Sigma 30mm f/1.4 Vs. Canon 35mm f/2

[eos 10 fan]

I saw this and thought some of y'all might be interested: <br>

<a

href="http://www.whichlens.com/index.php?blog=5&title=sigma_30mm_ex_dc_f1_4_vs_canon_ef_35mm_f&more=1&c=1&tb=1&pb=1"><b>Sigma

30mm EX DC f1.4 vs Canon EF 35mm f2</b></a> <br>

The test was done with a 350D. <p>--

[majid]

Interesting, if you compare the official Sigma MTF curve:<p>

 

<center><img src="http://www.sigma-photo.co.jp/lens/digital/img/mf_30_14.gif"></center><p>

 

With the one for the Canon 35mm f/2, you would think the opposite (keep in mind the scale on the 35mm extends full-frame).

 

<center><img src="http://www.usa.canon.com/html/eflenses/images/lenses/ef_35_2/ef_35_2mtf.gif"></center><p>

[majid]

Here is the comparison with the Canon curve stretched to make it really comparable. Either his 35mm f/2 sample was a lemon, or there is something else going on:<p>

<img src="http://www.sigma-photo.co.jp/lens/digital/img/mf_30_14.gif"><br>

<img width="420" height="237" src="http://www.usa.canon.com/html/eflenses/images/lenses/ef_35_2/ef_35_2mtf.gif">

[majid]

D'oh - of course, the MTF for the Sigma is at f/1.4, and the Canon is at f/2 and f/8.

[yakim_peled1]

If these samples are representative then this is indeed very impressive.

 

So, for double the price you get HSM and a one stop advantage but OTOH you will not be able to mount it on anything other than 1.6X DSLRs. Hummmmmm......... I'll have to think about this :-)

 

 

Happy shooting,

Yakim.

[dogbert]

Could someone please explain to me simply how to intepret these charts?

[david_lau3]

My Canon 35/f2 is much sharper than the sample shown in this test, especially performance at wide open. No sure I got a super copy or the tester got a bad copy.

[steve_dunn2]

<p><a href="http://www.photodo.com/art/Unde7.shtml">There's some information here</a> on how to read an MTF graph. The X axis represents distance from the center of the frame, and the Y axis represents how faithfully the lens reproduces the scene (it's a measurement of contrast and sharpness). For each pair of lines, you'd ideally like them to run pretty close to each other. And the higher they are on the graph, the better. A perfect lens (which doesn't exist in the real world) would simply have a straight line at 100%. MTF graphs don't measure distortion, and while various aberrations will tend to reduce the MTF scores, you can't tell how much of each type of aberration is present from the graph. Also, MTF is measured with the lens focused at infinity, and it's normal for a lens' performance to change somewhat at different distances.</p>

 

<p>One complication is that Canon's graphs show two pairs of lines, one for 10 lp/mm and one for 30 lp/mm, at each of two apertures (wide open and f/8), whereas Sigma's graphs only show two pairs of lines. Sigma's site says the lines are for 10 lp/mm and 30 lp/mm. It may say what aperture was used but I don't speak Japanese so I can't tell; I'm guessing wide open, as the results would be surprisingly poor if they were for f/8. If you want to compare these graphs to photodo's graphs, you get yet another complication: photodo uses 10, 20, and 40 lp/mm.</p>

[dogbert]

Thanks Steve. Very helpful.

[eos 10 fan]

I do not think it is wise to compare MTF graphs <br>

from different manufacturers or sources. <p>

 

--

[steve_dunn2]

<p>There can indeed be discrepancies between MTF graphs from different sources*. But since AFAIK there is no one source which has done MTF testing on both of these lenses, we don't have the option of comparing MTF graphs from the same source.</p>

 

<p>There's also the problem of sample variation. Someone has already posted that their 35/2 is much sharper than the one used in the test. Even if you can find one source which has done MTF testing on both lenses, you'd have the same problem; if they tested the reviewer's 35/2, they'd get different results than if they tested the one that someone said was better.</p>

 

<p>*: One such discrepancy is whether the results are actual or theoretical. AFAIK, Canon's MTF graphs are theoretical; rather than testing an actual sample of the lens, they have a computer calculate what the MTF graphs should be based on the design of the lens. I've also read that Canon's graphs do not show the results of diffraction, which would slightly reduce MTF at f/8. I'm not 100% sure but I think I recall reading speculation that Sigma's graphs are also theoretical. photodo's tests are done on actual lenses.</p>

[dogbert]

I don't see how one can claim a lens they own is sharper than an image shown on the web with any certainty. Pictures on the web are good for side by side comparisons of two lenses, but you can compare a compressed jpeg on the web and dismiss it as being worse than a lens in one's kit bag.

[yakim_peled1]

<p> From <a href="http://luminous-landscape.com/tutorials/understanding-series/understanding-mtf.shtml">Understanding MTF</a>: Canon's MTF charts are based on theoretical calculations used in the design of the lens, while some other manufacturers use actually measurements. <b></b> </p>

 

<p>Happy shooting, <br>

Yakim.</p>

[terry_smith2]

I think the Canon 28-1.8 would be a better comparison...

[kurt daghita photography]

Having an analytical background I fully agree that the testing of a single sample does not constitute a "valid" appraisal. Although the sample for the Sigma 30mm EX looks VERY VERY good, the sample for the Canon 35mm F/2 looks VERY VERY bad. What we need is a larger statistical base before we can come to any conclustions, but if the Sigma QC is good, I think that I for one will be seriously considering this lens.

[lu_yin]

You guys probably understand MTF better than I do. I am just curious as how accurate those measures really are.

 

Canon's own MTF charts are so optimistic. The photodo ratings, which are supposedly baesd on real lens measurements, tell different stories.

 

Sigma's MTF charts, on the other hand, are just confusing. For instance, I own both Sigma 24-70/2.8 zoom and 15/2.8 fisheye. From what I can see on Sigma's charts, 24-70 is much sharper across the whole lens with better contract wide open. I never performed MTF measurements, but I know from my experience that 15/2.8 is way sharper and contrasty than the zoom, as prime are generally better than zoom. Did Sigma do something funny with their MTF charts?

 

Finally, is there any web site that uses tools such as DXO and large sample of lenses to making meaninful measurements? There are so many rumors on the web. These opinions about lenses are generally unsupported statement based adversely selected users. Either by people who suffered from bad samples and bash the lens or from those whose sunk in costs provide much pride to praise the lens.

 

I am very interested in 30/1.4 for it's wide aperture. However, I must have some good reason to justify getting it on top of a Canon 35/2. It's quiet a high price to pay for a DC lens made by sigma. From my own experience, 35/2 is very sharp and contrasty. I am skeptical that 30/1.4 can be that much better.

 

Lu

[bens]

i like sigma, and see reasons to purchase the 30/1.4 over the 35/2. having said that, i have the 35/2 and it is awfully sharp, including at f.2, in my experience. sounds like you cannot go wrong with either lens.

[kurt daghita photography]

Hi all, I purchased this lens about 2 weeks ago and here are my findings:

1) Very Very SHARP at F/1.8 and up, sharp below F/1.8 even at F/1.4

2) Focus accuracy is very good to excellent

3) Build quality is very good

4) Fit and handeling is excellent

5) Focus ring is a bit stiff

6) Colors are very warm to hot (I like the canon better in this area)

 

I posted a single sample (full size 8 MP image) and will have more soon.

 

My conclusion:

If you need a lens in this range and have about $400.00 to spend and can live with the warmer colors, this lens should be at the top of your list. Hope this helps...

[jack_kessler]

I think Kurt Daghita may have inadvertently mis-stated his meaning. It is the Canon 35mm f/2 that looks very very good - unrealistically so. The chart for the Sigma 30mm f/1.4 lens seems unrealistically poor given the impression based on actual results that were just offered by an actual user of the lens.

[jack_kessler]

A closer analogue would be a lens with the same aperture and almost the same focal length, e.g. the Nikkor 28mm f/1.4. To make a meaningful comparison, the area to the right of about 14mm should be covered. Nikon's MTF graph is here:

http://nikonimaging.com/global/products/lens/af/wideangle/af_28mmf_14d/index.htm

 

The implication is that either Sigma has built a very poor lens or is using some very conservative measurements. If the latter, it is much to their credit, but confusing all the same.

[bard_fosse]

To compare these two lenses objective would be.

 

Sigma more than twice the weight, f/0.6 brighter, more than twice the cost, only for 1.6 crop sensor, HSM focus, and minimum twice the close up distance...

 

I personally think you get a lot of lens on both, depending on how much you're using the lens and not the least if you ever think you will use a larger than 1.6 crop sensor. I do not use the standard 50mm equivalent much, I remember even when I used film, it was classed by everyone as the most boring lens to own, everyone got one when they bought a camera so not much hip hurray for that, most people bought a 35mm to use as standard due to a tad more interesting view. For me the wide angle is most used, for photo on the street, travel, landscape... And my Canon 85mm f/1.8 is for portraits.

 

For me the Canon 35mm f/2 makes more sense because I can close focus it more (almost half) it's light, and does not burden me much when I travel, it is very sharp compared to the kit zoom and therefore fills my need for the limited amount of shots I do with it.

 

If I was mostly using a 50mm equivalent I would go for the Sigma, and it is a good value for that purpose.