Is there any software that can add raw files

Discussion in 'Digital Darkroom' started by thomas_lozinski, Jan 29, 2013.

  1. I'm wondering if there is any software available that can add 2 raw images together to create a new raw image. Would like this to add exposures together. Additional math functions such as averaging 2 raws would be great. The whole point would be to have the final image still in raw format to provide the flexibility of raw.
     
  2. You can open many kinds of files in Adobe Camera Raw - for example, in Adobe Bridge™ by right clicking on the file and choosing "open in ACR" (will do jpgs as well as tifs, etc.). To keep the maximum information just save it in a non-lossy format.
     
  3. Depending you your software, you might need to save to DNG to get close to Raw, but not quite. You're not going to lose a lot with an uncompressed tiff.
     
  4. There is a way of doing this, although a bit roundabout. You have to own Adobe Photoshop to make it work. If you have access to Adobe Photoshop, you have also access to its raw converter, called Adobe Camera Raw (ACR).
    After importing your raw image into ACR, you upload it into Photoshop as a SMART OBJECT. This means that the raw image is preserved in Photoshop as a layer. Nothing prevents you from up-loading another raw image as a SMART OBJECT into another file in Photoshop. Therafter it is a piece of cake to move the second smart object on top of the first in the first file created. Averaging can be done by choosing an appropriate opacity in the second smart object layer.
    I hope this helps!
     
  5. Once you've manipulated it it's not a raw file any more. But you're not really asking for that, just the functionality of it.
    Most raw files are based on the tiff file format. So if you save your edited file as a 16 bit tiff then you'll be very close to having flexibility of a raw file. You'll lose a little bit as you won't be able to have a camera manufacturer's raw converting program read it in some cases, but if you're going for Photoshop or another type of pixel editor it will fit that purpose.
     
  6. I'm wondering if there is any software available that can add 2 raw images together to create a new raw image​
    Nope. You can of course combine two rendered versions from each raw. But raw is read only and doesn't look like any kind of image you'd want to look at. Kind of like asking if you can combine two color negs (well in the darkroom you could but why?). You can expose a single piece of paper with two color negs but the negs themselves are not combined.
     
  7. Once you've manipulated it it's not a raw file any more.​
    Not true, at quite a fundamental level - you can manipulate the pants off a Raw, and it never stops being Raw.
    All Raw converters are "non-destructive" by their nature (with the very slight exception that some converters might write to the file's Exif) - you can therefore manipulate a Raw file until you're blue in the face and the Raw itself doesn't change - the changes are stored in a database or in an xml or other "sidecar" file associated with the Raw.
    The manipulations are committed to a physical image file only at the point of conversion to a tiff, a jpeg or whatever. And the Raw stays Raw.
     
  8. Thanks for the help but this still won't allow saving in raw format. It seems like this should be easily accomplished, just by adding the pixel values but I've never seen any software that does it.
     
  9. Thanks for the help but this still won't allow saving in raw format. It seems like this should be easily accomplished, just by adding the pixel values but I've never seen any software that does it.​
    You probably need to search and read up on what a raw file is (and isn't). It is partially processed data, you can't save out that kind of data any more than you can combine two raw files. And no, it isn't easily accomplished.

    http://en.wikipedia.org/wiki/Raw_image_format#Sensor_image_data
     
  10. I'm wondering if there is any software available that can add 2 raw images together to create a new raw image. Would like this to add exposures together.​
    Isn't this functionally what is done with the HDR method? Do you perceive some advantage in having the summed image in raw format (assuming it were possible to accomplish)?
     
  11. Isn't this functionally what is done with the HDR method?​
    No. The images have to be rendered first. Raw isn't a rendered image. It isn't demosaiced.
     
  12. Andrew-
    I recognize that. Key word is "functionally". The OP wants to add two exposures. Is this not what HDR does, in effect?
     
  13. Key word is "functionally". The OP wants to add two exposures. Is this not what HDR does, in effect?​
    Yes (partially). Directly from raw, no. The OP seems to be slightly in need of looking into the differences between raw data and rendered data. I don't see how bringing HDR into the mix helps in that respect.
     
  14. I understand that raw files can be partially cooked and compressed but besides that they must contain the luminance values for each
    photo site. I just want to add these values. My Nikon cameras can do this ( when doing multiple exposures) so why can't I do it on my
    computer?
     
  15. I understand that raw files can be partially cooked and compressed​
    But they are not partially cooked, they are raw. You have to cook em (render them) to do anything with the image, in fact to build an image. They do contain luminance values but only one pixel which contains only one 'color' filter. They have to be demosaiced to produce RGB data.
    Here's what a raw file looks like:
    http://digitaldog.net/files/raw.jpg
     
  16. I think everyone is missing my point. I am a photographer, not a software engineer but I do understand what a raw file is. I know that they need to be converted to be viewed. I just want to add 2 non converted files and have the result in the same nonconverted format. Like I said, my Nikon cameras can do this, why can't my computer? It's obvious that the file formats are able to be decoded otherwise lightroom etc. wouldn't work. If you're wondering what the practical implications/ uses of this are here is a scenario: You take 1 shot, well underexposed, take another at same exposure and continue. Now in post you can add these up until you have the optimum exposure with no blown highlights and still have all the capabilities that the raw format affords you. Another capability would be to allow extremely long exposures if you could average the raw files. I would also like to be able to take the top half of one exposure and the bottom half of another exposure with a gradation and still have RAW (OK I'm asking for a little bit more now.) I know you can do all this in photoshop but then you lose the raw format. I think the benefits of the raw format have been discussed ad nauseum so I won't get into that. And Yes, sometimes raw files are partially cooked (but don't take my word for it. Dxomark understands this stuff much better than I do. http://www.dxomark.com/index.php/Publications/DxOMark-Insights/Half-cooked-RAW If there is a software engineer that can explain why you can't average 2 raw files I'd really like to understand this.
     
  17. Thanks for posting that link. Did you notice that they include the output from scanners as part the definition of RAW?
    A camera raw image file contains minimally processed data from the image sensor of either a digital camera, image scanner, or motion picture film scanner...Flatbed and film scanner sensors are typically straight narrow RGB or RGBI (where "I" stands for the additional infra-red channel for automatic dust removal) strips that are swept across an image. The HDRi raw data format is able to store the infrared raw data, which can be used for infrared cleaning, as an additional 16-bit channel. The remainder of the discussion about raw files applies to them as well.
    http://en.wikipedia.org/wiki/Raw_image_format
    (Emphasis added.)
    Apologies to the rest of you for going off topic, but Andrew knows what the reference is about.
     
  18. I think everyone is missing my point.​
    I'd agree I'm not at all certain what you're asking for or to do.
    I just want to add 2 non converted files and have the result in the same nonconverted format. Like I said, my Nikon cameras can do this...​
    It can? How (and why)? Can you explain exactly what you are hoping to produce? I understand you want to do something with two raw files. Can you describe what you hope to do?
    You take 1 shot, well underexposed, take another at same exposure and continue. Now in post you can add these up until you have the optimum exposure with no blown highlights and still have all the capabilities that the raw format affords you.​
    Again, once you render the image, this can be done.
    Exactly how can the Nikon do this without rendering the data? That's a new one for me (but I'm not a Nikon user).
     
  19. Thomas, you're describing (for the most part) HDR. High Dynamic Range. Have a look at http://www.hdrsoft.com/ or http://www.fdrtools.com/front_e.php or http://www.oloneo.com/ or search google for HDR. But you do need to not get hung up on the raw file format. Any program that will let you do this (and more) will let you save in a number of formats but a raw file isn't one of them. Not unless you plan to write your own code. Even DXO isn't going to write to a raw file. But to be blunt there's no reason to write back to a raw file type. The advantages can still be had with a tiff or other lossless format. Jpegs are lossy, raw files are not, neither are a host of other file types. In fact, once you start getting into HDR, you don't even need to shoot in raw, just shoot jpegs and cover the dynamic range you want to end up with. When you combine those jpegs in HDR software you wind up with something that is greater than the sum of it's parts.
    Most of us have found that we need to just ignore Andrew. He gets off on a tangent and takes the discussion in a different direction. His insistence on what a raw file is for example. Yes, he's correct and posted links to support his argument, but he's not really answering your real question. He's more interested in proving his point. Wayne is just following up on a previous "discussion" with Andrew about something else. http://www.photo.net/digital-darkroom-forum/00avaH
     
  20. Spearhead

    Spearhead Moderator

    he's not really answering your real question.​

    He answered in the 12:48 post above. That's where this could have ended, but it seemed that some people weren't happy with the answer.
     
  21. I just want to add 2 non converted files and have the result in the same nonconverted format. Like I said, my Nikon cameras can do this, why can't my computer?​
    I think this is a reference to the HDR imaging capability of the newer Nikon cameras (D800 etc.). In fact this process does not use the raw images but rather only the jpeg images; so it is working with converted images.
    Now in post you can add these (raw files) up until you have the optimum exposure with no blown highlights and still have all the capabilities that the raw format affords you.​
    This discussion continues because of a conceived (by the OP) advantage of summing two raw files to produce a file that remains in raw format. Assuming it could be done (and we all agree, it cannot), what capabilities do you assume such a raw file would afford that could not be achieved by first converting the raw files to 16 bit tiff format, and then performing whatever summation operation you choose?
     
  22. Here is an example of an unfinished (unfortunately) project, consisting on an extension to DCRaw, by Manuel Llorens, which was aimed at what you are looking:
    Synthetic raws and raw stacking
    You might want also to take a look at Astrophotography software, which performs those kind of operations.
    This software, Images Plus, is highly regarded for data manipulation, although I have not personally used it (too expensive for my needs). I don't think that you can output a raw from images plus.
     
  23. Thomas,
    You mention that your NIKON can stack raw exposures, average them out and produce the result as a raw file. As far as I understand, the result is produced as a JPEG file.
    If my understanding is correct, this same operation can suitably be replicated by the process I described earlier. After having taken a burst of exposures with your NIKON, to be outputted by NIKON as raw files, you import this burst of raw files into ACR. From there, you upload them individually into Adobe Photoshop, each on its own file as SMART OBJECT. From there you just move those smart objects as individual layers into one file. Thereafter you can average out those layers and output the result as Tiff, PSD or JPEG, whatever is your fancy.
    The advantage with applying this cumbersome procedure is – precisely as you intend – that you can go back to each raw file picture freely and change the parameters without any pixel losses. It is only the output that is being rendered as Tiff, PSD or JPEG, the individual raw files are preserved as layers in the file: they can be accessed and re-processed loss-free as many times as you wish, and the output will be adjusting itself accordingly.
    I content that you cannot, with this procedure, get an output that is in the raw format and can be re-processed again loss-free as such. But as long as you can re-process the underlying raw files (smart objects), there is no lossy processing involved. In my opinion, this procedure works as well as the one desired by you; your proposed procedure will possibly be unobtainable to photographing amateurs. I am sure that data experts can construct a scheme to accommodate you perfectly, but will it be workable as easily as the procedure described by me?
     
  24. Francisco, thank you for understanding what I was saying. I don't know how I was misunderstood by so many. That is
    unfortunate that the project is unfinished.

    To the others, this has nothing to do with HDR, I despise it. I appreciate the smart object work around but it still isn't
    perfect, raw files are large enough, I don't want tiff files eating up my hard drive space. If I could average 10 raw files I
    would be happy to delete the originals. The smart object technique makes more files instead of less. I don't think I need
    to explain the advantage of having a file in raw fomat.

    And yes my nikons can stack 2 to 9 raw files either averaging or summing them resulting in a raw file. I do it all the time
    and I am positive the result is in raw format. I am talking about multiple exposure mode NOT HDR! You can not do this
    after the fact but have to set the camera before acquiring the images.
     
  25. I don't know how I was misunderstood by so many.​
    You weren't "misunderstood", Thomas - you were understood perfectly well. You asked a specific question and got a number of specific and correct responses to that question.
    What you did was ask the wrong question. It's nobody's fault but yours that the answers you got answered the question you asked rather than the question you meant to ask.
     
  26. Thomas, there is potential issue even if a software like that were available. NEF files aren't "pure" raw, as in a sensor dump. They have some processing that cannot be overriden by the user. This might reduce the effectiveness or precision of such operations.
    Some "known" manipulations:
    - White Balance pre-conditioning: The Red and Blue channels are multiplied (by a different factor each and they vary depending on certain variables like ISO) after they are converted to digital. This can be easily shown by gaps at constant intervals in raw histograms.
    - Black point offset
    - Noise reduction when using long exposure times (this is not long exposure Noise reduction or LENR, which is basically a black frame subtraction), also known as the "star killing algorithm"

    In the case of in camera multiple exposure, it might be possible that the addition or averaging (I think it is an exposure adjustment of each capture depending on the number of shots, not a real averaging) is performed before other manipulations as those described before. I would guess it is a similar case with LENR, where the user don't have access to the black frame.
     
  27. I still don't understand the benefit. I mean, I get the idea that the image can be 're-manipulated without loss at a later date.' But since Raw stacking requires converting to a non-standard format that cannot be easily uploaded to the internet or sent to a printer, what exactly is the benefit of using this technique as opposed to running a program like Noise Ninja and then saving as a stack of smart objects in the PSD format?
    Even your default Raw files already have their 'changes saved', provided you don't delete the linked .xmp file.
    It just seems like a lot of work to get a result that is already possible by existing methods.
    You're not a Linux user by any chance, are you? ;)
     
  28. I am talking about multiple exposure mode NOT HDR!​
    Well you're really not explaining it very well, nor are you showing anything as an example of what you're after, nor did you actually say "Multiple Exposure Mode" until now. So I googled it. Are you talking about something like this http://sherrimeyer.com/Blog/?p=1072 If so, buy Photoshop and forget the whole "advantages of raw over other formats" thing. There isn't any.
     
  29. Most of us have found that we need to just ignore Andrew.​
    Wow, someone woke up on the wrong (troll) side of the bed today...
    Yes, he's correct and posted links to support his argument, but he's not really answering your real question.​
    And then:
    Well you're really not explaining it very well, nor are you showing anything as an example of what you're after, nor did you actually say "Multiple Exposure Mode" until now.​
    I already asked the OP for clarification way before you came onto the scene and added nothing useful but your negative bias towards another member.
    You have any better idea of what he's asking for? Because the initial question (I'm wondering if there is any software available that can add 2 raw images together to create a new raw image.) was answered way, way back in the first posts.
    he's not really answering your real question.​
    Neither are you.
     
  30. Dan, yes I'm looking to do something like the link you provided but again, I want the result in raw file format (.nef or .dng preferred.) The link that Fransisco provided http://www.rawness.es/extender-dcraw/?lang=en also demonstrates what I would like to do. Although I've been told this can't be done, there seems to have been this dcraw project that was started and not finished, and frankly I don't believe it can't be done (especially because I know that my Nikon can do it.) My question may have been answered but I don't believe some of the answers. Perhaps I should have rephrased my question as adding 2 raw images to provide a new image in a .nef or .dng or .crw etc format to get away from the fact that raw implies original.
     
  31. Dan, yes I'm looking to do something like the link you provided but again, I want the result in raw file format (.nef or .dng preferred.) The link that Fransisco provided http://www.rawness.es/extender-dcraw/?lang=en also demonstrates what I would like to do. Although I've been told this can't be done, there seems to have been this dcraw project that was started and not finished, and frankly I don't believe it can't be done (especially because I know that my Nikon can do it.) My question may have been answered but I don't believe some of the answers. Perhaps I should have rephrased my question as adding 2 raw images to provide a new image in a .nef or .dng or .crw etc format to get away from the fact that raw implies original.
     
  32. I want the result in raw file format (.nef or .dng preferred.)​
    NEF is a proprietary Nikon raw format. So unless Nikon builds this for you, not going to happen.
    DNG can contain actual raw data. Or it can contain rendered data (you can save a TIFF or JPEG as a DNG, and doesn't make that raw data!). So be careful, you might find someone who will write you a DNG that is nothing more than two (or more) renderings saved within the DNG format. DNG is just a file container based on TIFF! It doesn't equal raw although it may.
    Perhaps I should have rephrased my question as adding 2 raw images to provide a new image in a .nef or .dng or .crw etc format to get away from the fact that raw implies original.​
    Raw should imply simply that, raw data which for some reason you want to end up with. So you're going to now render that data to get an actual, usable image? Don't see the point when you could do exactly the same thing, end up with the same results with two separate raw files you render and combine. The net results would be the same.
    I don't believe it can't be done​
    The answers here haven't said it can't be done, it isn't done and probably for sound reasons: you'd have to render the raw anyway so combining two raws into one, or rendering and then combing results in the same in the end. A bit like asking to combine two color negs into one. You could do this in the darkroom and end up with a 2nd generation single neg. Or you could just place one into the enlarger, make an exposure, then pop the 2nd in there and make an exposure and the results would be the same. Without having a 2nd generation neg and with a lot less work.
     
  33. You don't have to render the raw for what the OP is asking. It is a basic mathematical operation, pixel by pixel, similar to other operations performed (as in the Nikons) before writing the propierary raws.

    Let's take an example, A specific RGGB group has the following values in one capture:
    R=512; G1=824; G2=825; B=238
    Then, the same RGGB group has the following values in another capture:
    R = 724; G1=1520; G2=1519; B=680
    If we take just the basic addition we end with:
    R=512+724=1236
    G1=824+1520=2344
    G2=825+1519=2344
    B=238+680=918
    These will be the values for that RGGB group in new "raw", no need to render or demosaice. Any mathematical operation could be performed, but you could get into issues like weird colors when processing the syntetized raw. If it makes sense or not is another story.
    An example of where you could use this is for star trails images. You could combine thousands of captures into one raw much faster and taking less disk space than rendering each individual image and combining them in Photoshop.
    Issues like propietary formats complicate the matter in order to generate a new raw. It might also be impractical or of very limited use.
    Here you can read about another unfinished project that tried to combine raw files with a raw output.
     
  34. If it makes sense or not is another story.​
    The crux of the matter! That we have two unfinished processes that may not even be appropriate for the processing of photographic, not scientific imagery seems to point to this not making much sense (for the OP).
     
  35. I follow what the OP wants to do. I just pulled out a Nikon D800 and following the instructions in the manual (PGS. 195-
    200 in the English version) with the camera set to shoot 14 bit lossly compressed NEFs shot a two shot multiple
    exposure. The result is a single NEF raw file, not a JPEG or a TIFF. You don't end up with the individual frames and the
    multiple exposure, just the one file.

    From the manual: "Multiple exposures can make use of RAW data from the camera image sensor to produce colors
    noticeably superior to those in software-generated photographic overlays."
    used. Live View cannot be used.

    Here's what I think is going on inside the D800 and similar cameras: the true raw from each exposure is actually being
    used - held in a cache before all of the exposures are combined by he EXPEED 3 processor into a single processed NEF
    file.

    I can see this as being handy for shooting fireworks or any other situation where you have multiple elements ( like a
    strobe lit gymnast moving against a dark background for example. There are some limits: only up to 10 exposures can be

    Perhaps (and this a big speculation on my part as I don't use the program) Nikon's Capture NX2 may let you produce a
    multiple exposure NEF as well but if my hypothesis is correct that will not be possible either.

    You might try (this is a long shot) a program like PTGui or a focus stacking program but the result will not be a raw file.
     
  36. I wouldn't mind the same thing. I'd like to take red channel from one RAW, green from another and blue from the third and build the RAW from them. Why? In order to scan negatives with camera without blue filters. Setting the white balance from the resulting file would be much easier and it should avoid the noise problems resulting from pure white balance approach.
    I don't think the same can be achieved with PhotoShop as Bayer interpolation messes things in this case.
     
  37. I'd like to take red channel from one RAW, green from another and blue from the third and build the RAW from them.​
    Channels? Raw is essentially a Grayscale, single "channel" file until it's rendered.
     
  38. Very funny. You know quite well that different grey values in there represent different colors in the sensor and those can just as well be referred as "channels".
     
  39. Very funny.​
    Glad you were amused not that it was my intent.
    You know quite well that different grey values in there represent different colors in the sensor and those can just as well be referred as "channels".​
    That I do. What I don't know is what you are trying to achieve. After dozens of posts taking us all down a rabbit hole thanks to a ambiguous question/request of the OP, let's not go farther.
    You want to do exactly what when you say you want to take the red (just those pixel sites and none others), green and blue (which isn't red green or blue yet) and do what?
    http://www.cambridgeincolour.com/tutorials/RAW-file-format.htm
    Demosaicing and white balance involve interpreting and converting the bayer array into an image with all three colors at each pixel, and occur in the same step.​
    Is that incorrect? And if not, you want to take ¼ of the pixels that have a luminance value and color filter data and do what? That data will not look different from the illustration I provided above?
     
  40. I follow what the OP wants to do. I just pulled out a Nikon D800 and following the instructions in the manual (PGS. 195- 200 in the English version) with the camera set to shoot 14 bit lossly compressed NEFs shot a two shot multiple exposure. The result is a single NEF raw file, not a JPEG or a TIFF. You don't end up with the individual frames and the multiple exposure, just the one file.​
    That's probably the most useful and interesting information in this entire thread. Dang! I didn't know battery powered in-camera processors could create a Raw file like that, Ellis. Thanks for posting this. Technology keeps moving forward.
    Wonder what the final Raw NEF looks like with regards to noise as a starting point to edit in a Raw converter? Or can you only take advantage of the expanded dynamic range in Nikon's own converter?
     
  41. ...with the camera set to shoot 14 bit lossly compressed NEFs shot a two shot multiple exposure. The result is a single NEF raw file, not a JPEG or a TIFF. You don't end up with the individual frames and the multiple exposure, just the one file.​
    Now the interesting question is: If you have blown highlights in one exposure but not in the other, what do the highlights look like in the combined NEF?
     
  42. Mike Blume: they look blown.
     
  43. Thanks Ellis. That suggests the two raw images are not being "averaged". So I am still left wondering what the OP wishes to accomplish by his request, assuming it could even be implemented.
     
  44. Tim: "Wonder what the final Raw NEF looks like with regards to noise as a starting point to edit in a Raw converter? Or
    can you only take advantage of the expanded dynamic range in Nikon's own converter"


    No idea about the noise. I imagine this may be how dark exposure noise subtraction works if you are doing very long
    exposures and have the camera's long exposure noise reduction feature turned on.

    As far as I can see in my tests the multiple exposure function does not increase dynamic range beyond 14 bit. At any rate
    unless your camera is locked down and the position of everything in your composition is absolutely fixed.
     
  45. So is it a useful feature or not, Ellis?
    Have you compared and noticed any increase in the ability to recover more or less highlight and shadow detail on the extreme ends on a single exposed NEF vs a combined bracketed?
    A thought crossed my mind what might be happening in camera is averaging and applying two exposure/rendering instruction sets and embedding those averaged instructions so NEF converter will read and map tones accordingly on a single exposed sensor dump. They actually take two exposures but gauge the rendering instructions accordingly without combining two actual linear sensor feeds.
    But the alternative would be when there's an increase in light between two exposures for each pixel site on the sensor you'll have two different linear pixel dumps going through the A/D converter to combine digitally with rendering instructions for the Raw converter to read. Some linear pixel data has to be kept and some disposed of to combine into one linear Raw file. Not sure. Probably a lot more complicated and proprietary than I actually realize.
    I take it this is nothing like Nikon's D-Lighting.
     
  46. It us useful for making in camera multiple exposures.

    I haven't tried shooting an exposure bracket. I used it to make multiple exposures of disparate subjects, as opposed to
    working with disparate subjects as masked layers in Photoshop. At any rate what you propose would only work with
    subject matter that isn't moving.

    I suspect this feature has been included as a special effect for things like sports photography where you might have a
    gymnast or a baseball player at bat where there parts of the subject stay in basically the same place. That implies that
    some sort of algorithm exposure averaging is going on. Does that mean there's also an expansion of the camera's
    native dynamic range? I don't know the answer to that and haven't explored it.

    I also haven't tried it as a focus stacking technique.

    In camera raw processing is of course what happens when you set your camera to record a photo as a TIFF or JPEG. All
    digital cameras do that. Nothing new there.

    "They actually take two exposures but gauge the rendering instructions accordingly without combining two actual linear
    sensor feeds."

    Could be. Only an EXPEED engineer at Nikon would know that.

    I really think this is something that over a beer or a bottle of Sake somebody said to someone else "Wouldn't it be cool
    if?"
     
  47. It us useful for making in camera multiple exposures.​
    And other than that (asked a Canon shooter who knows no better)?
    I also haven't tried it as a focus stacking technique.​
    Can you do that with the raw's and combine them as you can in say Photoshop, or do you have to render?
     
  48. The OP was trying to combine two Raw files, one exposed for highlights, the other for shadows and blend together. Of course that can't be done at the Raw converter stage outside the camera.
    I gathered from what you've indicated in EXPEED, Ellis, is that the two Raws are actual separate sensor feeds sent through the A/D converter and combined into one Raw image in-camera, right? I guess that is the part we aren't sure on. One single set of default/or camera setting specific rendering instructions would still have to be embedded into the combined single Raw for the NEF converter to read.
    I'm trying to distinguish what that EXPEED thing is doing from D-Lighting which from what I saw from dpreview sample images all it did was to just embed the instructions to bump up the shadows after the normally exposed sensor data passed through A/D converter. And what I saw was it bumped up a lot noise in the process. I remember there's an image of an anchor on a shipping dock on dpreview showing the D-Lighting lightening the shadows and increasing noticeable noise.
    IOW the EXPEED feature looks like it's doing what the OP is wanting to attempt.
     
  49. Op here. The intent was not to combine shadows and highlights. I'm not trying to do anything like HDR.
     
  50. "One experiment is worth a thousand expert opinions."

    Evidently that's an ild saying but the first person I heard say it was former NASA Astronaut Alan Shepherd when I
    photographed him for a Fujifilm USA ad back in the late '80s or early '90s.

    But none of us are experts and are really just guessing in the dark.

    My joining this thread and initial response to the OP was because he wrote that "his Nikon camera was capable of
    blending two or more (up to ten) exposures into a single raw file. I decided to RTFM and try it and both he and the manual
    were correctly describing what a modern Nikon DSLR is capable of. I don't know how long this feature has been part of
    the Nikon EXPEED tool set. I suspect the D4 has it but the D800 is the one I have in house. It may be that Canon EOS
    1DX and 5D Mk III are capable of it as well.

    About 8 years ago I reviewed several 22mp MFDB backs and a few of these had a micro stepping feature that could
    move the sensor in half or whole pixel increments and make four or sixteen exposures so you had a full set (2 green, 1
    blue, 1 red ) set of pixel values at each point in the frame. Of course the camera and subject had to be motionless for it to
    work but it did work (color was extraordinary) and you did not have to adjust exposure to compensate for all of the piling
    up of exposures. dynamic range was maybe slightly increased but not by very much.

    Back to the OPs original question:

    Is there a raw processing program or post raw processing program that can take multiple raw exposures and blend them
    into a single raw file?

    I do not know of one but have a suggestion for a possible workflow if he is using Lightroom. Get Timothy Armes "Enfuse"
    plug-in for Lightroom. Take a raw frame that has a very large dynamic range and make a virtual copy of it. Process one
    copy for the highlights and one for the shadows. Use Enfuse to blend the two into a 16 bit per channel TIFF in the Pro
    Photo RGB color space. Enfuse produces a very naural looking result at its default settings. After that there are two
    possible options. Either convert it to a Smart Object to get ACR controls in Photoshop or reimport it into Lightroom.
    There's another more work intensive set of options well involving layers, layer blend modes, making the layers above the
    background clipping masks, and selections but you probably don't want to go there unless you really need to or are a
    masochist with a perfectionist control freak streak.

    You boys have fun speculating about what's going on inside a camera's processor programming. This Angelo's stepping
    off the pin head.
     
  51. Thomas you posted while I was writing.

    The short answer is: I don't know of any. Why does the result need to be in a raw (proprietary or DNG) format? If you can
    do it with layered smart objects and if you need to, use blending modes and clipping masks to achieve a result you are
    satisfied with, save a flattened version back into Lightroom and export it as a DNG would satisfy your criteria?

    If you go the in camera multiple exposure route what I see is that the camera is making decisions for you regarding how you want the relative weight of different subjects to balance against each other. There's a large degree of serendipity involved which can be a wonderful thing.
     
  52. Op here. The intent was not to combine shadows and highlights. I'm not trying to do anything like HDR.​
    So I repeat a question I asked several pages back. What are you trying to accomplish? Exactly what problem are you trying to solve that such software (assuming it existed or could be written) would be useful? Many good folks here have offered a variety of options but none of these seem to address your concern.
     
  53. Mike, here is an example of what I want to do. . . I want to do a very long exposure, about half hour in daylight without nd
    filtration. Let me average 1000 raw images and delete the originals. Now I have a 20mb file with everything I need.
    Sure smart filters would work but then I'm using 20gb of disk space. I can do 10 in camera so why not 1000 on my
    computer? If you have any alternative suggestion for a half hour exposure without filtration at non diffraction limited
    apertures during daylight with all the flexibility of raw format I'd be glad to hear it.
     
  54. Thomas,
    So it is clear that you wish to capture several exposures over a long period of time, without using nd filtration, and with a non diffraction limited aperture, and finally somehow combine these exposure to obtain a file in raw format. Unfortunately the general consensus is that the last of these desires is likely not possible.
    What you have not made clear is what ultimate goal you wish to achieve with the final image. (You have stated that you are not seeking to increase the dynamic range, but this only tells us what you are not aiming to accomplish.) There are lots of very bright people on this site, and lots of ways to skin a cat. So if you can be more explicit about where you want to go rather than how you wish to get there, I am sure someone can help.
     
  55. Thomas, I'm not sure why you got so much pushback and so many bad answers on here. You're right that it should be possible to add two RAW files and perform operations such as averaging, since the values are linear (usually). Just because popular tools don't support it doesn't mean it's impossible. It's more accurate to do it at the RAW stage than after demosaicing and tone curves have been applied, since the tone curve has to be re-converted back into linear values to do a true average.

    Stacking an hour's worth of shots into a single averaged long exposure is something I would like to be able to do as well. I think the reason we don't see this feature in RAW converters is that almost no one does really long exposures, so producing the software would not be economical. It would be even easier and more practical to have the feature in-camera (it would avoid writing and transferring gigs and gigs of RAW files), but we don't see that either except with a few manufacturers' limited multiple exposure modes.

    I am a software engineer, and I occasionally write software to do things like this. If I ever make a tool that can do RAW sum or average, or come across one, I'll be sure to let you know.
     
  56. Sam, Thanks for understanding. I have notifications enabled so if you update this I'll see.
    I'd be happy to test whatever you write!
     
  57. I know this an old thread, but every so often I look around for raw averaging tools and this thread continues to be one of the top links every time.
    So for those that come here looking I just wanted to say that there IS software to do something like this, Pixel Fixer at pixelfixer.org . It does median and maximum and also dark frame subtraction with raw files outputting raw files (among other tools). He hasn't added average but given a poke I think he would.
     
  58. Thanks Edward Downer! Especially since it's such an old thread. That looks great. If I understand correctly the output is still a raw file?
    Really looking forward to experimenting with it later.
     
  59. Indeed it is still a raw file. I am hoping he will add other math operations like mean/average. While I don't see myself using say a minimum operation you never know and it shouldn't be too hard to add.
     

Share This Page