1. This site uses cookies. By continuing to use this site, you are agreeing to our use of cookies. Learn More.

AP test of the Sigma SD14

Discussion in 'General Equipment Chat & Advice' started by huwevans, May 5, 2007.

  1. huwevans

    huwevans Not Really Here

    I would have tacked this onto the thread lower down, but the results in the test are so shocking I thought a new thread was merited.

    The direct resolution comparisons with a Nikon D200 were what really took me aback. It must be said that a completely greyscale target is the worst case scenario for the Foveon sensor, whislt being simultaneously the best case scenario for a Bayer sensor, but nevertheless I would not have expected to see images so dire coming from the Sigma. At worst I would expect maybe a 30% or so reduction in linear resolution for the Sigma, and probably rather less than that, because of its not having an anti-aliasing filter. But the images shown suggested something far, far worse.

    I suspect Angela will be having to field a few concerned questions from potential Sigma buyers, about whether anything other than the sensors might have been responsible for the Sigma's poor showing.
     
  2. daft_biker

    daft_biker Action Man!

    Not seen that yet but will look out for it...always good to know the grass is greener in the Canon camp ;)

    Would be nice if some of these alternative sensors did something useful...does it?
     
  3. Moos3h

    Moos3h Well-Known Member

    Wow, I've not had the chance to read the article, but this doesn't seem good.

    I was blown away by the results of the Foveon sensor when the SD9 came out (to rival the D30 and D1) and I was seriously expecting good things...

    How strange, perhaps there is a firmware issue that can resolve these problems?

    Cheers,
    James
     
  4. huwevans

    huwevans Not Really Here

    Well, I hate to break it to you but it was a Nikon that ended up on top of the podium this week! ;-) (D80, with the 30D in second, followed by the SD14.)

    But anyway, I do suspect that something wasn't quite right with the resolution test - I really would not have expected to see such awful performance, and I've seen some full resolution SD14 shots, so that's not purely going on theory.

    Anyway, sadly there seemed to be quite a few things it slipped up on - white balance wasn't great, colours always seemed to need tweaking in software after shooting, AF wasn't exactly brisk, and there was some lag before it actually began working, apparently. It did get a thumbs up for comparing well with the Fuji S5 at the lower end of the dynamic range test, though it couldn't compete at the upper end. And it got praise for its nice bright viewfinder, although the rangefinder-style extra view around the frame is gone, which might disappoint some SD9/10 users.

    On balance though it was a really poor showing. I don't normally make too much of the final percentage marks awarded, but it is perhaps notable that the SD14 was given a mere 72%, and I can't immediately recall any DSLR ever doing quite so badly in an AP test from that point of view.

    So, not good then, but I suspect we've not heard the end of it.
     
  5. huwevans

    huwevans Not Really Here


    I still suspect some problem with the testing - maybe a bad sample of the lens (Angela used two copies of the Sigma 180mm macro lens - one in the Sigma mount, and one for the Nikon, obviously), or some problem with the focussing calibration in the Sigma body, maybe, or whatever - I'm guessing of course, but something doesn't seem quite right.
     
  6. Learning

    Learning Ethelred the Ill-Named

    In spite of the delays, has the camera been released a tad too soon? If the raw is just about ok then in a final firmware the jpeg should be ok. In its present form, if the D80 beats it so would the D40x. On functionality the D40X is not significantly deficent on spec, and on image quality is as good as the D80 and D200.
     
  7. daft_biker

    daft_biker Action Man!

    Have seen some nice pics from the sigma dSLRs too...maybe they were too heavily processed to make any sort of assesment :D

    Not that I'm interested from a buyers point of view but it's nice to know how and why cameras do what they do.
     
  8. Ian_Gianni

    Ian_Gianni Well-Known Member

    I too was surprised by how bad the results are, expecially since the 3 Mpix SD-9 did hold its own against the 6 Mpix Canon D60, I would have expected th SD-14 to be in the same league of current 8-10 MPix DSLRs

    Following the Sigma forum at DCReview I discovered that there seem to be quite a few bad samples sent out by Sigma, one user claims he got a very good one, which unfortunately stopped working a few days later, and that the replacement he got had dreadful image quality (he claims his digital P&S put that second camera to shame the same way as his first sample pur his P&S to shame) .

    I wonder if the unit Angela used is one of the bad smaples, or mabe it was the 180mm macro lens that is defective.

    Maybe AP could convince Sigma to send them another unit just to confirm the results.

    BTW there is one small piece of incomplete information in the article, the "glass dust protector" mentioned in the "nice and clean" text box is actually a hot mirror, you take it out and you can shoot infrared.
     
  9. aquatarkus

    aquatarkus Member

    did you mean the Sigma forum on DPreview not DCreview?
    wonder if the sd14 they used had the original firmware v1.0 as Sigma now have 2 firmware updates which seem to correct quite a few things. Also the B/W tests will favor the D200 much more than the SD14 would much rather see colour resolution tests than b/w ones. the reviews just done by Digital Camera mag and the British Journal of Photography seem to have obtained much better results than AP has. yes the camera is probably not upto the handing of the D80/ D200 but the images i've seen taken with it are pretty stunning. Images on the Sigma stand at Focus on Imaging were all taken by members of the sigma forum on DPreview and were very impressive. have also downloaded sample raw files from some of the sigma forum members and processed with SPP v3 and must say was very impressed.
    Still think Foveon should be looking to tie up with Nikon or Canon to really develop the sensor as unfortunatly just don't think Sigma has the resorces to develop this technology and if the foveon sensor disapeared would be a real shame for everyone who takes image quality seriously

    Keith.
     
  10. Ian_Gianni

    Ian_Gianni Well-Known Member

    Yes sorry

    I can't help thinking that something happened that affected the review, while not a Sigma owner I am interested in the technology.
    Those results seem to point to a SD-14 that is worse than the old SD-10, which does not make sense to me.
     
  11. aquatarkus

    aquatarkus Member

    Yes i realy feel that something either was wrong with the camera or something went wrong with the test, Prehaps AP should request a second camera and redo the tests. I myself aren't to bothered about resolution tests and taking pictures of charts, more how the camera takes photos that i want to print and still feel that the Foveon sensor captures the sharpest images of any DSLR sensor.
     
  12. Learning

    Learning Ethelred the Ill-Named

    I still expect that the try layer is the ultimate way to go. It just seems right. Perhaps Foveon has some way to go or someone else will do it. For now I am content with the Nikon Beyer sensors which I understand are manufactured buy Sony.
     
  13. Dave_Cox

    Dave_Cox Well-Known Member

    The SD14 was top of my 'wish list' until I read the test - I'm both surprised and dissappointed so I hope that it's down to a duff camera and not the technology. As different coloured light has different wavelengths the Foveon sensor looked as if it should give better resolution than the more usual types.

    Mind you, I'd need another mortgage to buy one with all new lenses as well so perhaps it's no longer on the list.
     
  14. Ian_Gianni

    Ian_Gianni Well-Known Member

    WIthout looking to buy an SD14, I am very interested in the technology beacuse of the DP1, a digital pocket camera with an APS sized sensor and a 28mm wide lens looks extremely attractive on paper, that's why I really hope AP was given a bad sample to test.
     
  15. Alex1994

    Alex1994 Well-Known Member

    'Bad samples' of either the camera and/or lenses would put me off, as it does raise questions about Sigma's quality control ...
     
  16. aquatarkus

    aquatarkus Member

    think if you read the replys to the threads by goactive and carauction on the sigma forum on http://forums.dpreview.com/forums/forum.asp?forum=1027 you'll see that the supposed faulty cameras weren't faulty at all but simple user error and also being used to using Canon or nikon DSLRs where the work flow is quite a bit different. please have a look at the images taken by people like Olga, Chunsum, Kendal etc on the sigma forum to see what the camera is really capable of. as i said before i really hope that Sigma supply a second SD14 with the updated firmware for AP to reappraise and that people forgett about b/w resoluton charts which favour the bayer sensor type cameras but don't really tell us what the colour images will look like.
     
  17. huwevans

    huwevans Not Really Here

    For the record (because I'm not sure from what you're saying that you've actually seen the test report we're on about) there were no 'resolution charts' in the test. The target for the resolution test shots just happened to have all greyscale tones - it was something like newspaper or magazine print, rather than a test chart. But of course non-colour tones exist in 'real-world' images, just as coloured ones do, so it's at least as relevant as a test which includes all primary colours. It certainly doesn't tell the whole story but it does tell part of it - provided, of course, that there wasn't something wrong with the equipment, or whatever. We're still obviously waiting to find out if that might have been the case.
     
  18. FujiSigmaNolta

    FujiSigmaNolta Well-Known Member

    Hi all! Yes, yes, Im back. See a lot of new faces, hope everybody is alright and still enjoying their photography.

    Back on topic:

    Yes, I think that may well be the case. I remember there used to be the issue of dust with new SD10s when they came out (even though they have dust covers) and some of their lenses tend to be hit and miss. I currently have a Sigma 18-125DC which I have on my S3 pro and the thing is horrid. The worst softness in the corners I have ever seen. On the other hand, a friend of mine has the same lens and the softness is much less.

    In any case, I think the Sigma is brilliant, I have seen the samples of the SD14 and I recently bought (finally bought that is) an SD10 (second hand with 4 lenses). Even with the average lenses my kit came with, I am stunned by the image quality. I also learned that ISO 800 with this camera is not as bad as a lot of reviews make you think it is , given that you know how to process your image in Sigma Photo Pro.

    I have a really nice ISO800 example at my blog:

    ISO 800 sample

    In fact, I found the grain quite film like.
     
  19. Angela_Nicholson

    Angela_Nicholson Well-Known Member

    Hello

    Sorry I haven't responded earlier, I was let out of the office for a brief spell of R&R on holiday and didn't have Internet access. Your comments are all really interesting and its gratifying to see that we seem to have had some similar thoughts.

    We did actually have two SD14 camera bodies in for testing. The second one was sent because of my concern that the first one wasn't perfoming to standard. As it turned out both cameras had the same performance. It wasn't quite what we were all hoping for. The colour casts seen in the grey card shots appear to be the result of ugly splodges of colour which are visible at 100% on the computer screen. At a more usual printing resolution they 'merge' to give the colour cast.

    The camera test used formware Version 1.01. The most recent update to 1.02 came about as a result of me pointing out a fault with the camera.

    I have no reason to believe that the Sigma fit 180mm lens was not up to scratch. Aside from my concerns about the SD14's AF system it perfomed satisfactorily and achieved sharp focus.

    My overiding impression of the SD14 is that it is fundamentally intended for use in the raw mode. The SD10 and SD9 didn't have a JPEG format mode and perhaps this was a good thing for image quality.

    As for the scoring. We have changed our scoring system giving greater emphasis to image quality. You might have noticed that the Canon EOS 30D and Nikon D80 were both rescored to come into line with the new system and as a result got lower scores than they did originally. The Sigma SD14 cannot really compete in terms of specification with these two cameras - or in fact many other current DSLRs - so this did reduce its score dramatically. While I like the relative simplicity of the camera, it still had a couple of handling niggles so it acheived a score of 15/20 for build/handling. As you have noted the image qulaity wasn't quite up to that of a 10MP camera and this also kept its score down. I would say it is nearer to an 8MP camera than the 14MP that Sigma claims.

    Thanks for all your comments. Please keep them coming.

    Angela
     
  20. huwevans

    huwevans Not Really Here

    Thanks for the additional information, Angela. Two bodies giving the same poor result is quite compelling evidence. It's disappointing, any way you look at it - I guess we have become used to most DSLRs giving really quite respectable performance, even at the lower end of the price scale. I dare say the Sigma does that if one doesn't pay too much attention to the hyped 14MP figure, but all in all it's beginning to look very expensive for what it is.
     

Share This Page