Undecided on lenses?

[Milan Deo]

Hey guys so I've had my 600D with the twin lens kit for about a year and a half and I feel its time to upgrade to some better quality glass!!

So I mostly want to do rolling shots of cars during both the day and night, as well as your standard static shots and interior shots. So I was thinking of getting an ultra wide zoom lens as well as a standard zoom lens. I've got a rough budget of about $1100 Aud.

I would prefer one of the lenses to be at least F2.8 or even both. I've been looking at the Sigma 17-50mm F2.8 OS HSM and the Sigma 10-20mm F3.5 HSM how ever I've read that the Sigma 10-20 has some significant distortion at 10mm and also some cases of inconsistant sharpness? Would I be better off and wait a bit longer and getting the Tokina 11-16mm?

[nimrodisease]

As far as the UWA goes, I would say go for the Tokina. The Sigma is a very good lens, but the Tokina is better if you don't need the extra focal length. And it's only an extra ~$90 for the old version of the Tokina (which you can still buy brand new, and it's just as good as the new version).

[Bax]

I have a Sigma 10-20, and it's definitely a little soft around the edges and a little heavy on the distortion. I haven't seen the 11-16mm but if it's any good then it'd probably be a good one to have. I'm generally locked at 10mm regardless.

[Fruengalli]

Said it in the Canon forum...the Tokina is MUCH better (wait for arthurking83 to comment) Another worth looking at is the Tokina 12-24 also better than the Sigma. I've got a 17-40L & I reckon the 11-16 is a better lens. DxO rate the Tokina very well for a wide zoom.

[Milan Deo]

I guess I'll wait and go with the Tokina :) I rented one last weekend and it didnt disappoint as far as shooting a couple of my friends cars, I found it had minimal distortion at 11mm and close to none at 16mm, thought I'd get views of people who might've owned the sigma (I tend to get over excited and a tad impatient when buying gear or car parts ;) ) How ever the slow bursts of the 600D was starting to irritate me with some of the panning shots and rolling shots so I've decided to part with it and go with a 7D solely for the auto-focus and its 8fps shooting. Thanks for your input guys :)

[arthurking83]

Distortion on the Sigma is definitetly there, but never really a problem.

You're more likely to get bad distortion from bad use of the lens, than just the lens model type you choose anyhow.

I have no experience with the Sigma 10-20/3.5, and only some with the f/4-5.6 version
I can't see the benefit of the f/3.5 aperture to be honest. That's only 1/3 brighter than the cheaper version at 10mm, and not enough of a benefit at 20mm compared to the Tokina at 16mm!

Tokina seems to be the better option

(but I'll give a quick users point of view about the slower Siggy(hopefully this can also translate to the faster aperture version in some way).

From my understanding on the Sigma(f/4-5.6) is that distortion is much more noticeable on Nikon/Pentax than on Canons anyhow.
Where moustache distortion comes into the image on a 1.5x crop sensor, is slightly lessened by the 1.6x crop (narrower field of view) of the Canon format.

In over 7 years of using the Sigma myself, I never once had to look for software to un-distort any images .. (on Nikon cameras too tho!)
Just knowing that the lens can distort an image in a specific way at whatever focal length, empowers you to set up the shot so that this distortion is not going to ruin the image in some way.

Basically, it's not a major issue if you take some precautions.

BUT! as you have a preference for the faster aperture, then for sure the Tokina is the better lens .. as long as it doesn't cost an arm and a leg more than the Siggy!
if you can justify any price premium of the Tokina over the Sigma UWA .. then do so.
I also wouldn't worry about the shorter focal length range.

My only concern about the Tokina(cf the Sigma) lens is CA .. a lot more purple fringing in high contrast edges on the Tokina.
Can be easily processed out, but that may also introduce other image issues.

If people are having issues with edge sharpness on the Sigma, it's probably due to not stopping down far enough, or not focusing appropriately.


sample image:
http://www.ausphotography.net.au/forum/dbtgallery.php?do=gallery_image&id=3596&gal=gallery&type=full

crop:
http://www.ausphotography.net.au/forum/dbtgallery.php?do=gallery_image&id=3597&gal=gallery&type=full
(approximately of the red area)

This is captured on the D800E in Fx mode. The crop(where the lens was focused) is actually outside of the Dx circle .. ie. where the lens is supposed to create images.
While the crop doesn't look remarkable in any way in terms of detail, it must be remembered that its producing any detail out of the bounds of the lenses supposed working area.
And this is shot wide open, and with a CPL attached too.

Note tho, edges can be rendered unsharp if the lens has been knocked about a bit too. I guess lens elements can be skewed out of alignment over time if the lens has seen heavy use.

I know mine has, both .. heavy use, and now has some edge rendering issues.
One side is definitely sharper now than it used to be(as I remember) .. one day I'll take it to Sigma for a CLA.

Also, a small side note:
Which ever lens you choose, and subsequently decide you also want a CPL filter, keep this in mind.
(from what I've seen) the Tokina 11-16/2.8 benefits more from using a slim type filter than the Sigma 10-20/4-5.6. I got a slim type CPL for the Sigma, and it made no difference to vignetting at all(ie. a bit of a waste of money)!

So if you get the Tokina and then want a CPL think about the possible issue of vignetting and get a slim type CPL. If you get the Sigma, then a normal CPL should suffice.

I realize that my comments are not about the 10-20/3.5, but I can't imagine them being too much different in the overall scheme of things.

Tokina still sounds like the better option simply due to the advantage of aperture.
If 10mm is still the best wide angled FOV choice, I'd be inclined to forgo the extra 0.3 - 1.3Ev aperture advantage .. simply as not enough of an advantage considering the added burden(eg. 82mm filter size).

[Milan Deo]

Distortion on the Sigma is definitetly there, but never really a problem.

You're more likely to get bad distortion from bad use of the lens, than just the lens model type you choose anyhow.

I have no experience with the Sigma 10-20/3.5, and only some with the f/4-5.6 version
I can't see the benefit of the f/3.5 aperture to be honest. That's only 1/3 brighter than the cheaper version at 10mm, and not enough of a benefit at 20mm compared to the Tokina at 16mm!

Tokina seems to be the better option

(but I'll give a quick users point of view about the slower Siggy(hopefully this can also translate to the faster aperture version in some way).

From my understanding on the Sigma(f/4-5.6) is that distortion is much more noticeable on Nikon/Pentax than on Canons anyhow.
Where moustache distortion comes into the image on a 1.5x crop sensor, is slightly lessened by the 1.6x crop (narrower field of view) of the Canon format.

In over 7 years of using the Sigma myself, I never once had to look for software to un-distort any images .. (on Nikon cameras too tho!)
Just knowing that the lens can distort an image in a specific way at whatever focal length, empowers you to set up the shot so that this distortion is not going to ruin the image in some way.

Basically, it's not a major issue if you take some precautions.

BUT! as you have a preference for the faster aperture, then for sure the Tokina is the better lens .. as long as it doesn't cost an arm and a leg more than the Siggy!
if you can justify any price premium of the Tokina over the Sigma UWA .. then do so.
I also wouldn't worry about the shorter focal length range.

My only concern about the Tokina(cf the Sigma) lens is CA .. a lot more purple fringing in high contrast edges on the Tokina.
Can be easily processed out, but that may also introduce other image issues.

If people are having issues with edge sharpness on the Sigma, it's probably due to not stopping down far enough, or not focusing appropriately.


sample image:
http://www.ausphotography.net.au/forum/dbtgallery.php?do=gallery_image&id=3596&gal=gallery&type=full

crop:
http://www.ausphotography.net.au/forum/dbtgallery.php?do=gallery_image&id=3597&gal=gallery&type=full
(approximately of the red area)

This is captured on the D800E in Fx mode. The crop(where the lens was focused) is actually outside of the Dx circle .. ie. where the lens is supposed to create images.
While the crop doesn't look remarkable in any way in terms of detail, it must be remembered that its producing any detail out of the bounds of the lenses supposed working area.
And this is shot wide open, and with a CPL attached too.

Note tho, edges can be rendered unsharp if the lens has been knocked about a bit too. I guess lens elements can be skewed out of alignment over time if the lens has seen heavy use.

I know mine has, both .. heavy use, and now has some edge rendering issues.
One side is definitely sharper now than it used to be(as I remember) .. one day I'll take it to Sigma for a CLA.

Also, a small side note:
Which ever lens you choose, and subsequently decide you also want a CPL filter, keep this in mind.
(from what I've seen) the Tokina 11-16/2.8 benefits more from using a slim type filter than the Sigma 10-20/4-5.6. I got a slim type CPL for the Sigma, and it made no difference to vignetting at all(ie. a bit of a waste of money)!

So if you get the Tokina and then want a CPL think about the possible issue of vignetting and get a slim type CPL. If you get the Sigma, then a normal CPL should suffice.

I realize that my comments are not about the 10-20/3.5, but I can't imagine them being too much different in the overall scheme of things.

Tokina still sounds like the better option simply due to the advantage of aperture.
If 10mm is still the best wide angled FOV choice, I'd be inclined to forgo the extra 0.3 - 1.3Ev aperture advantage .. simply as not enough of an advantage considering the added burden(eg. 82mm filter size).

WOW! Thanks for being so comprehensive!! I do like to do rolling shots of cars at night which i think ( correct me if I'm wrong ) the 2.8 vs either the 3.5 or the 4-5.6 would make a difference in favour of the tokina in terms of possibly keeping the iso lower than the sigmas? (I'm still learning my ways around post processing so I try stay away from bumping up the iso too high)

[ricktas]

Next question...

Are you considering, or might you be considering upgrading to a full frame camera in future? If so, then you need to take into account than most of the UWA lenses are for crop sensor bodies only. If you are going to go the full frame route at some stage, then look at the Sigma AF 12-24mm f/4.5-5.6 EX DG HSM (the DG notation means full frame lens).

[Milan Deo]

I very much doubt I'll go full frame untill I'm out of uni (3 more years left) Can't justify spending on a 5DMkiii as of yet as this is stilll a hobby :)

[swifty]

I do like to do rolling shots of cars at night which i think ( correct me if I'm wrong ) the 2.8 vs either the 3.5 or the 4-5.6 would make a difference in favour of the tokina in terms of possibly keeping the iso lower than the sigmas? (I'm still learning my ways around post processing so I try stay away from bumping up the iso too high)

Do you mean the streaks of front/tail lights of cars with long exposure night photography?
If so, for a given ISO, increasing the aperture from something like f4 to f2.8 would allow for a faster shutter speed (assuming the same exposure level). You may not want a faster shutter speed though.
Shutter speed should be chosen to allow enough cars to pass by in that time to create the streaks of light you're after. So for these sort of photography where the shutter speeds are usually a few seconds, the ISO is usually not a big limiting factor and can be set to the base ISO or maybe a stop higher. The lens can also be stopped down to a more optimal aperture for better corner performance.

The faster wide angle lens have an advantage in eg. astrophotography where star trails are not wanted. A fast wide angle lens would allow lower ISO whilst keeping shutter speeds fast enough not to induce motion blurring in the stars from the earth's orbit.

Of course, as already mentioned above, the corner performance are generally better on these higher performance fast UWA's too especially when stopped down to equivalent max apertures of the slower lenses.

[Milan Deo]

Do you mean the streaks of front/tail lights of cars with long exposure night photography?
If so, for a given ISO, increasing the aperture from something like f4 to f2.8 would allow for a faster shutter speed (assuming the same exposure level). You may not want a faster shutter speed though.
Shutter speed should be chosen to allow enough cars to pass by in that time to create the streaks of light you're after. So for these sort of photography where the shutter speeds are usually a few seconds, the ISO is usually not a big limiting factor and can be set to the base ISO or maybe a stop higher. The lens can also be stopped down to a more optimal aperture for better corner performance.

The faster wide angle lens have an advantage in eg. astrophotography where star trails are not wanted. A fast wide angle lens would allow lower ISO whilst keeping shutter speeds fast enough not to induce motion blurring in the stars from the earth's orbit.

Of course, as already mentioned above, the corner performance are generally better on these higher performance fast UWA's too especially when stopped down to equivalent max apertures of the slower lenses.

By rolling shots I was more of reffering to shots such as this one http://i1284.photobucket.com/albums/a576/MilanDeo/IMG_3156_zps23030c2a.jpg (http://s1284.photobucket.com/user/MilanDeo/media/IMG_3156_zps23030c2a.jpg.html) This was take at 1/15sec F3.5 ISO-1600, I'm wondering if the 2.8 would help bring the ISO down by one stop to maybe somewhere around 800?

[swifty]

Ahh.. yes. You're referring to panning at slower hand-held shutter speeds.
With f2.8, you're better off by 2/3 stops compared to f3.5 so yes, for equivalent exposures you could reduce your ISO down to around 1000-ish from 1600.
But you should remember that you don't really have fine detail in panned shots so noise is not really a big problem. Even chroma noise isn't really an issue and you can get away with more NR since you're not too worried about removing a lot of fine detail.

[jjphoto]

Ahh.. yes. You're referring to panning at slower hand-held shutter speeds....

No. I believe he's referring to rolling or tracking shots which are car-to-car shots (ie cars driving at the same speed, one the subject the other the tracking car). Google it or have a look at http://photocornucopia.com/1042.html . Panning shots are from a stationary position. They're two different techniques with different results.

[Bax]

Ultimately rolling shots should be done with a rig. Then aperture won't even come into it.

Sent from my Nexus 5 using Tapatalk

[swifty]

Yes, sorry. My mistake. Please disregard my previous posts.

[arthurking83]

..... I'm wondering if the 2.8 would help bring the ISO down by one stop to maybe somewhere around 800?


As swifty already said, your light gathering advantage will technically be 2/3s of a stop .. but this is just the hypothetical numbers. They don't take into account T-stops, or transmission of light through the lens.

All lenses are different, and will have different T-stop values.

So(as an example) if the f/2.8 lens has a T-stop of say f/3.8 or something, and the f/3.5 lens has a T-stop value of f/3.5 .. then in terms of actual exposure you've gained nothing anyhow!

Aperture value is only a true indication of expected DOF rendering .. and sometimes not even that's as accurate as the manufacturers lead us to believe!

A word of waring tho, if you go out looking for info about the T-stop values of any lens you may be researching!

The T-stop is measured for the exposure of the entire frame, not just the brightest areas or the darkest areas.
As faster lenses vignette more than slower lenses, the inherent transmission values will generally be lower compared to a slower lens(and this is possibly only because of the vignetting!)

For an accurate idea of how well one lens works for you compared to another lens .. you should also look for a review done on both lenses, by the same sources.

A useful(but not the only) resource that could be of help in deciding some of the swings and roundabouts is DxOMark.

Just had a quick peek and compared the two lenses you have earmarked.

T-stop for the Tokina is 'measured' at T3, whereas the Sigma is T3.9. That is, the Sigma is slower(in exposure terms) than the Tokina by a larger margin than the aperture value would suggest!

So the Tokina loses 0.2Ev stop of exposure(the difference between f/2.8 and T3), and the Sigma 0.4Ev T3.9 vs it's f/3.5 aperture.
According to their graph of the lenses performance, at some points in the focal length range, they seem to match, but overall, the Tokina is a faster(or less optimistic) lens.

So the natural conclusion you would arrive at is that the Tokina is the better lens(in some ways at least) .. but if you keep looking through the comparisons at DxO .. they list the Sigma as being sharper at f/3.5, than the Tokina is at f/2.8 now!
Then, you hit the interactive widget and set the Tokina to f/4 and it's now sharper than the Sigma is going to be at f/3.5 .. and even f/4.
So now you know the Tokina is sharper, but only at f/4 .. so the benefit of the faster f/2.8 aperture has basically come to nothing! :p
Oh! ..... and not forgetting CA. The Tokina has a much stronger CA issue than the Sigma does .. so it may require more than a simple click to process out on one lens than the other.

swings and roundabouts ... and your task is to decide which of the lesser of two evils you have to put up with :D

[Milan Deo]

As swifty already said, your light gathering advantage will technically be 2/3s of a stop .. but this is just the hypothetical numbers. They don't take into account T-stops, or transmission of light through the lens.

All lenses are different, and will have different T-stop values.

So(as an example) if the f/2.8 lens has a T-stop of say f/3.8 or something, and the f/3.5 lens has a T-stop value of f/3.5 .. then in terms of actual exposure you've gained nothing anyhow!

Aperture value is only a true indication of expected DOF rendering .. and sometimes not even that's as accurate as the manufacturers lead us to believe!

A word of waring tho, if you go out looking for info about the T-stop values of any lens you may be researching!

The T-stop is measured for the exposure of the entire frame, not just the brightest areas or the darkest areas.
As faster lenses vignette more than slower lenses, the inherent transmission values will generally be lower compared to a slower lens(and this is possibly only because of the vignetting!)

For an accurate idea of how well one lens works for you compared to another lens .. you should also look for a review done on both lenses, by the same sources.

A useful(but not the only) resource that could be of help in deciding some of the swings and roundabouts is DxOMark.

Just had a quick peek and compared the two lenses you have earmarked.

T-stop for the Tokina is 'measured' at T3, whereas the Sigma is T3.9. That is, the Sigma is slower(in exposure terms) than the Tokina by a larger margin than the aperture value would suggest!

So the Tokina loses 0.2Ev stop of exposure(the difference between f/2.8 and T3), and the Sigma 0.4Ev T3.9 vs it's f/3.5 aperture.
According to their graph of the lenses performance, at some points in the focal length range, they seem to match, but overall, the Tokina is a faster(or less optimistic) lens.

So the natural conclusion you would arrive at is that the Tokina is the better lens(in some ways at least) .. but if you keep looking through the comparisons at DxO .. they list the Sigma as being sharper at f/3.5, than the Tokina is at f/2.8 now!
Then, you hit the interactive widget and set the Tokina to f/4 and it's now sharper than the Sigma is going to be at f/3.5 .. and even f/4.
So now you know the Tokina is sharper, but only at f/4 .. so the benefit of the faster f/2.8 aperture has basically come to nothing! :p
Oh! ..... and not forgetting CA. The Tokina has a much stronger CA issue than the Sigma does .. so it may require more than a simple click to process out on one lens than the other.

swings and roundabouts ... and your task is to decide which of the lesser of two evils you have to put up with :D


I would have a rather have a sharper lens in either case, however (again correct me if im wrong) the tokina 11-16 dx ii was supposed to correct the CA issues it had with its mrk i previous version? Im not too sure about its sigma counterpart however. Since its just a hobby idk i'm not that fussy on the CA aspects of the lenses as long as its crisp sharp. I dont mind paying the extra for a sharper lens as you stated the tokina is sharper than the sigma when not at 2.8 (only time i'd be relying on the f2.8 is at night and that to not being images that id be making any money off of - just looking for good quality images. So in terms of sharpness at 2.8 you would suggest id' be better off with the sigma f3.5 as my uwa? Just curious as to which one would be sharper at f8/f16?

I am aware of the vignetting of the tokina at 11mm and 90% of the time i dont mind it in terms of what i shoot so thats not an area that im too fussy about. Btw thanks so much for you input everyone :)

[ricktas]

I would have a rather have a sharper lens in either case, however (again correct me if im wrong) the tokina 11-16 dx ii was supposed to correct the CA issues it had with its mrk i previous version? Im not too sure about its sigma counterpart however. Since its just a hobby idk i'm not that fussy on the CA aspects of the lenses as long as its crisp sharp. I dont mind paying the extra for a sharper lens as you stated the tokina is sharper than the sigma when not at 2.8 (only time i'd be relying on the f2.8 is at night and that to not being images that id be making any money off of - just looking for good quality images. So in terms of sharpness at 2.8 you would suggest id' be better off with the sigma f3.5 as my uwa? Just curious as to which one would be sharper at f8/f16?

I am aware of the vignetting of the tokina at 11mm and 90% of the time i dont mind it in terms of what i shoot so thats not an area that im too fussy about. Btw thanks so much for you input everyone :)

so it would seem you have already made your decision, but don't really know it? And now that you have been given some great help, how about giving it back, with some CC of other members photos?

[Milan Deo]

so it would seem you have already made your decision, but don't really know it? And now that you have been given some great help, how about giving it back, with some CC of other members photos?

Yeop, I'll be sure to put in my fair share of input :)

[basketballfreak6]

as a previous owner of the tokina 11-16, i will tell u it's a very good lens, but do keep in mind though it is VERY flare prone which used to frustrate the hell out of me as i do a lot of landscapes

if 2.8 is not a must, i say have a look at canon's own efs 10-22 which is a very good lens

[Bennymiata]

I have the older version of the 11-16mm Tokina, and have found it to be excellent, not only in it's sharpness etc., but even more importantly, in it's great colour and contrast.
I bought it for my Canon 60D, but it also works very nicely on my 5D3 at 15 and 16mm (any shorter and it vignettes pretty seriously), but 15mm on a FF camera is very wide and wide enough for interiors etc.