Jump to content

Recommended Posts

I have recently acquired the EOS 5DS R camera.  However, all images that I submit using this camera will always be rejected by Alamy's QC stating the reason "Soft Due To Size".  I have never interpolated the image file nor have I increased the file size. 

 

Similar images submitted using my existing EOS 5D Mk II (21 mp) will always pass Alamy's QC.  My questions:

 

i) Is this camera allowed by Alamy?

ii) How can an image shot by this camera be "Soft Due To size"?

iii) Any ideas what should I do next if I want to submit images using this camera?

Share this post


Link to post
Share on other sites

I think you need to list to Canon very carefully when using the 5DS camera. They say that due to the large sensor size, it is either necessary to use a very high shutter speed or a tripod. 

 

Alternatively, you could interpolate images down to a smaller size so that any movement will not be apparent any more. (But don't tell Alamy.)

Share this post


Link to post
Share on other sites

Hi there,

 

Have to jump in here as your post is a bit misleading.

 

Looking at your account, some of your images have been failed due to over-sharpening and others have been failed as soft due to size. The soft due to size failures were for images sized between 120mb and 130mb.

 

For more information, please email memberservices@alamy.com.

 

Thanks

 

Alamy

Share this post


Link to post
Share on other sites

I have recently acquired the EOS 5DS R camera.  However, all images that I submit using this camera will always be rejected by Alamy's QC stating the reason "Soft Due To Size".  I have never interpolated the image file nor have I increased the file size. 

 

Similar images submitted using my existing EOS 5D Mk II (21 mp) will always pass Alamy's QC.  My questions:

 

i) Is this camera allowed by Alamy?

ii) How can an image shot by this camera be "Soft Due To size"?

iii) Any ideas what should I do next if I want to submit images using this camera?

Given Alamy's answer, you must have upsized a great deal.

My exports from 16MP range from 4-7MB. So an uncropped jpeg from 50MP should be somewhere around 20MB, depending on the detail.

  • Upvote 1

Share this post


Link to post
Share on other sites

Dear all,

 

Thanks very much for your replies.

 

I can assure you that the soft images were not caused by camera shake. All my shots were shot using a very sturdy Gitzo tripod.

 

As per Alamy's reply, all my failed images were:

 

1) Soft due to size for images between 120 MB - 130 MB. These were the images that was shot using the 5DS R. The images were initially uploaded to Alamy's QC unsharpened and not upsized or interpolated. After the failed QC due to "Soft Due To Size" reason, i then submitted another batch of images (also shot with the 5DS R) with slight sharpening. These images were then failed due to the "Excessive Sharpening".

 

2) I then selected another batch of my images shot with the 5D 2 - all unsharpened and not resized (downsize or interpolated) and resubmitted to Alamy. These images had all passed Alamy's QC without any problems.

 

Due to reasons still unknown to me, I think the files of the 5DS R is too large to produce a "sharp" image at the particular file size. In my next submission using images shot using this camera, I will try to select the Medium RAW or Small RAW settings to get a smaller file size.

 

Any comments regarding the above method greatly appreciated. Anyone else in this forum also using the 5DS R?

Share this post


Link to post
Share on other sites

Have to agree with Mark about file sizes. They do seem to be large for a compressed JPG. (Not a Canon user, so have no idea what size the uncompressed size, RAW or JPG, the camera produces).

 

If it is not a technical problem with the camera itself, are you upsizing the images submitted?

 

Krisken

Share this post


Link to post
Share on other sites

Dear all,

 

Thanks very much for your replies.

 

I can assure you that the soft images were not caused by camera shake. All my shots were shot using a very sturdy Gitzo tripod.

 

As per Alamy's reply, all my failed images were:

 

1) Soft due to size for images between 120 MB - 130 MB. These were the images that was shot using the 5DS R. The images were initially uploaded to Alamy's QC unsharpened and not upsized or interpolated. After the failed QC due to "Soft Due To Size" reason, i then submitted another batch of images (also shot with the 5DS R) with slight sharpening. These images were then failed due to the "Excessive Sharpening".

 

2) I then selected another batch of my images shot with the 5D 2 - all unsharpened and not resized (downsize or interpolated) and resubmitted to Alamy. These images had all passed Alamy's QC without any problems.

 

Due to reasons still unknown to me, I think the files of the 5DS R is too large to produce a "sharp" image at the particular file size. In my next submission using images shot using this camera, I will try to select the Medium RAW or Small RAW settings to get a smaller file size.

 

Any comments regarding the above method greatly appreciated. Anyone else in this forum also using the 5DS R?

 

I would not change the RAW setting.

 

As Mark said, how do you get a 120mb jpg? I can see a 120mb RAW, but not a compressed jpg.

 

With a 50mp camera, you should be able to downsize your images to improve the sharpness and still have a nice size image for sale.

 

I'm sure checking a 120mb RAW for flaws is a daunting task, probably one of the problems with taking images that size.

 

But I would never ever change the RAW setting to lower quality in my camera. Do that in PS.

 

Jill

Share this post


Link to post
Share on other sites

What size, in MB, not MP, are the rejected images? Look at the actual files you've exported from whatever software you are using.

Share this post


Link to post
Share on other sites

What lenses are you using? Some of the Canon L-lenses could have difficulties to follow the increased sensor size. I have heard some preferring a specific Sigma lens with this camera.

 

Increased sensor size - especially the sizes they are increasing to now - requires increased skill and perhaps even better lenses.

Edited by Niels Quist

Share this post


Link to post
Share on other sites

Do you have a sample at 100% somewhere?

Did you do a focus test? Any back/front focusing? Did you focus with live view x10?

It should be the logical next camera for people like myself who own a lot of Canon lenses. So I am interested in the real world results.

I noticed that you do twilight shots, do you need to bracket with the 5ds?

 

wim

Share this post


Link to post
Share on other sites

Canon has a list of their lenses recommended for use with the 5DS and Sr: http://www.canonrumors.com/2015/06/canon-releases-recommended-lenses-for-eos-5ds-eos-5ds-r/

 

50 megapixels will tax lenses and technique and unless you're in the business of producing ultra large reproductions, unnecessary. It's overkill for stock.

  • Upvote 1

Share this post


Link to post
Share on other sites

Dear All,

 

Once again, thanks very much for all your comments above.  Having said that, I am now convince that the fault lies somewhere in my workflow.  One reason I suspect is because I am now converting my RAWs using the new Canon's DPP 4 which was bundled together with the 5DS R.  I was previously using DPP v1 which was bundled with the 5D2.  I guess I had set the conversion settings in DPP 4 to automatically resize (interpolate) the images when converting the RAWs into TIFFs.  That could be the reason why I do not face these problems with the 5D2 since most of the RAWs was converted using the earlier version of the DPP software.  I am now relooking at the conversion settings in the DPP 4 to see where the fault lies.

 

Sincerely apologies if this topic seems to suggest that Alamy is biased towards the 5DS R.  Not my intention at all to suggest that. 

 

As a new contributor in Alamy, I do appreciate all your comments and assistance. 

 

Some of my images can be seen at my Flickr site - www.flickr.com/photos/amrilizan

 

Very much appreciate your kind comments.

Share this post


Link to post
Share on other sites

As Jill said, don't change the RAW setting. Adjust your workflow afterwards.

Share this post


Link to post
Share on other sites

Thanks very much for the comment. 

 

The 5DS R is a fantastic camera if you can control and managed the Dynamic Range (DR).  Having said that, I do not bracket any of my shots.  I tend to rely in the histogram when shooting my images.

Share this post


Link to post
Share on other sites

Thanks very much for the comment.

 

The 5DS R is a fantastic camera if you can control and managed the Dynamic Range (DR). Having said that, I do not bracket any of my shots. I tend to rely in the histogram when shooting my images.

Then I'm all the more impressed. Brilliant shots in your Flickr pages!

We share at least one subject ;-)

 

E9ERR2.jpg

 

E9BDHF.jpg

 

Let us know if your images have improved using the newer DPP and if they are being accepted.

Maybe it's too early to choose the Sony A7R2 over the Canon 5Ds.

Btw nobody has mentioned it in this thread but it is of the utmost importance to look at all the images at 100%, because Quality Control will do exactly that.

 

wim

Edited by wiskerke
  • Upvote 1

Share this post


Link to post
Share on other sites

It seems pretty certain that some misunderstandings have been propagated through this thread. I've read it a few times and nowhere does the OP or Alamy say that the JPEG size is 120-130 Mb on disk. Alamy must be referring to the uncompressed file size in Mb. If nothing else, the uncompressed size (pixel dimensions) of a 120 Mb JPEG would be in the region of 300 Mb or more which would be automatically rejected on upload as the maximum size is 200 Mb (I know this from experience uploading panoramas and forgetting to downsize).

 

So, assuming that there are no problems with focus and the lenses used are up to the 50MP job, then the simplest solution would be to either downsize the 50MP images or to do what the OP suggested and shoot at a smaller RAW size. I don't see why this would necessarily be a bad thing and it would certainly be worth testing.

 

My experience is with Nikon D800 series 36MP cameras and not with Canon but similar principles apply - you need to use very good lenses more than anything else and careful focusing. I normally use a tripod but have no problems handholding as long as use a suitably high shutter speed and there are no problems. However, the further jump to 50MP is likely to require even more careful technique and lenses than with the D800 series. I know for sure that if Nikon decided to follow suit with a 50MP machine, then I would not be upgrading - 36MP is more than adequate for anything I will ever do with a camera.

  • Upvote 1

Share this post


Link to post
Share on other sites

It seems pretty certain that some misunderstandings have been propagated through this thread. I've read it a few times and nowhere does the OP or Alamy say that the JPEG size is 120-130 Mb on disk. Alamy must be referring to the uncompressed file size in Mb. If nothing else, the uncompressed size (pixel dimensions) of a 120 Mb JPEG would be in the region of 300 Mb or more which would be automatically rejected on upload as the maximum size is 200 Mb (I know this from experience uploading panoramas and forgetting to downsize).

 

So, assuming that there are no problems with focus and the lenses used are up to the 50MP job, then the simplest solution would be to either downsize the 50MP images or to do what the OP suggested and shoot at a smaller RAW size. I don't see why this would necessarily be a bad thing and it would certainly be worth testing.

 

My experience is with Nikon D800 series 36MP cameras and not with Canon but similar principles apply - you need to use very good lenses more than anything else and careful focusing. I normally use a tripod but have no problems handholding as long as use a suitably high shutter speed and there are no problems. However, the further jump to 50MP is likely to require even more careful technique and lenses than with the D800 series. I know for sure that if Nikon decided to follow suit with a 50MP machine, then I would not be upgrading - 36MP is more than adequate for anything I will ever do with a camera.

Agreed. When I bought my 36mp D800, I came from 16mp. I would have preferred 24mp, but Nikon didn't offer that range then. 24mp is perfect in my world. The Nikon 24-70 is my go-to lens with my D800 although the 105 macro works well, too. I can shoot those lenses hand-held. That is, when I pick the Nikon up. I seem to really like mirrorless now.

And the Fuji lenses are special. If only the X-T2 would be 24mp, I'd be on cloud 9.

Betty

Share this post


Link to post
Share on other sites

I am shooting it myself!  look! I also use a HD5 ( hasselblad) for commercial work, its the same there. Lenses!!  don't take Canons word for it when it comes to optics,  the ONLY lenses I have found suitable for the new DS Canons are Primes, fixed lenses, all the zoom, bar the 70-300L ( white one) gives in fact slight soft appearance.

 

Downsizing is not a very good option either, it can cauese artifacts, haze, etc, depending on subject-matter. The DSr also lacks an aliasing-filter, which if unlucky can cause certain issues. 

 

You would in fact have been much better of getting the 5DMIII, perfect stock-camera! the new DS range 50MP cams are for high-end commercial, studio and landscape work, not really for stock as such, in which case most photographers will use medium-format anyway. These are strictly MP loaded cameras, nothing else.

 

If possible I would change the DSr model to the other DS model, its much friendlier and slightly more forgiving with the optics.

 

best. and enjoy your camera.

Edited by christian58
  • Upvote 1

Share this post


Link to post
Share on other sites

I have been wrestling with this issue as I am the upgrading from a 5D11 and deciding between either a 5DS or a 5DSR. 

 
When viewed at 100% an image should look sharp. I does not matter if the  image was taken on a 18, or a 22, or a 38, or a 50 megapixel camera, it should look sharp when viewed at 100%.
 
The larger the image, the smaller part of the image that you are viewing at 100%. An 18 megapixel image can look as sharp as a 50 megapixel image when both are viewed at 100%. There will not be as much detail in the 18 megapixel image, but it can still look as sharp when viewed at 100%.
 
The buyer expects the image to look very sharp at 100%. That is why Alamy gives a file size like 3400 x 2852 px 11.3 x 9.5 inches (300dpi)27.7 MB. This means that this smaller sized 15 megapixel image will be sharp at 100%, as long as you do not use it larger than 11.3 X 9.5 inches.
 
A Zeiss lens on a 22 megapixel 5D11 image looks very sharp at F5.6. F5.6 is the sharpest aperture. It still looks sharp at F11, but there is a very slight degradation of image sharpness due to diffraction. On a 22 megapixel 5D11 viewed at 100% this slight degradation at F11 is hardly noticeable, but may be very noticeable on a 50 megapixel 5DSR when viewed at 100%.
 
Therefore if I purchase a 50 megapixel 5D SR and shoot at F11 I would probably downsize to 38 megapixels because the image at 50 megapixels would be "Soft Due To Size”. It would be a 50 megapixel image with the sharpness of a 38 megapixel image. Not good.
 
Tripod at F5.6 ISO 100 submit at 50 megapixels. Tripod at F11 ISO 100 submit at 38 megapixels. Handheld at F11 higher 400 ISO submit at my old 5D11 size of 22 megapixels.
 
Evaluate sharpness on screen at 100% and, if necessary, downsize until the image looks sharp.
 
I am concerned about moire on the 5DSR, and often shoot handheld at ISO 400 and F11. I am leaning to downsizing a 50 megapixel file from the softer 5DS to at least 38 megapixels.
 
There is a point of diminishing returns as the camera megapixel count rises. You have to either upgrade your shooting technique when practical, or downsize the file until it looks sharp at 100%.

Share this post


Link to post
Share on other sites

Dear All,

 

Just to provide an update since my last posting.  I have since resubmitted the image which was previously failed by Alamy's QC due to "Soft Due to Size".  Glad to report that this image had passed Alamy's QC and is now accepted for sale.

 

Just to recap on what had transpired:

 

- I had initially shot the image using the 50.6 mp EOS 5DS R.  After doing my normal editing, I had submitted the image to Alamy for their QC.  However, it was rejected due to the above reason - Soft Due to Size.  I was initially perplexed as how could an image from the 5DS R be "Soft Due to Size".  Thinking that it had something to do with my conversion settings in the new DPP 4, I had relooked at all the settings (and my workflow) to ensure that I had not accidentally set any settings to automatically up-size or interpolate the image during the RAW conversion.  After checking all the settings in DPP 4, I can confirm that none of the settings had caused the RAW image to be up-sized or interpolated.  FYI, I was shooting at the highest RAW settings which would give me the full resolution that the camera and my lens would allow me to (although, I do not think it will give me the full 50.6 mp).

 

- After reading some of your comments above, I was convinced that the problem is not due to the software or my editing workflow - as neither any of these actions had interpolated the image during the RAW conversion.

 

I then decided to downsize the image using the steps below:

 

- Highest RAW settings from the 5DS R would give me an image size of 8688 pixels (Width) X 5792 pixels (Height).  Since this image size was "Soft" according to Alamy, I had resized the same RAW file to 5616 pixels (Width) X 3744 pixels (Height) which is the same image size that my other camera, the 5D Mk II would give.

 

- this image had passed Alamy's QC earlier this evening.  I have yet to prepare this image for sale at time of posting these comments.  Therefore, I can conclude the following:

 

1) As Bill Brooks had correctly pointed in his posting above, the problem of diffraction magnifies in a 50.6 mp file.  It is not too obvious in a 21 mp file.  I read somewhere that diffraction kicks in at f/8.0 in the 5DS R.  I was using an aperture of f/20.0 when shooting this image.

 

2) Lens choice - I was using the 24-70 mm f2.8 (L) lens (Mk I)  for this and not the Mk II version.  As Canon had pointed out, you will need the Mk II version or newer L lenses to realise the full resolution of the R.  The Mk I lens had caused some image softness in the 50.6 mp file.

 

3) You will need to downsize the RAWs from the 5DS R to avoid "Softness Due to Size" issues, unless you are using the Mk II L lens, primes or the OTUS lenses. 

 

4) Having said that, I will try to upload another batch of images taken using the 5DS R later tonight using the downsizing method above. 

 

Thanks a million guys!! :) :)

 

 

 

-

  • Upvote 1

Share this post


Link to post
Share on other sites

Sorry Amril but this doesn't sound right I sometimes supply pictures from a HD5 with a giant 60MP Phaseone back, 4 times the size of my canon 50DSr and I have no problem. Reading the Canon guys who also bought this cam, sure some complain about a soft look but when viewed at 100% its then tack sharp. The 24-70 M1 is known for artifacts such as chroma, CA, distortion, etc, but I have never experienced the M1 giving a soft look. Although I have heard some say the other, the 50DS gives a sharper file out of the camera. Don't know really.

Share this post


Link to post
Share on other sites

I would look at the f=20 as the main possible culprit.
This would be a simple test for diffraction. I suggest doing it for every lens you use.

Print 5 of these Siemens stars; stick them to a wall and put your camera square and plumb in front of it at about 3m / 10ft on a very sturdy tripod. Be sure it can not move a hair. (Tape it to the floor; use sandbags, whatever works.)
Put one star exactly in the middle and do not move it. Pin the others so that they are in the corner of your image and reposition them for a different focal length.
Focus using live view at 10x on the center star.
Critical: make sure you do not have mirror or shutter slap interfering with your results.
Use mirror lock up or live view and a remote or 10sec self timer. For most Canons live view produces less vibration than MLU.
Critical: manual focus only.
Critical: no IS or VR.

Use the main focal lengths, in this case 24; 35; 50 and 70mm.
Do a complete cycle at real f-numbers: 1.4 - 2 - 2.8 - 4 - 5.6 - 8 - 11 - 16 - 22 - 32 for each of them.
Load the results for each focal length in a viewer exactly on top of each other. In the right order ascending or descending.
For instance as layers in Photoshop.
Now cycle through the layers at 100% looking at the center of a Siemens star. Do it for each of the five stars.
And observe.

Do this with your mk2 (if you still have it) and the same lens.
If you have other cameras with different sensor sizes like a point and shoot, test them as a reference.
With legacy lenses without coupling have a post it note with the lens name and the f-number. (I use a white key card and a whiteboard marker.)
Here's a different test target (actually a bit too small).
Here's a description of how to square up a camera fast, using 2 cheap mirror tiles. (Actually not that critical for testing diffraction in one lens)
Here's a good setup, although it's meant to find other things. They do not look at the actual recorded image, but at an enlarged live view image on a monitor.
Great for a quick assessment like in the classroom, but worthless for future reference for yourself.

For deciding between lenses and cameras or systems, there's Imatest.

wim

 

edit: typo

Edited by wiskerke
  • Upvote 2

Share this post


Link to post
Share on other sites

I would look at the f=20 as the main possible culprit.

This would be a simple test for diffraction. I suggest doing it for every lens you use.

 

Print 5 of these Siemens stars; stick them to a wall and put your camera square and plumb in front of it at about 3m / 10ft on a very sturdy tripod. Be sure it can not move a hair. (Tape it to the floor; use sandbags, whatever works.)

Put one star exactly in the middle and do not move it. Pin the others so that they are in the corner of your image and reposition them for a different focal length.

Focus using live view at 10x on the center star.

Critical: make sure you do not have mirror or shutter slap interfering with your results.

Use mirror lock up or live view and a remote or 10sec self timer. For most Canons live view produces less vibration than MLU.

Critical: manual focus only.

Critical: no IS or VR.

 

Use the main focal lengths, in this case 24; 35; 50 and 70mm.

Do a complete cycle at real f-numbers: 1.4 - 2 - 2.8 - 4 - 5.6 - 8 - 11 - 16 - 22 - 32 for each of them.

Load the results for each focal length in a viewer exactly on top of each other. In the right order ascending or descending.

For instance as layers in Photoshop.

Now cycle through the layers at 100% looking at the center of a Siemens star. Do it for each of the five stars.

And observe.

 

Do this with your mk2 (if you still have it) and the same lens.

If you have other cameras with different sensor sizes like a point and shoot, test them as a reference.

With legacy lenses without coupling have a post it note with the lens name and the f-number. (I use a white key card and a whiteboard marker.)

Here's a different test target (actually a bit too small).

Here's a description of how to square up a camera fast, using 2 cheap mirror tiles. (Actually not that critical for testing diffraction in one lens)

Here's a good setup, although it's meant to find other things. They do not look at the actual recorded image, but at an enlarged live view image on a monitor.

Great for a quick assessment like in the classroom, but worthless for future reference for yourself.

 

For deciding between lenses and cameras or systems, there's Imatest.

 

wim

 

edit: typo

I think Wim is absolutely right about f20.  Such a small aperture is very likely to cause diffraction.  I never use a smaller aperture than f16 for 35mm and only then if I really have to. I much prefer the results from f11 for my Canon 5D3 whatever the lens.

 

Pearl

Share this post


Link to post
Share on other sites

same here, I prefer f11, max f16. However in this case there are several reports from pro photographers that the 50DSr give a soft result, even giving the impression of being unsharp and on many various appertures. Might be something in it?

Took a few pics this morning just as a test and yes, looking slightly soft but of course after processing, etc, its tack sharp. We are talking straight out of the camera, raw.

Share this post


Link to post
Share on other sites

Create an account or sign in to comment

You need to be a member in order to leave a comment

Create an account

Sign up for a new account in our community. It's easy!

Register a new account

Sign in

Already have an account? Sign in here.

Sign In Now

×
×
  • Create New...

Important Information

We have placed cookies on your device to help make this website better. You can adjust your cookie settings, otherwise we'll assume you're okay to continue.