p.4 #1 · “Full Frame Equivalence” and Why It Doesn’t Matter
carstenw wrote:
Anyway, the 1 2/3 stops is also imaginary, and doesn't apply to any actual cameras with very few exceptions. If you have an interest in specific cameras, the numbers are probably quite different. The noise comparison between the OM-D and the 5DII, to take two very popular cameras in this forum, representing MFT and FF, is nowhere near 1 2/3 stops, and is probably less than 1 stop, with the OM-D being better in some aspects, probably including low-ISO noise floor.
If we include all the sensors found in cameras of the most poluar brands, in current production until at least a year ago,
the span is from about 1 1/3 stops (worst FF vs best u4/3) to about 2 1/3 stops (best FF vs worst u4/3).
p.4 #2 · “Full Frame Equivalence” and Why It Doesn’t Matter
If I could choose only one system, then it would be FF, but because of the amount and the quality of depth of field, not noise issues.
The last time I was concerned about noise was when I was a Canon 5D and Leica M8 owner. When I bought my D3, and also now with my D800 and the Olympus E-PL3, noise is simply not an issue. In fact, I almost always use my Olympus in lower light than my D800, and I am perfectly happy with the results, even though there is noise. The noise from my E-PL3 is about a thousand times more pleasant than the noise I would get with my Canon 5D, for example, which was sprinkled with colour and very splotchy, horrendously ugly, and the Leica M8 was also not pleasing above ISO 640, and I didn't even like it that much at 640. In fact, I even quite like a little random noise in many images, as long as it has no colour sprinkles and isn't splotchy. It removes the squeaky clean digital feel and adds a little atmosphere.
I use my D800 to maximise the image quality for wide open or near wide open shots, and work on a tripod. I also don't use it all the time, but often enough for it to be of primary importance. For this the larger formats give more satisfying results. I also strongly prefer working with an optical viewfinder, flaws and all, for creative work. Electronic viewfinders function as well, but they don't feed into my creative gland like an optical viewfinder can.
Your example sounds as if you have the same concern as I, i.e. image quality with narrow depth of field. I don't think that any format has any particular advantages when stopped down, possibly other than sheer resolution.
p.4 #3 · “Full Frame Equivalence” and Why It Doesn’t Matter
alundeb wrote:
If we include all the sensors found in cameras of the most poluar brands, in current production until at least a year ago,
the span is from about 1 1/3 stops (worst FF vs best u4/3) to about 2 1/3 stops (best FF vs worst u4/3).
That's only noise performance. There is also colour fidelity, not to mention how pleasing the colour palette is, dynamic range, highlight rolloff, and so on, and FF doesn't always win, depending on the camera. As an example, Canon's current and recent lineups are quite weak in noise floor and dynamic range. You can't pull out noise and say that this is the most important and just ignore all that other stuff. The picture is incredibly complex and highly personal. To me noise has ranked near the bottom of the importance list for a few years already. Almost anything else is more important.
Could you point to some images which noise has ruined, and which would have been saved and looked nearly identical with a larger sensor camera?
p.4 #4 · “Full Frame Equivalence” and Why It Doesn’t Matter
carstenw wrote:
Could you point to some images which noise has ruined, and which would have been saved and looked nearly identical with a larger sensor camera?
I have no clue where you want to get with that question. I have never said anything close to implying such a thing.
p.4 #5 · “Full Frame Equivalence” and Why It Doesn’t Matter
I like the article, Jordan - well written and concise. Not sure why some folks are getting their knickers in a twist over it. It's a well considered opinion from a knowledgeable and talented photographer.
This is, after all, a creative endeavour for all of us. There aren't really any "rights" or "wrongs", and as Carsten says, use whichever system works for your intent.
I went through my own obsession of chasing ultimate IQ, and similar to Jordan, have since backed off to an OM-D system that I'm very happy with, but size/weight were a major factor for me. I agree that you can get "better" IQ in some instances from full-frame (not all instances though, as he described in the article), although the differences are surprisingly small.
A significant part of the perceived benefit of FF is almost certainly viewing bias. A blinded comparison would, IMHO, be much more of a wash than some might guess. That more 'relaxed' feel could easily be a subjective interpretation that we apply to the photograph, knowing in advance what gear was used to take the shot.
Jordan - I appreciate the time/effort you take in making your website and putting your thoughts down for all of us to read. I'm always impressed with folks that can manage this along with the rest of life's demands. There are lots of other blogs written by people that use/favour bigger FF gear, so yours is a valuable and useful resource that fills a niche very well.
p.4 #7 · “Full Frame Equivalence” and Why It Doesn’t Matter
bmwrider75 wrote:
A significant part of the perceived benefit of FF is almost certainly viewing bias. A blinded comparison would, IMHO, be much more of a wash than some might guess. That more 'relaxed' feel could easily be a subjective interpretation that we apply to the photograph, knowing in advance what gear was used to take the shot.
I agree with everything else you wrote, but this really not. If you look at images from the Zeiss 35/1.4 (Contax or ZF/ZE version), or the Leica 35 Lux-R, or the 50 Lux-M or 50 Lux-R E60, or the 105 on the Pentax, or the 110 on the RZ67, not to mention portraits with large format, then you will see a look which you just can't get with smaller formats. Look in the medium format or regular film threads, the Zeiss 35/1.4 thread, and the Leica R thread, and you will see some examples. If you don't find any, I can chase down a couple for you.
p.4 #8 · “Full Frame Equivalence” and Why It Doesn’t Matter
I know what you mean, and I do agree. I probably worded it incorrectly, and didn't mean it to sound like the two were "equivalent". I have shot with my own Leica R lens on crop and FF, and Zeiss C/Y, and there is a look that I can't replicate on m4/3rds for sure, and I haven't even used the more boutique lenses like you have described, which would be superior.
What I did mean was that those situations/looks are more rare than some might imagine, and that in many instances, it would be hard to tell the difference. Not all cases, of course, but for me, I was surprised to find that I was very happy with m4/3rds more than I was expecting (and I was expecting only the size/weight to be the boon, and ready to take a major hit in IQ).
I do love those Leica R and Zeiss threads though. Some wonderful images in there, although in large part their quality should also be attributed to the talent of the person pulling the trigger, not just the gear!
p.4 #9 · “Full Frame Equivalence” and Why It Doesn’t Matter
carstenw wrote:
Well, if noise doesn't ruin images, then it has zero importance, or?
I have no clue where you want to get with that question either. All I am saying that if you want to compare shutter speeds, and possible "advantages" of such in one system over another, in any meaningful way, you cannot ignore noise.
Oh, one exception. The larger sensor can take longer exposures before saturation with the same DoF that a smaller sensor. Nice if you don't want to use a ND filter
p.4 #10 · “Full Frame Equivalence” and Why It Doesn’t Matter
carstenw wrote:
(...)
Your example sounds as if you have the same concern as I, i.e. image quality with narrow depth of field. I don't think that any format has any particular advantages when stopped down, possibly other than sheer resolution.
I agree. Maybe there is something to gain with bigger formats and big fat sensors when it comes to heavy post processing. I can't say I'm concerned with that very often though.
p.4 #11 · “Full Frame Equivalence” and Why It Doesn’t Matter
carstenw wrote:
then you will see a look which you just can't get with smaller formats.
Well, you CAN get that look...it just takes a LOT of work and a very still subject, and it's pain in the butt.
For instance, these were taken with m4/3....but has a medium format level of DOF and blur, using the Brenizer method. The original image on the second one is 465 megapixels and is a stitch of 64 images with the Rokinon 85/1.4 on the GH2.
p.4 #12 · “Full Frame Equivalence” and Why It Doesn’t Matter
pdmphoto wrote:
I've shot 4x5 to 645, 35mm, APS-C and most of the smaller formats. While "full frame equivalence" is a nice term, it is just that - based on the math. I don't find it holds up well in real world terms with digital sensors. There is a reason I still hold onto my 5DII and large lenses. Sure I can get the same equivalent result with another format, but it doesn't give me the same quality results across the board.
This points to one important effect that tends to break "equivalency" between systems: lens quality/aberrations. It's easier to make a smaller-coverage lens with lower aberrations than a larger one, but not necessarily by as much as the difference in format sizes.
Look at MTFs for medium-format lenses, and they typically look like those of usable but unexceptional 135 format lenses in terms of contrast at a particular number of line pairs per millimeter. But once you consider that MF has up to twice as many millimeters to work with (so the 20lp/mm MTF curve for a MF lens should be "equivalently" compared to 40lp/mm in 135 format), it's clear that using a mid-grade lens on MF will blow nearly every "equivalent" 135 format lens out of the water in image quality (only a very few of the highest quality 135 lenses are anywhere near competitive).
Similarly, there are great lenses for smaller-than-135 formats, that can resolve line pairs per millimeter as well or better than the best 135 format lenses --- but when considering resolving power per sensor size, quality is likely to take a hit in the smaller format.
p.4 #13 · “Full Frame Equivalence” and Why It Doesn’t Matter
Format discussions seems to be the most heated debates just like brand wars / fanboy guerrillas. This has been the case for ages already. I remember reading one of my first photobooks when I was about 12 years old and a great part of the book was to "defend" Leica and the small 35mm format against 8X10 and "standard" 120 formats like 6x6 and 6x9. I remember another article in a photomagazine about the quest for the "ideal" format that was in the opinion of the writer the 6x7 format and the Pentax 67 the best camera.
It seems nothing has changed. Except, that the formats have become smaller.
The arguments have always been the same too: bigger formats are able to generate better image quality, smaller formats are easier handling, and offer their quality at lower prices. The details of that arguments go down to:
• better image quality means: more resolving power, less noise, better DOF control, better tonal response and dynamic range, and the criteria to measure them and margins that are still visible for the viewer, etc.
• Easier handling goes mostly down to more subjective criteria, but the objectives ones are size, weight, camera stabilization etc. Subjective criteria always have been about: quality of the viewfinder systems, ergonomics and handling of all kinds of settings etc.
• Arguments about lower prices often go about the apples vs oranges element in the comparisons.
This was true for film and formats and it still seems to be true in the digital era. I am waiting for the day that new technologies will make the format discussion obsolete for good. When 3D imaging techniques are going to be used to emulate large format image effects on a small sensor (narrow DOF, bokeh on demand, profile based tonal response etc.) I guess when we are gonna see that innovation this decade, the discussion will go around the argument whether syntethic large format effects are as good as real large format sensors.
p.4 #15 · “Full Frame Equivalence” and Why It Doesn’t Matter
alundeb wrote:
I have no clue where you want to get with that question either. All I am saying that if you want to compare shutter speeds, and possible "advantages" of such in one system over another, in any meaningful way, you cannot ignore noise.
Sure you can, if you don't get extreme about it. Noise is rarely an issue these days, sensors are good enough. Making it the primary discussion point makes no sense when there are much more important image quality effects resulting from aperture, colours, and so on.
If you are using any camera system seriously, then the primary interest is to get the photos you want, and your primary tools are angle of view, depth of field, and exposure. Noise is a secondary thing, and was more important in the days of film and the early days of digital. If you couldn't even get the depth of field your shot needed, noise didn't even enter the picture, pun intended. Once depth of field is equal across systems, the other differences diminish in importance. You can ignore depth of field equalisation only when it isn't important to you, which is very rare. Portraiture often means narrow depth of field, architecture and landscape usually means large depth of field, and so on. You will almost always choose depth of field before anything else, other than possibly shutter speed for shots where the amount of motion blur is crucial.
p.4 #16 · “Full Frame Equivalence” and Why It Doesn’t Matter
Jman13 wrote:
Well, you CAN get that look...it just takes a LOT of work and a very still subject, and it's pain in the butt.
Right, I love that look myself But it is a lot of work. At least with the D800, the necessity of doing exposure bracketing has diminished dramatically.
Btw, I prefer not to call it the Brenizer method, but boke panoramas. The method was not invented by him. I went to his blog some time ago to figure out when he found it and why he calls it the Brenizer method, and he essentially said that he experimented with it, and then he asked a few of his friends if they had seen that before and they said no... what a wiener. There were a few people right here on FM who were experimenting with it publicly before Brenizer did, one of which was Daniel Buck. Here is one thread with his most memorable work, and another more recent thread:
p.4 #19 · “Full Frame Equivalence” and Why It Doesn’t Matter
I have a question on this whole 'equivelence' thing. Maybe it's the question everyone ought to be asking, because it seems like people are just playing around with semantics.
If I have a lens, like let's take the Canon 85mm f/1.8 used on a 5d ('full frame' camera). It's used for portraits all the time, so often narrow DOF and pleasing 'bokeh' matters.
Now if there were to be a micro four thirds lens made, to match the same size and bulk and field of view of this lens, ie. a 42.5mm f/??, would it match the photos taken with the 85mm/FF setup in terms of DOF and 'bokeh'?
I only have a 45mm f/1.8 lens for the u4/3 system and of course it does not deliver the same narrow DOF that the 85mm f/1.8 does on the 'full frame' camera. But it's also considerably smaller. What if the 45mm for the u4/3 was the same size as the 85mm for the full frame, and hence had a considerably lower aperture? I don't know whether this would be one to two stops lower, but think of both cases.
Another comparison would be the 50mm f/1.8 on a full frame versus the 25mm f/0.95 on a micro four thirds. I know they are different lenses, hard to compare or not quite the same in terms of dimensions, but it's close to a 2-stop difference. What can one expect in terms of DOF and how quickly areas go out of focus there (I've learned not to trust the DOF calculators online because most don't take into account the real effects of background magnification).
p.4 #20 · “Full Frame Equivalence” and Why It Doesn’t Matter
For the 45, to get the same FOV and same blur/DOF as on full frame you'd need a 42.5mm f/0.9. Voigtlander has announced a 42.5/0.95 that will get you darn close, and Panasonic will have a 43/1.2 out later this year that isn't that far off.
The 50mm question is similar...the 25/0.95 will be essentially identical to a 50/1.8 in terms of DOF.
