sRGB Vs Adobe RGB in Canon
- Thread starter AndyWest
- Start date
Dino f
Wiiiiilmmmaaaaaaaa
- Messages
- 2,711
- Name
- Dean Feltimo
- Edit My Images
- Yes
If im correct, its all about colour space. Argb has a larger colour base than Srgb, but, if a colour on you picture is outside the space Argb will change all the colours until it reaches the nearest colour available, whilst Srgb will only change the unavailable colour. Meaning more of the colours on you picture will be truer than if you were using Argb.
If that makes sense.
Dean
If that makes sense.
Dean
- Messages
- 5,288
- Name
- Andy Jones
- Edit My Images
- Yes
Use sRGB. AdobeRGB is only needed for high end reproduction with images that require the extra saturation the colour space provides. For web and "normal" printing with companies like Photobox, DSCL, etc. sRGB is the correct colour space to use.
The golden rule here is anyone who needs to be using AdobeRGB won't need to ask. Unfortunately a lot of magazines say it has more colours but the articles are written by people who really don't have clue one about colour management, colour spaces or profiles
- Messages
- 2,683
- Name
- stewart
- Edit My Images
- Yes
I agree with px8. Try it take the same photo on sRGB and then AdobeRGB. If you do not then process the adobe one correctly it will likely appear very flat. Adobe compresses the colours for later processing which is why the claim to record more colours comes about
- Messages
- 5,288
- Name
- Andy Jones
- Edit My Images
- Yes
It's not strictly true that is has more colours, an 8bit file has a bandwidth of 16.7m colours regardless of the colour space. AdobeRGB gives more saturation but at the cost of tonality. A red value of 255 in AdobeRGB is more saturated than a red value of 255 in sRGB. For the AdobeRGB to match the sRGB it would only need a red value of 219. So, if the shot doesn't need the extra saturation then you're basically wasting around 15% of the colour values available, instead of having more "colours" you've actually ended up with less - about 10m (220x220x220) instead of 16.7m (256x256x256).
Apart from the image not needing the extra saturation the output device (monitor/printer) probably won't be able to display it anyway. To get the most from AdobeRGB you either need a true 16bit workflow which means a printer capable of handling 16bit files (and not converting back to 8bit internally) or use CMYK which introduces a whole new set of problems not unlike a square peg and round hole.
Ideally the working space should always be the smallest one possible that encompasses the colours in the image and the profile for the output device should be slightly larger than the source. That way you've represented the saturation with the maximum number of tones but not exceeded the capability of the device displaying the image.
Apart from the image not needing the extra saturation the output device (monitor/printer) probably won't be able to display it anyway. To get the most from AdobeRGB you either need a true 16bit workflow which means a printer capable of handling 16bit files (and not converting back to 8bit internally) or use CMYK which introduces a whole new set of problems not unlike a square peg and round hole.
Ideally the working space should always be the smallest one possible that encompasses the colours in the image and the profile for the output device should be slightly larger than the source. That way you've represented the saturation with the maximum number of tones but not exceeded the capability of the device displaying the image.
- Messages
- 7,645
- Name
- Lewis
- Edit My Images
- Yes
Does the colourspace setting make any difference for shooting in RAW?
- Messages
- 910
- Name
- gary
- Edit My Images
- Yes
anyone understand this ?:bonk::bonk::bonk::shrug::shrug:
- Messages
- 1,017
- Name
- David
- Edit My Images
- Yes
someone get Adel in her she knows all about it 
- Messages
- 5,288
- Name
- Andy Jones
- Edit My Images
- Yes
Hence the comment about "anyone who needs AdobeRGB won't need to ask". If you don't understand just stick with sRGB and you'll be fine
- Messages
- 5,288
- Name
- Andy Jones
- Edit My Images
- Yes
Well a picture is worth a thousand words so here's 3,000 
This shows the relative difference of saturation between sRGB and AdobeRGB. Both sides have the same RGB values, the gradient runs from 255 to 127 but the different colour space means the result is very different. The AdobeRGB version is "clamped" at the top because the saturation is beyond what my monitor can display.
This shows how the image would look when printed by DSCL compared to the sRGB version. Again the difference in saturation is clear, the colour space of the printer is smaller even than sRGB. Hopefully this makes clear why AdobeRGB doesn't give you "more colours", it's how big the colour space of the output device that's important. If that can't handle the colours you'll never see them and the values set aside for those colours are wasted. By using a smaller colour space that's closer to the output device you're making better use of the values for smoother gradients.
Another common problem with AdobeRGB is when they're posted online, or even printed at a lab such as Photobox or DSCL. Unless the image is converted to sRGB first you end up with this:
The image was originally in AdobeRGB and only the middle part has been converted to sRGB, the outsides are AdobeRGB but being displayed in sRGB without conversion (note in particular the pink stripe on the little girls dress). This is what would happen if the image was printed by a lab such as PhotoBox, DSCL, Boots, etc. as they expect the image to be in sRGB and don't colour manage. The same happens if you view the image in a web browser or other application that isn't colour managed because the s/w isn't converting the RGB values so they are correctly displayed on your monitor.
Hopefully this makes things a little clearer. Colour management, profiles and calibration is a complex subject but get your monitor calibrated stick with sRGB throughout your workflow and you can't go wrong - images will look good on screen (and pretty close even in an unmanaged app) or print. When working with raw files the colour space for the image is decided by the raw app, not the camera setting so make sure it's exporting images in sRGB as sometimes this setting can be hidden - Adobe Camera Raw for example defaults to AdobeRGB but the option is tucked away.
This shows the relative difference of saturation between sRGB and AdobeRGB. Both sides have the same RGB values, the gradient runs from 255 to 127 but the different colour space means the result is very different. The AdobeRGB version is "clamped" at the top because the saturation is beyond what my monitor can display.
This shows how the image would look when printed by DSCL compared to the sRGB version. Again the difference in saturation is clear, the colour space of the printer is smaller even than sRGB. Hopefully this makes clear why AdobeRGB doesn't give you "more colours", it's how big the colour space of the output device that's important. If that can't handle the colours you'll never see them and the values set aside for those colours are wasted. By using a smaller colour space that's closer to the output device you're making better use of the values for smoother gradients.
Another common problem with AdobeRGB is when they're posted online, or even printed at a lab such as Photobox or DSCL. Unless the image is converted to sRGB first you end up with this:
The image was originally in AdobeRGB and only the middle part has been converted to sRGB, the outsides are AdobeRGB but being displayed in sRGB without conversion (note in particular the pink stripe on the little girls dress). This is what would happen if the image was printed by a lab such as PhotoBox, DSCL, Boots, etc. as they expect the image to be in sRGB and don't colour manage. The same happens if you view the image in a web browser or other application that isn't colour managed because the s/w isn't converting the RGB values so they are correctly displayed on your monitor.
Hopefully this makes things a little clearer. Colour management, profiles and calibration is a complex subject but get your monitor calibrated stick with sRGB throughout your workflow and you can't go wrong - images will look good on screen (and pretty close even in an unmanaged app) or print. When working with raw files the colour space for the image is decided by the raw app, not the camera setting so make sure it's exporting images in sRGB as sometimes this setting can be hidden - Adobe Camera Raw for example defaults to AdobeRGB but the option is tucked away.
- Messages
- 245
- Name
- Bradley Smith
- Edit My Images
- Yes
Depends what's in your photos really. If you're doing portraits then sRGB is better isn't it? With more bits used to represent the skin tones. If it's a landscape with a large amount of cyan present (as an example) then Adobe will be better. DS Colour Labs do have profiles for their printers that you can use as well but most labs will assume sRGB and if you try being fancy by using Adobe RGB (or any other profile) then your pics'll print up worse than if you stuck to sRGB.
- Messages
- 6,955
- Name
- Tel
- Edit My Images
- No
Blimey i didn't realize this :bonk: so how do you go about changing it ?
- Messages
- 5,288
- Name
- Andy Jones
- Edit My Images
- Yes
No, not without a 16bit pipeline you don't get more "cyan" you get more saturation. In 8bit you lose the tonality and if anything you're more likely to notice that with a landscape and the subtle gradients in the sky.
Bottom left there's a check box for workflow option, tick it to expose them and make sure the profile is sRGB.
- Messages
- 6,955
- Name
- Tel
- Edit My Images
- No
Thanks just did it as it was set to Adobe, maybe this would explain why i had trouble with a photobook recently, is there anything else that would need resetting in there ie 8bit or 16bit
- Messages
- 6,955
- Name
- Tel
- Edit My Images
- No
- Messages
- 5,288
- Name
- Andy Jones
- Edit My Images
- Yes
The printer driver will convert to 8bit so it's better to have control over that step and be able to preview the results before printing. 16bit printer drivers do exist but only for high end printers. You also also buy special s/w for printing 16bit but it's limited to high end printer and is pricey.
thanks px18 for your explanations of this topic.I think I have got the gist ofwhat you are
explaining. This matter has been bothering me for a while now. I have allways had my
camera set to adobe rgb as this is recomended in the D 300 digital field guide, however I always shoot in raw & use elements 5 with adobe camera raw plug in for all shots.I did convert to srgb only when saving to web pics or for sending to dscs col lab. however, I can not see how to set cam raw to Srgb output . can you point me to this optoin.regards
explaining. This matter has been bothering me for a while now. I have allways had my
camera set to adobe rgb as this is recomended in the D 300 digital field guide, however I always shoot in raw & use elements 5 with adobe camera raw plug in for all shots.I did convert to srgb only when saving to web pics or for sending to dscs col lab. however, I can not see how to set cam raw to Srgb output . can you point me to this optoin.regards
- Messages
- 5,288
- Name
- Andy Jones
- Edit My Images
- Yes
Profiles do exist for cameras, after all the raw s/w has to use one to decode the image, but it's not an easy task to relate the profile to what the camera sees. But as a rough guide camera profiles are smaller than ProPhoto which is what LR uses as the working profile when editing the raw image.
It might be interesting to know what extremes of saturation the camera can record but ultimately the useful information is what the output device can display because that's the best you'll get. What use is it to know the camera can record red saturation at 506 if the printer/screen can only display 300? (note figures are plucked from the air and don't mean anything). With colour management you always work to the limits of the output device in order to get the best results.
It might be interesting to know what extremes of saturation the camera can record but ultimately the useful information is what the output device can display because that's the best you'll get. What use is it to know the camera can record red saturation at 506 if the printer/screen can only display 300? (note figures are plucked from the air and don't mean anything). With colour management you always work to the limits of the output device in order to get the best results.
That's what I'm interested in.
If the actual colour value in real life is X and the camera can record X but that value is being converted to Y for the output device, how do we know that's being done in the manner that we would want?
Taking your theoretical example a bit further, let's assume we've shot something red that has that value of 506. But without knowing the extent of the scale that number is part of, it's meaningless. It could be 506 from 506. It could also be 506 from 5600. Similarly, when that value is being converted to the 300, the same questions apply. Is that 300 from 300, or from a larger number?
Is it a logical conversion based on a ratio or is something much more complicated going on within the conversion process? And for that matter how are the decisions being made?
I'm extremely curious to find out more about how we get from what we see in real life to what we see on our output devices and what goes on in-between.
- Messages
- 5,288
- Name
- Andy Jones
- Edit My Images
- Yes
It depends on the rendering intent you select when printing, proofing, etc.
Good article here on how the different methods affect the results:
http://www.cambridgeincolour.com/tutorials/color-space-conversion.htm
Good article here on how the different methods affect the results:
http://www.cambridgeincolour.com/tutorials/color-space-conversion.htm
- Messages
- 788
- Name
- Jamie
- Edit My Images
- No
What made me think that choice of colour space mattered even for RAW only shooting is that the image gets named differently e.g IMG.... or _MG...
I'm sure this may have conveyed to others and not just me that there is a difference in the respective raw images. After this discussion, this now seems entirely superfluous or is there a point in the naming I'm still missing?
Jamie
That makes for an interesting read, but it's not really telling me anything I don't already know.
I'm still interested in knowing what's going on at the actual moment of the image being recorded. And what, if any, colour space conversions are going on before we even get the RAW file on screen in front of us.
- Messages
- 5,288
- Name
- Andy Jones
- Edit My Images
- Yes
You can't really compare raw data to a colour space in the same way you can compare sRGB and a printer profile.
The raw data is read by LR say and it applies its own camera profile which defines tone curves, colours, etc. The image this generates is stored internally using the ProPhoto working space and displayed on screen using Perceptual intent for your monitor profile. Other apps may do it differently.
As to what the sensor is capable of it's really down to dynamic range of the pixels, the www.doxmark.com has a whole bunch of data about what each camera can record.
The raw data is read by LR say and it applies its own camera profile which defines tone curves, colours, etc. The image this generates is stored internally using the ProPhoto working space and displayed on screen using Perceptual intent for your monitor profile. Other apps may do it differently.
As to what the sensor is capable of it's really down to dynamic range of the pixels, the www.doxmark.com has a whole bunch of data about what each camera can record.
- Messages
- 5,288
- Name
- Andy Jones
- Edit My Images
- Yes
A raw file isn't an RGB image, it's only after it's been decoded by s/w using a profile that it's an image. The profile can be changed (picture styles for canon, camera profiles in LR/ACR, etc.) and the examples in your link are comparing the profile from CaptureOne's s/w. It's not definitive, it's just one way of decoding the raw data.
- Messages
- 4,240
- Edit My Images
- No
Thanks for the info/links PX18, very useful...
So it seems that the best option is just to allow clipping of colours to take place "at source", in other words when importing the raw file into ACR (by using Srgb colourspace).
Reading the article, even though I can prevent clipping by using Prophoto for editing when I convert to 8bit for printing using "Perceptual" rendering I run the risk of actually reducing the impact of ALL the colour in the image, if I were to use "absolute colourmetric" it just clips the out-of-gamut colurs anyway.
Back to the simplified Srgb workflow I think
Simon
So it seems that the best option is just to allow clipping of colours to take place "at source", in other words when importing the raw file into ACR (by using Srgb colourspace).
Reading the article, even though I can prevent clipping by using Prophoto for editing when I convert to 8bit for printing using "Perceptual" rendering I run the risk of actually reducing the impact of ALL the colour in the image, if I were to use "absolute colourmetric" it just clips the out-of-gamut colurs anyway.
Back to the simplified Srgb workflow I think
Simon