RobertLynn wrote:
255X3 is white, it's when there's no detail to recover that it equals blown.
Exactly - things are only unrecoverable when all three channels are at 255.
There's a lot of software out there that can recover detail from situations where not all channels are blown - and some can extrapolate and "recreate" data for the blown channel from the information remaining in the channels that haven't blown.
Even the free Raw Therapee can do this - very well too, as it happens.
Shooting with this in mind is a perfectly valid way to maximise DR, and unless all three channels are completely cooked or as near as dammit, you haven't got "blown highlights".
keithreeder wrote:
I agree - they're very nice images. They certainly don't show - or prove - any problem with your 7D...
Thank you for the nice comment-
The photos demonstrate my point. The 7D works well in low light.( my definition) lol
I shoot in natural light using fill when needed. When it is bright MY camera did not produce photos to my standards. If I spend enough time in post I can get close but WHY?
I prefer files from my 20 30 & 50D over the stuff I got from my 7D under bright conditions. To be fair I normally do not shoot when it gets bright but when testing new gear I shoot under extreme light.
I complain about color & blown highlights in bright light & some post examples of how wrong I am but the examples are shot in moderate light & have blown highlights as posted.
Stumped wrote:
...
When it is bright MY camera did not produce photos to my standards. If I spend enough time in post I can get close but WHY?
I prefer files from my 20 30 & 50D over the stuff I got from my 7D under bright conditions.
Technically speaking, what could cause this? Is there something different in the 7D vs. the others cameras mentioned that would theoretically change bright light performance? Would the new exposure metering that now uses color as one of the inputs be a possibility?
dwweiche wrote:
Technically speaking, what could cause this? Is there something different in the 7D vs. the others cameras mentioned that would theoretically change bright light performance? Would the new exposure metering that now uses color as one of the inputs be a possibility?
Quite possibly a metering issue and the camera may intrinsically overexpose.
I can control blown highlights with camera adjustments but something has changed not to my liking. What I saw was brighter highlights(blown), darker shadows & poor color? I can view web photos & pick the 7D from other cameras by seeing the very thing I did not like about my files. Again this pertains to stuff shot in bright light.
Much of this is subjective so YMMV.
Guys I do not want to rain on anyone’s parade.
The 7D is a great camera for most. It is a big improvement over previous models.
In low & moderate light it excels. The new focus system is excellent.
I love the features but it did not work for me.
Stumped wrote:
I can control blown highlights with camera adjustments but something has changed not to my liking. What I saw was brighter highlights(blown), darker shadows & poor color? I can view web photos & pick the 7D from other cameras by seeing the very thing I did not like about my files. Again this pertains to stuff shot in bright light.
Much of this is subjective so YMMV.
Guys I do not want to rain on anyone’s parade.
The 7D is a great camera for most. It is a big improvement over previous models.
In low & moderate light it excels. The new focus system is excellent.
I love the features but it did not work for me.
I'd be interested to see if you could really tell the difference between two web photos from the different cameras, and pick which shot what. Sometimes with web files, I can't even tell the difference between my 5D and 7D, and I'm the one who shot them! haha.
I'm still confused as to where the color went awry, and citing a shot in an arena filled with compact fluorescents isn't telling me anything, because like i said, I can reproduce that with any camera I own/owned. Even the 5D which is well documented as having a pleasing reproduction of color can do it.
I may go out and shoot some photos with both cameras, same conditions (outdoors, bright light), and post them. I've done multiple online galleries between both, and I don't see sucky color coming from the 7D, and I'm also seeing your photos posted with the 7D, and I'm not seeing sucky color there.
As for blown highlights, that's not the camera's fault. If you follow the meter, it's just trying to average the scene out (in eval. metering) to 18% grey.
If you blow the highlights with the 7D, you'll blow it with most other Canon cameras at the same setting.
The only variable I can see is the ISO sensitivity of the camera (meaning it's ISO is really like 125 by ASA/ISO standards), but AFAIK Canon cameras are constant in having the same rating.
just a note on comparing noise at the pixel level, and the use of the term "scientific":
the most basic question is: what is the response you are interested in maximizing and therefore measuring? If "IQ at the pixel level" is what is important to you, that's fine. However, it's not the response that most folks care about. Most folks will care more--and in most cases, exclusively--about the quality of the final product (and visible/objectionable noise contained therein). The final photo being, for most of us, the print or the image as viewed via another medium (screen, etc).
If you see a print, and you'd like to evaluate the amount and character of noise in that print, do you ask what the image resolution was (when shot? prior to printing? during printing?), so that you can pull the appropriately-powered loupe out of your pocket? Evaluation of the final image--of the response of interest--should not and cannot hinge on characteristics of the tool(s) used to create it. That would be extremely unscientific. (and it is shameful that respected sites would measure noise this way)
If the response of most interest to you is "noise in the final product", but you're employing noise-at-the-pixel as a proxy because it's (much) easier, logistically, to compare noise at that level, then you'll have to present a powerful argument to convince me/us it's a valid proxy. I am not well-versed on that subject, but I haven't read such an argument myself.
so to sum (hopefully) if you'd like to measure noise at the pixel level, then measuring noise at the pixel level is an appropriate method (duh), and quite scientific. However, if you'd like to predict noise of final output by measuring noise at the pixel (as a direct proxy), you are likely introducing an inexcusable bias into your analysis.
DXO Mark tests show superior 40D performance at high ISO with 18% gray target, viewed on screen. But tests on print are quite similar at high ISO, with edge to the 7D.
The real problem here is not the cameras, but the fact that there is no universally accepted method for measuring noise and making camera-to-camera comparisons.
Until those standards are decided upon, we will continue to have such arguments.
wilt wrote:
DXO Mark tests show superior 40D performance at high ISO with 18% gray target, viewed on screen. But tests on print are quite similar at high ISO, with edge to the 7D.
RobertLynn:
That's because the people at DxO do not know how to compare two different size sensors.
Robert, 7D and 40D have the same size of sensor. If you use DxO, you may always choose "print" option. I do not see anything wrong comparing sensors this way, do you?
mfurman wrote:
Robert, 7D and 40D have the same size of sensor. If you use DxO, you may always choose "print" option. I do not see anything wrong comparing sensors this way, do you?
You need some serious assurance, don't you?
Answer me this: use DXOMark to compare 7D vs 50D. Go through the 18%Gray, DR and color sensitivity pages. Look at BOTH screen (pixel level) and print (image level) tabs. Notice how closely they follow each other? Then look at the score page. The 7D gets ~ 800 for low light ISO and 50D ~ 700. Why? How do you explain this self-contradiction?
Even John Sheehy thinks there is something seriously wrong with DXOMark. See this:
"Something is wrong with the 7D used by DxO. It has far too much read noise at high ISOs. Almost every figure I've seen regarding 7D read noise at ISO 1600 averages somewhere around 19 ADU, and the 40D is around 22 or so. With 1.8x as many pixels, the image-level read noise of the 7D should be (19/22)(1.8^0.5) = .64 or about 2/3 stop more DR. The DxO "Print" for DR shows the same results as the 40D for high-ISO ... this doesn't make sense.
I've seen a lot of attempts over the years of people trying to push the 5D to ISOs like 50K, and the results are poor, compared to the 7D. What the 5D is good at, compared to APS-C cameras, is shot noise (but not as good as the 5D2 or other new FF cameras). As far as read noise is concerned, at the image level, the 5D is showing its age, and the banding at high pushes is strong."
mfurman wrote:
Robert, 7D and 40D have the same size of sensor. If you use DxO, you may always choose "print" option. I do not see anything wrong comparing sensors this way, do you?
I'm certainly aware the 7D and 40D have the same size sensor.
If you look at the final output of the cameras, there is no problem at all comparing them this way.
If you take the same picture, same settings, same evertything just different camera (40D/7D) take ALL of EACH camera's pixels, and do 2 things. First, print an image at say, 8x10 (one that each camera can easily do). Secondly, print an image at say, 16x20.
If you can't see the difference that the 7D clearly beats the 40D, there's one of 3 things going on.
A; You're blind.
B; Your 7D has issues.
C; You're blind and in denial.
Take those same two photos that we used for print. Now downsize the 7D file to the native size of the 40D file. Tell me which is cleaner/crisper. Now, take both of those files, and down them to a display resolution of 1500x1000. Again, the difference is there.
I have extensively used the 40D. I have processed my files every way that my peanut brain could figure out how to do. I know the limits of the 40D. The 7D picks up where the 40D left off and runs all the way down the field for the touch down, and is kicking the extra point as we speak.
I don't want ANYONE to think I'm bashing the 40D, 30D or 50D (or 20/10 for that matter). I'm simply stating what I've proven to be a fact. Again, another forum member was as so gracious to take one of my files, crop it, and upsize it to 16x20. The 7D held detail, and the 40D just lost all detail.
I just sold my 40D, and truth be told, if I didn't use full frame so much, I'd sell my 5D and buy another 7D.
roberto1979 wrote:
How did nerds argue before the internet?
I'm definitely not a nerd, as I don't really understand all of the technical aspects of sensels, pixels, sensors and density.
I know 1 thing, and that's the end product.
The 40D has made me more than the money it cost, but I needed to be able to go further, and the 7D looked better. I got it, and shot both sie by side when I owned them both (just shipped out the 40D on Thursday so I no longer have access to it) and the 7D beats it at the categories I look at.
High ISO noise (and in my opinion all ISO noise/control)
and OUTPUT.
roberto1979 wrote:
The fact that you've been arguing on the same topic for 7 days is pretty nerdy. I have both cameras. The both take great pictures. End of story.
If you don't like it, don't post, there's a hide me button if you don't like what I have to say. I certainly may use it. if you think the story ends at both taking nice pictures, that's your perogative. If you deny the 7D outperforms the 40D, you're just trying to see things the way you want.
