Just got my new 7D yesterday. So far, I'm very happy with the first bird pics I've taken. Now comes the learning curve. Any tips??
Printable View
Just got my new 7D yesterday. So far, I'm very happy with the first bird pics I've taken. Now comes the learning curve. Any tips??
I really got a lot out of the article BPM member Doug Brown wrote about setting for the 7D for bird photography. You can read it at
http://www.deepgreenphotography.com/...-new-canon-7d/
And or my 7D User's Guide here. In addition, Dan Cadieux is working on an article that will be published both on my blog and at BPN detailing how he makes such clean images with his 7D while others are complaining about image quality... I am sure that a lot of it will deal with exposing to the right. See here for more on that.
I would also add never globally sharpen a 7D file even when using capturing sharpening during conversion. I just got done testing a 7D for a few days that I had the option of buying and while I decided to keep my 1D III the 7D really is a fine camera and I had no real issue with noise as long as the the files are properly exposed and handled in post Here is an image I posted made with a 7D that shows how clean the files can be this was shot at 400 ISO http://www.birdphotographers.net/for...7104-Decisions
The 7D files tend to be noisy. You can apply some basic rules;
-try to keep the ISO below or equal 400 (I have seen some excellent pictures taken at ISO 800 and 1250. It would be nice to know how they managed to)
-try to push the histogram to the right as much as it is possible
-use DPP as a raw converter. It does a good job in keeping the noise level down
-use a noise reduction plug in. There are several good ones out there.
Good shooting.
Karl Egressy
Karl, If what you say is true, why do Dan Cadiuex's 7D images always look great and never look noisy?Quote:
Originally Posted by Karl Egressy
This winter it seemed like my default ISO was at 800, and I've been very happy with it. Even this past weekend it was ISO 800 all the way and I wasn't even bummed about having to use it. I'm a bit puzzled at what I often see as people saying as long as it is not set above 400 then you are OK. Exposing to the right helps, especially with darker tones. At "high" ISO I apply NR to the background only (Noise Ninja PS plug-in for me), and then sharpen the whole image. P.S. I use LR3 for raw conversion.
More to come soon......:S3:
Hi Artie,
I was talking about what comes out of the camera.
In comparison to Canon 1D Mark III I used before and in comparison to Canon 1D Mark IV I use now, it is noisy.
I can deal with it since I was PP. hundreds and thousands of my wife's pictures and a few of mine coming out of 7D.
I have the second copy of the 7D now, but I use it during my trips as a backup almost exclusively.
We had many discussions about it and the general concesus has been that the 7D is noisy above ISO 400.
(I'm mostly participating in NSN)
Having said that, there are shooters whom I admire for their work, one of them is Daniel the other is Glenn Bartley, who shoot with 7D and always come up with excellent pictures.
I hope it is clarified my hones intention in trying to give some basic advice to someone who is new to 7D so he won't be scared when he sees some more nise before PP as I was and returned my first copy.
Best Regards,
Karl.
I had a 7D for 2 year, and now I have a 1D IV, 80% of my shots in the 7D where taken at ISO 800 without complain, dark areas a lit bit more noise, but not a big problem. I never needed to use any NR.
My default for shooting my 7D with my 500mm f/4L IS seems to be ISO 800. I'm hand holding the 500mm, so I usually try to get a little faster SS and still stop down to f/7.1 or f/8 and ETTR. Occasionally I'll have an abundance of light and shoot at 200 or 400, but ISO 800 is really my starting point. I shoot RAW and use DxO Optics Pro for RAW conversion.
I have realised that when many people talk about noise, some are talking about noise at 100% view, some are talking about noise in print, some about noise in resized images, some about noise in underexposed images, some about noise in dark areas, some are talking about noise before doing NR, some about noise after NR. Everyone's tolerance is different. Hence, we get so many different opinions. If one were to post a 100% crop and do a poll with simple yes-no to 'do you find this noise objectionable?', we can find out. And the same poll can be done on the same shot that is processed and resized. Then, everyone would be talking about the same thing. Otherwise, one person is talking about apples and the other about oranges.
I dont have a 7D...but on my 40D, I dont like the noise even "at ISO 640 in properly exposed-to-the-right images at 100% view". But I dont find it an issue after I resize for web or make prints. But I dont like to crop those images a lot. Net net, if I have a properly exposed shot with high pixels-per-duck at iso 640 or iso 800, I like it....not the 100% view...but what I get after resizing and cleaning BG noise.
Excellent points well made KD. In addition, with the 7D, what some folks are seeing as noise, in the sky for example. are actually just those tiny pixels...
Thx Artie. You hear some folks say 'I dont find the noise objectionable at ISO 1600 in 7D images'...and then you see someone else say 'I absolutely hate the noise I see at as low an ISO as 400 in 7D images'. I refuse to believe they are talking about the same thing. I understand everyone's preference is different and all...but still.
I have to agree after following my usual post processing routine i find little difference between the files from my current 1D III my old 40D and the 7D files up to 400 ISO and both the 7D and 1D are better at 800 and higherQuote:
Originally Posted by Kaustubh Deshpande
Very intelligent summation of the whole, Kaustubh.
I want to mention only one more thing that came to my mind which is actually part of you summary but a bit embedded.
The way you see noise or the lack of it also depends on;
a. the quality of you monitor you use
b. the quality of your vision which usually depends on your age
Karl, agree with the monitor. If pixels-per-inch is high, you are going to see less noise and better sharpness. good example would be an iphone.
This is a big problem with internet( not so much on this forum though, I must say. Out here, we usually add a lot of verbage)...someone makes a blanket statement and someone else counters it. Both might be talking about different things and both might be right...but arguments go on without details.
You can say the same about another controversial topic...AF. The person who says 'AF sucks' might be talking about focussing on a least sandpiper flying in front of reeds with center point. The person sho says 'AF rocks' might be talking about focussing on a pelican in front of blue sky with all AF points active....but they dont mention that :-) I much prefer the discussions we have here where everyone expands a bit on what they are trying to say.
I have been using the 7D for two years now and only complain about the noise when I am working on processing images taken in low light. In good light the 7D performs very well. Here is something I took last summer at ISO1250 in the evening, http://jamiedouglasphotography.com/p...ff1bb#h39ff1bb . Great IQ and noise levels if you make sure to expose to the right. You will enjoy the 7D Darrell and make sure you check out the links posted above.
I am happy shooter with 7D.
Earlier even I used to have issue with noise at ISO 800 and above but once I understood how exposing to right and some correct PP can help I am doing pretty well.
If Noise is an issue I highly recommend Arash's High ISO noise reduction guide.
I just have to say, the 7D is my first Dslr and though I can't compare it to anything else I absolutely love it. Like others mentioned here I think alot of folks here, amatuer, pro-amatuer to pro's are testimonies to not only the 7D but that the camera is only a part of it. I think Glenn Bartley & Daniel Cadieux is shining examples of great skill behind a great Pro-sumer camera.
Doug Brown's article for setting up the 7D for BIF was an insightfull article as well.
I highly recommend Arthur Morris's 7D User's Guide. It explained to me exactly how he set's up the camera and why, covering every feature of the camera. It was a fantastic introduction to the camera for me, that along with Birds As Art I&II and the bullitens & blogs it's been a wealth of knowledge & a blessing to me. Heck I never even used a camera with AF until I got the 7D. I used a Minolta X-700! How to use the varying focusing modes, exposure modes, metering modes, how to read a histogram to evaluate exposure, highlight alerts, ETTR, & all the whens & whys for all the above. Thank you Artie!! I can't take your workshops but you've been my teacher all along Sir. I'm learning to create images like the bird photographer "I always wanted to be when I grow up!"
As far as ISO issues, for me it's mostly from user error, bad exposures. For me, I find now that I'm shooting mostly in manual mode my exposures are alot more consistant.
PS - I'm really looking forward to reading Daniel's upcoming article. I also use the 7D & the 100-400IS L lens and (although I don't mean to sound like a giddy teenager meeting Justin Bieber for the first time), I am very inspired by the work Daniel consistanly produces.
Thanks.
I have been using Topaz Labs DeNoise5 to clean up 7D images and like it so much, I sold my Mark IV. I currently use (2) 7D's for most of work although I now have the 5D Mark III and have been experimenting with it as well. I also like the 7D because it is a light-weight camera. Hiking with lots of heavy-weight equipment is not any fun and now with the new (soon-to-be) Canon 600mm II lens, the 7D and 600mm II lens should make an excellent, light-weight combination for BIF.
My goal for this past year is to reduce equipment weight and the 7D is part of the equation. I really don't miss my Mark IV other than its slightly faster fps.
Here's a low-light, in-flight shot of a Great Blue Heron, ISO 800, 7D:
http://iwishicouldfly.com/iwishicoul...685_small2.jpg
Alan
www.iwishicoudlfly.com
And looking to the future Alan- a new 7D running at say 24mp with better noise characteristics than the present one, and the 1Dx AF system. Stick a relatively light 300/2.8 on and you have even more portability!
Well...maybe later this year? That would be a nice combo to be sure!Quote:
Originally Posted by John Chardine
Alan
I doubt we'll get such high pixel density, but the computing power of the Digic 5 for improved moire and noise handling and the new AF system will be enough for me. :c3:Quote:
Originally Posted by John Chardine
<embed id="application/x-exifeverywhere" type="application/x-exifeverywhere" width="0" height="0">I didn't use my 7D a whole lot, but ISO 800 was never a problem for it. Not Mark IV quality to be sure, but definitely usable. Even 1600 is decent with a good exposure. Here's a 100% crop of an ISO 1600 frame with LR default sharpening and NR, but no additional processing.
Attachment 111635
I would be very happy with a 1-stop ISO improvement in the 7D along with the new 5D3 AF system. However given the 5D3's image quality compared to the 5D2 and 1DIV it doesn't seem like Canon is making much improvement in technology. I'm not sure they could get 1-stop of noise improvement with the current pixel density of the 7D.Quote:
Originally Posted by David Stephens
Quote:
Originally Posted by Doug Schurman
I've written quite a bit about this subject, so I'll not go too deeply into it here. There is a trade in pixel size and detail on the subjectand the light available with a given lens. Just like adding a TC to get more reach and pixels on the subject, Decreasing the pixel size does the same thing, but without the added TC optics. So asking for better ISO improvement is like asking for better ISO improvement with TCs. The pixel efficiency of the 7D (e.g. photons captured per square mm on the sensor) is equal to or slightly higher than the 1DIV, and higher than a 5DII.
Some recent threads: comparison of the same subject with 1DIV, 5DII and 7D as a function of ISO.
http://www.birdphotographers.net/for...-Little-pixles
Don't like the high ISO noise? Simply average pixels to make effectively a larger pixel (example given on the above page).
One other reference:
http://www.clarkvision.com/articles/...m.performance/
I have 3 DSLRs that I use regularly: 1DIV, 5DII, and 7D. On my next African safari, I'll take the 1DIV and 7D.
Roger
Roger, I think I have read about every post you have made on this subject in the last year or more. Although I'm sure I did not fully understand everything you wrote.
What I'm suggesting is that we all expect technology to improve over time. For example, the 7D has almost double the pixels of the 40D and also handles noise better. The 5D2 is similar compared to the 5DC (original). In seeing those advances I think I lot of people were hoping for the 5D3 to be noticeably better in noise handling than the 5D2 since it have been over 3 years. Some might argue there is some noise improvement but I think most say it is not that significant especially below ISO 6400.
The question is has Canon reached a plateau on how much they can make improvements on gathering light versus amount of noise (I'm sure there is a more technical way to say this). Some seem to think Nikon is more advanced in this area.
Anyway, everyone seems to want less noise on greater pixel density. My question to you is do you think there is more room for improvement?
Thanks
Doug
Hi Doug,Quote:
Originally Posted by Doug Schurman
There has actually been little improvemnt in the sensor sensitivity (quantum efficiency) over the years. Most improvements have come from fill-factor and micro lense improvments to gather more light. From the Canon 10D 5D, 1D Mark II eres to today, the measured improvemtn in light gathering is just a little over 2 (one stop). But fill factors are effectively not about 100% so no more improvement is possible there. What has improved, and probably what makes most people see as improved high ISO performance is the read nosie and fixed pattern noise which has improved perhaps 4x (2 stops) over the last decade. Read noise likely won't get any lower (it is already uner an astonishingly low 2 electrons. Canon can certainly make improvements in fixed pattern noise as they are behind Nikon in that regard. The latest Nikon DSLRs show litle to no apparent fixed pattern noise from the samples I've seen. The fact that canon's latest, the 5DIII did not significantly improve the fixed pattern noise over the 5DII is a big disappointment to me. So yes, we can expect/hope to see improvements, but it is likely only in the low end fixed pattern noise, which is pretty godd compared to a decade ago. The 1DIV has very very little fixed pattern noise, and that is the reason I choose it for low light wotj over a 5DII.
What I'm surprised we haven't seen is special high end DSLR-s with back-illuminated sensors that would be a nice 2x sensitivity improvement over the latest sensors. Back-side -illuminated sensors are being produced in small sensor P&S cameras. In a full frame DSLR it could add a couple of thousand $ to the price, but is possible.
But for any given imaging scenario, it is the lens + camer, working as a system, that determines noise in the final image. If we equalize the system: the lens aperture, the number of pixels on the subject, and the exposure time, there is no advantage to larger pixel cameras at higher ISOs (400 and above). (Ignoring fixed pattern noise issues.) The 7D performs very very well compared to larger pixel cameras whenthe lens + camera system (the Etendue) is equalized.
Roger
I see I had quite a few typos which make it hard to understand. For a few days now, BPN is flagging every work in my posts as misspelled, which makes it harder to spot the real misspelled ones.Quote:
Originally Posted by Roger Clark
So here is a repost with (I hope) the typos fixed. Plus I added a bit.
There has actually been little improvement in the sensor sensitivity (quantum efficiency) over the years. Most improvements have come from fill-factor and micro lens improvements to gather more light. From the Canon 10D 5D, 1D Mark II eras to today, the measured improvement in light gathering is just a little over 2 (one stop). But fill factors are effectively now about 100% so no more improvement is possible there. What has improved, and probably what makes most people see as improved high ISO performance is the read noise and fixed pattern noise which has improved perhaps 4x (2 stops) over the last decade. Read noise likely won't get any lower (it is already under an astonishingly low 2 electrons). Canon can certainly make improvements in fixed pattern noise as they are behind Nikon in that regard. The latest Nikon DSLRs show little to no apparent fixed pattern noise from the samples I've seen. The fact that Canon's latest, the 5DIII, did not significantly improve the fixed pattern noise over the 5DII is a big disappointment to me. So yes, we can expect/hope to see improvements, but it is likely only in the low end fixed pattern noise, which is pretty good compared to a decade ago. The 1DIV has very very little fixed pattern noise, and that is the reason I choose it for low light work over a 5DII.
What I'm surprised we haven't seen is special high end DSLR-s with back-illuminated sensors that would be a nice 2x sensitivity improvement over the latest sensors. Back-side-illuminated sensors are being produced in small sensor P&S cameras. In a full frame DSLR it could add a couple of thousand $ to the price, but is possible.
But for any given imaging scenario, it is the lens + camera, working as a system, that determines noise in the final image. If we equalize the system: the lens aperture, the number of pixels on the subject, and the exposure time, there is no advantage to larger pixel cameras at higher ISOs (400 and above). (Ignoring fixed pattern noise issues.) The 7D performs very very well compared to larger pixel cameras when the lens + camera system (the Etendue) is equalized.
One improvement we should see soon (I hope) is 16-bit A/D converters. Then with the elimination of almost all fixed pattern noise, we can choose ISO in post processing, or at least only have to choose a low versus high ISO in the camera, then refine the gain (ISO) in post processing. Scientific sensors (including systems being sold to amateur astronomers) have been 16-bit back-side illuminated for many years. It is time for consumer and pro cameras to catch up.
Roger
Let's see if I understand this with a practical example. Assume the ratio of pixel pitches between a 1D Mark IV and a 7D is such that from the same distance a 7D with a 400mm lens puts the same number of pixels a subject as a 1D Mark IV with a 500mm lens. (I think that's true within a few percent; but if I'm misremembering it, assume it for the sake of argument.) Furthermore they do this at the same aperture. (True? I'm not sure about that.)Quote:
Originally Posted by Roger Clark
Given that assumption, and assuming proper focus, there's no major image quality difference between the two setups at ISO 400 or above.
The 7D/400mm would be lighter; and the 1D Mark IV/500mm would have better autofocus so pick your preference. However a theoretical 7D Mark II that autofocused as well as a 1D Mark IV would eliminate this distinction. Furthermore, one of the 500mm or 400mm lens models is likely to be a better manufactured, sharper lens (better MTF chart) so that might also affect your decision; but there's no fundamental reason a 500mm or 400mm would necessarily be the winner here.
Second question: why does this change below ISO 400?
A key factor in this comparison is the lens aperture (the diameter, not f/stop). Modern use of aperture as an f/ratio is leading to more confusion. The area of the lens determines how much light is being collected. So in your example, you specify a 400 and 500 mm lens. If the lens diameter is the same (lets call it 71 mm) then the amount of light collected is the same. By your definition, the pixel size (1DIV+500mm and 7D+400mm) is the same, so the light per pixel is the same for a given exposure time, the number of pixels on subject is the same, the photon noise is the same, and the depth of field is the same. The two cameras produce exactly the same image (if their electronics are equally good) where the sensors cover the same area. Even the diffraction is exactly the same.Quote:
Originally Posted by Elliotte Rusty Harold
Correct. And if the electronics are equal, at any ISO.Quote:
Originally Posted by Elliotte Rusty Harold
Correct, if the lenses were equal in quality, the AF systems and other electronics are equal in quality, then there is no advantage to either system concerning fine detail, noise, depth of field, and pixels on subject given the same exposure time at any ISO.Quote:
Originally Posted by Elliotte Rusty Harold
The 7D electronics are not a good as those in the 1DIV, and the 7D suffers from some banding issues, especially at ISOs below about 400. That is why I specified above 400.
In the example above with 71mm aperture, 400 mm with 71 mm aperture is f/5.6 and 71mm aperture on 500 mm FL is f/7. So yes the 400 f/5.6 lens would be lighter and cost less. The 500 mm f/4 lens would have the advantage of opening up for more light, to 125 mm aperture, but then the comparison is not fair. One could counter with a 400 mm f/2.8 lens on the 7D, and the aperture jumps to 143 mm, so would give more light. Aperture is key, to not only delivering more light, but also better contrast in fine detail because diffraction is less (assuming other aberrations do not swamp the improvement due to aperture).
Hope this helps.
See also this extensive thread from earlier this year:
7D-or-1DIV-better-noise?
http://www.birdphotographers.net/for...noise?p=759867
I posted examples where the systems were equalized and the images are very close. Arash posted an example of the 7D fixed pattern noise.
Roger
Roger - So, if (as I think you say), for example, I have a 200mm lens with diameter of 80mm and a 400mm lines with diameter of 80 mm, they both transmit same number of photons to sensor, right? So I can shoot both wide open at the same ISO and shutter and get same image exposure. ("Exposure", wrong/bad word here?) Yet the wide open 200mm is an f2.8 while the 400mm is an f4 ( or thereabouts). But the 400mm will give the image one stop less exposure, a darker image than the 200mm. I'm lost, help!!!
Tom
Hi Tom,Quote:
Originally Posted by Tom Graham
There are many confusing things, including what is exposure. You talked about the lens, but one must equalize the field of view of a pixel too so that the systems (lens + sensor) are see the same amount of light and the same detail. So on the camera attached to the 400 mm lens, the pixels must be twice the size of the pixels on the camera attached to the 200 mm lens. Then the pixels in each camera see the same field of view (same detail on the subject). When they see the same detail on the subject with the same lens diameter, both cameras record the same amount of light. But the camera's reported exposure will be different because exposure, the way it is defined by camera manufacturers and the ISO standard, has confused the issue making ISO relative to the full capacity of the pixel. It is like two different sized buckets of water. They can hold the same amount of water, but the larger bucket has the capacity to hold more. So instead of saying there are 2 liters of water in the bucket, they will say one bucket is full and the other bucket is half full. That is how ISO is defined, rather than absolute amount of light. When you specify exposure in absolute amount of light, e.g. number of photons, then one can see what is really going on and what the noise will be. Photography has evolved using ratios (f/ratio and ISO) which confuses what is really going on. A paradigm change is needed.
Roger
Think I got it, thanks Roger. It is "exposure" that is confusing me.
But how better can we do it, i.e. talking about ISO and aperture and shutter relationship, and yet keep it simple? We do not need a forth parameter - I hope.
Why did this not confuse us (me) in old film days, was film grain size equivalent to sensor pixel size?
Tom
Hi Tom,Quote:
Originally Posted by Tom Graham
Nothing has changed: select the largest lens you can, and use the longest exposure time you can with that lens and adjust the ISO to not overfill the data range in order to maximize signal-to-noise ratio. One only needs to calculate the system throughput if one wants to compare to another system. When we see people say camera A is noisier then camera B, we know it is incomplete because it doesn't include the lens + camera system.
Regarding film, the same physics applies with similar confusing factors which have been propagated into the digital era. For example, film speed. We had different film "speed" but the quantum efficiency was always the same. A photon was absorbed in a grain and created the excited stare that after developing turned the grain black. In a faster "speed" film, the grains were larger, so when the grain was turned black, it had a greater visual effect. It was really more like a gain factor much like ISO in a digital camera has a gain factor. But film didn't have different sized pixels, so digital adds a new variable, which is confusing in the old paradigm.
Roger
Thanks. I now see there's one thing in my "spherical cow" example that isn't actually true in practice. I was thinking the 400mm f/4 on the 7D/4.3μm pixel had the same "aperture" as the 500mm f/4 on the Mark IV/5.0μm pixel rig. However in fact the 500mm f/4 has a larger absolute aperture diameter than the 400mm f/4, even though they both have the same aperture ratio. Thus the 500mm lens will deliver more light to each Mark IV pixel than the 400mm lens will deliver to each 7D pixel; and therefore have a higher signal to noise ratio (or a lower exposure time?). Is that correct? Or have I missed something again?Quote:
Originally Posted by Roger Clark
Hi Elliotte,Quote:
Originally Posted by Elliotte Rusty Harold
Yes, you are correct.
Note the 1DIV has 5.7 micron pixels.
The field of view of a pixel on the 400mm +7D = 206265 * 0.0043 / 400 = 2.22 arc-seconds
The field of view of a pixel on the 500mm +1D4= 206265 * 0.0057 / 500 = 2.35 arc-seconds
so they are very close. (The 206265 factor is the number of arc-seconds in on radian.)
The ratio of light per pixel if the aperture diameters were the same in a given exposure is the angular area ratio of the pixels: 2.35*2.35 / (2.22*2.22) = 1.12, so the 1DIV would get 12% more photons per pixel and less pixels on subject by the same amount (total area). But with a 400 f/4 and a 500 f/4, the ratio of the area of the lens is 125*125/(100*100) = 1.56, so the lens delivers 1.56 times more light per pixel. in a given exposure these two factors multiply 1.56 * 1.12 = 1.75, so the 1DIV+500 f/4 gets 1.75 more light per pixel (with slightly less resolution on the subject).
A 400 f/3.2 on a 7D would match the light per pixel as a 500 f/4 on a 1DIV (within 12%). And a 377 mm f/3.0 lens on a 7D would exactly match the detail on a subject, light per pixel, and depth of field as a 500 f/4 on a 1DIV.
Roger
Roger,
About the new 500 and 600mm series II on a 1D IV, the 600 as largers lens diameter, but the MTF charts, show that the 500mm is a almoust perfect lens, in noise and picture quality, both in a 1D4, who will be the winner?
Romy Ocon has a pretty quick and easy guide. I use it and I never have had problems since. I love the 7D for birding. Actually the only camera in Canon mount I own.. :)
Detail is quite amazing and a very tough camera I think.
Hi Gary,Quote:
Originally Posted by gary1952
I tried to find Romy Ocon's write up on 7D.
I checked his web page but couln't find it. Could you please post a link to it?
Thanks.
