November 01, 2014, 01:31:04 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] 2 3 ... 300
1
I hear what your saying and I agree with a lot of it. I just think the deisign is over complicated for what it needs to do and will be replaced by mirrorless designs. Right now I am using a 6d and 70d I have a 7dii on order. When it comes time to replace the 6d I will likely get an a7 type camera, hopefully from canon if they get on the mirrorless ball. In a few years when I switch out the 7dii I reckon mirrorless will have progressed to the point where dslr's are obsolete.

I don't think mirrorless is better because it can be smaller, I think its better because its simpler, it's got a ways to go but its coming.


How are you defining simpler, though? I've used the A7r, and while mechanically it lacks the mirror, outside of that it really isn't that much "simpler" of a camera to operate than a regular old DSLR. There are also some significant problems with the EVF that often present at the most inopportune times (such as severe EVF or LCD stutter).


There is significant value to be had with an OVF. I honestly do not believe that, regardless of how responsive EVFs get, that they will ever be as responsive as an OVF, and for some types of photography, that kind of instantaneous response is critical.


I see mirrorless and DSLR as different options, not one that will replace the other or one being better than the other (in the long run...mirrorless designs still have a lot of growing to do.) The other "complexity" of DSLRs, such as all the buttons and configurability, are IMHO their greatest STRENGTH. For the kind of action-centric photography I do most of the time, having as many buttons as possible to instantly access the majority of the critical camera functionality is HUGE to being able to effectively use the camera in difficult situations. There is some programmability to the A7r, however it's smaller body size (which is pretty endemic to mirrorless designs so far) doesn't leave much room for having lots of buttons that give instant access to important functionality. The larger, more "complex" bodies of DSLRs offer plenty of room for buttons and dials...it's their strength, not their weakness.

2
No, Dslr's are dinasaurs with roots in the film days. The design made sense then but it doesn't now. Mirrorless is the way to go. It's good enough for everything except sports and wildlife now. I expect the 7dii is the last dslr I am going to buy.


This is entirely a matter of preference.


It's always been curious to me how mirrorless lovers just assume that everyone wants something uber-tiny, or to give up their OVF for an EVF, or what an electronic shutter, or anything like that. Personally, I think the DSLR form factor...it's size, shape, and ergonomics, are IDEAL for photography. Doesn't matter if I'm using a giant 600mm f/4 lens, or a tiny 50mm f/1.4 lens, either way, a larger DSLR fits my hands WAY better than a tiny mirrorless. And I speak from first hand experience, I spent the greater part of a week with the A7r recently. I love the IQ, however I really don't' like the body. It  is just too compact. If I was casually shooting things on a vacation, the smaller form factor could be nice, especially with a compact lens, but the darn thing just does not fit my hands.


I may eventually warm up to EVFs for some things, however even with the best of them on the market, they are still, IMHO, radically inferior to OVFs as far as responsiveness and detail and all that goes, and really need to come a LOOONG way before I think they could possibly be ready to actually, en-mass, replace OVFs.


For me, I'm part of the group that will stick with DSLR to the utter end. You can basically pry it from my cold dead hands...right along with my optical view finder...which is what die-hard "eliminate DSLRs, mirrorless for everything!" guys like you are literally going to have to do in your crusade to expunge the universe of what you personally, and IMO incorrectly, believe is an inferior camera design. I refuse to give up my DSLR, and I personally hope they stick around for the next, oh, forty years or so (depending on how long I live. :P)

3
One of the reasons I prefer exposure stacking over HDR is that, for me, it looks more natural than HDR

Exposure stacking is HDR.


I think he means stacking multiple identical exposures, which technically speaking does not actually require something like a 32-bit float "HDR" file or anything like that. Exposure stacking, when you have still scenes, can reduce random noise (although it will reinforce non-random noise, which is bad), and improve dynamic range. If you stack a bunch of 16-bit TIFF files converted from 14-bit RAWs, you can gain more dynamic range, for sure, but in general I wouldn't normally call that "HDR"...you don't get a minimum of 20 stops of DR like you usually do with a true blend to 32-bit float.

4
Lenses / Re: Why does a 2x TC lose 2 stops?
« on: October 29, 2014, 01:39:47 AM »
Time for the day's stupid question:

It is a known fact that all 2x teleconverters lose 2 stops of light, and all 1.4x TCs lose one stop of light. Why? What if the glass in the TC was twice as large? ... it would let in more light, therefore the light loss would be less. Now I realize this logic is somehow flawed, but I can't reason why. Anyone?


Because stops are RELATIVE. Stops are ultimately based on a ratio, such as f/4. That 'f' stands for 'focal length', and it's a simple mathematical formula. So, it really means 'focal length divided by four'. If you have a 400mm f/4 lens, your entrance pupil (the physical aperture as viewed through the front of the lens at "infinity") is 100mm in diameter. The f-ratio is actually determined by dividing the focal length by the entrance pupil diameter, so the 'f/4' really comes from 400mm/100mm, which equals 4. 


If you slap on a 2x teleconverter, you are adding on what is effectively a magnifying glass to the end of your lens. In this case, it magnifies your subject by a factor of two, which is the same as doubling your focal length. So, now you have an 800mm lens. Thing is, the lens did not get wider...it's the same diameter, which means it has the same entrance pupil. The entrance pupil of the lens, when observed from "infinity", is still 100mm. That means that your f-ratio is now 800mm/100mm, which equals 8. You've lost two stops of light (f/4 -> f/5.6 -> f/8).


I'm sure there is still the same question, though...why? First, the entrance pupil is ultimately what determines the amount of light that will produce the image circle at the sensor plane. The AREA of the opening in the diaphragm, to be specific, is what determines the amount of light. So, a 100mm entrance pupil means you have pi(d/2)^2 area, or 7854mm^2 circular area. Attaching a TC to a lens is adding a magnifying glass to the BACK of the lens. You've already limited the amount of light, since by the time it reaches the TC, it's already passed through the aperture. A 2x TC is effectively taking the center portion of circle of light produced by the aperture, 1/4 the area or 1964mm^2, and enlarging it to the same 7854mm^2 (which is then projected by the rest of the optics in the TC to the standard image circle size for the camera, say 44mm diameter for FF; it's really a bit more complicated than that, but for casual conversation sake).


It's taking a smaller quantity of light from the center area of the image circle and spreading it out over a greater area. In other words...one quarter the quantity of light spread out over four times the area that light initially covers results in two halvings of the intensity of light at any given pixel.

5
with HDR... the image should still look better than pushing the blacks and the shadows... right? 

Nope.

I absolutely disagree.

Pushed and pulled, but particularly pushed, areas of an image will never be able to stand up to HDR/tone mapped/luminosity masked multiple images, not until we get 20 bit sensors that is. It goes to the very core of the difference between how our eyes automatically put a 'gamma' curve onto what we see and the necessity for a gamma curve to be applied to a linear captured digital image. The shadow tonality is stretched out already, stretch it out even more and you just end up with blocky colour tinted crap. Sure it is way more impressive than Canon's noisy blocky tinted crap, but it is still crap.


This I agree with. If you have a still scene and blend, you can get deep shadows that are as clean as bright midtones when using HDR.


However, sometimes you don't need 20 stops...you just need 13 or 14 stops, and just need to lift a little...and don't want banded or blotched noise showing up when you do. That doesn't invalidate the value of using HDR...it's just an in-between area where having cleaner read noise is valuable, and simpler than doing HDR to get excellent results.

6
I voted no, but considering that many of the 1D bodies were awfully close to $8k at release, $10k may not be much of a stretch.

Seems that high ISO usability and the interface have changed the most over the years. Good images taken today with a 2002 vintage 1D still look good.

I bet that if Canon could merge the best of the 1DX with the best of the 1DsIII, many would pay the asking price.


I think the biggest reason they were so expensive in the past was the cost of manufacturing the larger sensors. It used to be extremely expensive to manufacture a full frame sensor. That's also the reason MFD used to $60k and up in the past, instead of ~$10k or so. There is very little reason for DSLRs with FF sensors to be expensive anymore...manufacturing larger sensors is a lot more efficient (especially on 300mm wafers). I think a lot of the $6700 price premium of the 1D X is just the prestige of the line, and a more hand-crafted (or rather hand-assembled and optimally tested) touch.

7
Landscape / Re: Deep Sky Astrophotography
« on: October 28, 2014, 01:12:14 PM »
I botched my eclipse photos.  It was right at sunset on the east coast, I had to drive around a bit before I finally found a place with a low enough view to see the sun during the eclipse, and even then I just had a couple minutes to try and dial things in.  I should've planned ahead a little better, and I should've experimented with my equipment ahead of time.  Anyway, I got so frustrated with my results, I went out the next day to see what kind of sun shots I could get when I wasn't rushed.


Sorry you missed the eclipse. Image looks nice, though...that one cluster of sunspots is just giant.

8
Landscape / Re: Deep Sky Astrophotography
« on: October 28, 2014, 01:04:35 PM »
That sunspot is HUGE!!! it must be 10 times the size of the Earth. Is that with the 600mm?


Yeah, gargantuan. It was an X-class flare level spot. They were expecting a LOT of energy from it, but I think it was pointed away from earth when flares occurred.


It was indeed shot with the 600mm. I should have tried imaging with the 2x TC...but for some reason I didn't think to try until the eclipse was over.


Here is mine from last year solar eclipse.

3/11/2013 15:54

Partial Solar eclipse as seen from Limassol,Cypus

Canon 600D + 70-200/f2.8 is II + 2xTC + UV Filter + CPL Filter + 10 stop ND Filter — in Limassol, Cyprus.


Very nice. It's pretty amazing how much a 10-stop ND can do for imaging the sun. :D

9
Landscape / Re: Deep Sky Astrophotography
« on: October 28, 2014, 01:00:43 PM »
That and the Pleiades photo are really impressive pictures, what sort of lens you need for that?


Thanks. :) I use my EF 600mm f/4 L II right now, which doubles as a very high end telescope. It's similar in IQ to the Officina Stellare Hiper APO 150mm, which is about $11,500.
Hi,
    Hmm... I thought most super telephoto lens are not as good as telescope for imaging... how's the edge performance of the EF 600mm F4 L II??


As I stated, the 600/4 II has stellar corner performance. The field is extremely flat, a kind of flat that you have to spend tens of thousands of dollars for with a real telescope. The Canon superteles are used very often for astrophotography. There are even ultra sensitive, ultra high speed imaging arrays that use dozens of Canon EF 400mm f/2.8 L II lenses with FLI cameras to image some of the faintest stuff in the universe.


Canon L-series superteles make excellent telescopes. I don't know about other teles or superteles, but Canon's L-series superteles are extremely good.

    Anyway, I saw my friend's 130mm f6 StarFire and it's super sharp right to the edge, but the waiting list is 7 to 10 years...   :o

   Have a nice day.


The StarFire is an Astro-Physics scope. Astro-Physics makes excellent quality products, however they are pretty costly. From what I understand, AP is no longer making that particular scope...they stopped making it years ago (maybe even over a decade ago), so there isn't a waiting list, they simply aren't available unless you buy them used (and owners don't generally sell them). There is a newer 130mm f/6.3 StarFire that is currently in production, however it doesn't take years to get them...if it did, no one would even bother, that would just be ludicrous.


AP mounts are some of the best in the world, and in such demand and built literally with hand-made quality, that their order-to-delivery lead time is around eight months right now. (I.e. if you order an AP mount today, it wouldn't even start being manufactured and assembled until about six months out or so, and wouldn't be shipped until at least eight months out.) I figure the StarFires have a similar lead time, but I don't know of anyone who has had to wait over a year for any AP product lately. They just don't mass-manufacture, they don't even start manufacturing until they get an order, and it is truly hand-made quality (assembly is all done by hand, each and every product made is manually tested for quality and optimized for best performance, especially for mounts with absolute encoding, etc.) It's the price you pay to get high end, high quality astronomy and astrophotography equipment. It is not easy or quick to ramp up new people with the skill level required to deliver that kind of quality (although from what I have heard, companies like Astro-Physics are training new people to improve their turnaround.)

10
Nope. $10k for a DSLR, regardless what it does, is utterly ludicrous. I'd NEVER pay that much, not even if money was no object. If I was to spend that kind of money, it would need to be something with a much larger sensor than 35mm FF. For $10k, I would really expect, these days, to get a full 55x44mm sensor in an MFD with an excellent selection of lenses, and the option for interchangeable backs. You can already get a 44x33mm sensor MFD for as little as $8k.

11

it's not possible to select the 7D in their new comparison tool on DPR.


?  If you click on the OP's link, it's one of the three cameras selected to compare with the 7DII (it's right next to it).

sometimes you gotta dump your javascript caches?...
I was getting totally bad data from dxomark a couple years ago, thinking it was real.  Made a fool of myself here posting it.
turned out my browser was displaying incorrect data - different computer, same page, same time, was displaying correctly.


Odd, I have a similar issue. On my laptop, same browser (Chrome), the 7D is not listed. I cleared the cache and all, doesn't seem to matter what I do, 7D isn't listed. On my desktop system here, also with Chrome, the 7D is listed. Not sure what the deal is. As far as I can tell, both systems are running the same version of Chrome.


On  my desktop here, at ISO 6400 RAW, the 7D definitely seems to have more chroma noise than the 7D II. The 7D II still seems to have a less fine-grained, more blotchy distribution of chroma noise, where as the D7100 has a finer-grained, more random and less blotchy chroma noise.


@DanielW: I could take screenshots and post them, but we should all be looking at the same data. My screens are all calibrated, at a brightness of 120mcd. The lower brightness helps reveal detail, where as higher brightness, such as 400mcd or brighter which is often the default on many screens these days, could very likely wash out detail. If you can, reduce your screen brightness, and calibrate your screen.


Then, compare the light blue swatch on the color checker card at say ISO 6400 RAW, the one just above the black swatch in the lower right corner. Also, make sure you are in print mode. I've been comparing the 7D II and the D7100. The difference should be pretty obvious. I have also been comparing the grayscale swatches...on my calibrated screen, the 7D II seems to have more pronounced color noise, and the color noise seems to "clump" into groups of many pixels. The D7100, while it still has color noise, takes on a finer and more random characteristic which, to me at least, is more pleasing. I wouldn't say it has less noise...just a better characteristic in my opinion.


For a more real-world expression of how the characteristic of noise can affect an image, move the zoomed in view box to the picture of the guy in the lower left, and position it such that you can see a bit of the dark background behind the picture, as well as the guys chin and throat. The color noise in the background of the 7D II image takes on a rougher texture, and seems more pronounced. Also, you can see color noise in the guys neck. The D7100 has a finer and more random color noise texture, and I can hardly see any color noise at all in the guys neck.


As far as how readable the text is, that seems to be related to AA filter strength and pixel count. Sensors with more pixels generally have more readable text in the color wheels and in the text blocks in the middle.


I should note that all of these differences are more pronounced and obvious on my laptop screen. That has a 3200x1800 pixel screen, however it still renders everything at the same general scale as my desktop (I set the DPI up on the laptop to render everything using more pixels...so text is roughly the same size in absolute terms as my desktop, but more pixels are used to render it. Same goes for images...more pixels are used to render each image, or pixels are sub-detail on my laptop, which makes it easier to see that detail without seeing RGB subpixel elements.) The differences in text sharpness, color noise, etc. are very easy to pick out on my laptop. Yellow swatches in the color checker card, for example, pop right out as having a more blotchy and rougher characteristic than the D7100. I can still see the same issue on my desktop, but it is not as pronounced.


I was using my laptop when I wrote my prior posts, so just to be fair, I wanted to make sure that context was shared...I was very, very clearly seeing differences in noise characteristic.

12
The noise is not blotchy: conversions from some converters are blotchy. And this is exactly as true of D7100 conversions too, depending on converter.

It's also painfully obvious that Nikon uses on-chip NR to deal with Chroma - I've been saying for years that the results I get with the default Chroma NR in Capture One are identical to how Nikon files look: I actually once asked Phase One whether they'd licenced their Chroma NR algorithm to Nikon.

But - to repeat - the character of the noise you're seeing has practically nothing to do with the camera and almost everything to to with the converter.

And in the tests I've done comparing 7D Mk II files against D7100 files, converting in Photo Ninja and in Raw Therapee (remember, the D7100 is a supported camera, the 7D Mk II is not), the D7100 loses out: and it is of course prone to banding/pattern noise in pushed shadows to an extent an order of magnitude (yes, I know what that means) more than is the 7D Mk II.

But - really? - Complaining about the "quality" of the Chroma noise? The single easiest thing to fix (by a country mile), of all the things that might impact on image quality?

Are you really that desperate for something to bash the 7D Mk II about?


At the moment this is purely anecdotal. I'm going off of real images that can be viewed and compared by anyone. Based on the data at hand, the data linked in the original post of this topic, I can draw no other conclusions other than what I have, based on the observations I am able to make.


I am currently evaluating Capture One 8. I haven't had a chance to do a full blown comparison between it and Lightroom. You may well be right, C1 may indeed handle NR better. I have no data from which to base an opinion yet. Even when I do, all I can do is base an opinion of how my own data from my own images compares. I cannot take my own images and use them as a basis for comparison with DPR's data, though.


Within the given context, which in this case, in this thread, is DPR's sample images, the Canon conversions, whatever was used to make them, end up looking worse. Given that very significant number of photographers (probably a, likely strong, majority, topping DPR, DXO, and C1) use Lightroom to edit their RAWs, I think the comparable data DPR provides is very indicative of what people in most cases are going to get. MAYBE it's because of LR. Until there is an extensive and reasonable set of conversions done from all the same cameras that DPR lists, with a similar kind of test scene, as an analytical person, I have to base my opinions in a thread like this off of what everyone else can see and base their own opinions on.


There is no value in making the aggressive and combative claim that it's the person sitting behind the keyboard, or the archaic software they are using, that's the problem. Not unless you can provide irrefutable evidence of such a claim, and demonstrate how a tool like C1 can improve results. (If it does indeed, then I'll do what I can to provide such evidence myself...I'm all for getting better results, and if C1 can do that for me, then hell, I'll ditch LR, or at the very least, use C1 to do my base RAW edits, and use LR to manage the library of original RAWs and conversion TIFFs.)

13
EOS Bodies / Re: 4K Products Coming From Canon [CR2]
« on: October 24, 2014, 11:34:47 AM »
"You can pick an A7r up for as little as $1700 used, A7s' used for as little as $2000."

Ever wondered why these owners are selling?


Actually, those prices were from the LensRental used equipment store. They buy a bunch of copies of things, and sell a LOT of used stuff. I haven't seen those kinds of prices elsewhere, and I honestly haven't seen nearly the volume of used Sony cameras as I do used Canon cameras. That's expected, though, given the ratio of Canon sales to Sony sales. I don't think used sales can in any way be used as a gauge of which brand is better, you would need impeccable statistics about how many used items from each brand are on sale, and a damn good idea of why they are on sale, to make any kind of assessment as to why they are being sold.

14
It's no quirk of screen, browser, or anything else. It's a matter of characteristic...it's not the amount, it's how the noise presents. I'll produce some direct comparison images and GIFs from DPR data so people can judge on their own screens. Canon sensors are still more blotchy in most instances. Even the a6000 has a better noise characteristic than most Canon crops, with the exception of the reds and brown/tan swatches. It does have more color noise, however it's cleaner, random color noise with more per-pixel frequency and characteristic...it's not blotched.


I use that term very explicitly and specifically...blotched. That refers to characteristic, not amount. Canon color noise has a nasty characteristic. It's one of the things hate bout the 5D III at low ISO...it's the same blotchy characteristic in the shadows. Noise character matters just as much as noise amounts. As far as amounts go, there isn't a full stop difference between any one of the APS-C cameras. At most, on a normalized basis, there may be a third stop difference, but that's to be expected...barring color noise characteristic, the amount of noise overall is ultimately determined by total sensor area, Q.E., and maybe fill factor.

15
Pretty much comports with what I've been saying for a while about the 7D Mk II based on my own Raw conversion tests. That the 7D Mk II gets within a stop of the FF cameras is damn' impressive.

And - again - the 70D, and the 7D Mk II, show no appreciable pattern noise in pushed shadows.


I wouldn't say that at all. At "Full Size", that may be true, however once you switch to "Print", FF pulls ahead again, with considerably lower noise in all cases.


I would also strongly dispute the notion that the 7D II does better than the D7100. Again, at "Full Size", the gap is small, however at "Print" size, the D7100 exhibits far less color noise. The 7D II still suffers from color noise problems. This is most obvious in the monochrome color swatches of the test image, but yellow, orange, green, and blue swatches also suffer from blotchiness due to color noise. Yellow and blue and maybe purple seem to be the most hard hit of the color swatches.


The third gray swatch from the right edge of the color checker demonstrates the differences in color noise best. I am actually rather impressed with the 6D noise...very clean.


The grain pattern at ISO 3200, 6400, and on most color swatches 12800 is a far more pleasing random mostly mono grain with the D5300, D7100, D810, 6D, 5D III, A7r, A7s, etc. than it is with either the 7D II or the 70D (or any other Canon crop). Color noise is still a key problem for Canon crop sensors at high ISO (and I suspect with shadow lifting.)


The D5300 seems to perform a bit more poorly on the brown and tan swatches...color noise is worse in those, for some reason, however its quite excellent in all the rest. The D7100 does not seem to have that problem...all of the swatches look excellent on the D7100 at Print size.


Overall, I don't see a whole lot of difference in luminance noise levels overall between any of the crop cameras. Luminance noise seems to be about the same on a normalized basis. The key difference is color noise and how it presents. Canon's characteristic blotchiness is still in play. I am not sure how it compares to older cameras...it's not possible to select the 7D in their new comparison tool on DPR. I suspect the 7D and 5D II are worse, but Canon crop is still worse than everyone else when it comes to color noise (and the 5D III and 1D X are still worse than the D810).

Pages: [1] 2 3 ... 300