Leaf Aptus 75 versus Nikon D3x

[eronald]

D3 and D3x are two different animals when it come to the sensor AFAIK. And colors, IR sensitivity  ...

Edmund

I looked at the raws with ACR ( not good with Nikon I think ) and C1.

It was not that I did not like the files. I just think there is a big difference
between Leaf and D3X ( and between H3DII-31 and D3X ).

I am very surprised if people do not see this difference  

And I do not think there is much to gain from going from D3 to D3X besides
the file size.

[hubell]

Thank you Simon,

I processed the Nikon file in Raw Developer and the Aptus one in Leaf Capture and loaded 2 sets of JPEGs and a ReadMe file to here. I wanted to see how each file at its native size compares to the other one when scaled up/ down and also to confirm that in this particular case Leaf Capture works better than ACR. There is no sharpening applied to the downscaled Leaf file.

Yair

FWIW, I processed the Leaf file in both ACR and Raw Developer just to see how ACR handles a medium format file, using minimal adjustments in the conversion and  then opeining them in CS4. (My Haaselblad raws are not supported in ACR.) I was quite surprised by how superior the RD conversion was to that from ACR. The ACR TIFF was lifeless and and flat and had lousy skin color, whereas the RD TIFF had a really nice skin color and depth/three dimensionality to it. At least from that limited evidence, I cannot imagine how anyone would use ACR/Lightroom. The workflow with ACR/Lightroom is surely better, but it appears that you are sacrificing so much in IQ after spending a serious amount of money to get that IQ.

[paulmoorestudio]

Gidday,

Below is a link to down load the Raw Leaf and Nikon images.

  https://rcpt.yousendit.com/670715809/e1aa97...48fcddfb789367a

Look forward to every ones opinions.

Cheers

Simon

thanks for taking the time to do this..
I would much prefer to live with the moire on the leaf..  I am a studio still life guy and just don't think the dx3 is up the demands...I was wondering what the files looked like and again thanks for posting them.. I think I will be sticking to mfdbs for awhile.

[yaya]

FWIW, I processed the Leaf file in both ACR and Raw Developer just to see how ACR handles a medium format file, using minimal adjustments in the conversion and  then opeining them in CS4. (My Haaselblad raws are not supported in ACR.) I was quite surprised by how superior the RD conversion was to that from ACR. The ACR TIFF was lifeless and and flat and had lousy skin color, whereas the RD TIFF had a really nice skin color and depth/three dimensionality to it. At least from that limited evidence, I cannot imagine how anyone would use ACR/Lightroom. The workflow with ACR/Lightroom is surely better, but it appears that you are sacrificing so much in IQ after spending a serious amount of money to get that IQ.

I agree 100% about RD. With Leaf files you can also set it to work with the factory's input profiles and curves if you like.

Yair

[bcooter]

I agree 100% about RD. With Leaf files you can also set it to work with the factory's input profiles and curves if you like.

Yair

Raw developer probably is the best processing software on the planet though the interface needs a lot more development.

It can actually make a Phase (kodak chip camera) file mimick the film like look of a leaf file, which is not easy to do, but RD will do it.

Yair, we've talked about this before, but I've never understood why Leaf or Hasselblad didn't buy Brian's software, throw some serious development money at it and have a C-1 competitor, or with RD's processing, a C-1 beater that would process all files.

Some companies talk about open formats, but not when it comes to processing and if you want to be the standard of the industry, the software must process all files.

Only RD and LR/Photoshop does this.

 (This is why every retoucher  uses photoshop to process in).

[Panopeeper]

I processed the Nikon file in Raw Developer and the Aptus one in Leaf Capture and loaded 2 sets of JPEGs and a ReadMe file to here. I wanted to see how each file at its native size compares to the other one when scaled up/ down and also to confirm that in this particular case Leaf Capture works better than ACR. There is no sharpening applied to the downscaled Leaf file.

The color of the D3X rendering is way off. Had you picked WB on the white or light gray patch, it would look very differently. I don't have Raw Developer, but even with ACR the rendering of the shirt is much better than what you uploaded (at least it looks better, but I don't know how it is in reality). The skin color depends on the selected profile.

The garment looks the best in the full-size Aptus shot as long as there is no moire. However, the moire removal removes the fine structure as well; on that area the D3X shot is better than the full-size Aptus.

I wonder what the excuse is for using MFDBs for anything moire-prone, except for the lack of skill in processing the raw files.

[Willow Photography]

D3 and D3x are two different animals when it come to the sensor AFAIK. And colors, IR sensitivity  ...

Edmund

I have seen some side bye side comparison and I could not see any significant difference between D3 and D3X.

But if you can backup your word with some images, I am willing to reconsider.  

[Dustbak]

I wonder what the excuse is for using MFDBs for anything moire-prone, except for the lack of skill in processing the raw files.

Multishot. There is nothing that compares to a multishot file when dealing with moire prone garments.

[Panopeeper]

Here is an ACR rendering of the D3X image with the color profile Camera D2X Mode 2:

http://www.yousendit.com/download/UmNJN3RRTXZ6NEozZUE9PQ

[Panopeeper]

Multishot. There is nothing that compares to a multishot file when dealing with moire prone garments.

1. How many cameras offer this choice, and at what extra price?

2. What if someone is wearing that garment? Model deep frozen before the show?

[Dustbak]

Yikes,

That rendering is awful! Funny thing is that when I balance on the neutral patches of the card with Yairs file it doesn't alter but neither does it when I do that on yours. The background is a dreadful yellowish thing and there is an awful cast over his skin making you wonder how long he is dead. Both files (Panopeepers & Yairs) show way too much magenta in the skin to be pleasant IMO.

I most definitely prefer the Leaf file but I do see why the D3x file might be considered adequate enough for many.

Nowadays multishot backs are not that much more expensive than single shot. 50% of all MFDB brands offer it (that sounds much better than there are only 3 current models ).

First the clothing in this case hangs and is not worn. Even with this model I could take a single and a multishot and blend the 2 together which is actually something I quite regularly do with models. Nowadays much clothing is shot either styled flat or on dolls (much cheaper than models ) in which case it is a complete non-issue.

Most people that have never worked with multishot underestimate the range you can put it to use.

[yaya]

Here is an ACR rendering of the D3X image with the color profile Camera D2X Mode 2:

http://www.yousendit.com/download/UmNJN3RRTXZ6NEozZUE9PQ

Why use a D2X profile on a D3X file? The sensors are very different so what you have done was not just killing the model but also killing any detail that the D3X was able to show in the Nikon T-shirt.

If you read the ReadMe file that I've posted, I said that the D3X's files are quite new for me but even so, in RD I think that I've managed to keep the detail in, keep the background White (although I guess it was Yellow as the flash duration was fairly short, or shorter than what 1/125 can record) and make the skintone look real, if only a bit red-ish.

As for the Moire tool in LC, you are welcome to download the software from our website so that you can try it for yourself and see that it does not affect detail in any way whatsoever. I did say in my ReadMe that there is pattern Moire that cannot easily be removed from the raw file.

BTW RD is available as a free, fully functioning DEMO version and the licence is $125 so it is worth your investment for getting the most out of your photos.

Can we agree that since we are both not photographers (at least I'm not), we should let the photographers here judge the results photographically and not by reading numbers off a non-photographic analysis tool.

Kindest

Yair

[eronald]

FWIW, I processed the Leaf file in both ACR and Raw Developer just to see how ACR handles a medium format file, using minimal adjustments in the conversion and  then opeining them in CS4. (My Haaselblad raws are not supported in ACR.) I was quite surprised by how superior the RD conversion was to that from ACR. The ACR TIFF was lifeless and and flat and had lousy skin color, whereas the RD TIFF had a really nice skin color and depth/three dimensionality to it. At least from that limited evidence, I cannot imagine how anyone would use ACR/Lightroom. The workflow with ACR/Lightroom is surely better, but it appears that you are sacrificing so much in IQ after spending a serious amount of money to get that IQ.

ACR is a mediocre converter with good workflow. Adobe is succeeding in establishing itself as the Microsoft of graphics software.
Adobe do listen to customers, but  something is going wrong here.
RD has does a stellar, superb job on most camera files so far, but IMHO it is below its usual quality on the D3x.
C1 seems to be the converter to beat for D3x, I think, although Gimp (UFRaw) is actually quite good, and NX2 is the reference implementation with all the strange Nikon color twists.


Edmund

[Panopeeper]

Why use a D2X profile on a D3X file?

Because this demonstrates, how much effect different color profiles can have on the ACR rendering; there are nine profiles, and ACR now supports one's own color rendering.

As to "killing the model": that's the question of taste, for we don't know the reality. Who said you that the gentleman has a tanned color? I myself despise tanning.

The details on the shirt are not lost due to the color profile, I guess you know this. Aside from the raw converter, the contrast setting plays a huge role; I did not turn up the contrast. Beside, I would not be surprized to be shown in a competent comparison of the shot between different converters that ACR's result is inferior for the D3X. I am rather a critic of ACR, but a fair comparison of the rendering is more laborous, and for the comparison of cameras one has to evaluate the best way to process the raw files.

As for the Moire tool in LC, you are welcome to download the software from our website so that you can try it for yourself and see that it does not affect detail in any way whatsoever

I don't know who you want to fool: yourself or me. The moire is in the raw channels, it is not caused by the raw conversion. Just as the lack of the AA filter has its advantages, simulating that effect has the disadvantages as well. You can't have your cake and eat it.

[yaya]

Because this demonstrates, how much effect different color profiles can have on the ACR rendering; there are nine profiles, and ACR now supports one's own color rendering.

As to "killing the model": that's the question of taste, for we don't know the reality. Who said you that the gentleman has a tanned color? I myself despise tanning.

The details on the shirt are not lost due to the color profile, I guess you know this. Aside from the raw converter, the contrast setting plays a huge role; I did not turn up the contrast. Beside, I would not be surprized to be shown in a competent comparison of the shot between different converters that ACR's result is inferior for the D3X. I am rather a critic of ACR, but a fair comparison of the rendering is more laborous, and for the comparison of cameras one has to evaluate the best way to process the raw files.


I don't know who you want to fool: yourself or me. The moire is in the raw channels, it is not caused by the raw conversion. Just as the lack of the AA filter has its advantages, simulating that effect has the disadvantages as well. You can't have your cake and eat it.

A "color profile", be it an ICC input/ output profile, a matrix or a simple set of "instruction" can contain a contrast curve that will affect the appearance of detail without touching neither contrast nor sharpness settings. Try a raw converter that supports input profiles like C1, RD or LC and you will see it straight away. Speak to Edmund if you want further explanation on how profiles work.

The Moire tool in LC works on the raw data and not on the conversion. The upside is that you can create a new, "corrected" raw file that can be then converted in ACR or any other converter. It does not simply simulate an AA filter but uses the old patented Leaf MagicAl algorithm to identify the affected pixels and to adjust the hue/ saturation of each one of them without moving them or changing their luminance values. Again you do not have to trust me on any of this you just have to try it before dismissing it.

Yair

[Alex MacPherson]

Is there Raw Developer for PC? I see it for Mac but not pc

[Panopeeper]

A "color profile", be it an ICC input/ output profile, a matrix or a simple set of "instruction" can contain a contrast curve that will affect the appearance of detail without touching neither contrast nor sharpness settings

Yes. As I wrote, the details are an issue of contrast. It is not bad to remember: the color profile is primarily, secondarily and tertiarily about color. The effect of the profile on the contrast can be enhanced, mitigated or eliminated any time later, unlike the effect on the color.

The Moire tool in LC works on the raw data and not on the conversion. The upside is that you can create a new, "corrected" raw file that can be then converted in ACR or any other converter. It does not simply simulate an AA filter but uses the old patented Leaf MagicAl algorithm to identify the affected pixels and to adjust the hue/ saturation of each one of them without moving them or changing their luminance values. Again you do not have to trust me on any of this you just have to try it before dismissing it.

I won't install the program for the purpose of a one-time evaluation, but if you happened to have done this on the present sample raw image and have the adjusted raw file, I would happily take a look at it. This notwithstanding the fact, that the range of correctable moire is quite limited, judged from the sample.

[Ray]

A "color profile", be it an ICC input/ output profile, a matrix or a simple set of "instruction" can contain a contrast curve that will affect the appearance of detail without touching neither contrast nor sharpness settings.

This is not true, the way you've expressed it. 'Without touching neither' means 'touching either'. You've got a double negative. You presumably meant, ...'will affect the appearance of detail without touching either contrast or sharpness settings...", or you meant,.....'will affect the appearance of detail touching neither contrast nor sharpness settings...'

Sorry to be such a pain, but one needs to be precise in such matters   .

[eronald]

Ok, here is my own C1 conversion of the D3x image. I have done ONE single skin tone edit using the C1 Pro color editor.

By the way, the magenta skin issue is interesting, people I see in public transportation often have bright pink faces; however I do agree that it looks horrid in print. Which is why I fixed it to yellow. A pair of color-aware female eyes to whom I showed the edited AND unedited images said the difference is -to her- quite subtle, and that with this red-haired type the magenta rendering is not shocking,  quote "with that bottom part, the top part is to be expected". I'm sure glamour shooters have heard similar remarks

Frankly, if this is the worst that can happen to D3x color, I can live with it.

As for the difficulties we are all having with raw conversion on this image, I think there are some lens-color flare effects, possibly, which account for our problems with background color and  with setting the shadow point.

Edmund

[eronald]

Because this demonstrates, how much effect different color profiles can have on the ACR rendering; there are nine profiles, and ACR now supports one's own color rendering.

As to "killing the model": that's the question of taste, for we don't know the reality. Who said you that the gentleman has a tanned color? I myself despise tanning.

As long as you don't confuse ACR "profiles" and ICC profiles. I find it funny that Adobe, itself a founding member of the ICC cannot manage to work fully inside the consortium with regards t camera color.

Re. the model, Americans seem to dislike magenta skin tints a lot;  I think non-natives cannot "feel"  the reason, and need to just accept it intellectually.

Edmund