Luminous Landscape Forum
Equipment & Techniques => Medium Format / Film / Digital Backs – and Large Sensor Photography => Topic started by: bcroslin on November 23, 2007, 03:05:00 pm
-
I was able to test the Mamiya ZD back alongside a Leaf Aptus 22 today. I set up the camera on a tripod and simply shot a scene in overcast light with a Mamiya 120mm Macro lens. I included a gray card, some black fabric, some objects that show specularity and some fine detailed type.
Tests were at ISO 100 and 200. All images were shot within seconds of each other to prevent a change in lighting conditions.
I felt comparing the ZD back to the Aptus 22 was fair due to both sharing the same Dalsa sensor. No surprise to me that the images are virtually identical with neither looking better or worse than the other. My guess is that if this were a double-blind test no one would be able to consistently tell which file came from which back.
Tif images processed in ACR 4.3 as well as raw files can be downloaded here:
http://www.fourwestweddings.com/test/ZD_Leaf_test.html (http://www.fourwestweddings.com/test/ZD_Leaf_test.html)
Note - the images are full rez.
-
Forgot to mention the processed files have all sharpening and noise reduction turned off. They are processed straight out of the camera.
-
Here's the rez'd down ISO 100 tif files saved as jpg 12. File on the right is the Leaf and the ZD on the left.
(edited because file names were backwards)
-
My experience from both is that they are very close but:
The leaf has better color rendition and color uniformatity, try shooting a WHITE piece of paper with both backs full screen, the leaf has better uniformaty.
Also when you closely look at shadows parts the leaf has better detail rendition, this is rather logical because it's a 16 bits device compared to the 12 bits from the ZD.
The reason for me to switch was not the picture quality (when you have a good back (newer versions that is)).
But the tethered issue and the whole interface of the back like buffer, speed etc.
If it's worht the price difference is difficult to say.
It's like with very high end audio for a slight increase in quality be prepared to pay alot of money.
The Leaf is better but not day and night.
For my work the ZD was great but because I do alot darker looking work lately the more detailed shadows pushed me over for sure.
Also the added ISO25 is very handy with the leaf, would not want to miss it now.
-
Here's 2 crops from the tif files. Leaf on the right. ZD on the left.
(edited because file names were backwards)
-
Mamiya ZD a little more noisy in the blue channel.
-
Mamiya ZD a little more noisy in the blue channel.
[a href=\"index.php?act=findpost&pid=155285\"][{POST_SNAPBACK}][/a]
My bad - I had the files named backwards. This is a direct screen shot.
Both are noisy in the blue channel. The noise is a tiny bit smoother in the Leaf file.
-
You sure about that?
Leaf on the Left. ZD on the right.
[a href=\"index.php?act=findpost&pid=155286\"][{POST_SNAPBACK}][/a]
In attachement two details from the first images you posted (100 iso), mamiya is clearly more noisy but .. not too much.
-
Here's 2 crops from the tif files. Leaf on the left. ZD on the right.
[a href=\"index.php?act=findpost&pid=155284\"][{POST_SNAPBACK}][/a]
The Mamiya ZD shows a little CA in the white text (upper side). The first word "Skin" and a few words after in "poison control". The leaf is clean. Note that I'm not confusing it with the blue and pink dots.
Anyway, image quality is like 100% equal.
-
But couldn't you argue that the Leaf files processed in Leaf Capture will be superior in terms of detail and overall image quality compared to ACR, and since you can't process the ZD files in Leaf Capture, the Leaf Aptus wins, hands down?
See: http://www.outbackphoto.com/reviews/equipm...af_aptus22.html (http://www.outbackphoto.com/reviews/equipment/leaf_aptus_22/leaf_aptus22.html)
-
But couldn't you argue that the Leaf files processed in Leaf Capture will be superior in terms of detail and overall image quality compared to ACR, and since you can't process the ZD files in Leaf Capture, the Leaf Aptus wins, hands down?ht]
Not seeing your logic here. I did process the Leaf files through LC11 and didn't see any difference from what I got from ACR other than the LC11 conversions were flat.
-
The leaf has better color rendition and color uniformatity, try shooting a WHITE piece of paper with both backs full screen, the leaf has better uniformaty.
Done. Aptus 22 and ZD back shot on white background on lights. Both ISO 100 with 120mm macro and processed in ACR.
No difference that I can tell.
(EDIT)
I replaced both files with an Aptus 22 file run through LC 11 and a ZD back file run through Silkypix. All noise reduction and sharpening turned off.
-
Hi
Process the ZD files in Raw Developer & the Leaf files to. Also since the Mamiya/Phase announcement the ZD will be supported by C1. With that news you can bypass the Mamiya software. For the money the ZD back is a very good deal & there are a few Mamiya 645 AF cameras out there ready for a digital back.
-
Not seeing your logic here. I did process the Leaf files through LC11 and didn't see any difference from what I got from ACR other than the LC11 conversions were flat.
[a href=\"index.php?act=findpost&pid=155299\"][{POST_SNAPBACK}][/a]
Bob:
When I downloaded the Aptus 22 ISO 200 raw file, it downloaded as a tiff. I changed the suffix to .mos and then brought that file into LC11. I processed out and compared to the tiff ZD file and the .mef ZD file (which also downloaded as tiff, but I then changed it to .mef).
Here were my findings:
I viewed the ISO 200 ZD files in Photoshop - both the Tif and (ACR-converted) .mef.
I also viewed the Aptus 22 tiff file and the converted .mos (via Camera Raw) in Photoshop.
I also converted the ISO 200 .mos file through LC11.0.1 and brought it into Photoshop.
My results - the Aptus tiff and the ACR conversion both looked slightly noisy with a fair amount of color noise.
Ditto the ZD files, although the ACR conversion, while still a bit noisy, didn't exhibit much color noise at all.
The LC11 converted .mos file, opened up in Photoshop, was way cleaner than any of the other files. There was no comparison.
Something doesn't seem right with your results from the LC11 processed tiff.
Steve Hendrix
www.ppratlanta.com/digital.php
-
Something doesn't seem right with your results from the LC11 processed tiff.
Steve Hendrix
www.ppratlanta.com/digital.php
[a href=\"index.php?act=findpost&pid=155313\"][{POST_SNAPBACK}][/a]
I'm not quite sure what you're seeing - I processed everything through ACR 4.3.
I included the raw files so everyone can process them in their flavor of convertor. Process the raws and compare.
-
I'm not quite sure what you're seeing - I processed everything through ACR 4.3.
I included the raw files so everyone can process them in their flavor of convertor. Process the raws and compare.
[a href=\"index.php?act=findpost&pid=155315\"][{POST_SNAPBACK}][/a]
You indicated that you processed files through LC11 and they looked indistinguishable from the ACR conversion. That has not been the result I have seen with these files. The LC11 conversion is way superior to the ACR conversion.
Steve Hendrix
www.ppratlanta.com/digital.php
-
Bob:
When I downloaded the Aptus 22 ISO 200 raw file, it downloaded as a tiff. I changed the suffix to .mos and then brought that file into LC11. I processed out and compared to the tiff ZD file and the .mef ZD file (which also downloaded as tiff, but I then changed it to .mef).
Here were my findings:
I viewed the ISO 200 ZD files in Photoshop - both the Tif and (ACR-converted) .mef.
I also viewed the Aptus 22 tiff file and the converted .mos (via Camera Raw) in Photoshop.
I also converted the ISO 200 .mos file through LC11.0.1 and brought it into Photoshop.
My results - the Aptus tiff and the ACR conversion both looked slightly noisy with a fair amount of color noise.
Ditto the ZD files, although the ACR conversion, while still a bit noisy, didn't exhibit much color noise at all.
The LC11 converted .mos file, opened up in Photoshop, was way cleaner than any of the other files. There was no comparison.
Something doesn't seem right with your results from the LC11 processed tiff.
Steve Hendrix
www.ppratlanta.com/digital.php
[a href=\"index.php?act=findpost&pid=155313\"][{POST_SNAPBACK}][/a]
Hi
Try the ZD files through Raw Developer. You will notice that the in default the on RD the ZD files will be a little flat. There is enough controls to adjust the file before you process a tiff.
Thanks Denis
-
Done. Aptus 22 and ZD back shot on white background on lights. Both ISO 100 with 120mm macro and processed in ACR.
No difference that I can tell.
[a href=\"index.php?act=findpost&pid=155306\"][{POST_SNAPBACK}][/a]
I actually see a big difference here. The ZD back has like a green halo/cast at the edges and corners. Specially the left and right sides. Am I the only one seeing this?
-
Essentially identical. Any differences must be within sample variation.
Quentin
-
The LC11 conversion is way superior to the ACR conversion.
[a href=\"index.php?act=findpost&pid=155317\"][{POST_SNAPBACK}][/a]
Steve,
I see what you're saying. I went back and did another conversion with LC11 and you are correct - the Leaf file run through LC11 does look better than run through ACR 4.3.
LC11 is definitely a strong selling point for the Aptus and any Leaf product IMO.
It will be interesting to see if the new version of C1 will support the ZD back.
-
I added two new files to the download page.
The first is the Leaf Aptus 22 file processed through LC11 with all sharpening and noise reduction turned off. The conversion indeed looks better than the same file processed with ACR 4.3 but I'm wondering if some noise reduction and sharpening is still happening in LC11. The noise and sharpening in the blue channel looks as if it's not turned off all the way.
The second file is the ZD back file processed through ACR 4.3 with some minimal noise reduction and sharpening on. The file looks much closer to Leaf file processed through LC11.
I remember the older version of Capture One didn't allow you to turn sharpening off completely. Is this possibly the case with Leaf Capture?
Could that be the reason the ZD and Leaf files look almost identical processed through ACR 4.3 with sharpening and noise reduction turned off?
(EDIT)
I also ran the ZD image through Silkypix (which I had never used before) and I'm surprised at how nice the conversion is. It's definitely better, although flatter, than the ACR conversion.
I've linked it from the test page.
I'm beginning to wonder if the conversion software is the weak link with the ZD back.
-
Hi
There isn't much in it. I use RD for my ZD files.
-
Hi
There isn't much in it. I use RD for my ZD files.
[a href=\"index.php?act=findpost&pid=155400\"][{POST_SNAPBACK}][/a]
I ran the ZD file through RD with default noise reduction on and sharpening off and it looks pretty good - definitely better than the ACR conversion.
When I run the Aptus 22 file through RD it looks nearly identical with the ZD file.
Noise, sharpness and color look great in both with the ZD file trending towards being just a little warmer. The noise in the blue channel is just a tiny bit better than the ZD. I mean minutely.
-
How did you shoot the white ?
I shot a white piece of paper underexposed (to make sure nothing blew out) to app 180,180,180 with the leaf there was a good coloruniformity with the ZD I had a totally green cast on the lower part (portrait) of the white.
My first ZD back had this also but to a VERY MINOR ammount.
It would be a GREAT step forward when they have solved that.
-
I ran the ZD file through RD with default noise reduction on and sharpening off and it looks pretty good - definitely better than the ACR conversion.
When I run the Aptus 22 file through RD it looks nearly identical with the ZD file.
Noise, sharpness and color look great in both with the ZD file trending towards being just a little warmer. The noise in the blue channel is just a tiny bit better than the ZD. I mean minutely.
[a href=\"index.php?act=findpost&pid=155403\"][{POST_SNAPBACK}][/a]
Hi Frank
I know of the best retouching house in Sydney using RD as their developer. This is one hell of a developer & super fast as well.
-
How did you shoot the white ?
[a href=\"index.php?act=findpost&pid=155418\"][{POST_SNAPBACK}][/a]
I shot a white backdrop on lights at the metered exposure.
-
John,
I totally agree about the usability of the Aptus 22 versus the ZD back and really it's not my intention to pit one against the other. The buffer on the ZD makes it hard, but not impossible, to use in situations where you're trying to capture moments. In my experience the buffer can be filled on the ZD in ten consecutive frames. In normal fast shooting I would hit the buffer somewhere between 16-20 frames and then have to wait for it to clear in 3-5 sec's.
The Aptus buffer isn't really much larger but the Aptus will shoot in a compressed format that allows for the buffer to fill and clear faster. The compressed files have to be converted in LC11 or the Leaf Raw Converter tool thus adding a step to your workflow. This isn't a deal breaker by any means but I think it often gets left unsaid.
The LCD on the ZD works but is really best used to check the histogram. The LCD is probably my biggest gripe with the ZD.
To be totally honest, I was reluctant to compare the image quality of the ZD back to the Aptus 22 only because I figured the Aptus would blow it away. I was pleasantly surprised to see the ZD hold it's own.
For me it all comes down to ROI and as mainly an editorial shooter I can't justify $20k for a back and most of my shoots won't allow for an $800 MFDB rental fee. The ZD back for me is affordable and coincidentally since it's been released prices on refurb Aptus and Phase backs have dropped significantly.
My hope is that Mamiya is able to keep prices low on the new products to come out of their relationship with Phase.
-
I'm beginning to wonder if the conversion software is the weak link with the ZD back.
[a href=\"index.php?act=findpost&pid=155385\"][{POST_SNAPBACK}][/a]
It's the weakest link, IMO. The ZD back is capable of outstanding IQ, we all know that.
So far, the Mamiya software isn't too popular, while Adobe RAW converters don't really do the files justice, IMO. Raw Developer is your best bet right now. SilkyPix seems too watercolor-ish, IMO, losing the fine detail from smudging.
Just wait to see how CaptureOne processes ZD files, and you might feel like you're shooting with a whole new camera, but I doubt it will process any better than Raw Developer.
-
I actually did play with the Mamiya software and I was able to get really close to the Aptus file processed in LC11. The Mamiya software gets a bad rap IMO and while it's not as good as LC11 or RD it still does a very good job.
I added the Mamiya software conversion to the page of links:
http://www.fourwestweddings.com/test/ZD_Leaf_test.html (http://www.fourwestweddings.com/test/ZD_Leaf_test.html)
I'm convinced that noise reduction and sharpness is not totally turned off when you zero the settings out in LC11. I can nearly match the ZD back and Aptus 22 files in ACR and RD when I do some slight tweaking of the NR and sharpening. The file produced by LC11 looks very similar.
Yair, can you please offer some insight into how LC11 handles noise reduction and sharpening?
-
Thanks for posting these images. Interesting results. I was looking at the A22 at 200 and comparing it with the Mamiya at the same ISO in lightroom. I also checked out the tiff developed in Mamiya software at the same ISO.
The Leaf image is definitely more detailed. I wonder though if this could just be due to collimation variation (ie: chip placement). The place this really becomes noticeable is on the white balance reference card. In the word 'Bal' detail is being resolved by the A22, but the Mamiya just shows a grey tone.
I suppose noise reduction may also be reducing detail. The focus point looks to be pretty much the same in both images so I doubt if its that.
Not knocking the Mamiya by the way. For the price, this is a rather astounding product and it opens up MF to many more photographers.
By the way, I conclude the same about LC11 and its predecessors. It always feels that its doing some noise reduction no matter what. My workflows are
1) LC11 for tethered shooting
2) Lightroom for fast edits for rendering most client and casual work
3) Raw Developer for rendering for critical prints and personal work
The RD developer has created the best sharpening and file processing I've found. If only that could be plugged in to Lightroom!
-
I think that you see the difference between 12bits vs 16bits.
I found that shadow detail and highcontrast detail is better resolved by the leaf compared to the ZD.
I had both for quite some time.
For it's price the ZD BACK is amazing and will open up a new market.
I would however loved to have seen it in 16 bits and without tethered problems on PC.
-
The Leaf image is definitely more detailed. I wonder though if this could just be due to collimation variation (ie: chip placement). The place this really becomes noticeable is on the white balance reference card. In the word 'Bal' detail is being resolved by the A22, but the Mamiya just shows a grey tone.
I can easily get the same detail out of either file with the appropriate amount of noise reduction and sharpness. I honestly do not believe there's any more or less detail in either image. I can see the dot pattern resolved in the word Bal in both files.
By the way, I conclude the same about LC11 and its predecessors. It always feels that its doing some noise reduction no matter what. My workflows are
This has been confirmed to me. You have to set the "grain" setting to +50 to get as close to no noise reduction as you can get but there's still some happening. Also, there is some pre-sharpening that is happening dependent on scale even if you turn sharpening off.
Bottom line - same chip and in my testing the image is arguably the same. The most important factor is the choice of processing software with either file.
Once again I will say it for the record: I was only testing IQ with both backs. IMO the overall package of the Leaf Aptus 22 is a better product compared to the ZD but the ZD back is a better bang for the buck IMO.
-
I am new to the forum and this is my first post.
I am a pro photographer and I am looking into 4 different digital backs/cameras to photograph jewelry: The Aptus 22, ZD, Canon Ds mark III, and Better Light 6000-HS.
That’s the back story. I know that all digital backs/cameras have a low pass filter over the chip to control moray patterns by softening the data, if you will.
The ZD is the only digital back or camera I know of that lets you shoot without one, and I am very interested in seeing the difference with it off versus on .
The question to bcroslin is, ‘was your ZD test shot with or without the low pass filter?’
-
No, digital backs have no AA filter at all, just an IR filter.
I think the gold standard for jewelry nowadays are multishot backs as made by Hasselblad and Sinar.
Betterlight needs continuous light.
Edmund
I am new to the forum and this is my first post.
I am a pro photographer and I am looking into 4 different digital backs/cameras to photograph jewelry: The Aptus 22, ZD, Canon Ds mark III, and Better Light 6000-HS.
That’s the back story. I know that all digital backs/cameras have a low pass filter over the chip to control moray patterns by softening the data, if you will.
The ZD is the only digital back or camera I know of that lets you shoot without one, and I am very interested in seeing the difference with it off versus on .
The question to bcroslin is, ‘was your ZD test shot with or without the low pass filter?’
[a href=\"index.php?act=findpost&pid=187683\"][{POST_SNAPBACK}][/a]
-
No, digital backs have no AA filter at all, just an IR filter.
I think the gold standard for jewelry nowadays are multishot backs as made by Hasselblad and Sinar.
Betterlight needs continuous light.
Edmund
[a href=\"index.php?act=findpost&pid=187758\"][{POST_SNAPBACK}][/a]
Thank you eronald. Did some more reading and you are correct.
So the ZD is a Foveon sensor back with the option of using the same type low pass filter that is used with Bayer sensored cameras.
-
Thank you eronald. Did some more reading and you are correct.
So the ZD is a Foveon sensor back with the option of using the same type low pass filter that is used with Bayer sensored cameras.
[a href=\"index.php?act=findpost&pid=187831\"][{POST_SNAPBACK}][/a]
Incorrect, the ZD back also uses a CCD sensor with Bayer colour filter array. It is not a Foveon sensor. As far as I know, there are absolutely no available medium format digital back products with a Foveon sensor in it.
-
Actually, you can buy an AA/low-pass filter for the ZD (the filter is user-changeable, snaps in and out), but in stock trim it has no AA. Also, the old Kodak DCS back had an optional filter as well...
-
Incorrect, the ZD back also uses a CCD sensor with Bayer colour filter array. It is not a Foveon sensor. As far as I know, there are absolutely no available medium format digital back products with a Foveon sensor in it.
[a href=\"index.php?act=findpost&pid=187845\"][{POST_SNAPBACK}][/a]
Your help here has been wonderful!
No more CCD digital image sensors that are only capable of
recording just one color at each point in the captured image instead of the full range of colors at each location.
No more post capture interpolation or demosaicing with AA filter!
It's the Sigma SD14 for small work (even enlarging to the 5D raw output size the SD14 is still sharper and has more of a 3D feel then the 5D, and the Better light 6K for the big stuff.
Pure Image no compromises.
Thanks all.
P.S. Can’t wait for someone to make Foveon 22 meg med format back.
-
so with the zd back...what exactly is the reason it can't currently be used on a view camera body...is it just there is no way to fire it?
Is it simply waiting for the correct cable to be manufactured or is it more, such as an adapter plate, etc?
Marc