Canon slide copying set-up

[Harry Harrison]

2 minutes ago, MDM said:

Spending ages on noise or grain reduction does not make sense unless one really enjoys that sort of thing.

I know that you're not quoting anyone in particular but for me I was interested to see if there really was a phenomenon that explained why I was seeing more grain in prints from scans than in darkroom prints. The jury is probably still out on that one, at least until I've read the articles properly. If there was a slight change in technique that could avoid it happening, or as you say a simple Photoshop or even Lightroom fix then that would be good too. 

 

It would be very interesting to know how, say, Magnum go about scanning the negatives from their illustrious stable of photographers. Do they rescan as technology improves with digital backs etc.? I remember they once had a publicised tie-up with Imacon but who knows?

[MDM]

I was really questioning Mark’s idea of doing multiple shots and putting them in layers which seems like a lot of trouble when it can be done much more easily in the way I mentioned. The theory is interesting although I am more concerned with the practice myself. 

 

Edited February 19, 2020 by MDM

[M.Chapman]

9 hours ago, MDM said:

Also if you value your time then choose an instrument that is designed for the job as getting the film perfectly aligned is not an easy task and perfection is essential given the tiny depth of field.  

 

Indeed, I reckon the depth of field is less than 0.5mm in my setup (Lumix G7 MFT + 45mm lens @ f/8). With full frame higher resolution sensors it will be even less.

 

9 hours ago, MDM said:

Spending ages on noise or grain reduction does not make sense unless one really enjoys that sort of thing.

 

I'm just curious at the moment, but if it works and can be automated using a PS script or action then it might be worth doing. Personally I don't think it will work well enough, and the gaussian blur and downsize technique is quick. Although this probably works best when starting with a higher resolution sensor input (e.g. Nikon 850) than it does with my Lumix G7

 

Mark

[Harry Harrison]

8 hours ago, MDM said:

using a lens with a 67mm thread stepped down gives a tunnel effect so the idea of using step down rings may not be viable at all.

Do you mean internal reflections? It looks like the front element of the 100mm f2.8L is less than 52mm, or at least much less than 67mm so I thought it might be OK.

 

There's a really nice 55mm. Micro-Nikkor plus PK-13 on ebay for £189, ES-1 for £50 or less from Japan and a Novoflex EOS to Nikon adapter for £20. I rest my case.

Edited February 7, 2020 by Harry Harrison

[Harry Harrison]

43 minutes ago, M.Chapman said:

With full frame higher resolution sensors it will be even less

You must be right I think, it's certainly true that Micro 4/3 cameras have more depth of field in normal circumstances. However getting all the detail in a slide or negative down on to a sensor area around a quarter the size is quite a feat and I began to wonder about circles of confusion in Macro, but the real circles of confusion are in my head.

[geogphotos]

9 hours ago, MDM said:

I was really questioning Mark’s idea of doing multiple shots and putting them in layers which seems like a lot of trouble when it can be done much more easily in the way I mentioned. The theory is interesting although I am more concerned with the practice myself. 

 

Anyway more importantly is my discovery in the last few minutes that using a lens with a 67mm thread stepped down gives a tunnel effect so the idea of using step down rings may not be viable at all. Stepping up is ok but not stepping down in other words. 

 

 

Bad news for the Canon 100m L + EF-1 theory?

[Harry Harrison]

Quite a nice DSLR setup for getting through a lot of material:

 

https://medium.com/@dcbryan/digitizing-film-a-modern-approach-827a8cf019bf

 

 

...well mirrorless Sony actually

Edited February 7, 2020 by Harry Harrison

[M.Chapman]

In line with the technique in this article. I just tried an experiment. I took 5 RAW shots in quick succession of the same slide. I had AF on so each frame was refocussed, but IS was OFF. The idea being to get tiny amounts of movement  between each frame. I processed one in LR and synched the settings to all 5 and exported as PSDs. I then loaded the 5 PSD files into layers in PS CC.

 

Toggling layers on on off reveals a barely perceptible movement between frames but the "grain mottling" does indeed jump about quite noticeably suggesting aliasing is a factor and that averaging the frames may indeed give a benefit. So I then did an auto-align on the stack of layers and set opacities as (bottom layer to top) of 100%, 50%, 33%, 25%, 20% to combine (average) the frames.

 

It works. Grain appears reduced. With 5 frames the reduction isn't as quite much as doing a 0.5 pixel gaussian blur, but more detail is retained. I also tried doing a 200% upsize before aligning with bilinear resample, followed by a 50% downsize with bilinear resample. This gives an even smoother result, but there's a some loss of detail.

 

I think in general MDM's technique is best (Gaussian smooth applied to selected areas only) if you only need to reduce appearance of grain in areas where some loss of detail doesn't matter (e.g. the sky).

 

But if I needed to reduce the appearance of grain overall, I think the multi-frame averaging might be quite useful. I used a 5 frame average, but I think 10 would be better. I'll try producing some before and after examples to post on line tomorrow.

 

Mark

Edited February 7, 2020 by M.Chapman

[MDM]

On 07/02/2020 at 07:05, Harry Harrison said:

Do you mean internal reflections? It looks like the front element of the 100mm f2.8L is less than 52mm, or at least much less than 67mm so I thought it might be OK.

 

There's a really nice 55mm. Micro-Nikkor plus PK-13 on ebay for £189, ES-1 for £50 or less from Japan and a Novoflex EOS to Nikon adapter for £20. I rest my case.

 

 

 

With the 55mm Nikkor, a proper extension ring behind the lens would also be needed to get 1:1.

 

On 07/02/2020 at 08:00, geogphotos said:

 

 

Bad news for the Canon 100m L + EF-1 theory?

 

I think so. The Tamron 90 which I mentioned way back might work. I have one and will report when I get my Chinese delivery.

 

On 07/02/2020 at 06:41, M.Chapman said:

 

I'm just curious at the moment, but if it works and can be automated using a PS script or action then it might be worth doing. Personally I don't think it will work well enough, and the gaussian blur and downsize technique is quick. Although this probably works best when starting with a higher resolution sensor input (e.g. Nikon 850) than it does with my Lumix G7

 

Mark

 

Working in Photoshop you would be losing the benefit of working on raw. I think it would be an awful lot of work for little or no gain or possibly loss, moreover given the likely quality of the originals. I will try to post some new stuff later on Dropbox.

Edited February 19, 2020 by MDM

[M.Chapman]

11 minutes ago, MDM said:

Working in Photoshop you would be losing the benefit of working on raw.

 

Agreed, that's why I did most of the processing in LR starting from RAW.

 

Mark

[Harry Harrison]

34 minutes ago, MDM said:

The 100mm f2.8L has a 67mm filter thread so I presume the front element is the same or thereabouts

No, it's much smaller, as I've already pointed out.:

 

https://www.kenrockwell.com/canon/lenses/100mm-f28-is.htm

 

The difference over the non 'L' version is probably down to the IS.

Edited February 7, 2020 by Harry Harrison

[Harry Harrison]

35 minutes ago, MDM said:

With the 55mm Nikkor, a proper extension ring behind the lens would also be needed to get 1:1

My 'shopping list' included the PK-13.

[MDM]

21 minutes ago, Harry Harrison said:

No, it's much smaller, as I've already pointed out.:

 

https://www.kenrockwell.com/canon/lenses/100mm-f28-is.htm

 

The difference over the non 'L' version is probably down to the IS.

 

 

I'm confused. He actually says the IS version has a 67mm thread which agrees with the specs for the L version on Wex, for example. The non-IS version is a 58mm.

 

19 minutes ago, Harry Harrison said:

My 'shopping list' included the PK-13.

 

Sorry was in a rush and didn't spot that. That sounds like an exceptionally good deal, especially if it is really in mint condition as claimed.  

 

 

[Harry Harrison]

3 minutes ago, MDM said:

I'm confused. He actually says the IS version has a 67mm thread

Appreciate you're in a hurry, it has a 67mm thread but if you get a chance to look at the pictures, Rockwell has a good one, then the front element is much smaller than the filter ring. So I'm not sure that you'd get a tunnel effect, though internal reflections might be a possibility.

[spacecadet]

16 hours ago, Klinger said:

 

 

 

 

downsized for the forum, this is not the one I edited and not the one I want to upload to Alamy. Just an example.

 

 

Whilst you might get it through the archival route this is nowhere near the quality achievable with a proper macro or enlarging lens. The aberrations are terrible and the dynamic range very poor- the skies are blown out and the fringing on the gazebo is very bad. The duplicator optics just aren't up to the job.

I would not submit an image of this quality.

[MDM]

25 minutes ago, Harry Harrison said:

Appreciate you're in a hurry, it has a 67mm thread but if you get a chance to look at the pictures, Rockwell has a good one, then the front element is much smaller than the filter ring. So I'm not sure that you'd get a tunnel effect, though internal reflections might be a possibility.

 

Not so much of a hurry now and yes you are correct. Strange design actually. Even if it did work, one is still left with the problem of adding extensions in front of the lens though and  the overall length of these is unknown for any particular lens until it is actually tried out. They don't stock these things in the local camera shop if such exists so the solution is to order plenty from China or be prepared to wait weeks for a second order.  

 

Really it would be much more sensible to do what you said in your other post (the Nikkor 55 et al). Ian could probably buy all that and the Tamron 90 new as well for the same price as the Canon 100L. Metering is not important as it is easily done by trial and error. I don't meter at all now - somewhere between 1s and 1/4 s at f11 and ISO64 with the light between 20-40cm away. I shoot 3 or 4 for each slide and pick the best.

 

 

[Harry Harrison]

40 minutes ago, MDM said:

Really it would be much more sensible to do what you said in your other post (the Nikkor 55 et al)

The manual stop-down could be an obstacle, though more in the mind I think if you're used to using Auto, same with manual exposure as you describe. I think Ian probably would like to use the 100mm for other photography but unfortunately there's no way of really seeing how well it will work with the ES-1 without actually getting the stuff and trying it.  The guy in the forum had it working with the non-L version so I think it probably will. I suspect that they might be very similar optically.

 

We do know that it works with Phil's Castel Novoflex setup though but that will need some improvisation and cameracraft to make it work. The ES-1 (or ES-2) is the only 'off the shelf' product that just does the job, combined with the Micro-Nikkors of course.

Edited February 7, 2020 by Harry Harrison

[MDM]

10 minutes ago, Harry Harrison said:

The manual stop-down could be an obstacle, though more in the mind I think if you're used to using Auto, same with manual exposure as you describe. I think Ian probably would like to use the 100mm for other photography but unfortunately there's no way of really seeing how well it will work with the ES-1 without actually getting the stuff and trying it.  The guy in the forum had it working with the non-L version so I think it probably will. I suspect that they might be very similar opticaly.

 

We do know that it works with Phil's Castel Novoflex setup though but that will need some improvisation and cameracraft to make it work. The ES-1 (or ES-2) is the only 'off the shelf' product that just does the job, combined with the Micro-Nikkors of course.

 

I have never used manual stop-down but I presume it only applies to the metering. You just set the aperture and use the DOF preview (assuming the camera has one) to meter. Manual exposure is the only sensible way for copying anyway and not a big leap on the learning curve I expect. I do wonder if the Nikon AIS lens would work on a Canon camera though with the adapter (i.e. will the lens stop down when the shutter is released). I will take your word for it that it does. 

 

I am now lost as to which forum you are referring to where the non-L version works with the ES-1 if you could remind me please. Going round in circles here. 

 

The Phil method may be the best option if Ian is fixed on the Canon 100L lens. 

[Harry Harrison]

6 minutes ago, MDM said:

 

I am now lost as to which forum you are referring to where the non-L version works with the ES-1 if you could remind me please

Not to worry, you are quite reasonably coming from the standpoint of using the ES-1 on a Micro-Nikkor. On a Canon then you need the Eos-Nik adapter but then you have to manually stop down the aperture, the camera won't do it. In fact using Live View you can keep it stopped down and check the focus that way, should be easy.

 

This was the link using the non-L version, way back on page 4, more or less the method that we've both been suggesting.:

 

In this thread someone has used the 100mm f2.8 Macro with the ES-1 and describes how he did it:

 

https://www.fredmiranda.com/forum/topic/1185821

 

"I've used the Nikon ES-1 slide adapter on the Canon 100/2.8 (along with other macro lenses); it works fine. I found some cheap 52mm-threaded extension tubes on eBay (marked "for Olympus SP550 UZ 52mm") --- just a hollow metal tube with male/female 52mm threads on each end --- to add the extra extension needed in front of the lens. Then, a 58mm->52mm filter step-down-ring on the front of the lens, and you're set. Illumination by pointing towards a wall lit by a camera flash (manually set to consistently give the right exposure) worked well."

 

 

Ian, if you're reading this, it looks like you coud be a pioneer with the 100mm f2.8 L and the ES-1, but if MDM might give you some dimensions for the range of available extension of the ES-1, and Phil might give you the distance of the focused slide from the end of the lens then you should be able to work out which extension tube(s) you might need to give it a try. I think that without fresh input from someone who's actually done it then this thread has run its course.

 

[MDM]

18 minutes ago, Harry Harrison said:

Not to worry, you are quite reasonably coming from the standpoint of using the ES-1 on a Micro-Nikkor. On a Canon then you need the Eos-Nik adapter but then you have to manually stop down the aperture, the camera won't do it. In fact using Live View you can keep it stopped down and check the focus that way, should be easy.

 

This was the link using the non-L version, way back on page 4, more or less the method that we've both been suggesting.:

 

In this thread someone has used the 100mm f2.8 Macro with the ES-1 and describes how he did it:

 

https://www.fredmiranda.com/forum/topic/1185821

 

"I've used the Nikon ES-1 slide adapter on the Canon 100/2.8 (along with other macro lenses); it works fine. I found some cheap 52mm-threaded extension tubes on eBay (marked "for Olympus SP550 UZ 52mm") --- just a hollow metal tube with male/female 52mm threads on each end --- to add the extra extension needed in front of the lens. Then, a 58mm->52mm filter step-down-ring on the front of the lens, and you're set. Illumination by pointing towards a wall lit by a camera flash (manually set to consistently give the right exposure) worked well."

 

 

Ian, if you're reading this, it looks like you coud be a pioneer with the 100mm f2.8 L and the ES-1, but if MDM might give you some dimensions for the range of available extension of the ES-1, and Phil might give you the distance of the focused slide from the end of the lens then you should be able to work out which extension tube(s) you might need to give it a try. I think that without fresh input from someone who's actually done it then this thread has run its course.

 

 

 

You can get the extensions in 7, 14, 21 and 28 mm. Exactly what is required on a particular lens is another matter. Better to order more than less which is what I have done as they are cheap. Available on Amazon and ebay. I used gift cards from Christmas for mine. I had been going to buy a Kindle book or two but this is interesting. 

 

As for the thread running its course - surely not 😀. Here is something new. Yesterday I tried Topaz Denoise AI which has been sitting inactive on my computer (Black Friday madness a few years ago) on a few images and it certainly gets rid of the film grain, much more effectively than LR noise control. However it can introduce some artifacts and may give a plastic look on skin. I have only done a few so too early to judge properly as there are several controls and it takes ages but it could work well depending on the image. That said, the film grain can look good, particularly on black and white negs - proper grain  instead of artificial noise that is.

Edited February 7, 2020 by MDM

[Harry Harrison]

9 minutes ago, MDM said:

As for the thread running its course - surely not 😀

I was thinking in terms of hardware options for Ian, but no, onwards and upwards...

[Klinger]

19 hours ago, M.Chapman said:

 

A couple of comments on the setup. I think it's worth creating an opaque mask (e.g. sheet of card with 50mm square hole in it) and sticking to the diffuser (e.g. blu-tack). This will reduce the stray light entering the lens (an advantage of the Nikon ES-1...) and also helps locate the slide.

 

Also a bubble level isn't as accurate as using a mirror to set the camera square to the plexiglas. (Again something that the ES-1 does automatically... oh dear I'll have to get one....)

 

A comment on the digitised image - lots of blue fringing around the trees and some flare / CA around the marquee, presumably in the original slide? Likely to be rejected by Alamy QC so archival is probably the only option if the images are historically or culturally significant. You also mentioned that you'd downsized to 8MP, you can go down to 6MP on Alamy if you want.

 

Mark

 

Mirror, thanks. Yes the image is a quick one for comparison to the macro lens and Canon slide holder - the tube with slide holder - the standard lens with extension tubes. I need to do one slide, all three and variations to see which is best for what I have. That or buy a slide scanner (again?).

 

Did I mention 1963 slide, probably Ektachrome for speed, Kodak Retina II camera? Not the best but it is vintage? 😉

 

I did think of the mask, I need to make it. I have flat black foam for crafting, it's 1/8" or less.

 

5 hours ago, spacecadet said:

Whilst you might get it through the archival route this is nowhere near the quality achievable with a proper macro or enlarging lens. The aberrations are terrible and the dynamic range very poor- the skies are blown out and the fringing on the gazebo is very bad. The duplicator optics just aren't up to the job.

I would not submit an image of this quality.

 

Quote from page 9 to page 8, see if this worked?

 

Thank You, always looking to improve.

 

This was not the duplicator it was extension tubes on a standard lens. The photo example in the post was the slide with the camera. An EOS-M by the way.

 

But no disagreement on the quality. Might be the original, the age or just bad exposure because I wasn't manual or careful. It is an unedited snapshot!

 

Time to test all options, lens, camera, or copier and then find the Nikon Coolscan and see how that looks, 2700 ppi? I can also mount the 1Ds II (16MP) on the copy stand, but I'm not sure that will make anything better.  EOS-M 18MP but I used the M lens with extension tube adapters, not the EF 100MM L, and 60-D 18MP I have the manual extension tubes. Want one more for the collection? I have the EF to M adapter. Lots of possibilities to explore.

 

 

 

Edited February 7, 2020 by Klinger

[Klinger]

So posting from page 8 didn't show me that the reply was posted and stayed in compose? Sorry

 

Edited February 7, 2020 by Klinger

[Klinger]

  

17 hours ago, MDM said:

Anyway more importantly is my discovery in the last few minutes that using a lens with a 67mm thread stepped down gives a tunnel effect so the idea of using step down rings may not be viable at all. Stepping up is ok but not stepping down in other words. 

 

Can you explain tunnel effect please. Do you mean when adding to the front of the lens, instead of tubes to the back of the lens? I have the 58mm version the old 100mm macro Canon.

 

 

Edited February 7, 2020 by Klinger

[geogphotos]

3 hours ago, Harry Harrison said:

I was thinking in terms of hardware options for Ian, but no, onwards and upwards...Ian, if you're reading this, it looks like you coud be a pioneer with the 100mm f2.8 L and the ES-1

 

 

 

I will probably buy the Canon 100mm L lens next week and make a start with tripod/lightbox and see how I get on with that.

 

I am the very last person you want to be any sort of pioneer for this sort of thing! But will happily post back on the results of the above. I have the slide holder from my Minolta Dimage scanner and, if necessary, will figure out some system using card to place the slide in the same position beneath the camera/lens. And if that doesn't work out it looks like a lens that it will be very interesting to have anyway. 

Edited February 7, 2020 by geogphotos