Support NeoGAF

[Astral/H3X]

OK, maybe I'm misremembering, but I thought Tony's main point was that FF lenses on crop bodies don't give the sharpest results? Maybe there was a point about budget there too that I'm not remembering.


Cheaper, yeah, but I dunno about "better". A crop sensor on a full frame lens is using the best parts of the image circle of the FF lens.

They don't give the sharpest results, but that doesn't mean they aren't acceptable or good.

Using the "best part" of the FF lens still isn't as sharp as the whole part of an appropriate format lens. Vignetting is a very minor concern unless you are using a very cheap lens, which if you're using a full frame lens, probably isn't the case.

It's perfectly fine to use FF on APSC. Most photos have detail that people will never be able to see anyway. For most photographers it isn't something that will be noticeable or important in any way. But it is a truth, that if you take equivalent lenses, one for APSC and one for FF, the APSC one will perform better and be cheaper. Even if you have to adjust your vignetting.

EDIT:

You put a DX lens on your FX camera? The huge vignetting probably was screwing up the metering and color balance.

I actually use the Sigma 30mm eMount Art on my A7II. It's stupid sharp, as long as I crop out the corners and adjust the vignetting. Ends up being kinda like having a 35mm as opposed to 30mm, but the Sigma does a very good job of coverage.

 

[Ty4on]

[...]
The same thing happens with lenses. If one is rated at 36MP across a Full Frame, and then you put it on an APSC sensor, you're cutting out half of that 36MP, because a lot of the "pixels" that they are counting are just not being used. Technically, a 36MP APSC resolves detail greater than a 36MP FF lens.

Yeah, the pixel density on APS-C/MFT is much higher than most people think. The D810 with its "insane" 36MP is only ~15MP in APS-C crop while the 5DS is only 20MP (Canon APS-C is slightly smaller tho).

 

[Rentahamster]

Using the "best part" of the FF lens still isn't as sharp as the whole part of an appropriate format lens.

In my understanding, it usually is. A DX lens is going to have a smaller image circle to start with, and that will have the full vignetting and corner unsharpness that pertains to that image circle.

Take this for example. Thom's recommended lenses for DX bodies: http://www.dslrbodies.com/lenses/lens-databases-for-nikon/thoms-recommended-lenses.html

He doesn't even bother listing FX lenses because they're all good. The reason for the list is because there are enough DX lenses on the market that aren't that good.

I won’t get into FX lenses in this article; dozens of FX lenses are fine performers on DX cameras. The real issue is that there aren’t a lot of DX lenses that are delivering everything the sensor can resolve, so I’ll keep my comments just to those DX lenses.

 

[Astral/H3X]

In my understanding, it usually is. A DX lens is going to have a smaller image circle to start with, and that will have the full vignetting and corner unsharpness that pertains to that image circle.

Take this for example. Thom's recommended lenses for DX bodies: http://www.dslrbodies.com/lenses/lens-databases-for-nikon/thoms-recommended-lenses.html

He doesn't even bother listing FX lenses because they're all good. The reason for the list is because there are enough DX lenses on the market that aren't that good.

That's why I've taken care to say "equivalent quality" throughout. Yeah, anyone can make a shit lens, but typically shit lenses are made for the cheaper APSC "consumer" cameras. If you make a shit lens for a Full Frame, which is for professionals, you ain't sellin shit.

So if you take a good APSC lens, and a good FF lens, the APSC one will perform better.

And again, I don't worry about vignetting. Like, at all.

 

[JadedWriter]

You put a DX lens on your FX camera? The huge vignetting probably was screwing up the metering and color balance.

Oh I already knew about the vignetting, I wasn't expecting my whites to be yellow though.

 

[Skel1ingt0n]

Struggling to find a backpack.

Here's what I'm looking for:

-Very long lasting and "good looking." I'd love a Billigham or Filson or the like, but unfortunately, all their shit is supposedly very uncomfortable and doesn't hold everything well. This will be on me all day when traveling abroad, and I want it to be somewhat fashionable. STRONG build quality

-Ability to hold mirorless body, 3-4 primes, accessories, and an iPad

- weather resistant

- can carry a portable, carbon fiber tripod and ballhead either internally or on via straps on the outside - but tripod CAN'T bump into me while I'm walking

- backpack straps are wide at the top of the bag, and have some type of padding - many hikes are 5-10 miles; bag will often be on me for 15 miles a day. I'm 6'2"/240lbs, and need a larger bag for my larger frame

Nice-to-haves:

-full grain leather or leather accents
-strong, tin-cloth or similar build; thick zippers and quality hardware
-Handle on side or top to grab on the go
-Outside water bottle holder


I've looked at ONA (too small), Billinham (not comfortable), Filson (poor protection), Holdfast (ugly, too expensive for quality getting), Saddleback (way too heavy). I use an Ona Bowery in leather for my every-day-carry bag or on 3-day weekend getaways with my wife where we'll be walking around town and whatnot - perfect for my XT-1 an an extra prime. But it's just not enough when I'm gone all week.

What haven't I looked at?

 

[RuGalz]

What haven't I looked at?

My wife wanted to get me the Peak Design Everyday Backpack for Xmas but I turned it down because I just recently bought a Pacsafe bag for travel. Otherwise, it looks nice and seems very functional imo. https://www.indiegogo.com/projects/the-everyday-backpack-tote-and-sling-photography-bag#/

For longer hikes though, I still prefer the ones made for hiking and stuff the gear in inserts because they just tend to be more comfortable.

 

[Rentahamster]

That's why I've taken care to say "equivalent quality" throughout.

I'm not sure if Tony made that distinction though. I don't remember it, anyway.

So if you take a good APSC lens, and a good FF lens, the APSC one will perform better.

I dunno, there really are a lot of variables here, but if we're talking about a "generally speaking" sense, I still think the FF lenses are on par, or win out. I don't think many of them would "lose" to a DX lens, good or otherwise.

And again, I don't worry about vignetting. Like, at all.

It's not just vignetting. Corner sharpness too.

 

[Astral/H3X]

Struggling to find a backpack.

Here's what I'm looking for:

-Very long lasting and "good looking." I'd love a Billigham or Filson or the like, but unfortunately, all their shit is supposedly very uncomfortable and doesn't hold everything well. This will be on me all day when traveling abroad, and I want it to be somewhat fashionable. STRONG build quality

-Ability to hold mirorless body, 3-4 primes, accessories, and an iPad

- weather resistant

- can carry a portable, carbon fiber tripod and ballhead either internally or on via straps on the outside - but tripod CAN'T bump into me while I'm walking

- backpack straps are wide at the top of the bag, and have some type of padding - many hikes are 5-10 miles; bag will often be on me for 15 miles a day. I'm 6'2"/240lbs, and need a larger bag for my larger frame

Nice-to-haves:

-full grain leather or leather accents
-strong, tin-cloth or similar build; thick zippers and quality hardware
-Handle on side or top to grab on the go
-Outside water bottle holder


I've looked at ONA (too small), Billinham (not comfortable), Filson (poor protection), Holdfast (ugly, too expensive for quality getting), Saddleback (way too heavy). I use an Ona Bowery in leather for my every-day-carry bag or on 3-day weekend getaways with my wife where we'll be walking around town and whatnot - perfect for my XT-1 an an extra prime. But it's just not enough when I'm gone all week.

What haven't I looked at?

I have a Lowepro 400L AW or some shit that I love. Has a pull out rain sleeve that protects it, nice and sturdy, holds all my shit, and for bonus actually flips around and opens from the rear so I don't need to take the pack off to access my stuff. Love it.

I'm not sure if Tony made that distinction though. I don't remember it, anyway.


I dunno, there really are a lot of variables here, but if we're talking about a "generally speaking" sense, I still think the FF lenses are on par, or win out. I don't think many of them would "lose" to a DX lens, good or otherwise.


It's not just vignetting. Corner sharpness too.

He made that distinction when he defined "36MP FF lens" vs "36MP APSC lens". A lens that can sufficiently resolve a 36MP FF sensor will not be able to sufficiently resolve a 36MP APSC sensor -- it will only resolve somewhere around half of that. Whether you need or care about that extra resolving power is ultimately up to you, but to me, it makes a larger difference than corner sharpening.

Remember: we are talking about a theoretical case where the only differences are that one lens is FF and resolves 36MP, and the other is APSC and resolves 36MP. I'm sure it's not hard to find a FF lens that outresolves your APSC sensor, especially when most of them are 16-24MP, and many FF lenses are targetting 36MP of resolving power, but at that point you're hilariously blowing out your budget when you can just get a good APSC lens with the same quality and spend much less money, tiny easily correctable amounts of vignetting be damned.

Your argument hinges on "Well FF are higher quality than shit APSC lenses". Which yes, you're comparing to shit APSC lenses. But if you're willing to buy a new FF lens for your camera, you can save a lot of money getting a good *APSC* lens, and get a better picture. We aren't talking about random apsc lens you found in a beat up box at Best Buy.

 

[captive]

I have a Lowepro 400L AW or some shit that I love. Has a pull out rain sleeve that protects it, nice and sturdy, holds all my shit, and for bonus actually flips around and opens from the rear so I don't need to take the pack off to access my stuff. Love it.


He made that distinction when he defined "36MP FF lens" vs "36MP APSC lens". A lens that can sufficiently resolve a 36MP FF sensor will not be able to sufficiently resolve a 36MP APSC sensor -- it will only resolve somewhere around half of that. Whether you need or care about that extra resolving power is ultimately up to you, but to me, it makes a larger difference than corner sharpening.

Remember: we are talking about a theoretical case where the only differences are that one lens is FF and resolves 36MP, and the other is APSC and resolves 36MP. I'm sure it's not hard to find a FF lens that outresolves your APSC sensor, especially when most of them are 16-24MP, and many FF lenses are targetting 36MP of resolving power, but at that point you're hilariously blowing out your budget when you can just get a good APSC lens with the same quality and spend much less money, tiny easily correctable amounts of vignetting be damned.

Your argument hinges on "Well FF are higher quality than shit APSC lenses". Which yes, you're comparing to shit APSC lenses. But if you're willing to buy a new FF lens for your camera, you can save a lot of money getting a good *APSC* lens, and get a better picture. We aren't talking about random apsc lens you found in a beat up box at Best Buy.

you're making my head hurt. There is no such thing as lenses "resolving" *only* 36mp. Even film age lenses can easily "resolve" the 36mpx sensor.

furthermore a lenses resolving power doesnt change because you put it in front of a different or smaller sensor.

 

[Rentahamster]

He made that distinction when he defined "36MP FF lens" vs "36MP APSC lens". A lens that can sufficiently resolve a 36MP FF sensor will not be able to sufficiently resolve a 36MP APSC sensor -- it will only resolve somewhere around half of that. Whether you need or care about that extra resolving power is ultimately up to you, but to me, it makes a larger difference than corner sharpening.

Remember: we are talking about a theoretical case where the only differences are that one lens is FF and resolves 36MP, and the other is APSC and resolves 36MP. I'm sure it's not hard to find a FF lens that outresolves your APSC sensor, especially when most of them are 16-24MP, and many FF lenses are targetting 36MP of resolving power, but at that point you're hilariously blowing out your budget when you can just get a good APSC lens with the same quality and spend much less money, tiny easily correctable amounts of vignetting be damned.

Your argument hinges on "Well FF are higher quality than shit APSC lenses". Which yes, you're comparing to shit APSC lenses. But if you're willing to buy a new FF lens for your camera, you can save a lot of money getting a good *APSC* lens, and get a better picture. We aren't talking about random apsc lens you found in a beat up box at Best Buy.

I still feel like our terminology and definitions are not matching up with each other, which is why we don't see eye to eye on this.

 

[mrklaw]

Yeah, the pixel density on APS-C/MFT is much higher than most people think. The D810 with its "insane" 36MP is only ~15MP in APS-C crop while the 5DS is only 20MP (Canon APS-C is slightly smaller tho).

one of my favourite photos that we have printed out and framed in our house at 10x8 is a shitty old 2mp photo from a sony F505v. Looks fine.

I'm not sure I care much about MP anymore, we've had 'enough' for years now.

 

[Ty4on]

We can take this old chart:

That's line pairs per picture height where contrast is at least 50%. Remember to double it if you want to compare to pixels. For comparison a 24MP sensor is 4000 pixels tall. You can see more detail than that, but its contrast would be lower.

TL;DR, higher numbers means sharper lens.

Anyhow, this is 35mm full frame which is 24mm tall while APS-C is 15.6mm or 0.65x the height. So the hypothetical result for an APS-C (assuming mid is edge of APS-C sensor) would for the sharpest lens be:
1092 center
890 edge

And that's the point. A sharp FF lens will on APS-C perform like a softer FF lens would perform on FF. Lot's of APS-C lenses are sadly quite soft, but a lot of lenses (especially zooms) that are "good enough" for FF suddenly become quite soft on APS-C because of this.
It's a moot point if the sharpest lens happens to also be FF as well ofc.

The sweet spot argument also isn't clear cut all the time. Some lenses are really good until the very edge while some lose performance barely off center and maybe even gain a little in the edges. It's often a tradeoff between sharp center or decent corners.

Nikon: Decent corners
Sony/Canon: Sharp center

Here it's contrast on the Y axis versus distance from center on X axis.
The purple line is fine detail while the red line in more contrast and the others are everything inbetween. The red line is the "most important/noticable" in pictures, but the purple really stresses the flaws of the lens.
It's in line pairs per mm so using my old line pairs per picture height we can convert 50lp/mm into 1200lp/ph for full frame.

 

[Astral/H3X]

you're making my head hurt. There is no such thing as lenses "resolving" *only* 36mp. Even film age lenses can easily "resolve" the 36mpx sensor.

furthermore a lenses resolving power doesnt change because you put it in front of a different or smaller sensor.

When someone says "Wow this lens is sharp", they are referring to the resolving power, which is what companies/reviewers/etc are talking about when they say it's a "36MP" lens. When they are saying that a lens is "36MP" they typically mean that a lens is sharp enough, that a sharper lens won't have any noticeable change on that resolution of sensor. (to say nothing of contrast/colors/etc) I don't particularly prescribe to the idea, because you're right, lenses don't have pixels, but I understand the idea.

And no, the lenses resolving power *doesn't* change at all. But instead of using a sensor whose pixels are (totally random number) 12 microns wide, you're now using one where the pixels are (again totally random number) 6 microns wide. It's about the pixel density getting much tighter, thus softness will cover more pixels and thus make the overall image softer. It's hard to explain, but having great pixel density means you need to have a sharper image from the lens to get the same end result sharpness compared to Full Frame.

And again, none of this shit is anything I give a damn about. I use my canon 50mm for 85% of my photos. It's old as hell. It's a smidge soft. But it's sharper on my A7II than it is on my A6000, because my A6000 has a that higher pixel density.

 

[Ty4on]

When someone says "Wow this lens is sharp", they are referring to the resolving power, which is what companies/reviewers/etc are talking about when they say it's a "36MP" lens. When they are saying that a lens is "36MP" they typically mean that a lens is sharp enough, that a sharper lens won't have any noticeable change on that resolution of sensor. (to say nothing of contrast/colors/etc) I don't particularly prescribe to the idea, because you're right, lenses don't have pixels, but I understand the idea.

And no, the lenses resolving power *doesn't* change at all. But instead of using a sensor whose pixels are (totally random number) 12 microns wide, you're now using one where the pixels are (again totally random number) 6 microns wide. It's about the pixel density getting much tighter, thus softness will cover more pixels and thus make the overall image softer. It's hard to explain, but having great pixel density means you need to have a sharper image from the lens to get the same end result sharpness compared to Full Frame.

Let's make up a new lens called ABC which is a tiny bit soft on the Nikon D810, but fine.
We'll also invent a new Nikon camera, the APS-C and 15MP... D1500? D1500.

Anyhow, you put the new ABC lens on your D810 and take a picture. Then you crop the center portion so it's just 15MP big. Tada! You now have the exact same image as if you'd used the D1500. Not bad, but the D810 picture is probably way sharper in comparison.

I find that to be the easiest way to explain it.

I still feel like our terminology and definitions are not matching up with each other, which is why we don't see eye to eye on this.

I think some of the more expensive ones really do. Especially those made for APS-C only systems like Fuji's lenses.

 

[mmazurkiewicz]

anyone here know much about action cams? is there any discussion about em here?
or, what is the place to look to get info about em?

 

[Astral/H3X]

Let's make up a new lens called ABC which is a tiny bit soft on the Nikon D810, but fine.
We'll also invent a new Nikon camera, the APS-C and 15MP... D1500? D1500.

Anyhow, you put the new ABC lens on your D810 and take a picture. Then you crop the center portion so it's just 15MP big. Tada! You now have the exact same image as if you'd used the D1500. Not bad, but the D810 picture is probably way sharper in comparison.

I find that to be the easiest way to explain it.


I think some of the more expensive ones really do. Especially those made for APS-C only systems like Fuji's lenses.

Yeah, essentially, you're zooming in on the softness that's there.
Except to make it worse, the "D1500" would be 24MP or some shit, so pixel peeping would be that much worse haha.

 

[Ty4on]

Yeah, essentially, you're zooming in on the softness that's there.
Except to make it worse, the "D1500" would be 24MP or some shit, so pixel peeping would be that much worse haha.

The D1500 was just to make it apples to apples

A better example would be a ~56MP D900 and a 24MP D3400/D5500/D7200 which would have the same pixel pitch.

 

[Rentahamster]

Let's make up a new lens called ABC which is a tiny bit soft on the Nikon D810, but fine.
We'll also invent a new Nikon camera, the APS-C and 15MP... D1500? D1500.

Anyhow, you put the new ABC lens on your D810 and take a picture. Then you crop the center portion so it's just 15MP big. Tada! You now have the exact same image as if you'd used the D1500. Not bad, but the D810 picture is probably way sharper in comparison.

I find that to be the easiest way to explain it.

Err...So this fictional D1500 is a DX sensor at 15 megapixels. The D810 in DX mode is 15 megapixels too. This means that they are the same thing, ceteris paribus.

If I take a picture of an apple with the D810 using the ABC lens in FX mode, then take the same picture of the same apple with the same ABC lens with the same distance to subject in DX mode, when I zoom into 100%, both pictures should look exactly the same.

If I use the "D1500" instead, which has the same photosite density as the D810 in DX mode, it should look no different.

 

[Astral/H3X]

Err...So this fictional D1500 is a DX sensor at 15 megapixels. The D810 in DX mode is 15 megapixels too. This means that they are the same thing, ceteris paribus.

If I take a picture of an apple with the D810 using the ABC lens in FX mode, then take the same picture of the same apple with the same ABC lens with the same distance to subject in DX mode, when I zoom into 100%, both pictures should look exactly the same.

If I use the "D1500" instead, which has the same photosite density as the D810 in DX mode, it should look no different.

This is true. The problem comes up when that "D1500" ends up actually having much smaller photosites. In doing that, you get a much finer sensor detail... on the softness of the lens. So a 100% crop on your typical APSC camera, which has that higher pixel density, the 100% crop will be much softer.

 

[Rentahamster]

This is true.

Then this:

Tada! You now have the exact same image as if you'd used the D1500. Not bad, but the D810 picture is probably way sharper in comparison.

Is false.

 

[Ty4on]

Then this:

Is false.

I phrased that wrong. I meant if you used only the center part they'd be identical, but if you'd use all of the frame and then the d810 would be more than twice the resolution hence much sharper.

 

[Rentahamster]

No. You don't view images at 100%, you scale the entire image. They're "equally sharp" at 100%, but one is twice the resolution.

If you need to crop the center frame then and only then is the resolution identical.

They're both 15 megapixels, end to end. (D1500 and D810 in DX mode). Same pixel pitch.

If you need to crop the center frame then and only then is the resolution identical.

Shooting a DX 15MP sensor is the exact same thing as shooting in DX mode on a D810, which produces the exact same image as in FX mode, just cropped.

 

[Ty4on]

They're both 15 megapixels, end to end.

Second paragraph.

 

[Rentahamster]

Second paragraph.

This?

Anyhow, you put the new ABC lens on your D810 and take a picture. Then you crop the center portion so it's just 15MP big. Tada! You now have the exact same image as if you'd used the D1500. Not bad, but the D810 picture is probably way sharper in comparison.

And what I'm saying is the "D1500" is the same as the D810 in DX mode, which is the same as the D810 if I cropped it in the computer after, but in each case, they all look the same when at 15 megapixels.

It's the same pixel pitch, same megapixel count, same lens, same distance to subject.

 

[Rentahamster]

I phrased that wrong. I meant if you used only the center part they'd be identical, but if you'd use all of the frame and then the d810 would be more than twice the resolution hence much sharper.

If you're filling the frame, then you're changing the distance to subject, and now you're not comparing apples to apples anymore.

 

[Astral/H3X]

If you're filling the frame, then you're changing the distance to subject, and now you're not comparing apples to apples anymore.

If you're going to go off on that point, then it's impossible to do an apples to apples comparison, because FF and APSC are not both apples.

Long story short, a well designed quality lens targetting APSC will be sharper than an equally well designed lens targeting FF, because APSC's higher pixel density requires it to be so. Ergo, you will get the best sharpness using a high quality APSC lens if you are using an APSC lens.

We aren't talking about "Generally FF lenses are better than your average APSC lens". We aren't talking about same focus distance. We aren't talking about cropping your FF lens. Straight, apples to apples, good glass against good glass, the resolving power per micron of an APSC lens should be higher than FF.

I don't see why it's so hard to understand that APSC cameras have higher pixel densities than their FF counterparts. I don't see why it's so hard to understand that a lens being able to sharply resolve a FF 36MP sensor would not necessarily be able to sharply resolve a 36MP APSC sensor. It's easy. One has smaller pixels, the other doesn't.

 

[Rentahamster]

I don't see why it's so hard to understand that APSC cameras have higher pixel densities than their FF counterparts.It's easy. One has smaller pixels, the other doesn't.

We already agree on this. (although recently with 24MP crop and 50MP full frame, they're about the same on those particular sensors).

 

[JadedWriter]

Why don't Full Frame raws look as good as when you finally export them to Jpeg with all the changes?

 

[StopMakingSense]

Why don't Full Frame raws look as good as when you finally export them to Jpeg with all the changes?

What?

 

[Rentahamster]

Why don't Full Frame raws look as good as when you finally export them to Jpeg with all the changes?

If you're going from prophotoRBG to sRGB, that might be one reason. Other than that, it should look exactly the same.

 

[JadedWriter]

What?

They seem cleaner looking as jpegs than as raws. Maybe this makes no sense at all.

 

[Rentahamster]

If you're going to go off on that point, then it's impossible to do an apples to apples comparison, because FF and APSC are not both apples.

If you keep the lens the same, the aperture the same, the distance to subject the same, and only change the sensor, then you are doing a test that minimizes the variables and keeps the only thing that you are concerned about (the sensor) constant.

But fine, in real world usage scenarios, people will probably want to take pics that fill the frame (but then that makes it lose validity as an actual comparison test), but it still doesn't really make a difference, which I'll show later.

We aren't talking about "Generally FF lenses are better than your average APSC lens". We aren't talking about same focus distance. We aren't talking about cropping your FF lens. Straight, apples to apples, good glass against good glass, the resolving power per micron of an APSC lens should be higher than FF.

Doesn't that mean we are talking about FF lenses are (not) better since they don't have as high a resolving power? But anyway, we're mixing up that whole scenario that ty40n put out against what the original point is: the statement Tony Northrup said. Which is:

Use crop lenses on crop sensors for sharpest results. Which would imply that using full frame lenses on crop sensors does not give you the sharpest results. And as far as I know, that's not really true. And if it is in some cases, it's very rare.

But that's the point you agree on with Tony, and which you reiterate here:

Long story short, a well designed quality lens targetting APSC will be sharper than an equally well designed lens targeting FF, because APSC's higher pixel density requires it to be so. Ergo, you will get the best sharpness using a high quality APSC lens if you are using an APSC lens.

(I'm going to assume you meant to say "if you are using an APSC sensor" at the end there)

I went to DXOmark to see what the deal is, since even Tony says to go there himself, and the comparisons I did didn't make much of a difference in "perceptual megapixels".

Now, it's kinda hard to find "good" DX lenses since Nikon doesn't really make them, but I tried to work with what I had.

Example 1:

35mm FX prime vs 35mm DX prime on a 24MP body:

https://www.dxomark.com/Lenses/Comp...35-mm-f-1.8G-on-Nikon-D7100__1294_865_313_865

The FX lens has 4 more "perceptual megapixels" (whatever that means) than the DX version, which in DXO-speak would indicate that the FX version is sharper on the DX sensor.​

Example 2:

Both 24-70 f2.8 lenses vs 17-55 f2.8 lens on a DX body. (the 17-55 is Nikon's most expensive DX lens, I think)

https://www.dxomark.com/Lenses/Comp...D-VR-on-Nikon-D7100__173_865_173_865_1583_865

According to the DXOmark scores, the sharpness is the same.​

OK, so what about real world tests now? What if we use different parameters and keep the field of view constant (i.e. fill the frame the same way on FX and DX) ? Well, this guy did some, and his results indicate that there really isn't much of a difference.

Keeping the lens constant and changing the sensor (FX sensor vs DX sensor):
https://www.youtube.com/watch?v=YlT7jUtF1o0

Keeping the sensor constant and changing the lens (FX lens vs DX lens):
https://www.youtube.com/watch?v=00mNef6m5IE

 

[Ty4on]

The Nikon 17-55 is ancient (2003) and crap. The 18-55 kit lens also gets 9MP from DXO.

9MP on DXO is a pathetic result regardless. Kinda justifiable with the kit lens, but unacceptable from a 2 grand lens.

The Nikon 35mm result is interesting though. Guessing that's where the extra cost of the FX went into.

A lens doesn't turn useless on APS-C, but if the time and energy designing it for FF had been used to optimize it for APS-C it could have been better. Someone buying the 24-70mm would be really disappointed because they get a bad zoom range, not terribly fast lens and certainly not sharp for the price when the kit-lens is comparable. Nikon has a really bad DX lineup, but for the most part equivalent DX lenses are way cheaper.

Roughly same price class:

Half the price and twice as fast:

Roughly the same price new (and you get a zoom...):

They trade blows a bit, prime sharper stopped down, but the zoom is even sharper at 24mm f2.8. I think it's a no brainer though unless you get the prime very cheap.

The latter one is an older lens, but that is partly because people often put older, cheaper lenses on cheaper bodies. Cheap 50s etc. do great, but especially wider lenses from the film era do poorly by modern standards and aren't as cheap. Some stand out (and are really expensive), but some are just expensive for no reason other than people really liked it in the 70s.

 

[Rentahamster]

The Nikon 17-55 is ancient (2003) and crap. The 18-55 kit lens also gets 9MP from DXO.

Ah. Admittedly, I'm not very familiar with what the latest and greatest DX lenses are. I wanted to do the Sigma 50-100 f1.8 ART or whatever it is, but dxo didn't have the score ready for it yet.

 

[Rentahamster]

This one:

I guess it's a fair comparison on price if they're about the same, but is it a fair comparison on relative quality? The Sigma is a fixed aperture, while the Nikon is a variable aperture.

A lens doesn't turn useless on APS-C, but if the time and energy designing it for FF had been used to optimize it for APS-C it could have been better. Someone buying the 24-70mm would be really disappointed because they get a bad zoom range, not terribly fast lens and certainly not sharp for the price when the kit-lens is comparable. Nikon has a really bad DX lineup, but for the most part equivalent DX lenses are way cheaper.

Yeah, it's cheaper and all, but Tony's argument of "for sharpest results, use crop lenses on crop sensors" still doesn't hold much water.

 

[Ty4on]

I guess it's a fair comparison on price if they're about the same, but is it a fair comparison on relative quality? The Sigma is a fixed aperture, while the Nikon is a variable aperture.

Yeah, it's cheaper and all, but Tony's argument of "for sharpest results, use crop lenses on crop sensors" still doesn't hold much water.

Fixed vs variable aperture doesn't affect image quality, it's a design decision. Leica's 24-90mm is huge and really expensive and a variable aperture zoom. Nobody (I've seen) is complaining that it is soft. Consumer lenses tend to be variable because they tend to have a long range.

In a perfect world the crop variants would be sharper, but some (especially Nikon) focus much more on FX and really only make budget lenses for DX. The Sigma for example is significantly faster and sharper than any FF zoom. In DXO's testing the Sigma 24-35 is a little worse than the 18-35. That's really impressive for a FF lens, but for any APS-C shooter the 18-35 is a no brainer. Faster, cheaper, smaller and more zoom range.

 

[Rentahamster]

Fixed vs variable aperture doesn't affect image quality, it's a design decision.

It's not a direct factor of image quality, no, but generally speaking, fixed aperture zooms are of better quality than their variable aperture counterparts since it requires better materials and construction to achieve a fixed maximum aperture on all zoom ranges.

 

[Ty4on]

It's not a direct factor of image quality, no, but generally speaking, fixed aperture zooms are of better quality than their variable aperture counterparts since it requires better materials and construction to achieve a fixed maximum aperture on all zoom ranges.

Nah, that's just the design. Generally more expensive lenses have been fixed aperture while cheaper have been variable, but back in the early 70s almost every zoom was bad and fixed aperture.

This simple 25-100mm zoom is fixed aperture, but you can see on the wide setting how little of the front elements are used. With a variable aperture you can use some of that wasted glass.

That's a Cooke Triplet in the front. The Cooke Triplet was used as a cheap prime, but by moving the middle element you can create one of the simplest zoom designs.

 

[RuGalz]

This simple 25-100mm zoom is fixed aperture, but you can see on the wide setting how little of the front elements are used. With a variable aperture you can use some of that wasted glass.

Not sure what you mean by this part. The diagram shows 2 light rays but lights are going to fill the whole element regardless and will be projected onto the sensor. I think it's trying to show if focusing on the same object from the same distance using 50mm vs 100mm.

Tony's argument of "for sharpest results, use crop lenses on crop sensors" still doesn't hold much water.

I think it makes a bit more sense if it was "for the sharpest results, use lenses that are designed for your sensor density". There are so many uncontrollable variables that in real life it just doesn't matter that much as long as a lens is 'sharp enough'. I can't find the quote but I remember reading a Zeiss lens designer saying that people are so focus on sharpness alone these days that they sacrificed other characteristics for the newer modern design that himself really dislikes but no one notices because everyone just focuses on sharpness rating these days.

 

[Ty4on]

Not sure what you mean by this part. The diagram shows 2 light rays but lights are going to fill the whole element regardless and will be projected onto the sensor. I think it's trying to show if focusing on the same object from the same distance using 50mm vs 100mm.

The elements in the rear would cut those light rays off. Maybe a bad example because of the tiny aperture, but if you zoom any zoom you'll see from the front that the aperture appears to get smaller as the focal length is reduced.

Here's the source with plenty of animated examples where you can see internal vignetting making the aperture smaller.
http://www.pierretoscani.com/echo_telezooms_english.html

**** Very important, when I say aperture I don't mean f stop, but the physical size of it. F stop is the divisor you use for focal length so a 50mm lens with a 50mm aperture is f/1.

 

[RuGalz]

That page made a lot more sense. Thanks.

 

[brerwolfe]

Struggling to find a backpack.

I've looked at ONA (too small), Billinham (not comfortable), Filson (poor protection), Holdfast (ugly, too expensive for quality getting), Saddleback (way too heavy). I use an Ona Bowery in leather for my every-day-carry bag or on 3-day weekend getaways with my wife where we'll be walking around town and whatnot - perfect for my XT-1 an an extra prime. But it's just not enough when I'm gone all week.

What haven't I looked at?

My wife bought a bag for me from Kata in 2013. It's been through some shit and it still works amazingly well. Very strong, durable, manageable. It was pricey, but I'm sure I've gotten far more usage out of it than dollars spent.

 

[Rentahamster]

I think it makes a bit more sense if it was "for the sharpest results, use lenses that are designed for your sensor density".

That doesn't necessarily always give you the sharpest results, though. Nor is it always the best option in terms of budget/usability/longevity. As you said, there are many different variables at play as to what lens a person should buy.

 

[RuGalz]

That doesn't necessarily always give you the sharpest results, though. Nor is it always the best option in terms of budget/usability/longevity. As you said, there are many different variables at play as to what lens a person should buy.

Not sure how much more general you can get. Basically if a lens has higher resolving power it will be sharper for the area that it covers, doesn't matter what format the lens is designed for. (Or pretty much: sharper lens is sharper, well duh) Either way, I'm not the one obsess over sharpness or corner to corner sharpness. Most of the modern lenses are good enough to me so the other characteristics like color responses are more interesting to me.

 

[TerryLee81]

This photographer posted lots of sample pictures of the new Olympus lenses:

http://buchangrant.format.com/Blog/the

Looking good!

 

[Ty4on]

Such a useless thing in the grand scheme of things, but I like how quiet the shutter in the Fuji GFX seems to be. Especially after the Sony A7r showed us mirrorless didn't automatically mean silent

If the extra volume on those sensors can push prices down to ~5k it could be a decent seller. I guess lenses really make or break this, but Fuji did well with the X-mount. I don't think it'll be as popular with adapters as the A7 was (crop, much bigger adapters), but who knows.

 

[Skel1ingt0n]

Such a useless thing in the grand scheme of things, but I like how quiet the shutter in the Fuji GFX seems to be. Especially after the Sony A7r showed us mirrorless didn't automatically mean silent

If the extra volume on those sensors can push prices down to ~5k it could be a decent seller. I guess lenses really make or break this, but Fuji did well with the X-mount. I don't think it'll be as popular with adapters as the A7 was (crop, much bigger adapters), but who knows.

The best kept secret in photography right now - that more people are admittedly catching onto - is that Fuji is producing some of the absolute best glass on the market right now; and almost all their lenses are a good bit under a $1000 second-hand.

 

[Comalv]

Hey GAF!

I am struggling to decide on what compact camera to buy

The main things I am looking for is
- pocketability (doesn't have to fit in jeans pockets but in a jacket would be nice, as long as I don't have to have a separate bag/holster for it is small enough for me)
- fast lens/image quality so I can take RAWs with a decent amount of bokeh

Everything else is not as important (although unique features are welcome)

I have a Nikon D800 with the standard set of zooms and primes for everything serious, I am looking for this compact camera to take with me on small holidays and still be able to shoot a natural light portrait if I feel like it.

the main camera I was interested in was the Nikon DL but it keeps getting delayed and by the time it will actually hit the market it will be already old. I also probably need a compact before that happens (probably for november/christmas holidays)

The main contenders to that seem to be the following:

-Sony RX100 MkIV: good overall, image quality looks great, bokeh not so much but it seems usable, has some neat video features like super slo mo which might be fun once in a while. Has EVF and popup flash in case I need to trigger an external flash as optical slave

-Panasonic Lumix LX10 (LX15) - Announced at photokina, seems better than the RX100 mainly because it's newer so it has a better sensor, low light performance and lens. No EVF is a bummer, the problem is that it's good on papers but I don't know if there will be enough reviews by the time I need to actually make a purchase, if anyone tried it at Photokina please let me know

-Panasonic Lumix LX100 - similar Image Quality to the RX100, has a bigger sensor for better bokeh. It is however way to big for a compact from what I have seen, so that could be a deal breaker. It's also as old as the RX100

Let me know your suggestions!

 

[Skel1ingt0n]

Hey GAF!

I am struggling to decide on what compact camera to buy

The main things I am looking for is
- pocketability (doesn't have to fit in jeans pockets but in a jacket would be nice, as long as I don't have to have a separate bag/holster for it is small enough for me)
- fast lens/image quality so I can take RAWs with a decent amount of bokeh

Everything else is not as important (although unique features are welcome)

I have a Nikon D800 with the standard set of zooms and primes for everything serious, I am looking for this compact camera to take with me on small holidays and still be able to shoot a natural light portrait if I feel like it.

the main camera I was interested in was the Nikon DL but it keeps getting delayed and by the time it will actually hit the market it will be already old. I also probably need a compact before that happens (probably for november/christmas holidays)

The main contenders to that seem to be the following:

-Sony RX100 MkIV: good overall, image quality looks great, bokeh not so much but it seems usable, has some neat video features like super slo mo which might be fun once in a while. Has EVF and popup flash in case I need to trigger an external flash as optical slave

-Panasonic Lumix LX10 (LX15) - Announced at photokina, seems better than the RX100 mainly because it's newer so it has a better sensor, low light performance and lens. No EVF is a bummer, the problem is that it's good on papers but I don't know if there will be enough reviews by the time I need to actually make a purchase, if anyone tried it at Photokina please let me know

-Panasonic Lumix LX100 - similar Image Quality to the RX100, has a bigger sensor for better bokeh. It is however way to big for a compact from what I have seen, so that could be a deal breaker. It's also as old as the RX100

Let me know your suggestions!

Professional reviews are a bit more picky; but I don't think I've run into a single person with any variation of the Sony RX1000 that hasn't been very satisfied with their camera.

It's supremely portable, has great IQ, great build quality, and you can add to it as needed. You might be fine saving money on a Mk II or III, instead - just a thought. But if I were in your shoes, and I didn't want to spend Leica Q money, it would be my very first pick.