Velvia 100 vs Provia 100F

[jennym]

I use my local lab, Palm Laboratories in Birmingham. They have served me well, and I know they also do quite a lot by mail order. They charge £2.50 a sheet but there is no extra charge for push/pull and they give a 10% student discount. They will also do a one hour turnaround for no extra charge which is quite handy.

So Tim, if you do a test comparing labs it might be worth including them.

Cheers,

Jenny

[/quote]

[joolsb]

Yes ,I should have said that you have a consistent representaton of the colour and that if you are familiar with the film you are using you will have a good idea what the actual colour was like. Whereas if you introduce random colour casts, this is impossible.

A fair point. OTOH, so long as you can get a halfway decent scan, does it really matter? I mean, if you were doing commercial work, getting a company's signature colour absolutely perfect is highly critical but I think most of us here aspire to a more or less artistic interpretation of our subjects, in which case is absolute colour fidelity even relevant? Or should we all be including a colour-chart in our images to be 100% positive that we have accurate colour? I shoot NPS a lot and rely on memory and feeling to get colour that fits my creative goals. Even if I'm working on a Velvia scan, I usually don't consult the original while I'm editing unless something's gone badly wrong somewhere. Yet if I compare the final result to the original tranny, there usually isn't a huge difference.

As regards casts, I take your point, Tim. Removing them in PS is a real faff - which is why I've recently started importing scans into Capture 1. Being able to eliminate casts by clicking on a colour-neutral part of the image is a real time-saver.

[timparkin]

Yes ,I should have said that you have a consistent representaton of the colour and that if you are familiar with the film you are using you will have a good idea what the actual colour was like. Whereas if you introduce random colour casts, this is impossible.

A fair point. OTOH, so long as you can get a halfway decent scan, does it really matter? I mean, if you were doing commercial work, getting a company's signature colour absolutely perfect is highly critical but I think most of us here aspire to a more or less artistic interpretation of our subjects, in which case is absolute colour fidelity even relevant? Or should we all be including a colour-chart in our images to be 100% positive that we have accurate colour? I shoot NPS a lot and rely on memory and feeling to get colour that fits my creative goals. Even if I'm working on a Velvia scan, I usually don't consult the original while I'm editing unless something's gone badly wrong somewhere. Yet if I compare the final result to the original tranny, there usually isn't a huge difference.

As regards casts, I take your point, Tim. Removing them in PS is a real faff - which is why I've recently started importing scans into Capture 1. Being able to eliminate casts by clicking on a colour-neutral part of the image is a real time-saver.

Hmm... sort of yes but quite often the distance between "beleivable" and "doesn't feel right" can be quite small.. I've sometimes had one picture taken with Velvia where the colour is just as I've wanted it but the light wasn't in quite the right place and another in Provia where the light was bang on. I've tried converting the Provia one to look like the colour in the Velvia one and couldn't get it right.

Anyway - it's a matter of degree. If the inconsistent colour I get is still beleivable then fine. If it isn't beleivable then it's an issue. Would you agree?

As far as getting Pro160 to look like I want, well lets say I've wasted an enormous amount of time playing - graduall turning a picture from beleivable but shit colour to unbeleivable with the odd non-shit colour here and their. <sigh>..

By the way, if you use the go into curves and pick the color picker with the 'half full' dropper you can click anywhere to get back to a white balance.

Tim

[Patrick Dixon]

I've tried converting the Provia one to look like the colour in the Velvia one and couldn't get it right.

I'm sure I came across something on the web where someone reckoned that you could shoot provia (or was it a neg film?) and post process to get it to look like velvia.

If I can find it again I'll post the link.

[EDIT]

I think it was http://www.dl-c.com/Velvia%20vs%20Provi ... 0100F.html that I was thinking off - it doesn't exactly say how you get from provia to velvia though ...

[joolsb]

Hmm... sort of yes but quite often the distance between "beleivable" and "doesn't feel right" can be quite small.. I've sometimes had one picture taken with Velvia where the colour is just as I've wanted it but the light wasn't in quite the right place and another in Provia where the light was bang on. I've tried converting the Provia one to look like the colour in the Velvia one and couldn't get it right.

Anyway - it's a matter of degree. If the inconsistent colour I get is still beleivable then fine. If it isn't beleivable then it's an issue. Would you agree?

I think that's fair. Of course 'believable' is open to a degree of interpretation.

As far as getting Pro160 to look like I want, well lets say I've wasted an enormous amount of time playing - graduall turning a picture from beleivable but shit colour to unbeleivable with the odd non-shit colour here and their. <sigh>..

Interesting. I've had a fair amount of success with NPS, tbh. I start with the Silverfast 'approximation' and then apply my normal processes from there. Seems to work at least as well as with Velvia - if not better as NPS scans seem more malleable.

By the way, if you use the go into curves and pick the color picker with the 'half full' dropper you can click anywhere to get back to a white balance.

Ah, thanks for that. I'll give it a go. As a matter of routine, I do Curves adjustments only in luminosity blending mode so I can separate out contrast adjustments from colour ones (which I do using Selective Colour). This could be why I've never discovered that little trick.