Home > Contributors > Blog

Blog

Are you having issues passing Alamy Quality Control?

Permalink Comments (34)30 May 2008 at 12:33 by Patrick Ashby - QC team
Posted under Advice and tips

Blue toy robot wind up toy sparking - Image AR3A2A © Garry Gay
© Garry GayAt Alamy we like to have happy contributors as well as happy customers, which is why QC is such an integral part of the process. A happy contributor is one whose work passes our QC and a happy customer is one who can purchase images knowing they are of the highest standard.

Our QC Team (consisting of real people not robots, contrary to some rumours) have devised a short checklist with the aim of helping you improve your QC pass rate.

  • Make sure that you check every single image you submit.
  • View your images at 100%. It is important you look at your images at this zoom as it is the size at which we check image quality and look for any faults. There is no point in checking them at 50% or 200% as you won’t be able to see the same problems that you will at 100%.
  • Check your file size. Interpolate your images so they have an uncompressed, 8 bit tiff file size of 48-50Mb. If you have a camera that can produce an uncompressed file size of over 48Mb out of the camera then leave them at that size. We would rather see a higher quality 48Mb image, than the same image at 90Mb.
  • If you have failed QC go to Track submissions, within My Alamy to see why your image failed by clicking on the link for a more detailed explanation.
  • Don’t just resubmit a failed image or take the offending image out of the submission. Check the rest of your submission as there may be problems with other images in it.
  • Read the Submission guidelines on our website. Our QC staff have put these together to help you and they contain the answers to most QC questions that you may have and there are even visual examples of the failure reasons.

Still struggling? Then have a look on the Alamy forum where a fellow contributor will be able to give you some friendly help and advice. Also, don't forget to keep your eyes on the blog for regular updates from the Quality Control team over the coming months.

Top

Add your own commentComments (34)

  1. 30 May 2008 at 13:57 Kumar Sriskandan

    Please could someone tell me waht "Image processing error" means when an image is rejected? I've had several of these lately having previously had no problems -
    Are these a fault of my processing, or a problem at Alamy's end, and what (if anything) can I do about it?
    Thanks for all advice received!!

  2. 30 May 2008 at 16:28 John McKenna

    Hi Kumar,

    An "Image processing error" usually means that the image in question has been corrupted during the upload process and has nothing to do with the quality of the image or your work process. Uploads like this will normally sit at the can take 25 days to process stage for up to a couple of weeks, the uncorrupted images will pass QC but the batch will be marked as a partial fail due to the corrupt image. The reason for rejection is not necessarily your fault or Alamy's, these things can happen when sending files through the Internet.

    I hope this explanation is of some help to you.

  3. 30 May 2008 at 17:16 Kumar sriskandan

    Hi John - thanks for that - I wonder why there seem to be quite a few recently whereas I never had one till a couple of months ago....?

  4. 02 June 2008 at 10:34 Ed Book

    'seems like images from agencies aren't held to the same standards as individual contributors... for example images marked as having both model and property releases when there are no people in the image and no private property.

    Peace,
    Ed Book

  5. 02 June 2008 at 11:19 Tony lilley

    I also fully agree with Kumar and Ed. I too have had failures with images that I have checked at 100% as I do with all images. This is also very recently but OK in the past. Some Agency images I have looked at would fail even when viewed at 450 pixels wide!

  6. 04 June 2008 at 11:03 Simon king

    Are you saying then that files should be reduced to 8 bit before saving as tiffs?

    I just tested one and it was 72mb converted from a 16bit RAw file, but only 32 when converted from an 8 bit file. (no uprezzing)

    So are you saying you need files converted from 8 bit files- in which case I need presumably to uprez before converting to 8 bit then to tiff?

  7. 05 June 2008 at 11:55 Mary Clark

    Since February 08 I have experienced endless QC rejections for Interpolation Artifacts for which
    Alamy give the following definition.
    “ The image exhibits rather intrusive artifacts and degradation of detail resulting in an overall soft look, a “mottled” texture or "jaggies" on edges. These symptoms occur when you interpolate an image beyond the capabilities of the camera from which it came.”

    The example which is given is an all over “disturbance pattern “ which can be seen on the following link http://www.alamy.com/images/qc/interpolation-large.jpg

    I am using a Nikon D200 with a 70- 300mm VR Nikon lens .

    If I am pushing the processing of my images beyond the capabilities for this set up ,why do some images pass because they do not demonstrate IA and other's fail ,when they are all processed in exactly the same way ?

    Can the subject of the picture actually influence the processing so that it produces Interpolation artifacts ?

    Many Thanks

  8. 05 June 2008 at 18:55 Tony Watson

    Mary, You won't get an answer here, you need to ask you question on the alamy forum.

    Tony Watson

  9. 05 June 2008 at 19:56 Mary clark

    Thanks Tony, at least somebody found me.

    Whats this one for then ?

    I identify with the picture , cross , sparky and a bit stiff

  10. 05 June 2008 at 21:14 Kumar sriskandan

    Hi Mary - looking at the picture at full size, which you have posted I have to say that there is a latrge amount of what I think of as JPG artefact, which is most visible in the areas of one colour, eg. the sky or the grass, which looks like little "squiggles" in the image - I presume that this must be due to post-camera processing, but I'm not surprised it failed QC (sorry!) - I agree with Tony that this would be well worth posting in the Alamy forum
    Kind regards
    Kumar

  11. 06 June 2008 at 01:13 Mary clark

    Hi Kumar,
    Your absolutely right. The posted image is Alamy's example , and not my image at all .
    Its dreadful isn't it , quiet distinctive and not easily missed .
    Thats what is supposed to be present ,but only in some of my work , and not all ,which makes no sense to me , since it is all processed the same way .
    Thanks for your reply .

  12. 06 June 2008 at 01:26 Mary clark

    I also have a camera that can produce an uncompressed file size of over 48Mb out of the camera which I leave untouched .
    I don't knowingly do any interpolation .

  13. 09 June 2008 at 05:36 Carol Buchanan

    Mary, in comment # 12 you said your camera (a Nikon D200 - according to your post #7) can produce an uncompressed file of over 48mb out of the camera. This would be a 16 bit file, (about 60-62mb by my calculation for the D200 @ 10.2mp), which would reduce back to half when converted to an 8bit for Alamy, leaving you just 30-31mb.

    So you must be interpolating somewhere.....

  14. 11 June 2008 at 11:57 Mary clark

    Hi Carol ,
    Yes your right Photoshop CS3 ,apparently does it automatically.I had'nt twigged that was what was happening. It doesn't alter the fact that I cannot find the Interpolation artifacts as illustrated by Alamy.
    The all over jagged pattern is quiet distinctive ,and I would have thought something you could not miss.

  15. 12 June 2008 at 03:18 Carol Buchanan

    Mary

    CS3 doesn't do it "automatically" - you must be asking it to do it somewhere, or have something setup, somewhere to do it. You should try and work out how/where this is happening so you can regain control of your workflow and ensure your interpolation is done correctly.

    I use Genuine Fractals, but understand that bicubic smoother is as good, or better than GF.

    Re IA - they can be sneaky little blighters! And not always huge big jaggies!

    Good luck in improving your acceptance rate.

  16. 12 June 2008 at 10:18 Fen Oswin

    Mary,

    As Carol said CS3 won't do it automatically. Can I suggest that you write down your full workflow and post it on the Yahoo Forums? Then we can see:
    - A problem in your workflow
    - or An issue at Alamy.

    Regards,

    Fen

  17. 12 June 2008 at 10:38 Mary clark

    Thanks Carol,
    Good place to swap ideas .Yes the program can be set up to do it automatically .Its part of the preferences when you first setup the software .
    Enlarging an image is a straight forward technical procedure ,and nothing too mind boggling. Other's contributors use exactly the same method with no problems at all .
    In your experience how small can your "no go" jaggies get? and are yours an overall interference pattern in both dark and light areas , or something else .
    The Alamy example is very clear as to what they mean by IA , but there is nothing to illustrates all the other varied and differing forms IA it is suppose to be able to take.
    Identifying the little blighter's I see as the main problem .One mans jaggy is another mans, CA or spec of dust, which only adds to the confusion .
    The problem is not a technical one, its one of communication and it will be resolved when its becomes clear what the majority and those setting the standards mean by IA .
    After all faults can be corrected , but you have to be able to understand , see and know what people are taking about in order to correct them .
    I am wondering if IA has become by common usage , over time, a blanket term for many ills , and is not longer confined to just the illustration given .

  18. 12 June 2008 at 11:31 Carol Buchanan

    Mary, I would consider Fen's suggestion. When and how you are interpolating could perhaps make a difference. Re IA - one nasty pixel is all it takes - although I find it's rarely one. If there's something wrong with the image, you usually find a whole pile of problems.

    I check @ 100% before and after interpolating.

    I use a Canon 5D and can crop an image to about 1/2 size, resize to 5400px on the longest size (for another agency - and leave it the same for Alamy)and rarely encounter any problems.

    But, as I said, the image must be spot on first. If your RAW file is lacking, the end result will not be as perfect as it could be.

  19. 16 June 2008 at 11:00 Kenneth Peck

    I have had several rejections lately for Intopolation Artifacts, could the fact that the camera as been trough an X-Ray Machine several times have an effect. Ken

  20. 19 June 2008 at 10:17 Mary clark

    Hi Kenneth,
    I don't think so ,can you see any difference between the images before and afterwards ?

  21. 20 June 2008 at 09:47 Kenneth peck

    Hi Mary No I cant see any difference in the pics,it just puzzeld me why I got so many rejections one after the other. Ken

  22. 21 June 2008 at 09:50 Mary clark

    Kenneth ,your not alone .Join the club.
    Have your image got some subject movement due to the activity they are doing or movement in large bodies of water in them ? I haven't got any of this type of subject through QC for more or less the whole of this year.The images of stationary subject are fine ,but not the others.There is a definite pattern emerging .

  23. 23 June 2008 at 17:49 Michael Dutton

    A few comments about my own recent experiences with Alamy QC.

    After 47 consecutive "passed QC" submissions (4 Sep 07 to 25 May 08), sailing along very happily, I suddenly seem to have hit something of a brick wall with no less than 5 consecutive QC failures (since 25 May 08). Usually, the reason is given as the good old "Interpolation Artefacts" although a couple were due to "soft or lacking definition".

    Since Feb 2007 I have always used a 10 megapixel Nikon D200 with Nikkor AF-S DX 18-70mm f3.5-4.5 G IF ED lens, generating RAW images of file size 18.7Mb. I use Adobe Photoshop CS2 and, after conversion to tif, the image files are up to about 28Mb. After processing/editing I resize up to 5400px along the longest side (same as Carol Buchanan) leaving me with a file size of around 50-57Mb. Then, saving as jpeg at Quality 10 (as recommended by Alamy) I end up with 3-8Mb files for uploading to Alamy.

    I have not changed anything at all about the way I process/edit/resize the images so I'm really struggling to understand this sudden brick wall I've hit with the Alamy QC process. Surely Alamy must be imposing very much stricter criteria now - but I haven't actually seen any mention of that from Alamy.

    To make it even stranger, having gone back to the original RAW files for all my recent QC failures, I find that, where IA have (allegedly) been identified, the offending areas appear exactly the same in the RAW file - so not attributable to any interpolation since that would surely only be a factor after the resizing on CS2.

    Have any of you out there been suffering a similar string of QC failures? I guess I'll give it a couple more goes with further submissions but it's all beginning to get me down now and I think I may call it a day with Alamy (after about 1600 images passed). Incidentally, none of the three iamges of mine that have sold in the last month would have had any chance of getting through the current Alamy QC process as it stands now.....

  24. 25 June 2008 at 10:07 Mary clark

    Hi Michael,
    Yes I have exactly the same experience with the IA ,most recently some muddy water has been mistaken for Interpolation Artifacts .

  25. 26 June 2008 at 16:58 Shel

    Mary, you must post your queries/comments on the Alamy forum, otherwise you are almost just talking to yourself. On the forum you will get the advantage of many, many contributors who have never had an image fail, and they will be able to explain to you why alamy qc have rejected an image....no one has yet posted a link to a 100% crop of a failed image that hasn't been shown to have good reasons for failure :)

  26. 27 June 2008 at 13:25 Mary clark

    Hi Shel .Yes I am going to post a selection.I am finishing them now .
    I am having problems seeing my work on the web.Virgin are working on it as we speak.
    Hopefully you will be able to see 12 images which will give you the full picture.

  27. 02 July 2008 at 10:07 Mary clark

    My ability to tell which images would pass and which might fail Alamy QC , left me in February this year .
    I have put together a selection of my images on the web site
    http://www.trickofthelight.net/test/
    I am hoping to find contributors who can distinguishing those images passing from the failures , as they will have the understanding and knowledge I am looking for .
    If you would like to help , please can you select the images which you
    think passed Alamy QC from those failing and send me your findings. I have but a contact link on the index page .
    Each image is numbered , and there are two pages to look at.
    Please only record images which failed for either " soft and lacking "
    or having " interpolation artifacts " or both , and sort these from
    the ones you know passed .
    Click images to navigate , enlarged images are 1 monitor screen pixel
    = 1 image pixel [100%]
    There seem to be local problems with viewing the site in the UK , but
    if you use http://www.megaproxy.com/freesurf/ and type in the link all
    should be reveal to you .
    Many thanks for your help
    Mary Clark

  28. 04 July 2008 at 17:39 Mary clark

    James Allsworth QC

    Has just e-mailed me to say that they examine the images at 143% and no longer at the 100% 1 image pixel to 1 monitor pixel , so thats a big change being almost as half as much again

  29. 06 July 2008 at 13:24 Mark Amy

    Mary, that is simply not true! As per the thread you started in the forum, which includes the reply you received from James, you have clearly misunderstood his reply.

    Just because you personally interpolate your images by 143% doesn't mean that QC view them at 143%.

    They view your images, and the rest of ours, at 100% like they always have done.

    Mark

  30. 06 July 2008 at 20:30 Mary clark

    Hi Mark .
    It does seem to have started a rumpus.It has been very difficult getting people to understand that I needed to know the value of what Alamy was calling 100 % because I am working in software which gives another value to 100 % .Adobe 100% is larger than in Silkypix . I needed to establish whether an image of 3872 X 2592 was the same as Alamys 100 %,As it turns out they don't mean 100% in mathematical terms at all ,they mean that they open the 48-58 Mb file as it arrives to a one to one size .I leave that to you to work out what % that is. So if you like thats a double misunderstanding.
    Sadly this necessary clarification of terms has resulted in me getting banned from the Alamypro blog , because of the hight of feeling it has caused ,and there has been a great deal of personnel abuse posted against me on the Alamy forum as well .

  31. 07 July 2008 at 08:46 Mary clark

    100% at 1 image pixel per 1 monitor Pixel is a value set in stone,as apposed to an arbitrary ratio 1:1 which moves about and can mean anything size.

  32. 24 July 2008 at 21:05 flying dog

    thank goodness that is cleared up then - what a t'do! Oh how I don't miss the forum - oh is this not the forum...

    I think the rest is explained very well above in Alamy's blog by Patrick. READ IT CAREFULLY!

    Oh, Hi shel, ops not the forum, this old dog forgot...;-)

  33. 03 September 2008 at 14:04 Nigel Ollis

    I too suddenly started having problems with 'soft or lacking definition'. Peering at the images at 100% on my large monitor, I often had difficulty in establishing where this softness might be. Scanned medium format slides seem to sail through but I've had to call time on 35mm film scans. There's too much noise to enable compression down to a 20mb jpeg. Noise Ninja etc. softens the image even more and is not recommended. My 8mp Canon eos350d struggles to secure background detail; I assume this where the problem lies with digital submissions. I had a lengthy exchange of emails with Alamy which was inconclusive. I have since resorted to using a 28mm prime lens in order to get the requisite sharpness but first you have to establish the hyperfocal distance and set the lens accordingly. Some of the advice on the internet is misleading, Canon don't seem to know and Alamy won't comment. The various calculators on the internet will give you the approx hyperfocal distance for the lens and aperture. Measure this out and focus on an object at that distance. Then switch the lens to manual and start taking pictures being careful not to disturb the focus. I have since bought a new Canon eos450d. The improved spec seems to deliver better images with the above procedure. David Noton is going to do an article on this in one of the mags, so look out for that. If your a real photographer and have a Canon EOS-1Ds Mark III, ignore all of the above. Happy shooting!

  34. 03 September 2008 at 23:35 Joseph furtado

    Nigel,

    I'm not sure what you're shooting but certainly your choice of glass will have a huge impact on images quality. If your work entails a lot of small distant detail then a prime or L series lens is your best choice. The other thing which destroys a lot of detail is incorrect exposure. The sensor on your camera is more than capable of rendering fine detail provided your exposures are not too dark (which introduces noise when corrected in post, there by destroying detail) or too bright (reducing local contrast in areas of detail). There are some measures you can take to help deal with these problems but nothing beats a proper exposure.

    Also in reference to your comment about Noise Ninja, if used correctly it will in fact enhance detail, not soften it but you must use the proper profile for your camera.

    Hope this helps,

    Joseph.

    PS, while the 1Ds MkIII is a nice piece of gear I think it's over rated (unless you need massive files) ;-)

Add your own comment
Means that we require this information. We respect your privacy and store your details securely. See our privacy policy for more.
Security code picture
No picture? Reload page.
No picture? Reload page.