Canon slide copying set-up

[M.Chapman]

20 minutes ago, MDM said:

 

It's good you are finding Topaz DeNoise (let's abbreviate as TDN) useful. Would you care to elaborate a bit on how you have blended the images as this is a vital part of the process and it is not clear what you did.  

 

This is my workflow.

• Open RAW file in PS ACR RAW

• Apply my LED lightbox + Velvia preset

• Check and adjust WB

• Make alignment and crop adjustments

• Make all tonal and saturation adjustments

• Develop/tranfser into PS as 16 bit sRGB resized to 3000 x 2000

• In PS remove all dust spots and do cloning etc.

• Save as a 16 bit TIFF

• Open in Topaz Denoise AI

• Apply 15, 15, 0, 0.00

• Save as 16 bit TIFF

• Open denoised TIFF in PS select all and cut and paste as a new layer over original image layer

• Adjust opacity of top layer (Topaz denoised) to 50-75%  to allow 25-50% of original image to show through

• Apply 0.2 radius Gaussian blur to entire top layer (if some edges are too crisp)

• Save as jpg at quality level 12

 

For me I regard TDN as a final step in the process, a bit like sharpening before print. It also runs much faster on the downsized 3000 x 2000 tiff file. It also helps if the TDN window isn't maximised, so the preview generates more quickly. 

 

25 minutes ago, MDM said:

By the way, there is a Lightroom plugin installed on my machine (maybe not available for the trial version or maybe you didn't install it?) but it is limited as it can't work on the raw files. There is also a Photoshop plugin.

 

I can see the TDN plugin in LR, but I can't find it in PS. ISTR you're on a Mac too, where do you find the TDN plugin in PS?

 

Mark

[Harry Harrison]

28 minutes ago, MDM said:

3000 pixels will easily clear the bar

Ok, interesting, but that is the bar isn't it? This is Alamy so I suppose it's a judgement about whether a larger file would be more saleable and maybe it wouldn't be, I don't have enough experience here to know. It may well be that in order to clear QC by the normal route then all traces of the film origins have to be expunged which seems a bit Orwellian. Looking for visible grain at 100% is an unrealistically severe test in my opinion and doesn't translate into the useable size for reproduction in print. There were fantastic double page spreads and coffee table books in the days of film.

 

I used to be associated with a small picture library through the period of the transition to digital, not one that you would have heard of. We used to scan everything to 50MB, 5200 px as I remember for 35mm but the standard for supplying from their website was 3600 px, larger sizes were available on request.

Edited February 13, 2020 by Harry Harrison

[M.Chapman]

1 hour ago, Harry Harrison said:

I must say I was scared off by the blue spilling over on to MDM's mountains

 

I don't think TDN caused that. I suspect may have occurred when MDM enhanced the sky and part of the mountain was also selected. I've not seen TDN cause problems like that.

 

46 minutes ago, MDM said:

There is also the benefit that there is no introduction of artifacts at boundaries which happens in TDN (I have noticed that the slide mount can cause problems at the margins of the slide but artifacts can also be introduced at boundaries within the image).

 

Certainly TDN result needs to be checked for problems, especially around high contrast edges (I crop the mount out before applying TDN). I've found results to be reasonably reliable on 2000 x 3000 images using the default 15, 15, 0, 0.00 settings in TDN.  I had problems with higher resolution images (no idea why). TDN is not as "predictable/uniform" as doing adjustments to noise and detail sliders in ACR/LR. But when TDN works I think it gets better results than I seem to be able to get with sliders in ACR/LR. I must confess I'm not good with those sliders as I never usually touch them and there are too many that simply seem to cancel one another. I just use the LR/ACR defaults. I have put Martin Evening's book on my Amazon list 🙂

 

I found TDN useless on RAW. It was slow and wrecked the colour balance and I couldn't recover it.

 

Mark

Edited February 13, 2020 by M.Chapman

[MDM]

1 minute ago, M.Chapman said:

 

This is my workflow.

• Open RAW file in PS ACR RAW

• Apply my LED lightbox + Velvia preset

• Check and adjust WB

• Make alignment and crop adjustments

• Make all tonal and saturation adjustments

• Develop/tranfser into PS as 16 bit sRGB resized to 3000 x 2000

• In PS remove all dust spots and do cloning etc.

• Save as a 16 bit TIFF

• Open in Topaz Denoise AI

• Apply 15, 15, 0, 0.00

• Save as 16 bit TIFF

• Open denoised TIFF in PS select all and cut and paste as a new layer over original image layer

• Adjust opacity of top layer (Topaz denoised) to 50-75%  to allow 25-50% of original image to show through

• Apply 0.2 radius Gaussian blur to entire top layer (if some edges are too crisp)

• Save as jpg at quality level 12

 

For me I regard TDN as a final step in the process, a bit like sharpening before print. It also runs much faster on the downsized 3000 x 2000 tiff file. It also helps if the TDN window isn't maximised, so the preview generates more quickly. 

 

 

I can see the TDN plugin in LR, but I can't find it in PS. ISTR you're on a Mac too, where do you find the TDN plugin in PS?

 

Mark

 

It's under the Filter menu at the bottom under Topaz Labs. I have other Topaz plugins as well under Topaz Studio.

 

Thanks for posting the info on your workflow.  I would have to go with each to his own on that. Working on the noise reduction on the raw file is a must for me. Topaz recommend doing the noise reduction first as well. Rather than doing a global blend I think it is better to do localised blending. You are also downsizing at a very early stage presumably to have a smaller file for speed in TDN so ending up with a small file. I would rather produce a full size master and then downsize as required. 

 

As for using sRGB, well that is a whole new story. I can understand that if you are only ever intending the images to be uploaded to Alamy as it is at the moment but you are throwing away masses of information. Why work in 16 bit and throw away so much of the colour information. It doesn't have any effect on the workflow to use a bigger colour space and, as with resizing, export or save in a smaller space as a final step to suit the purpose.

[MDM]

13 minutes ago, Harry Harrison said:

Ok, interesting, but that is the bar isn't it? This is Alamy so I suppose it's a judgement about whether a larger file would be more saleable and maybe it wouldn't be, I don't have enough experience here to know. It may well be that in order to clear QC by the normal route then all traces of the film origins have to be expunged which seems a bit Orwellian. Looking for visible grain at 100% is an unrealistically severe test in my opinion and doesn't translate into the useable size for reproduction in print. There were fantastic double page spreads and coffee table books in the days of film.

 

I used to be associated with a small picture library through the period of the transition to digital, not one that you would have heard of. We used to scan everything to 50MB, 5200 px as I remember for 35mm but the standard for supplying from their website was 3600 px, larger sizes were available on request.

 

That was pretty much the case with Alamy until around 2012 I think. Everything had to be a minimum of 48MB and had to be upsized if the image was too small which made a lot of images from 12MP cameras borderline. Alamy adapted as things went more and more online where much smaller files sizes are fine. When we met in Ely, James Allsworth did say that most of Alamy's business nowadays is for online images as well so large files are not necessary. He also said that the QC people can recognise a scan from a digital camera shot and would take that into consideration. But as I said, why risk it when a 3000x2000 is fine and looks almost as good as a digital camera image (by my high standards that is 😀). I have not failed QC since 2012 just after I got my first 36MP camera so why start now?

 

I recall wim pointing out that large files can be required for projecting as well as large prints. I occasional see FS as a criterion bringing up some of my images in searches - not sure if I have ever sold one for that reason though.

 

 

 

 

 

 

[M.Chapman]

21 minutes ago, MDM said:

It's under the Filter menu at the bottom under Topaz Labs. I have other Topaz plugins as well under Topaz Studio.

 

I assume you mean this menu? Topaz seems to be missing from mine 😞 

 

 

21 minutes ago, MDM said:

Rather than doing a global blend I think it is better to do localised blending. You are also downsizing at a very early stage presumably to have a smaller file for speed in TDN so ending up with a small file. I would rather produce a full size master and then downsize as required. 

 

Yes, localised blending is also possible and relatively easy but I haven't found it necessary. I like the global blend as it adds a little noise into the sky  and takes a little sharpness of the edges and it's quicker. The final steps (TDN + blend) are quite quick so I'm happy for them to be seen as a "final touch" and leave my "master" (16 bit TIFF or the mulit-layer PSD) alone. Although TDN recommend doing the denoise early, that would be fine if they didn't overcook the sharpening at the same time. That level of sharpening is IMHO best kept to the latter stages of workflow. But there's no ideal here as noise reduction (and light "capture" sharpening) is best early in the workflow. 

 

21 minutes ago, MDM said:

As for using sRGB, well that is a whole new story. I can understand that if you are only ever intending the images to be uploaded to Alamy as it is at the moment but you are throwing away masses of information. Why work in 16 bit and throw away so much of the colour information. It doesn't have any effect on the workflow to use a bigger colour space and, as with resizing, export or save in a smaller space as a final step to suit the purpose.

 

I standardised on sRGB a while back because of the enhanced consistency it gives throughout my workflow. Histograms accuracy is preserved and WYSWIG is better between PS and the sRGB jpegs I produce. I also find (but I am red/green deficient ) the visual difference between correctly rendered sRGB and AdobeRGB images is minimal. I suspect a lot of the prejudice against sRGB comes from folks comparing incorrectly rendered images.

 

Mark

Edited February 13, 2020 by M.Chapman

[Harry Harrison]

9 minutes ago, MDM said:

When we met in Ely, James Allsworth did say that most of Alamy's business nowadays is for online images as well so large files are not necessary

Perhaps I won't put up those 10,000 px panoramics after all!

[MDM]

Yes it should be at the bottom of that menu. You might need to reinstall or perhaps it is due to you using the trial. 

 

Let's not go into the colour space thing again except to say that it is not what you are seeing that is important, it is about preservation of data in the image. Use Lightroom's softproofing to see some of what is happening. 

 

I recommend you to get Martin Evening's Lightroom book as, even if you are using ACR,  there is a huge amount of excellent info on raw processing - practical and theoretical about noise reduction, sharpening and colour spaces to name a few. 

[M.Chapman]

16 minutes ago, MDM said:

Use Lightroom's softproofing to see some of what is happening. 

 

Or take a look at my article and especially the images it links too (fx runs off and hides). 

 

Mark

Edited February 13, 2020 by M.Chapman

[MDM]

48 minutes ago, M.Chapman said:

 

Or take a look at my article and especially the images it links too (fx runs off and hides). 

 

Mark

 

Had a quick look but too busy right now to do read it properly and do it justice. I will in any case stick with the conventional wisdom - just because you can't see the information doesn't mean it isn't there and one should not throw it away just because it is not detectable to the subjective eye.   

[M.Chapman]

29 minutes ago, MDM said:

 

Had a quick look but too busy right now to do read it properly and do it justice. I will in any case stick with the conventional wisdom - just because you can't see the information doesn't mean it isn't there and one should not throw it away just because it is not detectable to the subjective eye.   

 

100% agree and that's why I keep my RAW files. But when when producing images for agencies which only accept sRGB (or convert to sRGB on receipt like Alamy) I will work in sRGB colour space so that I know what I'm editing/seeing on screen and histograms match how the final jpg will appear, IMHO it's easier and potentially more accurate than "soft proofing", especially when looking at clipping indicators (maybe Adobe have fixed that now?). But that's just my personal preference. Maybe PA media will move Alamy "upmarket" and allow sRGB or even ProPhotoRGB to be uploaded, stored and sold to the customer. I'd like that

 

Mark

Edited February 13, 2020 by M.Chapman

[MDM]

2 hours ago, M.Chapman said:

 

100% agree and that's why I keep my RAW files. But when when producing images for agencies which only accept sRGB (or convert to sRGB on receipt like Alamy) I will work in sRGB colour space so that I know what I'm editing/seeing on screen and histograms match how the final jpg will appear, IMHO it's easier and potentially more accurate than "soft proofing", especially when looking at clipping indicators (maybe Adobe have fixed that now?). But that's just my personal preference. Maybe PA media will move Alamy "upmarket" and allow sRGB or even ProPhotoRGB to be uploaded, stored and sold to the customer. I'd like that

 

Mark

 

Yes but if you are doing any work on the raw images then you have to do it all over again. Even spotting an image twice could be a big pain and I often do extra work in Photoshop which I definitely do not ever want to have to redo. For me the jpeg is the final stage and I don't usually keep them. I keep the raws and the psds.

 

Having said that I understand what you mean. I work in AdobeRGB in Photoshop as I have a wide gamut monitor. I used to work in ProPhoto before I got the new monitor in 2014 but changed to AdobeRGB then for similar reasons to what you say. However, AdobeRGB is huge in comparison to sRGB. Also ProPhoto files can look pretty awful outside of a colour managed application if I forget to convert.

 

Nowadays a lot of devices such as iPhones and iPads are able to display colours well outside of sRGB space as well and most people seem to be using phones and tablets rather than computers now. I don't convert to sRGB at all anymore for pics I am putting on my website. Having checked on various devices I think AdobeRGB is better than sRGB on phones. You can't please everyone and you have no control over what device somebody uses to view your images. My wife's work Win 10  laptop dispays everything in a horrendous shade of blue as the default white balance is up in the 9000s so images look quite horrible. 

 

 

 

Edited February 13, 2020 by MDM

[Inchiquin]

7 hours ago, Harry Harrison said:

 

I used to be associated with a small picture library through the period of the transition to digital, not one that you would have heard of.

 

 

Try us...

 

Alan

[geogphotos]

10 minutes ago, Inchiquin said:

 

Try us...

 

Alan

 

 

Nope, never heard of them.

[M.Chapman]

20 hours ago, M.Chapman said:

I assume you mean this menu? Topaz seems to be missing from mine 😞 

 

 

20 hours ago, MDM said:

Yes it should be at the bottom of that menu. You might need to reinstall or perhaps it is due to you using the trial. 

 

Mmm... I've closed PS, uninstalled TDN and reinstalled it, still no joy (even though during installation it asks me a question "Install Photoshop Support") and, at that point, I am installing the full version (not a trial)  and I've also searched through the Library/Application Support/Adobe folders and can't find any sign of Topaz putting anything in there. I've read the online support which seems to refer to earlier Topaz Lab only. So I'm baffled.

• If you run your Denoise filter from within PS, on a multilayer image, is TDN only applied to the selected layer? If so that's exactly what I want to be able to do so it's worth me pursuing.
• Is the Denoise filter that runs the new "AI" version and not the earlier Denoise 6 that came with the Topaz stuff  you had installed previously?
• If it is, are you able to check where the plugin is stored and what it's called? Possible location is here.

Photoshop CC 2019 - Macintosh HD/Library/Application Support/Adobe/Plug-ins/CC/

 

Sorry to trouble you, if you haven't got time I'll try the Topaz forum.

 

Mark

Edited February 14, 2020 by M.Chapman

[M.Chapman]

18 hours ago, MDM said:

AdobeRGB is huge in comparison to sRGB.

 

That surprises me. I thought the difference was relatively small and largely affected the greens, see here. Obviously these figures are not a realistic indication of the colours that are missing from sRGB colour space because all the background colours are probably generated in sRGB space anyway. So these figures significantly exaggerate the visual appearance of the colours are missing. A comparison of Pointer's gamut (natural surface colours) versus sRGB and AdobeRGB is interesting too (See paper here and figures Pointer's Gamut vs sRGB figure and Pointer's Gamut versus AdobeRGB). Although, given the trend towards more (over?) saturated/colourful images, what is "natural" is perhaps less relevant 

 

Mark

Edited February 14, 2020 by M.Chapman

[MDM]

2 hours ago, M.Chapman said:

 

 

Mmm... I've closed PS, uninstalled TDN and reinstalled it, still no joy (even though during installation it asks me a question "Install Photoshop Support") and, at that point, I am installing the full version (not a trial)  and I've also searched through the Library/Application Support/Adobe folders and can't find any sign of Topaz putting anything in there. I've read the online support which seems to refer to earlier Topaz Lab only. So I'm baffled.

• If you run your Denoise filter from within PS, on a multilayer image, is TDN only applied to the selected layer? If so that's exactly what I want to be able to do so it's worth me pursuing.
• Is the Denoise filter that runs the new "AI" version and not the earlier Denoise 6 that came with the Topaz stuff  you had installed previously?
• If it is, are you able to check where the plugin is stored and what it's called? Possible location is here.

Photoshop CC 2019 - Macintosh HD/Library/Application Support/Adobe/Plug-ins/CC/

 

Sorry to trouble you, if you haven't got time I'll try the Topaz forum.

 

Mark

 

Yes it only works on the active layer. I can't envisage how it would work any other way.

 

Yes I have the AI version.

 

There is an alias to a plugin in that location in a TopazDeNoiseAI folder. The actual plugin is held within the app package. Right click on the Topaz Denoise AI app, show package contents - Contents - Resources - PS_Plugins - TopazDeNoiseAI.plugin. Any further info you should really ask on the Topaz forum as it is well above my level of knowledge.

 

As I said above I have no intention of getting into a discussion about color spaces. I have neither the time nor inclination to do so. If you are happy with sRGB then that is fine. I recommend you read Martin Evening and other authoritative,  peer-reviewed authors like Jeff Schewe. 

[M.Chapman]

20 hours ago, MDM said:

 

Yes it only works on the active layer. I can't envisage how it would work any other way.

 

Yes I have the AI version.

 

There is an alias to a plugin in that location in a TopazDeNoiseAI folder. The actual plugin is held within the app package. Right click on the Topaz Denoise AI app, show package contents - Contents - Resources - PS_Plugins - TopazDeNoiseAI.plugin. Any further info you should really ask on the Topaz forum as it is well above my level of knowledge.

 

As I said above I have no intention of getting into a discussion about color spaces. I have neither the time nor inclination to do so. If you are happy with sRGB then that is fine. I recommend you read Martin Evening and other authoritative,  peer-reviewed authors like Jeff Schewe. 

Thanks!

That gave me enough clues to work out what was missing. I had none of the expected Topaz directories in either Library/Application support folder. So I reran the TDN installer script using terminal with admin rights / root privileges and now everything has appeared and the filter is available in PS. No idea why that made a difference. I had no error messages previously except the Mac OS asked me if I was happy to run a program downloaded from the internet. Maybe it blocked some parts of the install even though I was logged in as Admin? Mmm... 

 

Martin Evening's PS book is in my shopping basket.

 

Thanks again,

 

Mark

 

Edited February 15, 2020 by M.Chapman

[M.Chapman]

Now that I've got Topaz Denoise AI working as a filter in PS, the workflow becomes slightly simpler.

 

• Open RAW file in PS ACR RAW

• Apply preset (applies WB of my Lightbox and reduces contrast to suit Velvia)

• Check and adjust WB

• Make alignment and crop adjustments

• Make all tonal and saturation adjustments

• Export to PS as 16 bit sRGB

• Resize to 3000 x 2000 (use bilinear - to avoid unnecessary sharpening)

• In PS remove all dust spots and do cloning etc.

• Duplicate layer and call TDN (Topaz De-Noise)

• Apply Filter>Topaz Labs > Denoise AI to TDN layer with default settings (15, 15, 0, 0.00) and click Save.

• Wait for filter to complete and return automatically to PS

• Adjust opacity of TDN layer to 50-80%  to allow 20-50% of original image to show through to retain some grain and soften edges a little

• Apply 0.2 radius Gaussian blur overall (if some edges are still too crisp)

• Save as jpg at quality level 12

 

Mark

[KevinS]

I've done some comparisons of Nikon & Canon lenses w/Nikon ES-1 Slide Copying Adapter. Since the ES-1 has 52mm threads, I was able to try it on a 55mm/2.8 Micro-Nikkor and a 50mm/2.5 Compact-Macro Canon EF. Disclaimer: Email and web browsing are still challenging to me at times. There are many on the forum who show a deeper understanding of tech issues, and I'm afraid I don't speak the language. Inspecting files at 100% is something I can do and the differences are visible. I used f/8 and f/11, aimed the rig at sunlight snow-covered front yard, live view to eliminate mirror slap. PP limited to a few DxO automatic settings and my manual adjusting of Unsharp Mask via sliders. Using a Canon 6D (20MP) and the Nikon lens + 27.5 mm of extension only fills about 80% of the frame. This is due to a difference in flange distance (index?), I think. The Canon lens + Life Size Converter shows the whole slide plus a narrow border of slide mount. The converter fills the frame nicely (makes the 50mm a 70mm), but no improvement in image quality over 25mm extension tube that I could see. I may try a cheap Nikon bellows to get the frame filled by the Nikon lens and see if the quality is maintained (I knew I shouldn't have sold those Nikon bellows!).

Surprisingly, the files from the ancient Nikon are preferable to the Canon.

 Observations: Focus is tricky, but essential. There is a big difference in how the out of focus areas render. Nikon is smoother, Canon shows sharpening artifacts. Increasing the sharpening on the Nikon works ok, but reducing the artifacts on the Canon file not easy to do. Lighting was good, I believe, but I did read an article about diffused vs collimated light for copying negatives and recalled that I much preferred the diffuse light when working in the darkroom.

 Conclusions: If carefully done, I believe the Nikon files would pass Alamy QC. Maybe Canon files would, but I'll move forward with the Nikon lens for that. All files would meet my needs for family photos and stock offered on my own website. The set-up used by Phil on page one is of great interest, but I doubt that the 100mm macro plus Novoflex gear will make my shopping list. Ian M might be exploring that and it might be really good as that lens cost many times what these did. The ES-1 is a huge time-saver as there is no way to become misaligned. Higher MP count might help, as MDM has alluded. Reversing the Nikon lens would be interesting, but would require yet another doo-dad to mount the ES-1. 

 Sorry for the rambling narrative, but know that there are many variables that I didn't anticipate, making a simple comparison impossible. If the image sizes were all consistent, that would have been simple. Canon and Nikon systems are different, and that is magnified when doing macro work. I'll have a look at DropBox this weekend and see if I can get some examples up.  

 

[Harry Harrison]

38 minutes ago, KevinS said:

Using a Canon 6D (20MP) and the Nikon lens + 27.5 mm of extension only fills about 80% of the frame

Thanks for doing that and posting it here. I don't quite understand what you mean when you say that the Canon lens is showing sharpening artefacts, they're processing issues rather than lens properties aren't they?  Presumably you used a normal off the shelf Nikon to Canon adapter? I'd have to check but I think I get 1:1 with my Micro-Nikkor on the 5D MkII. Is the Micro-Nikkor fully extended or is it that the ES-1 doesn't allow it to be and still be in focus?

Edited February 14, 2020 by Harry Harrison

[Harry Harrison]

1 hour ago, KevinS said:

the Nikon lens + 27.5 mm of extension only fills about 80% of the frame

Yes, I get a thin black border around a Kodachrome slide with the Micro-Nikkor 55mm plus PK-13, so since the slide mount crops the frame slightly I must be getting near enough 1:1.

[KevinS]

34 minutes ago, Harry Harrison said:

Thanks for doing that and posting it here. I don't quite understand what you mean when you say that the Canon lens is showing sharpening artefacts, they're processing issues rather than lens properties aren't they?  Presumably you used a normal off the shelf Nikon to Canon adapter? I'd have to check but I think I get 1:1 with my Micro-Nikkor on the 5D MkII. Is the Micro-Nikkor fully extended or is it that the ES-1 doesn't allow it to be and still be in focus?

Could be processing; but DxO does display those adjustments. It also applies lens corrections which might be a problem that I can't see by looking at the settings in DxO. What looks like artifacts show in the dark green areas of a flower photograph and are more pronounced with the Canon lens. I was very surprised to see this. Best to wait for examples (busy time for me now). 

Just checked and found I was using ES-1 incorrectly. Correct way is to focus the lens so it says 1:1, then visually focus the image by sliding the ES-1 in and out as needed. Doing it this way fills the frame. I was racking the ES-1 all the way out, then having to adjust the focus ring on the lens in a way that resulted in a smaller image. Thanks for your question; it helped me along in this endeavor. Stay tuned!

[Harry Harrison]

4 minutes ago, KevinS said:

Doing it this way fills the frame.

Glad it works, I couldn't work out what was going on!

[MDM]

I get full frame or thereabouts with the PK-13 and Nikkor 55 but pulling back a little is no problem. The file size might be a bit smaller but still more than ample. In fact it might even be a wise move if the lens being used for the copying is not from another world (such as the Nikkor 55) with excellent edge to edge and corner to corner sharpness. The lens is old in terms of its design but is still being manufactured by Nikon amazingly. I am still waiting for my stuff from China to try other lenses.