April 21, 2014, 09:14:29 AM

Show Posts

This section allows you to view all posts made by this member. Note that you can only see posts made in areas you currently have access to.


Messages - jrista

Pages: 1 ... 100 101 [102] 103 104 ... 217
1516
EOS Bodies / Re: Canon EOS 7D Mark II in 2014 [CR2]
« on: July 14, 2013, 01:09:51 AM »
I am reading (most of) these posts with interest. It may be that someone else has touched on this, since the number of posts is fairly high, but I´ll give it a go.

I am not an expert on all the technical aspects of the bodies we are discussing. I am first and foremost a photographer, trying to get the best images with what I have. Currently I use 1DX and 5DIII as my main bodies, but I also have 1DIV and 7D. And I am looking forward to the 7DII. But my main concern is AF and not IQ. For everything that stands still, I prefer to use FF bodies (a few thousand reasons have been posted before this one). I have lots of high quality glass to get the framing right, so my interest for a cropped camera is only for long reach.

As I see it, the real benefit of a 1.6x sensor, besides cheaper lenses, is in combination with a 400/500/600mm tele, possibly combined with extenders. The AF system will work on a closer framed image and thus give me a better opportunity to secure focus on the specific part I want in focus. When I crop an image from a FF, it´s more of a gamble. But currently, the AF systems on both the 1DX and the 5DIII are so much better than the current 7D, that you don´t really have an advantage with a cropped sensor. So at the moment my 7D stays in the bag (actually at home). But with the 7DII, that may all change.

So, I cross my fingers and hope for a best possible AF system and good high-ISO performance (rather than very high resolution) on the 7DII. I´d be interested if some of you more technically competent could comment on this.

/Eldar

Man you are reading my mind. Exactly same thing here. IQ at this point is last on my list. With AF being on top.
As others though am hoping for better IQ too?

I've made the same argument in the past. To me, both AF as well as frame rate come before sensor IQ. I don't see any real drawbacks to the 7D IQ as it is, though. I think my previous example posted a few responses up demonstrates that the 7D can extract more than enough detail to overcome any drawbacks such as higher noise (especially, as Eldar stated, when paired with a 400/500/600 and a TC).

I would go so far to state that one of the primary reasons the 1D X can do as well as the 7D is because of its AF system. The 7D's 19pt AF, while good, definitely has it's drawbacks. It's margin of error is rather large, and any time it decides to confirm focus, its really hit or miss...you might actually be missfocused enough that the shot (or even a whole sequence of shots) are a bust.

I think my top two biggest wants for the 7D II are 61pt AF and 10fps. Give me those, even with a lower noise 18mp APS-C sensor, and I'll be happy.

1517
Animal Kingdom / Re: Show your Bird Portraits
« on: July 12, 2013, 09:43:25 PM »
Snack time.

Wow...it looks like it got another bird of prey. Maybe a Kestrel?

1518
Animal Kingdom / Re: Show your Bird Portraits
« on: July 12, 2013, 08:30:13 PM »
Colorado Mountain Songbirds:






Click for larger, and to see more on my site

  • Canon 7D
  • Canon EF 600mm f/4 L IS II + 1.4x TC
  • Gitzo GT3532LS + Jobu Pro2 Gimbal

1519
This sensor is made from canon line 2 , APS line and compact camera line, nothing difficult with that


At the CMOS/electronics level, the 70D sensor is a 40mp sensor.  You'd better accept that.

It's hard to believe that Canon has managed to make a 40mp sensor on their old 500nm process.
But there are no clear clues at this time to confirm whether Canon has indeed switched to a new process.

We'll surely find out, as I'm sure Chipworks is eager to put this new sensor under their microscopes.


If Canon has used some kind of BSI process, I guess it's possible they stuck with 500nm. Even if they did that, I think it's still impressive...I don't think BSI has been used in a sensor as large as APS-C before. I do know that they actually have a patent for such a thing, however:

http://thenewcamera.com/canon-patent-back-illuminated-cmos-sensor/

1520
Landscape / Re: Beautiful sunsets
« on: July 12, 2013, 08:22:32 PM »
Big Bend National Park

Really love the first one! Awesome work.

1521
EOS Bodies / Re: Canon EOS 7D Mark II in 2014 [CR2]
« on: July 12, 2013, 06:29:20 PM »
Is Topaz much better than the NR in PS? I have never tried anything but PS.

It seems to be to me. I wouldn't say by a hugely significant margin, but it seems to do less damage to detail, and is definitely more configurable. It does noise reduction and detail recovery, as well as debanding in both horizontal and/or vertical. It has black level correction, as well as independent shadow, highlight, red, and blue channel fine tuning.

So far, I have not seen much need to really tweak much beyond the basic "RAW - light" or "RAW - moderate" presets. If, for whatever reason, I find detail loss in my subject to be too much, I simply mask off the subject in Photoshop, copy it to it's own layer in place, and run a preset on the base layer...and my critical subject detail remains 100% untouched. I rarely spend more than a couple minutes per image, five at most, getting results that to me seem to be ideal (something I never quite felt with just Photoshop or Lightroom noise removal.)

1522
EOS Bodies / Re: Canon EOS 7D Mark II in 2014 [CR2]
« on: July 12, 2013, 06:02:20 PM »
Since you insist on me demonstrating my own work...here is an example of how 7D noise can become a complete non-issue when using modern noise removal. Pre and Post Topaz DeNoise 5 for an ISO 800 shot that was accidentally underexposed by about 1 2/3rd stops, then lifted in post (so, roughly the equivalent of ISO 2500). Fine bird feather detail, at it's finest merely two pixels wide, is completely untouched. The background cleans up completely. Used one of the premade DeNoise profiles for medium noise removal:

Before:


After:


Final shot (Green-tailed Towhee, for anyone who's interested):



  • Canon 7D
  • Canon EF 600mm f/4 L IS II + 1.4x TC

1523
Then I have one answer, where are the high megapixel cameras?

The 70D has 40.4 million independent pixels. Each full bayer pixel is 4 microns in size, however each half pixel, which include independent readout, is a mere 2 microns in size.  Is that not high enough in terms of megapixels, pixel density, and readout wiring complexity for you?

nope
you have no data sheet about the read out noise etc from the sensor yet, it works in one way during AF and empty 2 channels/pixel regarding the read out S/N

Ignoring whatever the noise or S/N might be (I suspect it would be on part with the prior 18mp APS-C), that wasn't your argument...you were stating Canon hadn't demonstrated an ability to make small pixels. My point is that if Canon has produced a sensor with dual pixel FP-PDAF, then they have 40.4 million separate pixels, each with independent readout logic. That puts the height of each of those pixels at 4 microns, and the width at 2 microns. At 2 microns, that is smaller than any other APS-C sensor on the market so far. That demonstrates a fairly significant leap forward...over a factor of two smaller than Canon's previous smallest DSLR pixel size.

1524
Landscape / Re: Beautiful sunsets
« on: July 12, 2013, 05:45:02 PM »
south africa

Love that lightning! Great shot, well done with the silhouette.

1525
EOS Bodies / Re: Canon EOS 7D Mark II in 2014 [CR2]
« on: July 12, 2013, 05:15:17 PM »
Quote
" I'll see if I can muck with Roger Clarks images by upscaling the 5D II and 1D IV shots to 7D size to demonstrate my argument. "


I don't want an argument, I want your own real world images backing up your assertions.

Few people trust my opinion here and there is no reason they should, but I made a statement and backed it up with my own images. Many people here respect and trust Neuro's opinion, his experience tallies with mine.

Your same generation sensor pixel density meme is false unless you are bench testing, I don't, as a rule, frame and hang prints of bench tests.


If I had the capability to, I would. As I stated, I don't have  5DIII or 1D X in my possession right this minute. You can be as unreasonable as you want to, that's your prerogative. BTW, a LOT of my work IS the moon, and I find it to be an ideal subject to provide visual backing for an argument like this (can't get much more focal length limited than the moon). Soon as I have a 5D III in my possession however, I'll happily do some comparisons with birds as well.

I don't think I've ever posted any image here when trying to visually back up my claims that involved a technical bench test, any kind of test chart, etc. My examples have always been of my actual work (assuming I've had the capability of producing such.) I can't say I've ever framed an ISO 12233 chart sample either. ;P



So, not that it matters to you specifically, but here is a comparison of the 7D, 1D IV, and 5D II using Roger Clark's moon samples. The 5D II seems quite soft in comparison to the 7D. The 1D IV is also softer. There are nuances of fine detail that the 7D picks up that the other two blur over. Enlarged 400% for emphasis.


1526
EOS Bodies / Re: Canon EOS 7D Mark II in 2014 [CR2]
« on: July 12, 2013, 05:00:56 PM »
Now, do you have any example images that demonstrate your assertions or not?

I'd like to see that, as well.  But only as a comparison between APS-C and FF...not just how sharp a shot with the 7D can be.

I recall a couple of posts by AlanF, where he performed mathematical calculations to determine the theoretical resolution advantage of the 7D over the 6D, followed by shots of a feather with his 7D and 300/2.8L IS II + 2xIII, showing how the image approached the theoretical limit of resolution, concluding that the 7D must outresolve the 6D.

Then he got himself a 5DIII and compared the real world performance of the two.  He found that despite his earlier theoretical predictions and testing (testing only the 7D), the IQ of the cropped 5DIII image was equivalent to the 7D.  I found the same thing with my 1D X and 600 II, which is why I eventually sold the 7D (that, and the higher hit rate with the better AF system).

My conclusion matches that of PBD - when looking at real world images, the 'extra reach' of the APS-C sensor is an illusion, provided that the cropped image from the FF sensor gives you sufficient resolution for your desired output format.  Plus, that applies only at low ISO - at high ISO (>800), the IQ advantage of the cropped FF image progressively surpasses the APS-C image as the ISO increases.

At last, a voice of reason that many will respect. Thanks Neuro, I felt I was fighting a one man battle on occasions on this one.

As I have always said, there are very good reasons to own crop cameras, just not because you think you are getting a free TC.

Don't miss the critical qualification there:

My conclusion matches that of PBD - when looking at real world images, the 'extra reach' of the APS-C sensor is an illusion, provided that the cropped image from the FF sensor gives you sufficient resolution for your desired output format.

If all you ever do is downscale and drop your images online somewhere, then I have no argument. If you print large like I do, then I still disagree...the spatial resolution advantage of a cropped sensor is still valuable, even in the face of increased noise.

I would gladly do a comparison. I don't own a FF camera myself yet, as I've been waiting for an official 7D II announcement. As soon as I pick one up, or the next time I find a good reason to rent one, I'll provide as many visual comparisons as I can, at a range of ISO settings (because Neuro is definitely right that there is a threshold wherein the superior ISO performance of bigger pixels outweighs the spatial resolution advantage of smaller pixels.)

So you don't have any actual images to back that up? I did.

I print, my smallest print is 16"x24". I looked at the 7D specifically to print big and to give me more reach, it doesn't give you a focal length multiplier, enhancer, or anything else many seem to think it does. "Pixels on duck", when used in this context, is a fallacy.

Even at base iso when making big prints the 7D noise interferes with the detail, the 21MP FF doesn't have the ultimate detail, but it doesn't have the noise either.

Remember, I am not theorising here, I did the tests to see for myself.

Here is 200% crop of the APS-C and a 300%+ crop of the FF, no development processing just resizes to match pixels to each other. To my mind if you process the noise out of the APS-C you get the FF detail, or

Well, perhaps this is just me, but the hair is much sharper and better defined in the 7D enlargement there than the 1D enlargement. I'd print the 7D shot without any NR, as I prefer to have a bit of noise in photos I print anyway...avoids posterization. So I don't consider the NR softens detail argument to be an issue here.

Assuming I do apply NR, I always use Topaz DeNoise 5 these days. It takes about 10 seconds for most of my work, and automatically masks areas with detail beyond a certain threshold, where noise and fine detail are indistinguishable enough that noise is a non issue. Backgrounds, where Nero pointed out noise can be a real issue, clean up beautifully with one run through DeNoise. If there is any banding, it too can be removed, and a lot of DR recovered, using DeNoise as well, with practically no effort  and without affecting fine detail.

I won't disagree that at ISO 1600 and up, there isn't any comparing a 7D and 1D X. I'd take the 1D X every time. I figure the same would be mostly true with the 5D III as well. But with modern tools, I've found that noise on the 7D is no longer an issue, and that when removing noise, it no longer has to be the detail decimating process it used to be.

Last, I do have 20/10 vision with my glasses or contacts on. I am also plagued by hypersensitivity issues...hearing, sight, and at times touch (and most of the time, they are not a benefit...sharp vision is the one thing I consider a bonus). I guess it is entirely possible I am seeing something most people don't see. To my eyes, however, the 7D in all of your actual sample shots appears to be quite a bit sharper, and even more color saturated, than the 1Ds III.

1527
EOS Bodies / Re: Canon EOS 7D Mark II in 2014 [CR2]
« on: July 12, 2013, 04:40:34 PM »
Don't miss the critical qualification there:

My conclusion matches that of PBD - when looking at real world images, the 'extra reach' of the APS-C sensor is an illusion, provided that the cropped image from the FF sensor gives you sufficient resolution for your desired output format.

To clarify my "critical qualification," if you are printing at 16x24" or smaller, there is no difference. We aren't talking about 4x6" prints at Target. If you routinely print at 24x36", yes the APS-C sensor has an advantage, assuming the AF of the 7D is up to the task. But, that only applies at low ISO, i.e. in good light. Much of the time, my shots are not under those circumstances. 

@jrista - I encourage you to rent or borrow a 5DIII and compare it to the 7D head to head for yourself.

I print anywhere from 13x19" to 32x48", hence my long standing desire for pixel density. It is really more about that, than specifically crop factor (i.e. 47mp FF or 18mp APS-C, doesn't really matter to me, although the 47mp FF would be my pick for landscapes, obviously.) I'll see if I can muck with Roger Clarks images by upscaling the 5D II and 1D IV shots to 7D size to demonstrate my argument.

I intend to buy a 5D III soon enough, and if Canon doesn't announce a 7D II by fall, then I will.

1528
EOS Bodies / Re: Canon EOS 7D Mark II in 2014 [CR2]
« on: July 12, 2013, 03:28:07 PM »
Now, do you have any example images that demonstrate your assertions or not?


I'd like to see that, as well.  But only as a comparison between APS-C and FF...not just how sharp a shot with the 7D can be.

I recall a couple of posts by AlanF, where he performed mathematical calculations to determine the theoretical resolution advantage of the 7D over the 6D, followed by shots of a feather with his 7D and 300/2.8L IS II + 2xIII, showing how the image approached the theoretical limit of resolution, concluding that the 7D must outresolve the 6D.

Then he got himself a 5DIII and compared the real world performance of the two.  He found that despite his earlier theoretical predictions and testing (testing only the 7D), the IQ of the cropped 5DIII image was equivalent to the 7D.  I found the same thing with my 1D X and 600 II, which is why I eventually sold the 7D (that, and the higher hit rate with the better AF system).

My conclusion matches that of PBD - when looking at real world images, the 'extra reach' of the APS-C sensor is an illusion, provided that the cropped image from the FF sensor gives you sufficient resolution for your desired output format.  Plus, that applies only at low ISO - at high ISO (>800), the IQ advantage of the cropped FF image progressively surpasses the APS-C image as the ISO increases.


At last, a voice of reason that many will respect. Thanks Neuro, I felt I was fighting a one man battle on occasions on this one.

As I have always said, there are very good reasons to own crop cameras, just not because you think you are getting a free TC.


Don't miss the critical qualification there:

My conclusion matches that of PBD - when looking at real world images, the 'extra reach' of the APS-C sensor is an illusion, provided that the cropped image from the FF sensor gives you sufficient resolution for your desired output format.


If all you ever do is downscale and drop your images online somewhere, then I have no argument. If you print large like I do, then I still disagree...the spatial resolution advantage of a cropped sensor is still valuable, even in the face of increased noise.

I would gladly do a comparison. I don't own a FF camera myself yet, as I've been waiting for an official 7D II announcement. As soon as I pick one up, or the next time I find a good reason to rent one, I'll provide as many visual comparisons as I can, at a range of ISO settings (because Neuro is definitely right that there is a threshold wherein the superior ISO performance of bigger pixels outweighs the spatial resolution advantage of smaller pixels.)

In the mean time, I'll once again provide a link to the best visual evidence of the 7D's resolution advantage in a focal length limited situation (photographing the moon) performed by someone far more respected than myself:

http://clarkvision.com/articles/pixel.size.and.iso/index.html

I don't know how many more times I can post this link and have it be ignored, but it provides exactly the visual comparison, AT A RANGE OF ISO SETTINGS, that you've been asking for @privatebydesign. You seem to have conveniently ignored it the last several times I've linked it in relation to these kinds of discussions, both in this thread and others. I would really like to hear an actual response from you, as I don't know what better evidence you want than this.

1529
EOS Bodies / Re: Canon EOS 7D Mark II in 2014 [CR2]
« on: July 12, 2013, 03:23:46 PM »
...any serious or professional photographer who relies on actual real-time update and large, full-detail viewfinder won't be moving to EVF any time soon. There is just plain and simply no replacement for a prism.
Wouldn't a serious photographer know the benefits of a waist level finder and consider the prism a technical detail rather then a dogma? The viewfinder of a 1Dx feels rather tiny compared to a 645, and don't even think about anything large. And if one compares the mandatory ~40ms release lag to the less then 10ms you can get electronically...
Guess it will take some time for the pros who now shoot tethered and use the computers screen as VF to accept anything else then a prism. ;)
Just think of switchable glass, a surface that changes from clear to mirror depending on a static voltage applied to it, that allows for a hybrid viewfinder with a user selectable blending ratio. You'd still be limited in VF size though.

There are many full time pros using EVF's, Ctein and Kirk Tuck are very prominent ones that springs to mind. Neither is sponsored by any camera manufacturer and are both pro EVF's and have written many articles on their blogs pointing out how good they are. Ctein might not need ultrafast refresh, but Kirk Tuck is a very active general shooter often in theaters and poorly lite events. Not saying EVF's are for everybody, but blanket statements like jrista's are clearly unsupportable and easily shown to be false.

As an individual I can well understand

My statement wasn't blanket...it's limited to the original context that inspired it: high speed action photography. Sports. Wildlife. Birds. You could probably throw air shows in there as well. I am not saying that in every form of photography an OVF is superior. There are simply certain types of photography where an EVF has a LONG way to go before it even catches up, and from a technological standpoint, unless someone figures out a way to emit 700nm-550nm light from a 400nm aperture, they will never provide pixel-free viewing.

1530
EOS Bodies / Re: Canon EOS 7D Mark II in 2014 [CR2]
« on: July 12, 2013, 01:47:52 PM »
...any serious or professional photographer who relies on actual real-time update and large, full-detail viewfinder won't be moving to EVF any time soon. There is just plain and simply no replacement for a prism.
Wouldn't a serious photographer know the benefits of a waist level finder and consider the prism a technical detail rather then a dogma? The viewfinder of a 1Dx feels rather tiny compared to a 645, and don't even think about anything large. And if one compares the mandatory ~40ms release lag to the less then 10ms you can get electronically...
Guess it will take some time for the pros who now shoot tethered and use the computers screen as VF to accept anything else then a prism. ;)
Just think of switchable glass, a surface that changes from clear to mirror depending on a static voltage applied to it, that allows for a hybrid viewfinder with a user selectable blending ratio. You'd still be limited in VF size though.

Well, you've switched contexts from sports to studio photography, where use of a tethered laptop or computer is quite normal. The original context was sports and action photography, where the OVF and a dedicated AF unit still rules as king.

I would take my tethered Surface Pro any day over an EVF, though. There was a thread a while back where I computed the necessary pixel densities to make an EVF screen be high enough resolution for the average 20/20 viewer at a 25mm eye relief such that pixels were invisible. For 20/20 vision, you would need just over 5000ppi. To accomodate users who have better vision, or users such as myself who have 20/10 vision with contacts, you would need an insane 12,000ppi. With the average size of a viewfinder, 5000ppi is pushing the limit of how small pixels can be and still be transparent to light. At 12000ppi, you are already cutting off the longer frequencies of light, and therefor only able to pass greens, blues, and violets. And that isn't even touching DR, or the fact that even if the EVF supports high bit depth it is still limited by the camera's DR.

The day will never come when an EVF (or, for that matter, a tethered laptop screen) becomes superior to an optical view finder for action photography. There is no substitute for a truly real-time, high resolution, bright, optical prism based viewfinder. For action. Studio work is a different matter, but as you say, people have been tethering and using huge screens for a very long time in that industry, so they still have a superior tool than an EVF.

Not really sure what you mean about switchable glass. Sounds like you are talking about the piezoelectric effect,  however I am not really sure how that is much different than what Canon already has with their transmissive LCD that overlays their current viewfinders. It is fairly simple right now, but there is no reason Canon couldn't drop a whole ton of information into that screen with a selectable mode button...imagine seeing the histogram as black bars in the viewfinder...or the electronic level...or, any amount of information you desire, and still always have full use of the OVF.

Pages: 1 ... 100 101 [102] 103 104 ... 217