An experiment to compare a 400mm f/5.6 on a 60D vs. the same lens on a 6D with the 1.4X TC III (manual focus) ended up displaying a vastly different color balance. The 60D has much cooler (more blue) tones than the 6D with the TC and the same lens. I checked the Picture Style on both bodies. The 60D is on Standard (3,0,0,0) and the 6D is on Auto (3,0,0,0) which has the same values as Standard (3,0,0,0) I don't know what "auto" does vs. "standard" since the numbers are identical. I had to reduce the 6D files size in Photoshop to 60% so both files were 2.7 MB. I cannot imagine the 1.4X TC III is the cause of the tonal differences. I have never adjusted Picture Style settings on either body. Any suggestions as to what's going on here would be appreciated.
Color balance 6D vs. 60D
- Thread starter nc0b
- Start date
What happens if you put them both on Standard? I suspect "Auto" is doing something more than Standard, hence the difference. Testing should tell.
Upvote
0
Both shot at ISO 400 and with auto white balance. Shutter speed above 1/1000. I'll do another test with both on "standard" to see if that is the dominant factor. As far as the original purpose of the test, about half of my shots with the TC on the 6D (manual focus) were sharp. Would definitely prefer a body that could handle f/8 AF with center point if I was going to use the TC on the 5.6 lens very often.
Upvote
0
nc0b said:
That would most likely be "it", do check if the cameras chose the same wb setting on both shots - and for further comparisons, set a manual wb on camera.
nc0b said:
Probably also the result of different jpeg engine tunings of the cameras, I don't think jpegs are supposed to give identical results across cameras - better try raw, the raw converters (DPP, ACR, ...) might be give more consistent.
In any case, crop vs. ff, ancient 7d1-design 18mp vs. newest ff 6d are hardly to perform the same color-wise. The picstyles aren't meant to act like icc profiles or color calibrations, but just give different types of "pop" to your sooc shots. There's a reason ppl go to lengths with color cards and tuned workflows to get the actual colors of a scene.
Upvote
0
'Standard' picture style on the 6D produces a warmer, slightly richer colour than 'standard' on the 5DII. I presume Canon are continually tweaking these profile to produce a better OOC JPEG. Certainly that's one area where there has been continual improvement.
Sorry that should have been 'neutral' picture style. I very rarely use 'standard'.
Sorry that should have been 'neutral' picture style. I very rarely use 'standard'.
Upvote
0
beforeEos Camaras said:
No it won't, not unless you trick it into believing it's not f8 with taped pins or a Kenko tc.
Upvote
0
I just shot some comparisons with the 6D on picture style on "Auto" vs. "Standard". There is quite a difference with more chroma and a warmer hue on "Auto" than on "Standard". "Landscape" has a modestly bluer sky, but somewhat less chroma. "Neutral" is really bland, with "Faithful" a bit more vibrant than "Neutral", which I like the least.
Auto may exaggerate the chroma, but I prefer the overall appearance of the scene. The "Landscape" picture style option would be my second choice, particularly if one likes bluer skies. These subjective comments are made using an HP DreamColor Z24x display and nVidia Quadro K2200 video card using a DisplayPort connection.
As far as an option to use live view, I have never tried that with wildlife, particularly if they are on the move. The 60D with the bare 400mm f/5.6 focuses fine any way. It was the 6D, 1.4X TC III and the 400mm that won't AF. Even if a given body could AF with an f/8 lens combination, could live view track a running antelope or a flying raptor?
By the way, I get much better results with the 6D & 400mm with BIF than the 60D using the center focus point in either case.
Auto may exaggerate the chroma, but I prefer the overall appearance of the scene. The "Landscape" picture style option would be my second choice, particularly if one likes bluer skies. These subjective comments are made using an HP DreamColor Z24x display and nVidia Quadro K2200 video card using a DisplayPort connection.
As far as an option to use live view, I have never tried that with wildlife, particularly if they are on the move. The 60D with the bare 400mm f/5.6 focuses fine any way. It was the 6D, 1.4X TC III and the 400mm that won't AF. Even if a given body could AF with an f/8 lens combination, could live view track a running antelope or a flying raptor?
By the way, I get much better results with the 6D & 400mm with BIF than the 60D using the center focus point in either case.
Upvote
0
ok did not know that thought the differences was just a few more focus points and lens adjustment and the duel pixel thing for moviesMarsu42 said:
Upvote
0
Marsu42 said:
Are you sure?
Most of the current Canon cameras will AF at f/8 in liveview using contrast detection. They won't when using phase detect.
Upvote
0
nc0b said:
They are very different, the numbers are the same but that is additional tuning, like the bass and treble on a stereo that also has a graphic equaliser. The equaliser sets the overall tone for the style of music, you can then tune a particular album with the bass and treble. Standard and Auto are different base (equaliser) settings, you then get the Sharpness, Contrast, Saturation, and Color Tone to further customise from that base setting.
AWB is another killer. To compare true sensor differences you need to take two RAW files and put them into PS using the same settings in ACR.
Upvote
0
Mt Spokane Photography said:
Right, of course the 60d works in Live View - I though that was a given so my statement only was for phase af. Contrast af doesn't depend on the aperture, but simply on detecting ... well ... contrasts, so it fails only if the image drowns in noise.
Upvote
0
nc0b said:
If you shot both on AWB, then that right there can result in differences. Changes in light over even a few seconds can have an impact on color balance with AWB. It is also likely that both cameras actually have different tone curves in the picture styles, even though they may be the same two picture styles. To get a more consistent, comprable result, use a fixed WB setting (say daylight), use the same picture style (personally I use Neutral), take RAW images with each camera, and compare them at the same WB setting in ACR or LR. Then you should be able to get a better idea of what the differences in each camera are when it comes to color.
It should also be noted that color with cameras these days is more a matter of mathematics than sensor technology. The technology does play a role, and it is certainly nice to get better color strait out of the camera, but in the end, color is what you make of it with sliders and curves in post. With a program like LR, you can create user profiles to automatically apply certain settings to every image imported off of any camera. So you can tweak the color of a few images from each of your cameras, generate a user profile when you have the default settings you want for each one, and apply those profiles when you import. Your color would not only be consistent camera-to-camera, but it would also be YOUR color, every time, automatically.
Upvote
0
jrista said:
Generally I don't disagree, although I have a feeling that color quality out of cam might be the next rising DR discussion thing. At least I hope so. The hunt for good looking files at ISO 25 600+ has made reviewers out there color blind (no pun). Thinner CFAs on the sensors let more light in, but the color accuracy suffers. Most people probably don't think about it since the profiles today are poppy, contrasty and make most images fly with very little effort. But anyone with focus on color and especially those with older Canon generation experience will tell you that color and the subtle "look" has gone downhill in the Canon department lately. If it matters or not is very personal of course.
Needed or not, the upcoming 5Ds/r seems like an improvement in that area though.
Upvote
0
memoriaphoto said:
What utter drivel.
There is no such thing as "colour straight out of camera" the information contained in a RAW file is colour agnostic, it isn't until you demosaic it and render it that the rendering program assigns colour values to the pixels. Anybody that thinks "those with older Canon generation experience will tell you that color and the subtle "look" has gone downhill" clearly needs to learn how to make camera profiles, which takes about 15 seconds in total with a couple of seconds at capture time.
Colours are what you make them, you can tune your software to reproduce colours exactly as you want them and different camera models and brands (and lenses) can be used interchangeably with consistent colours.
Upvote
0
privatebydesign said:
I didn't say that you can't tune colors to whatever...I said that different models have different output - Lightroom, Aperture, C1 whatever. A custom profile is a good start but it doesn't change how the sensor/CFA handles the light. Newer models have less steep filters in order to let more light on the sensor. And, depending on how, what and where you shoot, this could be negative.
DXO actually has a section for this. And it is quite obvoius that newer Canons are not performing as well as older ones. Which many confirm in real life. However, as I said....most people probably don't care and it might be personal depending on what you shoot. But it doesn't mean that it isn't true...or, as you so kindly put it yourself..."utter drivel".
Upvote
0
memoriaphoto said:
This is what a RAW image looks like strait out of camera:
A bayer matrix. It doesn't really have "color"...to have color, it has to be demosaiced. Demosaicing, however, doesn't give you what most people think of as "strait out of camera" color either:
THAT is true "strait out of camera". GREEN. Twice as many green pixels as red or blue result in a very green image when all you do is demosaic and give the demosaicing algorithm a CFA pattern (RGGB in this case).
What do you do to fix that? Run the data through a bunch of complex mathematical algorithms. The whole entire notion of strait out of camera color or strait out of camera quality is pretty much a misnomer. The biggest camera factor in IQ is noise. The amount and characteristic of the electronic noise in a camera DOES play a role in IQ, but when it comes to color...that's pretty much all math.
This image required both demosaicing with a pre-made camera profile in Lightroom, which applies a complex multi-channel tone curve to the image. I then further processed color using saturation, vibrancy, and per-color channel sliders to get the color I wanted.
There is no such thing as out of camera quality. It's all an illusion produced by a bunch of math to hide the true nature of what data really comes strait out of the camera.
Upvote
0
"This is what a RAW image looks like strait out of camera"
It isn't, it has far too much luminosity, a 'naked' RAW file is much darker, your file has had a gamma curve applied whereas RAW files are linear gamma 1.0. I do agree with what you are saying re color though.
It isn't, it has far too much luminosity, a 'naked' RAW file is much darker, your file has had a gamma curve applied whereas RAW files are linear gamma 1.0. I do agree with what you are saying re color though.
Upvote
0
privatebydesign said:
True, I did "stretch" the data, as it was nearly black. It was not a gamma curve, however...I just manually pulled up the data until you could see the bayer pattern. PixInsight, the tool I used to process, doesn't apply ANY modifications to the data on it's own...all edits have to be made by the user.
This is the original RAW data:
No stretching, no gamma curve, no processing of any kind. Raw bayer matrix loaded without even any color channel data implied to the CFA.
Upvote
0
Similar threads
- Contains 1 staff post(s)
- Contains 1 staff post(s)
- Contains 1 staff post(s)