, D5100) has three options for Jpeg compression, the difference in size is 3-4 times: in maximum quality (Fine) the photo weighs 8-10 MB, in average quality (Normal) - 3,8-5 MB, in the minimum (Basic) about 2,5 MB (I don't know for sure, I never used it).
Accordingly, if the resolution of the matrices Nikon and Canon is the same, then Canon is more likely to press the photo, respectively, there are less details (especially smooth gradients).
I just exported one photo from RAW to JPEG in Lightroom,
quality,% - size, Mb:
100 – 4,45
90 – 2,42
75 – 1,1
As you can see, the difference is 4 times, and visually the differences between the photos 100% -75% are not visible (at 100% magnification). The difference between 75% and 50% is very big, the file weighs 580 KB. At 50%, artifacts are already very visible, 'squares' appear on smooth gradients.
Right crop 100%
Canon and Nikon are different manufacturers, so I would ask them. What algorithms do they use to get photos.
Of course, there is everywhere I set the highest quality. Therefore, in general, the question arose.
I see it makes sense to set the maximum quality only if further processing is planned (although jpeg is not the best choice in this case), otherwise you just waste space on your hard disk. And yes, I also noticed that Nikon's "hardware" Jpeg of maximum quality noticeably decreases in size when resaving to FS (even without editing it, and with maximum quality).
Perhaps VNX considers 100% quality to be 'maximum', and Photoshop, for example, 95%. Or maybe, after all, the compression algorithms themselves are used different (after all, in the FS you can select the type of compression, you will have to experiment)?
I'm also not sure that Adobe programmers, and even more so Nikon programmers, create their own compression algorithms, most likely they use ready-made algorithms developed by other people. An interesting question, it would be necessary to google))
tell me good people - after converting from ViewNX-I and Capture NX-D to TIFF 16-bit format and then opening it in the photo shop, in the exeffe there is the processing (conversion) date and not the shooting date
maybe someone knows how to fix it (cure) maybe where the "tick" is not worth it ???
"The essence of the experiment is very simple - you need to convert the same raw file using different converters and see what happens :)"
That’s what I always did with everything that I came up with the choice))) But for myself and for my Canon cameras, after testing probably all the products available on the market, I definitely chose Capture One, which is very different from the same Adobe raw- Converter much more detailed and more accurate color reproduction. I won’t write about functionality, but I use FS for many years only as a photo editor, but never as a converter.
http://www.ex.ua/736382335410, if interested, from this link you can download the rav and jpeg of one frame from Fuja S5pro. And in Lightroom (or another converter) try to make the same color from the ravah as in the on-camera jpeg, save it in the preset and use it.
I tried several editors, but I have been using ACDSee for about 5 years, and in the version number 9, the brightness equalizer (selective) helps a lot. Good and converts to jeep.
Good evening !!! Please tell me why after converting to viewNX 2 the color rendition changes? automatically makes the image with greenish tints ... thanks in advance
Such a thought, instead of converting to jpg, you need to do a printscreen of the image visible in view-nx :)
The resolution of the pictures is the same:
Here is a photo from a printscreen https://yadi.sk/i/UsObJmfLzZJbx
Here is a photo after conversion https://yadi.sk/i/TxaxJqbYzZJdA
It can be seen that the converter is soaping and the photo is losing in detail and sharpness
The difference is obvious ... How do you make a print screen of an image? if it's not difficult for you, please help ...
)))
The difference is obvious ... but how to make a print screen? if it's not difficult for you to explain ... thank you in advance)))
Arkady hello! I was looking for a rav converter for a long time, and came to Dxo optic pro, the colors are more correct in my opinion. Nikonovsky software the computer does not pull, I enclose a photo with Nikon D90 sigma lens 17-50 2.8
Respected Alexander writes the correct answer. Indeed, for files from the D90, and other Nikon files of that period, DxO Optics is the best converter. I personally remember how he helped me out when others gave just an uncorrectable result. There is only one rule - who set up and worked out the conversion algorithms for specific cameras. For example, I'll also say that Capture One handles files from the D700 just fine. Are Pentax and Nikon just OEM versions of Silkipix? a very bad program in general, and then some people even write with pathos that they "found" the best converter for themselves. It's funny, it just means that they simply agree to what they are offered without even suspecting that they are being offered a half-slag under the brand name. I also noticed that the same Capture one is not optimized in any way for Pentax cameras, the files from my K-01 are processed just awfully, but ACRs give excellent results. I also think that Capture One is sharpened for really professional cameras, but there is one thing - in no case should you use the built-in noise reduction in it, I think that just because it is sharpened for converting professional and partly staged files, this the module was simply not developed there, because from medium format cameras and even in the studio everything is filmed at the lowest ISO values, so if you have to torment yourself with jokes, convert it without using noise reduction, save it to a 16-bit tiff and then open it in Photoshop / This is where the “inexplicable “At first glance, the thing is that Photoshop does not directly open such a Tiff obtained from C1, but offers to convert it using ACR, like a real RAW file, and there it is enough to simply apply a very high-quality ACR noise reduction and everything falls into place. And since most of the files still have to be finalized in Photoshop (at least - converted to black and white in my case), then, in fact, a lot of additional time is not required to process the files.
Good day to all! Perhaps someone came across, I can not figure out what's the matter. I take a rav from 7100, open it in nx2 - I do whatever I want with it. In Picture Control, I expose the portrait, add saturation and contrast. After that, File-save as-jpg-quality 100. I open the file in the windows 10 viewer, and there is not the same color as nx2, I open that color in ViewNX-i again, I'm glad., I immediately thought that the color profile of Windows is possible displays something wrong, opened jpg on ipad retina, everything is ok with color rendition. But even there the colors are the same "not correct" as in the Windows viewer.
The question is how to save from nx2 in jpg, so that both windows and other devices have a color rendition, such as when viewing in nx2 and ViewNX-i, I give a screenshot below. where at the same time the same jpg is opened in Windows and ViewNX-i. Thank you, waiting for your answers.
Look at the color space - it should be sRGB. If you have AdobeRGB, this is the reason. standard viewers, as a rule, do not know how to work with this profile.
In the camera, most likely, you have ADOBE rgb space, put srgb and everything will work out :)
The joy was not long, I went to the Camera settings: Shooting mode menu -Color space-but I have sRGB there, as I understand this is the default value. Where else can I see?
The question remains open, and has not found a solution. Perhaps someone was doing the same experiment, it would be interesting to know the result. (Save the developed nef from capture to jpg, and simultaneously view the same jpg in view nx and windows viewer)
There will always be differences. I work like this:
1. Camera settings - AdobeRGB, RAW 14-bit
2. Developing Capture One - select “Built-in Camera Profile”.
3. For editing in Photoshop, when opening, I specify “Convert to AdobeRGB”.
With this technique, the color in these two programs is the same. All other programs show "who is good at what."
When printing on an Epson P50 printer, I indicate - Color is controlled by the printer. More or less the color matches.
Monitor NEC PA241W. Nikon D90 and D610 cameras.
Photolab is a separate song altogether! I tell them “no color correction”, get pictures, cry, catharsis ...
Still filmed a small video on simple development in Capture One Pro https://youtu.be/Yqac_fNG2Y4
Windows Viewer displays colors according to the monitor profile (others cannot), and photo editors according to the work profile specified in their color management settings. If you have a monitor with a wider color gamut (adobe rgb) than the working color gamut set in the editors (for example srgb), then in the Windows viewer, the picture will be richer, and in the editors - less saturated, corresponding to a narrower srgb color gamut.
This means that the windows viewer gives distorted colors, especially if you have a monitor with a wider color gamut than SRGB. It’s more correct to look at the photo in the editors, provided that you have correctly configured color management in them.
The settings in the camera of the color profile (SRGB, AdobeRGB), this is only for the in-camera jpg - Ravam does not care about them from the big bell tower, they have their own, much wider color gamut, ProPhoto RGB, if I'm not mistaken, and it is this huge color gamut when creating an intra-chamber jeep, it is compressed into either SRGB or Abobe by the fotik processor.
Therefore, if you need to get the maximum number of colors and shades, then shoot in raw and process in the widest possible color space (ProPhoto RGB or at least Adobe RGB) on a monitor with a wide color gamut.
When processing photos with wider color gamut (and it is better to process it in a wider gamut, there will be less clip clipping for complex editing), and then preparing the photo for posting on the Internet, the final conversion from a wide working color space is used (Melissa \ ProPhoto, Adobe) in the conventional, narrow SRGB, with preliminary proofing (soft-proofing). To prepare the photo for printing, the photo is also converted from a wide working profile, but already into the color profile of a particular printer, again with preliminary color-proof (soft-proofing).
When soft proofing is made for the edited image, the editor will display a simulated image on the monitor, how this photo will look in SRGB (what will be the final saturation, what colors and shades will merge, etc.). Then saturation and contrast corrections are made on the copy of the photo so that the photo in srgb visually is not much worse from itself in a wide working range. Then a copy with edits either to print or to the site.
As you can see, in this workflow there is no place for Windows viewers who are not able to display colors according to the selected color profile.
For more details, google Lynda color management essentials, in my opinion there is still one on the rucker.
An interesting way, I’ll try it today.
But all the same, it turns out that Father Nikon is making a fool, how is that ?! There is already a jpg 8bit file. -As far as I understand, it does not carry any specific information like nef. And it should display "almost" the same on the same monitor in the same operating system. But no, I open the same jpg in win and in the same win and nvnx, and the picture is fundamentally different. In vnx, a pinkish, natural complexion, in the viewer win and iPad Air 2 - frankly gray and dirty, which is actually visible in the screenshot.
this is not Nikon making a fool - it is the developers of formats and programs that make fun of. I think that the main thing is to select the color in: rav-converter, photoshop, print. but the fact that on some shit .. on the iPhone "not so" shows the color - it does not matter. you will not please everyone! you can't pick it up everywhere! then b / w do all the pictures. without toning. and I have been using this method for a long time. the only quality and at least somehow predictable. in lightroom, everything worked out gray-green for me. did not understand. put capture one.
by the way, you need to decide which color will be "correct". it's a matter of taste. canonists love carrots (excess vitamin a). I like more yellowish (hepatitis). someone rosy (baby). plus the principle of image formation on the screen and the print is different. the print is highly dependent on lighting.
when developing, I adjust the balance a little warmer. I look so that the whole picture is not covered in any shade. so that the colors are as if pure.
I am working in View NX2. For the selected photo, I try to vary the different engines. Then, WITHOUT SAVING the changes, I go to the next file. At the same time, a message is displayed that, they say, a backup copy of the previous saved file was created in the Originals folder, and you can change these settings in the Options dialog. And then, regardless of whether I agree or not, my original file is overwritten with these variations of mine, but the Originals folder is created with the original file. But I didn't set the Save command and didn't do the conversion! How can I get rid of the fact that when simply VIEWING a photo, it is overwritten? I looked through Edit-> Options but didn't find anything on this topic there. Tired of restoring the original files every time you view.
That just did not try. But, since I shoot mostly portraits, I eventually settled on RPP.
Tried: ViewNX 2; Capture NX 2; RawTherapee4.2.1; Zoner Photo Studio X; Adobe Photoshop Lightroom. But nothing beats ACDSee! It somehow automatically develops raw to an acceptable quality. It remains to correct if that ... -Everything is intuitive and simple. With the same - gemmorr: twist-twist the sliders ... well, it seems, that's it ... You compare it with ACDSee - and you want to spit - no color, no skin tone ...
I've always neglected Lightroom in the past, in constant search and testing of alternatives. But then I tried Adobe Lightroom Classic CC 2018 and was shocked - how he stepped forward. With a magnifying glass I looked out for the differences from Nikon Capture NX-D - everything is perfect! I'm not talking about the speed and usability - here Adobe is way ahead at times. The only thing that I did not immediately catch up was that it was necessary to change the camera calibration from Adobe Standard to Camera Standart / Vivid / Landscape, etc. - and for a long time I could not understand why the colors left my relatives so much. In general, after Lightroom, I delete all other converters, since nothing else is needed to be happy.
I'll have to check. Usually Lightroom gave out a clearly dirty color on Nikon and no shamanism could cure it. It is alarming that you only learned about Camera Standard / Vivid / Landscape now. Perhaps after choosing these presets you felt better. However, before it was such a difference - without these presets, it's full g..no, with these presets it's just g..no.
After reading you, let me, I think, I will check ... suddenly something has changed in a year. Everything is the same: Exp. climbs into terrible overexposures; the colors are hard. The portrait turns out to be lifeless.
In general, ACDSee rules. I don't know, can it be just me?
In general, of course, Nikon's trouble is the red channel. What did they do with him there, and most importantly why?
Arkady I am new to Raw. Please tell me if CaptureNX is better than the old version or the new free Capture NX-D. I have the old version on the disk with the recently purchased D300s. Is it necessary to additionally download View NX or the latest View NX 2?
CaptureNX and Capture NX-D are fundamentally different. The first was more functional. View NX is optional.
Thank you for your answer. I think the last free version of Capture NX-D is enough for a beginner. I was interested in the old version because of the Russian language, at least I read about it somewhere. But since it is no longer being updated, I think this option disappears.
View NX is not required, but I highly recommend it - it is very convenient to view and sort and reject RAVs in it: with pono-screen viewing, by clicking the mouse, the area where the cursor is hovered increases to 100% and in this way you can quickly check the sharpness.
Thank you. I’m downloading ViewNX 2. The main thing is that my laptop with 7 windows pulls all this. ViewNX 2 + Capture NX 2.
View NX is much, much lighter than Capture, NX2 and NX-D. Unless, of course, you try to edit the raves directly in it. However, according to experience, the load on the computer strongly depends on the megapixels of the camera and on the use of the "retouch" function. If you don't touch it, then 16 MP images can be processed quite comfortably on a 10-year-old laptop.
Thank you .. it means there is no need to worry about iron, but I already thought maybe a new laptop will have to be bought.
The biggest requests from these programs are for RAM. Unlike, for example, the programs of the Nik Collection suite, for some reason, on the contrary, there is almost no memory consumption, but the processor loads in full, natural stress testing.
I haven’t downloaded the converter yet, but I’m already wondering where is the exact standard, what should be the shot. As he showed a couple of photos, he said that they were all overexposed! As for me, everything was ideal!
The standard needs to be developed. The first 300.000 frames are the worst, and then, with any luck, an initial understanding of what is happening will come.
Then for a long time I was still studying, only 40 clicked on my D000.
Nobody says that it will turn out quickly :)
So it’s not in vain that he took an armor plate or the legendary shooter D300s and he will help me shoot 300😀
Certain standards exist. Google “Adams band theory”. It can and does cause controversy among prof. photographers, and to understand it, in relation to digital photography requires effort, but it is certainly much better than nothing, namely - the untrained eye of a beginner.
For a long time not relevant on the figure, if you shoot in RAW.
Since I had taken pictures before only in Jpeg and for home use I did not mess with the failed frames, I left it as it is. But since now I have D300s to grow with her ...
That's right, if you go to the ravine it’s already hard to drive back to the asshole, though if there is d40-d60 then you can not bother
Better - to work in several programs on one photo, and choose a better and more convenient one. Otherwise, you risk, at best, wasting time.
Earlier it was necessary - it helps a newbie to correct mistakes. Moreover, amateur cameras “provoke” mistakes. The horizon is constantly blocked up - it's hard to quickly find your bearings without marking on the focusing screen. Sometimes something gets into the frame that is not visible when shooting - the screen does not show the entire frame. And white balance and exposure - sometimes “unsuccessful” shots turn out to be simple movements, nothing at all. Native Capture NX-D is enough for me.
Thanks for the tips ..
Please advise RAW converter for Canon. Is there a native good converter, like Nikon? I won’t do business. Thank you in advance!
The native Canon Digital Photo Professional (aka Canon DPP) is quite suitable.
She is also free :)
Arkady, thank you!
It's not free - you paid for it with a Canon camera. :)
What is characteristic, it can be obtained from the official site by finding any Canon camera serial on the Internet. So did once.