30 July 2012

Is Velvia just Saturation – A Response to ‘Velvia Dissonance’

Yesterday Guy Tal wrote a blog post about the “The Velvia Dissonance” where he says that Velvia is redundant and offers no advantage over digital capture. Along the way he mentions that it is terrible for scanning and is effectively ‘instagram for lazy photographers’.

He also berates the ‘transparency is sacred’ commandment, bemoaning the lack of creativity of a canned colour profile. I sort of agree with the fact that treating the transparency as inviolate seems limiting however if some people wish to concentrate on composition, timing, artistic expression before capture and like the rendering of Velvia – why not. Just as there is nothing wrong with deciding that a transparency needs no improvement, in many cases an artistic hand can improve things massively (I would go as far to say ‘in most cases’ in my opinion).

“Enough of the agreeing already!” I hear you cry. So, s you would expect, I disagree with the premise that ‘better emulsions were introduced more than a decade ago’ as it depends on what you mean by better. Fuji Velvia did something that other films couldn’t and although saturation was part of the puzzle, it wasn’t the whole of it – as anybody who has tried to make a ‘velvia plugin’ would agree (and it seems most Velvia plugins just add saturation and contrast – very simplistic).

Velvia’s trick is to do more with separation of colour tonality than saturation – it has the ability to take a low contrast scene and add pizazz that no end of digital sourced simulation can manage. And in these days of good scans, the sacrosanct transparency can be just a start.

For instance, I just shot this image in the lake district and whilst the composition isn’t particularly interesting, at least it demonstrates how Velvia does this tonal separation. 

Of particular note to me are the neutral highlights in the ice and on the rocks and the way the rocks have a neutral to green tone (green caused by lichens) and the hillside and sky has a beautiful twighlight cast. Now here’s the 5Dmk2 version straight out of the camera

and with the ‘Alien Skin Exposure’ simulator plus some colour tweaking to try to get a similar colour overall without making any individual parts look ‘wierd’


Now obviously this is taken with a shorter shutter speed but ignore the water for a bit I for one much prefer the Velvia version. Even with half an hour of tweaks using the Velvia as a reference I get something that works OKish on first glance but still doesn’t have the punch of the Velvia. (if anybody wants to play with the digital original – here’s the raw file. I’ll be happy to publish people’s interpretations)

Guy also makes a point about Velvia being a crap film for scanning. Well, if you don’t have access to a good scanner it is but if you can drum scan or use a Nikon 9000 or Imacon it’s perfect. The tones spread across the whole dynamic range of the scanner, it’s almost like the film is tuned to the best DMax scanners can get. And if you’ve seen some of the post processing videos in ‘On Landscape’ you’ll see just what an expert (Joe Cornish) can do with a colour transparency scan.

It’s no surprise that a lot of people are playing with film again – they see something unique and different in it, not necessarily better but perhaps something that is not possible to capture in any other way. Guy mentions that perhaps it deserves a place in the history of photography alongside glass plates – well, like glass plates, Fuji Velvia will deserve both a place in history and in the hearts of many contemporary photographers who are doing unique things with it. (oh – and as for glass plates – Just check out Sally Mann, France Scully Osterman & Joni Sternbach and various other contemporary photographers whose work is probably irrelevant). It’s a shame we won’t be able to use these colour materials in another few decades unlike the above mentioned.

So – I cannot agree with the premise that Velvia was just about saturation. Perhaps that is all many photographers saw it as but for the majority that I know it’s a unique starting point to express an artistic vision.





Comments (skip to bottom)

14 Responses to “Is Velvia just Saturation – A Response to ‘Velvia Dissonance’”

  1. On July 30, 2012 at 12:13 pm Samuel West Hiser responded with... #

    Velvia abuse — such as that promulgated routinely and going back all the way by National Geographic and too many practitioners who do not LOOK — is the more objectionable phenomenon. I’ve found Velvia to behave true — even more so in low or less-contrasty light.

    Your example, above, is a good reflection of Velvia’s potential in mature hands; although, most caring photographers can do better with the digital example simply using Custom White balance.

    Most can’t be bothered, however, and Tal’s “Dissonance” is his own, for he has confused distinctions between different technologies for the difference between good and bad pictures.

    • On July 30, 2012 at 12:22 pm timparkin responded with... #

      Believe it or not I did using Custom White balance but it all depends on what part of the picture you click. I chose the closest to a neutral grey which was the rocks in this case.

    • On August 1, 2012 at 3:27 am David Leland Hyde responded with... #

      Excellent post, Tim. I like a contrarian position, though I’m not sure whether the contrarian is Guy or you, Tim, maybe a little of each. Your technical points all make sense. Also, I am quite sure that Rich Seiling of West Coast Imaging would disagree with Guy. With a staff of experts doing the scanning with the world’s best scanning software and drum scans, I am certain that Rich would agree with you about scanning Velvia. On the other hand, it is precisely the problems with over-saturation that are Velvia’s main drawback in my opinion. For that primary reason, it is widely agreed that Velvia changed landscape photography for the worse as I argue in my post:
      Furthermore, it is the attempts to match Velvia in Photoshop, that you have proven yourself here to render inferior results and that continue to send the landscape genre into nature-fake land.

  2. On July 30, 2012 at 7:26 pm Guy Tal responded with... #

    Thank you for the detailed response, Tim. I would like to point out that the sentence ‘just about saturation’ is not taken from my article, despite appearing in quotations. Also worth mentioning is the fact that my discussion did not make any value judgment about saturation at all, only about the (im)practicalities of using a film like Velvia.
    This response actually strengthens my argument that Velvia simply provides a canned solution for those who prefer to leave it to the film to do the heavy lifting rather than controlling it themselves. In that sense it is no different than JPEG algorithms built into practically all digital cameras. Very useful for commercial applications, but quite limited in terms of creative post-processing options.
    It is actually no surprise at all that no simple script can replicate the look of Velvia. To do so accurately, the software will need to know the response curve of Velvia, as well as that of every digital sensor, and have access to light measurements (RAW data) so it could successfully translate from one curve to another. That means that such a software will need to be a RAW converter and contain profiles for dozens (hundreds) of types of sensors. Quite impractical, in my opinion.
    When I can directly control pixel values and adjust individual colors and tones to achieve whatever result I want, why would I care how Velvia would have done it?

    • On July 30, 2012 at 7:42 pm timparkin responded with... #

      Sorry I put that in quotes – it did seem to me that the general point of the article that the only reason anybody used Velvia was for the saturation. Your comment “with digital capture, the entire niche of highly saturated positive film had become obsolete” suggests that it is the saturation that is irrelevant as it’s the only common factor that you use to collectively group the files together.

      A couple of things you said in the article – “color palettes are under the complete control of the photographer” – no they aren’t. Colour palette is also controlled by the spectral response of the individual parts of the camera too – The choices of colour palette in a picture are not infinite – they are constrained, a subset of all possibilities. Different cameras will provide different subsets. Velvia provides a subset that either not available at all or not easily available to digital sensors.

      If colour was just a ‘profiling’ thing then why do different cameras produce very different colours?

      The bottom line for me is that Velvia doesn’t provide a canned end point but a canned start point in the same way that your raw file gives a canned starting point. Just because your starting point is different doesn’t make it any more valid.

  3. On July 30, 2012 at 8:43 pm Guy Tal responded with... #

    Thanks for the clarification!

    I was just using the saturation reference to distinguish a general category of films (including Velvia, E100VS and others).

    Also, small correction to your point about color palette: once an image is in digital form and can be assigned a color space of your choosing, you DO have complete control of the palette. The sensor’s spectral response only determines what frequencies of light it is capable of measuring/recording. This is just a starting point (input) to processing, and does not limit the palette available to you to map colors into, so long as they are available in your chosen (output) color space. I would agree that dealing with light outside the visible spectrum may introduce some complexity but I assume that in all cases the output you’re after only uses colors within the visible spectrum, which essentially means having to re-map undesired color casts, if they exist.

    More to your point – I never said that color is a “profiling thing,” just that in order to accurately mimic the Velvia response curve, you will also need to know the sensor’s response curve and have access to the amount/frequency of light measured at any given point in the image so you can make an accurate mapping. A simple script applied to images already converted from RAW is highly unlikely to accomplish such mapping accurately. The practical question, though, is why you would want to force such mapping when you can custom-tailor one to the needs of each image.

    Comparing an already color-mapped image with an established contrast curve to a RAW file is really comparing apples and oranges, as I hope you’ll agree. A RAW file contains linear readings that can be turned into pixels using practically any contrast curve you may like. A Velvia capture, on the other hand, is far from linear data and you may need to “undo” some of the decisions made on your behalf to achieve a desired result that is different from its default response.

    • On July 30, 2012 at 9:27 pm timparkin responded with... #

      I have to still disagree about the colour thing.. I’ve studied, tested and written about this and found that colours are not canonical. You can’t take every image from one camera and make it look like the image was produced by another camera regardless of profiling/photoshop/etc.

      For example, many colours are actually a combination of spectral spikes of two different colours. If the spectral response isn’t identical to our eye, one of the spikes will influence the colour more and you’ll get a hue shift. But only for that material.

      Hence when you try and ‘fix’ that colour – you cause a problem with the same colour that has been correctly recorded elsewhere in the scene.

      Foliage is one of the classic metameric materials (that being the name of colours that are a detected as different by different spectral sensors). This is why foliage is one of the materials that gets so screwed up by different sensors.

      I particularly like the way that Velvia handles colour – it has spectral curves that are more comparable with the eye (at least my eye!!)

      • On July 31, 2012 at 10:43 pm Richard Wong responded with... #

        I agree with you Tim. I don’t know about the scientific aspect of cameras but I do know from having used a number of different digital cameras is that they all render things differently. You can set the same white balance across the board and play around but they never look identical.

        Your Velvia example looks great btw. I have no problem with people wanting to continue using it if the work looks good. People who do digital capture have done bad work as well, as evidenced by the HDR hack jobs and excessive saturation slider settings. One can look at HDR as a canned effect as well. Expose all over the map then let the software piece it together for you at home. It’s easier to pull that off in post-processing than it is to learn how to see light and expose for it.

  4. On July 31, 2012 at 12:14 pm Julian responded with... #

    Perhaps Guy is speaking with the zeal of the newly converted (he was shooting 5×4 until a couple of years ago, as I recall) but his point that Velvia (and similar reversal films) are a ‘canned’ endpoint rather than a creative starting point is ably disproved by those photographers who have mastered the dark art of Ilfochrome (Cibachrome) printing – photographers such as Christopher Burkett and Michael Fatali.

    He is right on the point about difficulty of getting a good scan, though. The ‘pepper grain’ effect was clearly visible (and very annoying) when I was using a Minolta Multi-pro to scan 645 transparencies. Using a diffuser between the light-source and film helps somewhat.but Velvia is still a swine in this respect. It was mainly the ease of scanning (and the clearly superior results from my humble hobbyist setup) that led me to move from reversal films to colour neg.


    • On July 31, 2012 at 1:27 pm timparkin responded with... #

      Indeed – scanning colour neg is results in very good results on quite poor scanners.

  5. On July 31, 2012 at 11:20 pm QT Luong responded with... #

    Some people assume that they can take a digital file and make it look like a film image shot with any emulsion. If they actually tried to do so, they’ll find out that it is not possible with common imaging tools such as PS. As to the grain of Velvia, I had not found it to be an issue with large format transparencies, but I agree that a drum scanner is key to getting clean shadows.

  6. On August 23, 2012 at 6:18 am Rick responded with... #

    Tim, I agree with you here and think although Guy has some interesting points he is missing some too.

    The Velvia scan is always a starting point and as you mention we can eliminate or drastically reduce the problems with a scan if we are using the right level of scanning equipment. I use a Nikon Coolscan 9000 and I think that really is probably the least you should try and work with to get a nice scan to start with.

    Personally although I said the scan can be the starting point for a final image I like to think of it as close to the result. As others have mentioned I prefer to work within the confines of the film so that I can truly try and ‘pre-visualise’ rather than ‘post-process’ the image. For me there is more ‘traditional photographic skill’ in producing as close to the final image ‘in-camera’ than in pushing pixels after the fact. This methodology often means I won’t take many or even any shots on days when the light is just not right but I sleep sounder in the knowledge that when I do get the shot it is as true to the original scene as I can make it.

    I know digital converts will always say that traditional photographers always pushed things in the darkroom but I don’t believe this was anywhere near the extent of what we are now seeing with some of the digital work being put forward. Almost always this was in the B&W darkroom and hence has no comparison to the colour manipulation we now see in digital. And although some film photographers did ‘fine-tune’ their Ilfachrome/Cibachrome prints I think you will find that again this was minor adjustments relative to photoshop capabilities…

    Anyway, just my 2 cents, I know others will disagree but I for one will really miss Velvia and although I only shoot 120 in 6×7 format (which I don’t believe is disappearing just yet) I was hoping to shoot the film in 4×5 and 8×10 one day…

  7. On November 8, 2012 at 4:11 pm Demian responded with... #

    Hi, Tim! I have played with your image and want to tell a little story. About two “old-school” developers, which aimed to get a true film-like processing the RAW images. And I think they almost reached result. Raw Photo Processor (RPP) includes few profiles of a real films (scanned and considered the actual color and tonal curves). I think it will be interesting for you – it’s free, but mac-only.
    What i did. I use the Velvia 50 v2 profile (in beta) and added one-only color-intervention adjustment layer. Clear profiled image sure has some color variance, but i guess you wouldn’t get two identically frames from two same Velvia films too. And in my case – I had a reference image, which i think, I achieved.
    Try it. It’s not a plug-in. RPP is hard to understand in first time (in second time too), but it is worth. You have my e-mail, just in case
    PSD file and


  1. The Ultimate End of Velvia? – Land & Colors - Landscape Photography, Blog, Fine Art Prints - September 9, 2012

    […] Parkin said it much better here: . He has put together some nice examples of color rendition of the same scene by Velvia and 5Dm2. […]

Add your response