burzum16 wrote:
Sorry but can you please elaborate a little bit more?
I am deliberating myself between using the Sigma 150-600 on my existing RP or buying a 80D/90D for extra reach and better focus for birding.
Thanks much!
The optical projection from a given lens will always be the same, no matter which body you mount it on. That will be your optical reach.
The pixel density of a given camera will vary the amount of refining divisions that optical projection is divided into.
Basically, it is either a (FF) 1000 piece jigsaw puzzle (of the same optical projection), or a (APS-C) 1600 piece jigsaw puzzle (of the same optical projection) when you move from FF to APS-C (Canon) for the same MP count of the two diff formats.
In either case, the optical projection remains the same ... only the degree of refinement divisions changes based on pixel density (with partial area of coverage cropped away on APS-C).
Mounting on a 90D will offer more refinement divisions than when mounting on the RP.
This is why the M4/3 using a 300mm with a 2X pixel density / crop factor, doesn't compare exactly to using a 600mm with a 1X crop factor on FF when considering the fine details (unless the resolving capability of the 300mm is double that of the 600mm). It still has an optical reach only 300mm ... just more divisions (assuming equiv resolving capability of lenses @ ceteris paribus).
TeamSpeed wrote:
However, you are not getting optical reach, just digital reach. Optical reach is almost always superior in IQ...
Perhaps its just semantics but I disagree with this assessment. The original 5D has an 8 MP sensor and the 5DSR has 50 MP; both are full frame and take the same lenses. Which has optical reach and which has digital reach? It comes down to the noise, color rendering, DR, etc. of each individual pixel (or set of input pixels). IMHO Canon has done a pretty good job of keeping the pixel level IQ at least equal in its progression of the 5D series. The APS-C or M4/3 sensors are smaller in area and designed to see a sub-area (crop) of the FF image circle but they are just collections of pixels reading projected light. As I understand it the 5DSR and 7D2 have the same size and generation of pixels but the 5DSR has a larger sensor and therefore more pixels. For a given lens which has more reach? Which has optical reach?
At the pixel level, the measurement accuracy depends on the pixel electronics quality and the lens projection quality no matter what size the surrounding image circle is. A pixel's size or its light gathering ability is one but not the only indication of its quality. Just as some lenses are better than others at coherent light projection. Advances in microlens design and electronics design have kept pixel measurement accuracy high while reducing the pixel's light gathering area. The result is greater image resolution which can be interpreted as effective reach.
So where is the crossover between optical reach and digital reach? As I understand Digital Zoom, from compacts and cell phones, digital zoom interpolates between pixels to add more pixels thus increasing the apparent resolution by computation not measurement. The same can be done in photoshop by manipulating the image size.
So no, increasing pixel density (resolution or effective reach) is a combination of digital (electronic) and optical (light projection) not one or the other. I'll add that I get what you are saying, that the 5D4 pixels are more accurate than 80D pixels when there isn't an abundance of light because of the 5D pixels having a larger light gathering area and relatively equal electronics. This has always been the APS-C tradeoff.
Your 35mm image has to be enlarged the same amount no matter the resolution. Optical enlargement via focal length/TC is not the same as digitally zooming into all those pixels especially when it comes to print. If you are a digital person, by pixel peeping or by cropping then yes you might have a different perspective, however accurate or inaccurate that assessment is.
Digital zoom is any time you are drilling into the pixels, real or extrapolated.
TeamSpeed wrote:
Digital zoom is any time you are drilling into the pixels, real or extrapolated.
You are mixing sensor size, sensor density, and post capture processing. We have different semantic definitions. I accept that for you. I find it misleading and confusing. Lets just say I disagree with your basic premise that longer focal length is always better than increased pixel density for increasing apparent object size.
Jeff Nolten wrote:
Lets just say I disagree with your basic premise that longer focal length is always better than increased pixel density for increasing apparent object size.
The math is a composite of the two (I forget the formula involving square roots, etc.). As a general premise (based on the assumption of reasonably equitable resolving capability of the different Focal Length glass), the optical reach is preferential. Of course, using a lesser resolving lens with a longer focal length MAY produce less resolution than the shorter FL with pixel advantage. Hence the qualifier regarding reasonably equitable resolving capability.
As to achieving greater optical reach with a TC ... well, that is a matter of diverging the optical projection. Depending on which angles are involved (centrally collimated angles are less effected, than wider ones), the losses may compound the native optic of the glass. So, another variable is introduced that can make real world outcomes slightly askew from the basic premise.
Simply put, the bigger the subject is in the frame, the better. The smaller the pixel size (density), the greater the refining detail.
Get both, when & where you can, the play well together. If real world constraints having you favor one over the other, so be it. They both have their place and work together. Also, it kinda depends if you are mostly interested in the capture as a whole vs. capturing a very small portion. The variable there will then place constraints on how large you can go with it.
Case in point regarding the 100-400 II vs. the Sigma 150-600 C. I had both and shot hundreds of side by side test shots of the 100-400 II + 1.4X TC III vs. the Siggy. At the end of three months, I had no clear winner, it was a dead heat, with tiny wins & losses on both sides of the column. So, while that isn't exactly the point regarding digital ... it is the point regarding different approaches can still be kinda close to quibble over. Now, against my Viv Ser. I 120-600 ... the 100-400 II + 1.4X TC III would be not so close.
Imo ... start with good glass (optical projection), get good light (effects your contrast resolving ability), and divide it in little pieces to your heart's content. If I can't afford to make the jump to the next level of optical projection, then I may make a jump in the little pieces side. The reverse can be true as well, but generally speaking, harnessing converging optical projection is where things really begin, imo. Once you tap out there ... start slicin' & dicin'.
Jeff Nolten wrote:
You are mixing sensor size, sensor density, and post capture processing. We have different semantic definitions. I accept that for you. I find it misleading and confusing. Lets just say I disagree with your basic premise that longer focal length is always better than increased pixel density for increasing apparent object size.
No I am not.
If you have two FF bodies, body A with 4x the resolution of body B, and you shoot with a 50mm on A, and then crop a 100mm view of it out, and then shoot with a 100mm on B, you won't get the same IQ results, provided both lenses resolve identically, and both cameras have the same AA filter and sensor profiles.
Pixel density is not a substitute for optical reach. One is superior to the other.
Also if you shoot a 100mm on both A and B above, they will look the same when enlarged for large prints because they have to be enlarged equally. In some cases, if the manufacturer reduces the AA filter of camera A, then it is possible it will have more fine detail, that is probably the most used advantage of pixel density (like the 5D3 vs 5DS, detail rendering better on the 5DS). Other advantages of higher pixel densities is the ability to clean up noise easier while retaining detail, possible cropping further of image while still maintaining a file that meets certain DPI requirements of a printing service, or image site, etc.
Using pixel density to zoom into an image by cropping or just viewing larger is digital zoom. There is a second level of digital zoom where your viewing medium has to manufacture data (like viewing at 200%+), and that is really bad for IQ. Many point and shoot cameras extend their optical zoom with digital zoom, which is nothing more than cropping of the image data further and further until the digital zoom factor has been max'd out.
People often say they want a crop to use instead of using a 1.4x TC on a FF because they don't lose a stop of light, but they still lag behind by about 1 stop on the APSC, almost equalizing the two anyways. I had a 7D2 for years coupled with my 5D4, but my 5D4 + 1.4x always produced better results than the 7D2, despite the 7D2's pixel density "advantage".
RustyBug wrote:
Old School ... Glass Rules. TeamSpeed wrote:
Pixel density is not a substitute for optical reach. One is superior to the other.
This is true. But also simplistic as your other paragraphs prove e.g. multiple use of providing. I prefer using a full frame sensor and the f4 L zooms including the 100-400. So I agree, get the biggest real-estate you can looking through the best glass. However its not that simple. If it were we'd all be using medium format sensors and the best professional primes or at least still using our original Canon 5D.
I've been using my G1X3 side by side with my 5D4+24-105 for close to two years. In probably 90% of my shots I can't tell the difference between which camera took which shot after post processing and displayed on my 5K computer. The 5D performs better in extreme light, but generally ... . Another case in point, the 5D or 6D + 100-400 performs better at AF tracking and low light than an SL2 + 55-250. But which would you rather carry on a hike where you might see a deer or squirrel? How much difference would there be in ultimate IQ for the majority of cases? How much would it impact your enjoyment of the pictures? Finally the 5D4 + 600 f4 is better optically than the 80D + 100-400, but for the cost difference one could pay for most of a trip to Tanzania and then there is the ease of getting the gear to the Serengeti. Again how much difference would there be in ultimate IQ for the majority of cases? My 80D + 100-400 did magnificently on my trip last year. Cameras are so good these days that in many cases you can exchange "digital zooming" for "optical zooming" with very little sacrifice of ultimate IQ. Get what you really need in terms of cost and performance.
That certainly brings in a bunch of other factors, sure, and different generation sensors with AA filters or lack of vs older all play a part. We can definitely agree there. But when I had my 7D2 and 5D4, I found it easier to just take my 1.4x with me and the 5D4, instead of both cameras (because each had their strengths), and thus why I finally, after a few years, got rid of the 7D2, and replaced it with the M50 (travel friendly, 10fps, focus peaking, AF at f13, etc), since it had things I could really use now.
I now shoot the M50 with my 150-600 and 1.4x or 2x every once in a while simply because I get the optical zoom, and I have AF working for me. I wouldn't hike with that combo, but in the back yard with our pond, it works fine. If Canon makes an M60 with 32Mpx, I would undoubtedly buy it.
I was about to add that there is a reason that Canon has been releasing Mk II and III versions of their L lenses, because the resolving power of the newest sensors are catching up to the older lenses. But you get that.
At one time I hauled around a 5D2 and original 7D. When the 5D3 added f8 AF I gladly gave up the 7D which I didn't like too much anyway. I was firmly in the bigger sensors with bigger glass camp. With my 5D3 and 7D I actually had to use equal ISO on both bodies (using the 1.4x on the 5D) to get equivalent apparent exposures. I compared the 7D2 to the 80D and chose the 80D as it best suited my needs. I can see that the 90D not being a 7D3 is not pleasing many. We are all individuals.
Thanks to Archibald for posting the link. I think the 90D is looking pretty good at ISO 3200 in these test shots considering I try to keep my APS-C bodies to a max of ISO 1600 and only go to 3200 on FF.
Jeff Nolten wrote:
I was about to add that there is a reason that Canon has been releasing Mk II and III versions of their L lenses, because the resolving power of the newest sensors are catching up to the older lenses. But you get that.
I think they're also seeing some true competition from the likes of Sigma and Tamron, as well as Sony becoming a major camera company now. Also, new design techniques and manufacturing precision have maybe made the extremely difficult a little less difficult. Just speculating.
At one time I hauled around a 5D2 and original 7D. When the 5D3 added f8 AF I gladly gave up the 7D which I didn't like too much anyway. I was firmly in the bigger sensors with bigger glass camp. With my 5D3 and 7D I actually had to use equal ISO on both bodies (using the 1.4x on the 5D) to get equivalent apparent exposures. I compared the 7D2 to the 80D and chose the 80D as it best suited my needs. I can see that the 90D not being a 7D3 is not pleasing many. We are all individuals.
Thanks to Archibald for posting the link. I think the 90D is looking pretty good at ISO 3200 in these test shots considering I try to keep my APS-C bodies to a max of ISO 1600 and only go to 3200 on FF....Show more →
I haven't had a cropper that can knock my full framers off the pinnacle as far as high-ISO and file "work-ability" is concerned. Granted, I haven't had a 7D2 - I did have the 80D and the M5, and this 90D/M6 II shows some promise. For me, it's always a competition between pixel density and the ability to keep a reasonable shutter speed in moderate to low lighting. The 5D4 keeps on winning that battle.
I pulled in the ISO-6400 test scene image from that site, in RAW (I have the latest DPP-4). It is surprisingly clean, but I'll still want to see some comparison shots against other bodies.
Jeff Nolten wrote:
This is true. But also simplistic as your other paragraphs prove e.g. multiple use of . I prefer using a full frame sensor and the f4 L zooms including the 100-400. So I agree, get the biggest real-estate you can looking through the best glass. However its not that simple. If it were we'd all be using medium format sensors and the best professional primes or at least still using our original Canon 5D.
I've been using my G1X3 side by side with my 5D4+24-105 for close to two years. In probably 90% of my shots I can't tell the difference between which camera took which shot after post processing and displayed on my 5K computer. The 5D performs better in extreme light, but generally ... . Another case in point, the 5D or 6D + 100-400 performs better at AF tracking and low light than an SL2 + 55-250. But which would you rather carry on a hike where you might see a deer or squirrel? How much difference would there be in ultimate IQ for the majority of cases? How much would it impact your enjoyment of the pictures? Finally the 5D4 + 600 f4 is better optically than the 80D + 100-400, but for the cost difference one could pay for most of a trip to Tanzania and then there is the ease of getting the gear to the Serengeti. Again how much difference would there be in ultimate IQ for the majority of cases? My 80D + 100-400 did magnificently on my trip last year. Cameras are so good these days that in many cases you can exchange "digital zooming" for "optical zooming" with very little sacrifice of ultimate IQ. Get what you really need in terms of cost and performance....Show more →
There's a HUMONGOUS DIFFERENCE between a conversation / question / answer regarding which one provides an objectively different outcome (major or minor diff notwithstanding), vs. which is preferred or more strategic or more cost effective, not meaningful for a given need, etc..
I believe the original context was predicated on an interest for the objective comparative. Introducing the plethora of subjective components makes the dialogue A) non-sequitur & B) exponentially more variable. That neither agree's with / nor disagree's with any of the mentioned points ... just, they are two very different conversations.
Well, at least I won't be anxious at my next shoot because I don't have the 90D, just yet. If AF performance gets heralded, it could pick back up to gas or Gas.
So the two cameras with the 32Mpix sensor (90D and M6 mark II) have been out for over a month. Is there a real improvement in dynamic range over the 24Mpix sensors (80D, M5,...)?
matejphoto wrote:
So the two cameras with the 32Mpix sensor (90D and M6 mark II) have been out for over a month. Is there a real improvement in dynamic range over the 24Mpix sensors (80D, M5,...)?
Even if the DR is the same, achieving it with 6 MP higher resolution equates to an improvement.
Sy Sez wrote:
Even if the DR is the same, achieving it with 6 MP higher resolution equates to an improvement.
... but as you probably know that 6 MP extra resolution (linear resolution is just 1.14x higher, almost imperceptible) is hardly visible in normal use, even with the very best lenses - only when you crop heavily it would probably show some minor advantage. Very minor increase in resolution, and probably a very minor increase in DR depending on how you use the camera and PP software; nothing like the huge improvements that were initially reported (as usual with new gear ...).
technic wrote:
Just to be sure, my 100-400II is in spec according to Canon service and perfectly fine for long distance shots.
By now I have verified the exact same problems on two other 80D / 100-400II combo's so I am absolutely sure my gear is representative, and quite a few others have mentioned on the forum they are noticing more-or-less the same problems. As I have mentioned before, the huge PDAF backfocus with 100-400II for closeups can depend on the AF system and is reported often by users with 80D, 70D and some Rebels. Maybe FF cameras don't have this problem to the same extent (it is probably partly related to the spherical aberration of the 100-400II for closeups, which may affect FF PDAF sensors differently). The "glow" problem will also be less visible on most FF bodies due to larger pixels, but I would expect it to be similar on the 5DSR. I have spotted the same glow problem in some images from people who are sure their 100-400II is perfect
The focus shift is a relatively minor issue that I would not complain about if the other problems were not present; it is only a problem because it makes the only theoretical solution for the PDAF backfocus - using Liveview focus - unreliable....Show more →
I know I am way late on responding to this, but I must. I bought the 100-400mk2 mainly for close-up type shots but also for BIF shooting, too. My copy just does not work correctly (missed focus on dragonfly heads, for example) for close up work, on any of my bodies (6D, 80D, 7Dm2, 5D4). Also, it is not very particularly sharp, either. So, I moved on to the 400 DOm2, and it is tack sharp on all of my bodies, even without AFMA. The same goes for the 500 f/4 M2 is recently got off this forum. There are definitely some not-so-great copies of the 100-400mk2 out there. The other folks that I regularly shoot with have good copies, though, so I don't believe these problems are characteristic of all of these lenses.
technic wrote:
... but as you probably know that 6 MP extra resolution (linear resolution is just 1.14x higher, almost imperceptible) is hardly visible in normal use, even with the very best lenses - only when you crop heavily it would probably show some minor advantage. Very minor increase in resolution, and probably a very minor increase in DR depending on how you use the camera and PP software; nothing like the huge improvements that were initially reported (as usual with new gear ...).
Digital image sensor technology is rather mature when it comes to efficiency and dynamic range. We gearheds tend to overemphasize small differences. Still these small differences can make different stories depending on what kind of confirmation bias we are seeking. No need to upgrade from the existing 24 MP APS-C sensor is one story. Getting your 500 mm lens croppable to almost 600 mm equivalent is another story. The resolution being so close to 36 MP FF sensors that the difference is hardly visible is a third story. Lastly, it is the sum of small differences between each generation that gets interesting. The 18 MP sensor seen in so many cameras until recently now looks dated compared to the new APS-C state of the art.
But also simplistic as your other paragraphs prove e.g. multiple use of 