Hi-res APS-C: Should images be downsampled before submitting?

[dbooksta]

I just submitted my first batch of four.  All were rejected, but only the first got a reason: "Soft or lacking definition."  This was on a 24Mpx shot from a Sony A77II (APS-C) at ISO 200 with f/7.1, and the focus was perfect.

 

Sony might be overly ambitious to be providing 24Mpx off that sensor at that size; it's certainly noisy even at ISO 200; so I'm wondering: Do successful contributors downsample in cases like this?  Any recommended practices for that?

 

[MircoV]

Hello,

 

It all depends. You really need to be 100 percent sure that focus is allright before you think to do next steps. Otherwise you will keep confronting the same problem. I think downsizing works only if the photo is for lets say 95 percent of the sharpness that is needed to get approved. If you downsize then a 24 MP photo to 12 MP you will get 100 percent sharpness. But this requires a minimum of sharpness discrepancy. In general it is not recommended to play with that. I would say create a new image and be sure everything is correct. I have myself many photos on Alamy taken with the Sony A77 first version at 24MP on iso400 and sometimes iso 800. All are accepted.

 

Good luck.

 

Mirco

[spacecadet]

It would help if you posted a 100% crop (nothing less will do) but downsizing won't make up for SoLD by itself. That camera is certainly capable of passing, depends on the glass of course. It shouldn't be noisy at 200- I've submitted images from the A55 (16MP) at 800, and occasionally 1600, with no or minimal NR.

[Craig Yates]

http://www.alamy.com/contributors/alamy-qc-failure-reasons.pdf

 

Have a look at the above, unless you already have.

 

Regards

Craig

[NYCat]

On your first submission Alamy looks at all four images and gives a reason for rejection on EVERY one that has a problem. So the three with no comment are OK. I don't downsize with my Nikon D7100 (24MP on a crop sensor).

 

Paulette

[DHill]

Considering it's your initial submission, you might as well be as safe as you can. You can gradually take more risks later as you get used to the process of preping for Alamy, including the pixel peeping that Alamy requires. 

 

No harm in downsizing - you could roughly half each dimension and submit a six megapixel image. 

[spacecadet]

Substantial downsizing just to pass QC won't necessarily tell you how to get past SoLD.

When we see the 100% crops we can offer an opinion. SoLD isn't just about focus.

[Inchiquin]

My approach has always been if it's not good enough at 100% it's not good enough. I must do better next time.

 

Alan

[dbooksta]

Here's a 50% downsample.

 

Here's the original.

 

Glass wasn't top-of-the-line, but at f/7.1 I've found my Tamron 70-300mm f/4-5.6 to be sharper than this sensor.

[spacecadet]

The 'original' isn't showing at anything like full size but based on the downsize I think you're off. The plane of sharp focus is on the twig on the right.

There's also some artefacting particularly around the head but that could be down to the downsizing.

Try something a lot simpler for your initial sub. Sunny day, tripod, f8 at 250th or something.

[Bill Kuta]

And also try something other than a 70-300mm at the long end. Also check the coverage of your focus points. If the focus points cover much more than the center point, then it focused on the nearest object, which, as Mark pointed out, is the twig at right.

[Bhandol]

The eyes & face  aren't very sharp, that's why it wasn't accepted

[Bhandol]

Even the bark on the thick branch is sharper than the bird. Wildlife is tricky, lots of varriable that can screw up your image (subject motion, camera shake, lens choice, weather, heat haize, autofocus missing the target), I would say your lens might have a focus issue and may need focus calibrating.

 

For initial submission I would shoot a table top still life with a prime lens.

[dbooksta]

Ah ha -- thank you guys!  Using the twig and bark as a reference I can see now that I did miss the focus.

 

I'm almost certain I wasn't running strict center autofocus like I should have been.  Though I'm surprised that at such a distance and aperture the focal plane is so thin!  Guess I'm still learning....

[spacecadet]

Now that's out of the way, yes, some of us do downsize a bit, usually after a run-in with QC. In my case, to 4000px long side. I just have it in my LR export dialogue.

But don't rely on it to get borderline images through QC.

[spacecadet]

Now that's out of the way, yes, some of us do downsize a bit, usually after a run-in with QC. In my case, to 4000px long side. I just have it in my LR export dialogue.

But don't rely on it to get borderline images through QC. You need Alamy sharp, not granny's birthday party snaps sharp.

[dbooksta]

Now that's out of the way, yes, some of us do downsize a bit, usually after a run-in with QC. In my case, to 4000px long side. I just have it in my LR export dialogue.

But don't rely on it to get borderline images through QC.

 

So you find the Lightroom export interpolation to be adequate?

 

And how do you decide when to downsample, vs. providing the original resolution?

[spacecadet]

 

Now that's out of the way, yes, some of us do downsize a bit, usually after a run-in with QC. In my case, to 4000px long side. I just have it in my LR export dialogue.

But don't rely on it to get borderline images through QC.

 

So you find the Lightroom export interpolation to be adequate?

 

And how do you decide when to downsample, vs. providing the original resolution?

 

Yes.

I don't decide. I downsize everything.

[Martin P Wilson]

My approach has always been if it's not good enough at 100% it's not good enough. I must do better next time.

 

Alan

 

Me too. If its not sharp at the camera's native resolution it is not sharp enough for me. let alone Alamy, or anybody else for that matter. I only downsize if I am using it on my own web sites.

[John Mitchell]

I look at each image individually and downsize only if I think it's absolutely necessary. I then try to stay as close to the native file size as possible. If, after downsizing, the image's appearance hasn't improved sufficiently for Alamy, I don't submit it.  Downsizing can be a very useful tool IMO. Just because an image is shot at the camera's full resolution doesn't mean it has to stay that way -- 24 MP, 16 MP, etc. are really just starting points. Purists will disagree, of course.

[Bill Kuta]

If it's Alamy that decides on "Best Answer", then we have their viewpoint on it. And I'm surprised by that.

[MDM]

If it's Alamy that decides on "Best Answer", then we have their viewpoint on it. And I'm surprised by that.

 

I would be absolutely amazed if Alamy gave that as a best answer. In effect downsizing is another form of sharpening (that again). If one is having to downsize every image to pass Alamy QC, then there is something amiss with equipment and/or technique, even more so if capture sharpening is being applied to raw images. While I think it may be necessary for some images to sharpen up edges or corners or to increase apparent depth of field with high megapixel cameras if lenses or focusing techniques are leaving something to be desired, it certainly should not be necessary to downsize to bring the main subject into sharp focus.

[spacecadet]

 

If it's Alamy that decides on "Best Answer", then we have their viewpoint on it. And I'm surprised by that.

 

I would be absolutely amazed if Alamy gave that as a best answer. In effect downsizing is another form of sharpening (that again). If one is having to downsize every image to pass Alamy QC, then there is something amiss with equipment and/or technique, even more so if capture sharpening is being applied to raw images. While I think it may be necessary for some images to sharpen up edges or corners or to increase apparent depth of field with high megapixel cameras if lenses or focusing techniques are leaving something to be desired, it certainly should not be necessary to downsize to bring the main subject into sharp focus.

 

I think they like it because it's succinct. The OP asked a quite specific question and I answered it. I've found it doesn't hurt my QC and I'm not advocating it, just describing it.

[dbooksta]

Heh, I wish Alamy was that active on these forums!  I (the OP) selected that as the "best answer" by clicking on the "Mark Solved" button under it.

[MDM]

Heh, I wish Alamy was that active on these forums!  I (the OP) selected that as the "best answer" by clicking on the "Mark Solved" button under it.

I was about to say I am truly amazed but I can now say that the universe has returned to normal. I did previously assume that It would be the original poster who would choose the best answer. Like many others here, I do not think that downsizing everything is the solution at all.

 

So if you are really finding that your images are not properly sharp when viewed at 100% or 1:1, then it's time for a rethink not a downsize. If it's technique, then it's easy to solve. If it's equipment, then it's either faulty or inadequate for professional purposes. In relation to the Tamron lens, I have been using Tamron since the early 80s and have never had a problem with a lens until recently, when I splashed out on a brand new top of the range 70-200 f2.8. I was most disappointed as there was a problem with the autofocus. Optically the lens was perfect as I was getting excellent results with manual focus but it was having serious back-focusing issues. Frustrating as it took up quite a bit of time figuring out that it was the lens and not me or my camera was faulty. Needless to say, it went back to where it came from. That was probably a one-off as my other Tamron lenses are excellent but it is always worth testing your lenses carefully.