Canon slide copying set-up

[M.Chapman]

2 hours ago, MDM said:

As I said, if I can get nice clean A4 prints and pass QC

 

I won't be producing A4 prints, but I do like "clean" images with a level of grain that is well below "obtrusive" and I also want my images pass Alamy QC. That's the bit that concerns me most. Looking at the files you uploaded here (A few downloadable examples), unless Alamy QC are significantly more lenient when it comes to what they perceive are slide copies, I would be quite worried about submitting PriestLeap032, and to a lesser degree Blackhead027 to QC. (Maybe I'm paranoid?)  If, as others have mentioned, Alamy apply lower QC standards to slide copies (which aren't submitted via the archival route) I don't understand their logic. If a customer selects an image submitted via the archive route, they are given a warning "This image could have imperfections as it’s either historical or reportage". No such warning will appear on slide copies submitted via the normal route. So surely the QC standard should be nominally the same as for straight digital camera shots? Otherwise don't they risk disappointing the customer?

 

Here are a couple of further examples I've produced (using the Topaz workflow I described earlier). These are, I hope, good enough for standard Alamy QC with no leniency since (to my eyes) they aren't easily recognisable as slide copies.

 

https://drive.google.com/open?id=1iFateIxBP1cRpLEphNsE_QKGorlMTfNj

https://drive.google.com/open?id=16-G_CdoLgYryXOlX0iDpMUKn9j9wDZaE

 

For proper inspection it's probably best to download as I'm not sure if Google Drive renders them 100% accurately.

 

Mark

Edited February 26, 2020 by M.Chapman

[Harry Harrison]

9 hours ago, M.Chapman said:

If a customer selects an image submitted via the archive route, they are given a warning "This image could have imperfections as it’s either historical or reportage". No such warning will appear on slide copies submitted via the normal route. So surely the QC standard should be nominally the same as for straight digital camera shots? Otherwise don't they risk disappointing the customer?

If one has gone to the trouble of meticulously 'scanning' and retouching a quality film original, in 35mm or medium format, then I would think it was actually misleading from the contributors' point of view to have to live with the byline "This image could have imperfections as it’s either historical or reportage". Putting QC to one side then I think it's easy to exaggerate the importance of seeing  film grain when zoomed at 100% providing it is representative of fine grain low ISO film stock, I can't imagine that it would be of any consequence in the majority of applications for which these images are sold. If it was a problem then presumably the buyer could get a refund as it wasn't fit for purpose but I don't recall anyone reporting that in my very limited experience here.

 

With regard to QC, Alamy clearly haven't had a consistent policy with respect to scans over time. I've seen a lot of images reported in sales that are scans of film originals uploaded through the normal route. Many of these were probably uploaded earlier in the history of digital but I've noticed that more often than not they are high resolution, often over 5000 px on the longest side and I've been told that there was even a stipulation for that in the past. They still sell.

 

I seem to recall on another library that images that started life on film were labelled as such, without any suggestion that they might be of inferior quality, but that was some time ago and I don't remember where I saw it or if it still happens.

 

Edit:

I've also noticed that when major collections are added to Alamy a large part may be from the film era, and so scanned, but there is no indication that they have been and they can be high resolution.

 

 

 

 

Edited February 27, 2020 by Harry Harrison

[geogphotos]

This why I started the thread about the size of Archive images. Unfortunately, Alamy have not responded to my email. 

 

Have any of you doing camera/scanner comparisons done one for an Archive type image ( scruffy, dusty, scratched image that would never pass QC but has historic value) using:

 

1) a scanner with dust removal and finished lightly in Photoshop

2)  a camera with dust busting done entirely in Photoshop?

 

I'm thinking about whether the time saved in camera 'scanning' is then lost in computer 'spotting'. ( though obviously it depends on the image)

 

Though I suppose you could just send the image unspotted under the Archive route with the overall proviso that it has imperfections ( and it really will have major imperfections!)?

 

Would you feel obliged to reduce the size of this Archive image because of the imperfections or just sent it in as it?

 

A film scan that can pass QC at circa 60 Mb is a totally different creature to one that can just about scrape over the Archive barrier of 5 Mb. Yet both might arguably be said to have imperfections.

Edited February 27, 2020 by geogphotos

[Harry Harrison]

One way for a buyer to differentiate between scans and digital originals is of course the date shot, in that anything before about 2000 is unlikely to be from digital, it gets more confused in the 10 years or so after that. I've noticed that sometimes the 'Date shot' is actually the date scanned as the contributor has left that in the EXIF, accidentally presumably, however they always make it clear in the caption and/or keywords. Most of us don't necessarily know precisely when images on film were shot unless they were of a particular event, I can probably pin it down to the year and of course Kodachromes give us the date processed on the mounts.

 

QC know none of this though when they are looking at our uploaded images.

 

Edit:

At the dawn of digital, say 1997 to maybe 2005 (roughly) a high res 50MB scan from film will almost certainly have been of better quality than a similarly sized digital original.

Edited February 27, 2020 by Harry Harrison

[spacecadet]

2 hours ago, geogphotos said:

I'm thinking about whether the time saved in camera 'scanning' is then lost in computer 'spotting'. ( though obviously it depends on the image)

 

Though I suppose you could just send the image unspotted under the Archive route with the overall proviso that it has imperfections ( and it really will have major imperfections!)?

If they're our own images, then  they've been carefully protected from dust in sleeves, of course!😉

Seriously, mine haven't been too bad, and they got an anti-static brush beforehand. With spotting you have to take a view, but I haven't had to spend more than a few minutes on anything. I don't seem to have many with more than about 25 spots and the average is much lower.

Although of course I could now go straight through on 5-star QC, I'm still dutifully going straight through on archival. So I can't find out if they would have passed QC even if I wanted to😉.

Oh dear, two winks in one post. Sorry.

 

Edited February 27, 2020 by spacecadet

[Harry Harrison]

2 hours ago, geogphotos said:

I'm thinking about whether the time saved in camera 'scanning' is then lost in computer 'spotting'.

You could take a view to downsize them so that at 100% there are no visible spots or other defects, that might be more satisfying as opposed to uploading stuff with defects that you can see. Clearly a 5500px image needs more retouching than a 3000 px one. Does your Minolta scanner make a good job of dust removal? Does it work with Kodachromes? Would the smaller size have an impact on saleability for that particular subject matter?

Edited February 27, 2020 by Harry Harrison

[geogphotos]

39 minutes ago, Harry Harrison said:

You could take a view to downsize them so that at 100% there are no visible spots or other defects, that might be more satisfying as opposed to uploading stuff with defects that you can see. Clearly a 5500px image needs more retouching than a 3000 px one. Does your Minolta scanner make a good job of dust removal? Does it work with Kodachromes? Would the smaller size have an impact on saleability for that particular subject matter?

 

The Minolta/Vuescan software does a good job even with Kodachromes - though probably softens them a  bit. 

 

I don't know whether it is advisable to adjust the size of the images - that is the question that I asked Alamy and have had no reply. In the past I have probably downsized too readily but who knows? 

 

Equally why spend time clearing up dust spots - apart from the very obvious big ones - when everybody acknowledges that Archive images have flaws?  Given the low RPI these days maybe it is better just to whack them in and not worry too much?

Edited February 27, 2020 by geogphotos

[Harry Harrison]

58 minutes ago, geogphotos said:

I don't know whether it is advisable to adjust the size of the images - that is the question that I asked Alamy and have had no reply

I woud think that it's going to depend upon the subject matter really, how big is anyone likely to want to use it, for what purpose? You're probably the only one in a position to gauge that. The normal minimum of 3000 x 2000 px is still 10" across @300 dpi. Nothing wrong with downsizing per se - or just scanning at a lower resolution in the first place which will most likely be a lot quicker if batch scanning (with a scanner obviously, not with a camera).

[MDM]

21 hours ago, M.Chapman said:

 

I won't be producing A4 prints, but I do like "clean" images with a level of grain that is well below "obtrusive" and I also want my images pass Alamy QC. That's the bit that concerns me most. Looking at the files you uploaded here (A few downloadable examples), unless Alamy QC are significantly more lenient when it comes to what they perceive are slide copies, I would be quite worried about submitting PriestLeap032, and to a lesser degree Blackhead027 to QC. (Maybe I'm paranoid?)  If, as others have mentioned, Alamy apply lower QC standards to slide copies (which aren't submitted via the archival route) I don't understand their logic. If a customer selects an image submitted via the archive route, they are given a warning "This image could have imperfections as it’s either historical or reportage". No such warning will appear on slide copies submitted via the normal route. So surely the QC standard should be nominally the same as for straight digital camera shots? Otherwise don't they risk disappointing the customer?

 

Here are a couple of further examples I've produced (using the Topaz workflow I described earlier). These are, I hope, good enough for standard Alamy QC with no leniency since (to my eyes) they aren't easily recognisable as slide copies.

 

https://drive.google.com/open?id=1iFateIxBP1cRpLEphNsE_QKGorlMTfNj

https://drive.google.com/open?id=16-G_CdoLgYryXOlX0iDpMUKn9j9wDZaE

 

For proper inspection it's probably best to download as I'm not sure if Google Drive renders them 100% accurately.

 

Mark

 

Perhaps we are placing a bit too much focus on Alamy QC. I was really thinking of getting images up to a standard where they are clean enough in terms of noise and sharp enough to pass QC as a general standard. I have no doubt that mine, including the ones you mention as being a bit worried about, and yours would have no problem whatsoever. Some of mine are slightly grainy but nothing that I would worry about at 100%. Anyway I uploaded a set, including the ones at that link, and they just went through without being checked but I would not have done it if I thought there was any chance of ruining my excellent QC record. 

 

I always download and open in Photoshop by the way. Your images look very clean indeed. 

 

 

10 hours ago, Harry Harrison said:

One way for a buyer to differentiate between scans and digital originals is of course the date shot, in that anything before about 2000 is unlikely to be from digital, it gets more confused in the 10 years or so after that. I've noticed that sometimes the 'Date shot' is actually the date scanned as the contributor has left that in the EXIF, accidentally presumably, however they always make it clear in the caption and/or keywords. Most of us don't necessarily know precisely when images on film were shot unless they were of a particular event, I can probably pin it down to the year and of course Kodachromes give us the date processed on the mounts.

 

QC know none of this though when they are looking at our uploaded images.

 

Edit:

At the dawn of digital, say 1997 to maybe 2005 (roughly) a high res 50MB scan from film will almost certainly have been of better quality than a similarly sized digital original.

 

When we met James A in Ely back in January, I asked him about scans and he said that the QC people are easily able to recognise a scan and did take that into account.

 

Your timing is probably about right on average although there will be lots of pics taken on earlier and generally very expensive DSLRs. I bought my first DSLR in 2005 (Canon 20D) and the images are a lot better in terms of noise than the LS4000 scans I was producing at that time. I have never uploaded any of thses as I used to sharpen them heavily and they would never have passed QC. They look awful in comparison to what I am producing now with the DSLR slide copies, which I don't think need any special considerations if they did get stopped in QC. Mark is the ultimate perfectionist 😀.

 

When I started on Alamy there was a 48MB minimum size and the Canon 20D was 8MP I think, so it was a big jump and the upsized images looked pretty soft. I am not sure if I actually ever uploaded any of those 20D pics, as by the time I started here I had moved up to the Nikon D700. Although this was only 12MP, it was a superb camera, especially with my old Nikkor AIS primes on board although the images still didn't look entirely sharp when upsized (I never sharpened the raws at all in those days)

 

 

Edited February 27, 2020 by MDM

[M.Chapman]

I've now submitted two batches of digitised 35mm slides to Alamy. The first batch went straight through (5 star QC rating). The second batch went through QC (random QC check) and passed.

 

One issue I ran into is the "Date taken" field. After I'd digitised the slides I adjusted the "Date Captured" to match the approximate date when the slide was taken, and then uploaded to Alamy. But when I checked the files in AIM I found the "Date taken" still showed the date that I digitised the slide. Mmmm... Turns out there are several different Date fields in the image file's Metadata. Alamy seem to use the one in the IPTC section of the metadata. Unfortunately that was the one date field that the software I was using (BreezeBrowser) didn't update. . Although I it was relatively easy to fix in AIM, I notice that the files provided for download are likely to contain the wrong date in the IPTC field of their metadata (it looks like Alamy don't strip this bit of metadata).

 

If I'd used LR to adjust the date (Metadata>Edit Capture Time...) I think I'd have been OK as it seems to adjust all the date fields.

 

Mark

Edited March 3, 2020 by M.Chapman

[Chuck Nacke]

Mark,

 

I now do all of my IPTC fields in PS and on scans I always make sure that the date the image

was shot is the date used, which I enter manually and it has stayed correct through AIM.

 

Chuck

[spacecadet]

19 hours ago, Chuck Nacke said:

Mark,

 

I now do all of my IPTC fields in PS and on scans I always make sure that the date the image

was shot is the date used, which I enter manually and it has stayed correct through AIM.

 

Chuck

Yes, otherwise it's disconcerting to see a "date taken" of 2016 on a photograph of a building that was demolished in the 90s.

[Chuck Nacke]

1 hour ago, spacecadet said:

Yes, otherwise it's disconcerting to see a "date taken" of 2016 on a photograph of a building that was demolished in the 90s.

I found an image of an actress on Alamy who died in 1989 and the "date taken" on Alamy was 2004......

Now that is a tick I would like to learn......

 

I am now including the year the original film image was shot in the Caption.

 

Chuck

 

Edited March 4, 2020 by Chuck Nacke
addition

[MDM]

22 hours ago, M.Chapman said:

I've now submitted two batches of digitised 35mm slides to Alamy. The first batch went straight through (5 star QC rating). The second batch went through QC (random QC check) and passed.

 

One issue I ran into is the "Date taken" field. After I'd digitised the slides I adjusted the "Date Captured" to match the approximate date when the slide was taken, and then uploaded to Alamy. But when I checked the files in AIM I found the "Date taken" still showed the date that I digitised the slide. Mmmm... Turns out there are several different Date fields in the image file's Metadata. Alamy seem to use the one in the IPTC section of the metadata. Unfortunately that was the one date field that the software I was using (BreezeBrowser) didn't update. . Although I it was relatively easy to fix in AIM, I notice that the files provided for download are likely to contain the wrong date in the IPTC field of their metadata (it looks like Alamy don't strip this bit of metadata).

 

If I'd used LR to adjust the date (Metadata>Edit Capture Time...) I think I'd have been OK as it seems to adjust all the date fields.

 

Mark

 

I have just keyworded a few in AIM and just set the year they were taken as I don't recall the exact dates anyway. I will see what comes through tomorrow when the database is updated. 

[M.Chapman]

23 hours ago, Chuck Nacke said:

Mark,

 

I now do all of my IPTC fields in PS and on scans I always make sure that the date the image

was shot is the date used, which I enter manually and it has stayed correct through AIM.

 

Chuck

 

That's good. Your news reporting days (presumably using IPTC fields) have helped you avoid my error.

 

I've done some more investigation on the subject of metadata dates and it's surprisingly complex. Over the years it looks like more and more fields have been defined and there's some overlap. It looks like using PS to set the IPTC date keeps Alamy happy, but it doesn't update the other (XMP) date fields.

 

Changing the date in LR (Metadata>Edit Capture Time....) does slightly more. This will not only adjust the IPTC DateCreated and TimeCreated fields, it also preserves the digital shoot dates as DigitalCreationTime and DigitalCreationDate fields and it also updates an XMP date field.

 

I think it's best to edit the date before uploading to Alamy (as you are doing), since that way AIM will show the correct date, and the file supplied to customer will also show the correct date. For Alamy I don't think it matters whether this is done in LR or PS, but I've found it doesn't work in BreezeBrowser (which I was using) as this leaves the IPTC DateCreated field unchanged.

 

Mark

Edited March 5, 2020 by M.Chapman

[M.Chapman]

New DPReview video on the Nikon ES-2 is here 

 

Mark

[Harry Harrison]

Interesting to see it in action, he mentions the new DPReview analog forum which I have also seen mentioned by Hamish Gill of 35MMC, he is one of the moderators, possibly one of the instigators in fact. More for people that are still shooting film I imagine.

 

https://www.dpreview.com/forums/1072

 

On his twitter account he has posted pictures of parts arriving for his pixl-latr DSLR scanning film holder, I imagine the aim is to launch at the Photography Show in the new Analog area.

 

This older video is more comprehensive, he uses something called a 'lume cube' as a light source, which I've never seen before.

 

https://www.youtube.com/watch?v=SNK24oQICUc

 

The Lume Cube has a high CRI and is £90 on Wex here

 

[MDM]

2 hours ago, M.Chapman said:

New DPReview video on the Nikon ES-2 is here 

 

Mark

 

Interesting video but clearly he has not spent a lot of time experimenting. I would take issue with a couple of things he says. He mentions the Nikkor 105 at the beginning as if it will work but doesn't mention that it would need extenders in front of the lens as it won't focus close enough otherwise, presumably because he has not tried anything else but the setup he uses. Also the idea of handholding might work in theory as the adapter is attached to the camera but there is a limit to the length of the shutter speed that will give a sharp shot. When I tried it handheld at about 1/2s I found there was significant shake. Not sure why but I would recommend a tripod anyway at all times. I think I will post a comment there in fact.

 

In relation to dates of uploaded images, the dates I set as year only don't show in the search engine results. I will try something else.

 

1 hour ago, Harry Harrison said:

Interesting to see it in action, he mentions the new DPReview analog forum 

 

https://www.dpreview.com/forums/1072

 

 

 

Looks like an interesting forum. Not that I am intending to start shooting film again just yet. 

 

 

[Harry Harrison]

28 minutes ago, MDM said:

Not that I am intending to start shooting film again just yet. 

Mind you, there is already a thread on scanning with the ES-2, quote...

 

"you may remember from discussions on the Medium Format Forum, I have been involved in drum scanning for many years.

Using a DSLR to scan film is capable of producing excellent results. Done with care it can rival drum scan quality."

 

https://www.dpreview.com/forums/post/63712007

[Harry Harrison]

36 minutes ago, MDM said:

When I tried it handheld at about 1/2s I found there was significant shake. Not sure why but I would recommend a tripod anyway at all times.

Tripod, in fact just a table tripod in this case, would be safer and more convenient anyway. In fact with his light source he was getting 1/40th at f8 on 100 iso, so that may have been OK I suppose.

 

Shame Nikon didn't let their Live View colour negative app work with RAW files, and also that they've seemingly dropped it from their mirrorless cameras.

[MDM]

22 minutes ago, Harry Harrison said:

Mind you, there is already a thread on scanning with the ES-2, quote...

 

"you may remember from discussions on the Medium Format Forum, I have been involved in drum scanning for many years.

Using a DSLR to scan film is capable of producing excellent results. Done with care it can rival drum scan quality."

 

https://www.dpreview.com/forums/post/63712007

 

Yes I just joined in there with a little bit of acquired wisdom about using short and medium telephotos lenses. I better be careful though. I spend too much time on here already without extra diversions.

 

13 minutes ago, Harry Harrison said:

Tripod, in fact just a table tripod in this case, would be safer and more convenient anyway. In fact with his light source he was getting 1/40th at f8 on 100 iso, so that may have been OK I suppose.

 

Shame Nikon didn't let their Live View colour negative app work with RAW files, and also that they've seemingly dropped it from their mirrorless cameras.

 

i just use a normal tripod and set the light on a table. Why go to all that trouble and go lazy on the tripod? It is not like I am carrying it up a mountain 😀.

 

Yes but I think one would still want to control the conversion on the computer so it is probably not a big deal. The in-camera conversion could be very useful for shooting old negs though. I have masses of old black and white negs and no contact sheets for them so being able to have a quick check in camera seems like a good way to do things. I have not tried negatives yet though but it is something for the future (assuming there is one 😀).

 

[Alan Gallery]

I made an adaptor, like the Nikon ES-2, for my 30mm Olympus micro lens that holds mounted slides at a distance that fills the frame.  It is 3D printed and I could make the files available.

[MDM]

1 hour ago, Alan Gallery said:

I made an adaptor, like the Nikon ES-2, for my 30mm Olympus micro lens that holds mounted slides at a distance that fills the frame.  It is 3D printed and I could make the files available.

 

Sounds fascinating. Can you explain how you actually made it. I know nothing much about 3D printing.

[Alan Gallery]

I used Fusion 360 to design the adaptor.  It is an engineering design program that can produce STL files that are the standard file type for all forms of 3D printing.   The next step is to run the STL files through a Slicer program to create a set of instruction for the specific 3D printer and the material that is to be used.  3D printing involves the printer laying down a layer of plastic (in this case) for each slice until each part is complete.   

[Harry Harrison]

10 hours ago, Alan Gallery said:

I used Fusion 360 to design the adaptor.  It is an engineering design program that can produce STL files that are the standard file type for all forms of 3D printing.   The next step is to run the STL files through a Slicer program to create a set of instruction for the specific 3D printer and the material that is to be used.  3D printing involves the printer laying down a layer of plastic (in this case) for each slice until each part is complete.   

Thanks for sharing that, I find it quite extraordinary what can be done, I came across this site where there seems to be a huge range of photgraphy related options:

 

https://www.yeggi.com/q/film+holder/

 

Is there much choice in the quality and properties of the plastic that can be used, strength, flexibility etc. and does that depend upon the type or expense of the 3D printer that you are using?