first comparo shots SD1Meos vs Sony A7r and canon 17mm ts-e image test shots

Started Dec 9, 2013 | Discussions
docmaas
Veteran MemberPosts: 4,792
Like?
first comparo shots SD1Meos vs Sony A7r and canon 17mm ts-e image test shots
Dec 9, 2013

As promised here are the first comparison shots between the SD1Meos and the Sony A7r.

The two stucco wall shots are f4 and f8 on the SD1Meos with the Canon 17mm TS-E. I don't see any evidence of problems with the lens or the SD1 from these shots. These were taken at 100 iso and as I have seen in the past on sigma cameras different apertures result in different colors. Neither one is close to the real color subtract which is a neutral beige.

Next 4 images at f4, f5.6, f8, and f11 from the Sigma. These images are jpgs created in spp and edited to match the wb of the sony including some desaturation.

To get these images on the SD1 it was necessary to shoot 3-4 shots at each aperture varying the focus from where the first confirmation beep occurs by a small amount each iteration until there is no beep.  Using that technique I was able to get one reasonably sharp image at each aperture setting.  The sony of course has live view with magnification and focus was easily accomplished.

All the files will be available in raw here:

https://www.dropbox.com/sh/4fkcsupknb3dmmc/-k7lsICS72

The sony images are ooc jpegs with no editing and particularly no sharpening.

Sigma SD1 Sony Alpha 7R
If you believe there are incorrect tags, please send us this post using our feedback form.
Kendall Helmstetter Gelner
Forum ProPosts: 18,637Gear list
Like?
A few questions
In reply to docmaas, Dec 9, 2013

1) How are you using the same lens on both cameras - did you convert the SD-1 to a Canon mount, or the lens to a SA mount and then use a Pentax adaptor on the Sony???  I just can't remember which way you did the 17mm conversion...

2) Why are the A7r shots so desaturated?  It seems to me like most cameras usually boost the saturation...

3) I was wondering what you used to process the RAW images.  The Sony f/8 leaves image seems not nearly as sharp, with a ton of noise reduction smoothing applied all over the place.

Thanks for providing the raw files though!

 Kendall Helmstetter Gelner's gear list:Kendall Helmstetter Gelner's gear list
Sigma 70-200mm F2.8 EX DG OS HSM Sigma 8-16mm F4.5-5.6 DC HSM Sigma 24-70mm F2.8 EX DG HSM Sigma 50-500mm F4.5-6.3 DG OS HSM Sigma 85mm F1.4 EX DG HSM +4 more
Reply   Reply with quote   Complain
docmaas
Veteran MemberPosts: 4,792
Like?
Re: A few questions
In reply to Kendall Helmstetter Gelner, Dec 9, 2013

The sony jpegs are ooc with all the internal settings as they came from the factory meaning neutral and with no preprogammed options as well.  Hopefully someone else with a later version of lightroom or ps will convert the raws to jpgs with appropriate processing.

The SD1Meos has a canon flange and contact block.

Mike

Kendall Helmstetter Gelner wrote:

1) How are you using the same lens on both cameras - did you convert the SD-1 to a Canon mount, or the lens to a SA mount and then use a Pentax adaptor on the Sony??? I just can't remember which way you did the 17mm conversion...

2) Why are the A7r shots so desaturated? It seems to me like most cameras usually boost the saturation...

3) I was wondering what you used to process the RAW images. The Sony f/8 leaves image seems not nearly as sharp, with a ton of noise reduction smoothing applied all over the place.

Thanks for providing the raw files though!

Reply   Reply with quote   Complain
Tiger1
Regular MemberPosts: 313Gear list
Like?
Re: first comparo shots SD1Meos vs Sony A7r and canon 17mm ts-e image test shots
In reply to docmaas, Dec 9, 2013

Can I ask why the raw file from the A7R has a size of 4800x3200???????? Is it the A7R or A7? Exif says A7R.  Either way the pixel count is too low.......

I was wondering why the files from the Sony were so small.  Most of the ones that are available to download are over 30Mb.

Can you please explain why this has occurred as it is VERY important.

 Tiger1's gear list:Tiger1's gear list
Sony RX1 Canon EOS 5D Mark II Fujifilm FinePix S5 Pro Sigma SD14 Sigma SD1 Merrill +8 more
Reply   Reply with quote   Complain
Kendall Helmstetter Gelner
Forum ProPosts: 18,637Gear list
Like?
Sony RAW files may be corrupted.
In reply to docmaas, Dec 9, 2013

docmaas wrote:

The sony jpegs are ooc with all the internal settings as they came from the factory meaning neutral and with no preprogammed options as well. Hopefully someone else with a later version of lightroom or ps will convert the raws to jpgs with appropriate processing.

Hmm - I can't open in my Lightroom trial, and the Sony RAW Viewer I downloaded can't read the ARW files?  Nor can Aperture, which is supposed to be able to handle the A7r files.  Perhaps they are corrupted?

I don't have Photoshop installed any longer, I'm trying to make a go without it... but since Lightroom can't read the files either not sure ACR can just yet.

 Kendall Helmstetter Gelner's gear list:Kendall Helmstetter Gelner's gear list
Sigma 70-200mm F2.8 EX DG OS HSM Sigma 8-16mm F4.5-5.6 DC HSM Sigma 24-70mm F2.8 EX DG HSM Sigma 50-500mm F4.5-6.3 DG OS HSM Sigma 85mm F1.4 EX DG HSM +4 more
Reply   Reply with quote   Complain
docmaas
Veteran MemberPosts: 4,792
Like?
Re: Sony RAW files may be corrupted.
In reply to Kendall Helmstetter Gelner, Dec 9, 2013

I just opened them with irridient.  try the latest release (Nov 13th).  If it works then the dropbox files are ok.  if not then they must have gotten corrupted.

Reply   Reply with quote   Complain
docmaas
Veteran MemberPosts: 4,792
Like?
Re: first comparo shots SD1Meos vs Sony A7r and canon 17mm ts-e image test shots
In reply to Tiger1, Dec 9, 2013

The A7r is set to aps-c output to match the SD1.  While the 17mm ts-e is a full frame lens the metabones ef-nex converter is version 1 and only supports aps-c and manual focus.  All the comparison shots including those from the 18-35 to follow will be aps-c size on both cameras.

I will be getting a metabones mk iii for the af in a couple of days and will be shooting some full size images from the 17mm ts-e but they won't be part of the comparison.

Mike

Mike

Tiger1 wrote:

Can I ask why the raw file from the A7R has a size of 4800x3200???????? Is it the A7R or A7? Exif says A7R. Either way the pixel count is too low.......

I was wondering why the files from the Sony were so small. Most of the ones that are available to download are over 30Mb.

Can you please explain why this has occurred as it is VERY important.

Reply   Reply with quote   Complain
Tiger1
Regular MemberPosts: 313Gear list
Like?
Re: first comparo shots SD1Meos vs Sony A7r and canon 17mm ts-e image test shots
In reply to docmaas, Dec 9, 2013

docmaas wrote:

The A7r is set to aps-c output to match the SD1. While the 17mm ts-e is a full frame lens the metabones ef-nex converter is version 1 and only supports aps-c and manual focus. All the comparison shots including those from the 18-35 to follow will be aps-c size on both cameras.

I will be getting a metabones mk iii for the af in a couple of days and will be shooting some full size images from the 17mm ts-e but they won't be part of the comparison.

Mike

Mike

Tiger1 wrote:

Can I ask why the raw file from the A7R has a size of 4800x3200???????? Is it the A7R or A7? Exif says A7R. Either way the pixel count is too low.......

I was wondering why the files from the Sony were so small. Most of the ones that are available to download are over 30Mb.

Can you please explain why this has occurred as it is VERY important.

Thanks Mike.

People can work out if this is a fair comparison then from a resolution point of view.  Of course colour will be unaffected but will depend on the RAW converter used too. I have Lightroom 5.3RC (free download), Iridient Developer and DXO 9.1 and all 3 give subtly different results.

 Tiger1's gear list:Tiger1's gear list
Sony RX1 Canon EOS 5D Mark II Fujifilm FinePix S5 Pro Sigma SD14 Sigma SD1 Merrill +8 more
Reply   Reply with quote   Complain
docmaas
Veteran MemberPosts: 4,792
Like?
Re: first comparo shots SD1Meos vs Sony A7r and canon 17mm ts-e image test shots
In reply to Tiger1, Dec 9, 2013

Those who have an interest will be able to use the raw files with Iridient to see what each camera is capable of for their needs.

Personally I have long wondered if the secret to the foveon accutance is the unique 3 level sensor or the simple fact that there is no aa filter.  I'm sure it is a combination of both of course but I've been inclined to think that the lack of aa filter was the predominate contributor to the sharpness.

Now that the sony and the Nikon D800E have used the 36mp sensor without an aa filter we should be able to get a better idea of the impact of the missing aa filter as opposed to the foveon sensor design because the pixel pitch between the two sensors is so close.

What is different here is that the same lens can be used on both cameras so the lens variable is eliminated and we are down to rest of the system of which the sensor is the main part.

Mike

Tiger1 wrote:

docmaas wrote:

The A7r is set to aps-c output to match the SD1. While the 17mm ts-e is a full frame lens the metabones ef-nex converter is version 1 and only supports aps-c and manual focus. All the comparison shots including those from the 18-35 to follow will be aps-c size on both cameras.

I will be getting a metabones mk iii for the af in a couple of days and will be shooting some full size images from the 17mm ts-e but they won't be part of the comparison.

Mike

Mike

Tiger1 wrote:

Can I ask why the raw file from the A7R has a size of 4800x3200???????? Is it the A7R or A7? Exif says A7R. Either way the pixel count is too low.......

I was wondering why the files from the Sony were so small. Most of the ones that are available to download are over 30Mb.

Can you please explain why this has occurred as it is VERY important.

Thanks Mike.

People can work out if this is a fair comparison then from a resolution point of view. Of course colour will be unaffected but will depend on the RAW converter used too. I have Lightroom 5.3RC (free download), Iridient Developer and DXO 9.1 and all 3 give subtly different results.

Reply   Reply with quote   Complain
docmaas
Veteran MemberPosts: 4,792
Like?
Sigma leads after these tests but 18-35 yet to be compared
In reply to docmaas, Dec 10, 2013

I've just converted all the images via Iridium adjusing exposure, using the walkway in the upper left corner to adjust wb and using iridium default sharpening and to me it looks like Sigma still has a pretty significant advantage.

When shooting the stucco wall yesterday I noticed a lot of noise on the edges of the sony image even at APS-C and iq was signifcantly degraded in those areas as well.  The sony nex sensors, even the aps-c size sensors, have had magenta noise problems on the outer edges of images with short focal length lenses from rangefinder cameras.  The A7 and A7r have been shown to also have this problem but I am a little surprised to see it with a retrofocal SLR lens like the 17mm.

I don't see it on the finished images but the I can see the resolution loss.  It will be interesting to see how it impacts full frame images.

Here is an image showing the noise along both sides of the image:

This image is at f4.  At full size I think I can see noticeable image degradation where the noise starts.  In the magenta problems on the older nex cameras they magenta could be eliminated with a small program called cornerfix but removing the noise left the degradation in place.  I don't see the noise in these comparison images either but I do think the degradation remains as the corners and edges never really do match the sigma other than in the middle of the image.

Mike

Reply   Reply with quote   Complain
Guido FORRIER
Senior MemberPosts: 1,497
Like?
Re: first comparo shots SD1Meos vs Sony A7r and canon 17mm ts-e image test shots
In reply to docmaas, Dec 11, 2013

hello Mike ,

this is not what i like to know . i have no canon lenses and when i buy a camera i also will use the best lenses made for this camera + full resolution + the best raw conversion + making large prints for exhibition and then compare  =  real world situation .

a lot of lenses are not suited for the sony A7R : see  sonyalpharumors and i will not use or buy them sorry  but this test is useless for me .

kindly,

guido

Reply   Reply with quote   Complain
D Cox
Senior MemberPosts: 8,706
Like?
Re: first comparo shots SD1Meos vs Sony A7r and canon 17mm ts-e image test shots
In reply to docmaas, Dec 11, 2013

Interesting to see the inaccurate colour of the creeper leaves on the Sony. I get the same too-blue result with these creeper leaves on the NEX 5N, using Auto WB. I think it is trying to compensate for the orange.

Next year I will try Daylight instead of Auto WB.

Reply   Reply with quote   Complain
D Cox
Senior MemberPosts: 8,706
Like?
Re: first comparo shots SD1Meos vs Sony A7r and canon 17mm ts-e image test shots
In reply to docmaas, Dec 11, 2013

docmaas wrote:

The A7r is set to aps-c output to match the SD1. While the 17mm ts-e is a full frame lens the metabones ef-nex converter is version 1 and only supports aps-c and manual focus. All the comparison shots including those from the 18-35 to follow will be aps-c size on both cameras.

I will be getting a metabones mk iii for the af in a couple of days and will be shooting some full size images from the 17mm ts-e but they won't be part of the comparison.

So the A7r is being cropped down from 36 Megapixels to 15.

At the same time you are quite correctly using the same lens. If you used it at full frame on the Sony you would get a wider angle of view, so the leaves type of shot would not be a fair comparison.

I think the way out may be to test on a nearly flat surface, which could be a textile with small colour stitches hanging on a wall (or a tapestry). You would then move the Sony camera closer so that the same area of the test subject fills the larger sensor.

The lens aberrations might be slightly different at different distances (especially curvature of field), but I think not by much.  Distances might be about 3 metres (SD1M) and 2 metres (A7r).

Does that seem reasonable ?

Reply   Reply with quote   Complain
docmaas
Veteran MemberPosts: 4,792
Like?
Re: first comparo shots SD1Meos vs Sony A7r and canon 17mm ts-e image test shots
In reply to Guido FORRIER, Dec 12, 2013

Guido,

There is no way I could possibly test all or even any of the lenses you might buy.  My test is intended to neutralize the lens as a variable and make it possible to evaluate the rest of the image path.

What you'll get here is an indication of how much better the sony or sigma is than the other is a limited set of circumstances.

This is substantially what I did with the Nex 7 vs DP2M but remember the lenses were different and I had to equalize the image size by adjusting the distance from which I shot the Nex 7.  Here the lenses are the same and so is the distance.  The only things different are the sensor and whatever else impacts the sensor.

The other test I can and will do is to use all the pixels of the sony to capture an image as close as possible in size to the same image from the Sigma.  The distance for the sony image should be 3/2 that for the SD1 image as the pixels are roughly the same size but the sensors are crop factor 1.5 for the sigma and 1.0 for the Sony.  I will do this test with the 17mm canon lens on both cameras. There will be unavoidable differences in the characteristics of the lens in one image as opposed to the other but the glass itself should be of minimal impact.  I couldn't do this test in the first round because I had the metabones mk 1 with no af support and only big enough for aps-c images.  I now have the Mark iii with af support and able to use with full frame cameras.

Mike

Guido FORRIER wrote:

hello Mike ,

this is not what i like to know . i have no canon lenses and when i buy a camera i also will use the best lenses made for this camera + full resolution + the best raw conversion + making large prints for exhibition and then compare = real world situation .

a lot of lenses are not suited for the sony A7R : see sonyalpharumors and i will not use or buy them sorry but this test is useless for me .

kindly,

guido

Reply   Reply with quote   Complain
docmaas
Veteran MemberPosts: 4,792
Like?
Re: first comparo shots SD1Meos vs Sony A7r and canon 17mm ts-e image test shots
In reply to D Cox, Dec 12, 2013

I think you are suggesting what I just described in my reply to Guido. Here is a quote:

"The other test I can and will do is to use all the pixels of the sony to capture an image as close as possible in size to the same image from the Sigma. The distance for the sony image should be 3/2 that for the SD1 image as the pixels are roughly the same size but the sensors are crop factor 1.5 for the sigma and 1.0 for the Sony. I will do this test with the 17mm canon lens on both cameras. There will be unavoidable differences in the characteristics of the lens in one image as opposed to the other but the glass itself should be of minimal impact. I couldn't do this test in the first round because I had the metabones mk 1 with no af support and only big enough for aps-c images. I now have the Mark iii with af support and able to use with full frame cameras."

Is this what you are suggesting?

Mike

D Cox wrote:

docmaas wrote:

The A7r is set to aps-c output to match the SD1. While the 17mm ts-e is a full frame lens the metabones ef-nex converter is version 1 and only supports aps-c and manual focus. All the comparison shots including those from the 18-35 to follow will be aps-c size on both cameras.

I will be getting a metabones mk iii for the af in a couple of days and will be shooting some full size images from the 17mm ts-e but they won't be part of the comparison.

So the A7r is being cropped down from 36 Megapixels to 15.

At the same time you are quite correctly using the same lens. If you used it at full frame on the Sony you would get a wider angle of view, so the leaves type of shot would not be a fair comparison.

I think the way out may be to test on a nearly flat surface, which could be a textile with small colour stitches hanging on a wall (or a tapestry). You would then move the Sony camera closer so that the same area of the test subject fills the larger sensor.

The lens aberrations might be slightly different at different distances (especially curvature of field), but I think not by much. Distances might be about 3 metres (SD1M) and 2 metres (A7r).

Does that seem reasonable ?

Reply   Reply with quote   Complain
Guido FORRIER
Senior MemberPosts: 1,497
Like?
Re: first comparo shots SD1Meos vs Sony A7r and canon 17mm ts-e image test shots
In reply to docmaas, Dec 12, 2013

many thanks mike for making a new test !

seems like the canon 17 mm is good for architecture :

http://blog.nicgranleese.com/2013/10/24/sony-a7r-review-for-architectural-photographers/

greetings ,

guido

Reply   Reply with quote   Complain
Keyboard shortcuts:
FForum MMy threads