Velvia 50 - A penny for your thoughts!

[steve m smith]

<blockquote>

<p>personally, i'd like to know how to how to simulate that digital 'look' with film</p>

</blockquote>

<p>Use a cheap scanner and blow out the highlights!</p>

[Rod Sorensen]

<p>Can you post a few of those digital images where you came CLOSE to Velvia 50?<br>

Thanks.</p>

[randyhargraves]

<p>Not sure if this helps, but I use DxO software and they have a preset that will emulate certain slide films. The Elite version has Velvia 100 as an option, and the "film-pack" add-on I believe has Velvia 50. I've tried them both and it's pretty good. Definitely gives you the pop at the expense of increased contrast and blocked up shadows. Although that's somewhat characteristic of Velvia. You might try the free trial download and see if that's anywhere near your tastes. Nothing beats actual slide film, but in my opinion that only holds true for viewing the actual slide. Scanned and/or printed it just loses something. </p>

[mfophotos]

<p>Did you check the Fuji spec sheet?<br>

<a href="http://fujifilm.co.uk/professional/films/pdfs/fujichrome_velvia50.pdf">http://fujifilm.co.uk/professional/films/pdfs/fujichrome_velvia50.pdf</a><br /><br /></p>

[StuartMoxham]

<p>Shoot some scenes with digtial and velvia put the velvia on the lightbox and adjust the digital image until it looks the same. Be sure to do it for shade, direct sunlight, varying times of day and different weather conditions. In the end you may be able to build up a collecting of adjustments to apply to your digital image to make them look close to velvia.</p>

[tim gray]

<p>It might be hard to exactly reproduce Velvia (as others have repeatedly said). I do think you could get in the ballpark though. Everyone has already told you how. Shoot some shots with your DSLR and the same shots with Velvia. Compare and adjust the DSLR shots to match. Save the adjustments as an action or something. Try a couple shots of a Color Checker as well if you want to know how the film reacts to specific colors.</p>

<p>As someone also already said, the high contrast of Velvia (and slide film in general) has a lot to do with the look as well. </p>

[alvinyap]

<p>Wow interesting thread, I was planning to do something similar, but with negative film (superia, portra). With slides we could do the side by side comparison between raw converter/lightbox, but how about negatives? Would a comparison scan do the trick?</p>

[akgraham]

<p><img src="http://gallery.me.com/akgraham/100114/2005-06-30%20at%2012-59-08/web.jpg?ver=12565303890001" alt="" width="1024" height="825" />To measure the delicious Velvia 50 colour-shift with long exposures, and also somehow plot that in a linear way for a post-processing action, would be quite an achievement.</p>

[apostolos_tournas]

<p>If you don't mind, why do want to emulate Velvia 50 digitally? If you see something in Velvia that deserves commitment, perhaps you should shoot Velvia along with your digital shots. I personally would prefer to see DSLR photographers leave film in peace.</p>

[m_dawg]

<p>One of the problems here is that you're trying to emulate what something looks like on a light table digitally. And that's tricky since you can't compare a light table and computer monitor very scientifically and even if you calibrate both...it's still subjective.</p>

<p>Emulating velvia scans is also tricky since every scanner sees things differently and then everyone treats their photoshop files differently to match how they saw the slide on a light table.</p>

<p>But I think matching digital files to velvia scans is very possible. Canon's picture style system is very powerful and you could shoot a color checker chart with that and then with 135 film and match the two. Without that, you might have to resort to curves and a photoshop action--way less scientific. But either should get you close. Shoot some slides yourself; no way will you get anything close based on descriptions of how other people think an emulsion looks... Velvia is also extremely high contrast, which is a little tricky since you'd have to recalibrate your expectations for a four-stop instead of nine-stop range. There's a LOT you can't shoot with it. You could finish off the effect by adding a little grain and a little wide radius usm to get a "sharp" look.</p>

<p>I may try some side-by-side tests. From what I've seen digital and velvia aren't that different but velvia does hang onto little bits of color really well (and also destroys skin tones).</p>

[h_s1]

<p>Well, thought I would also contribute a bit.</p>

<p>First of all, let me declare that I love Velvia. I am shooting Velvia 50 these days and just love the colors and contrast. It is mostly for landscapes ... autumn can a very very colorful season.</p>

<p>Now, regarding simulating Velvia 50 in a digitally captured image, the problem is not straightforward. The most intuitive method to do so has already been suggested a few posts above, i.e. shoot same scene with Velvia 50 and with your digital camera and do adjustments on your digital photo to make as close to Velvia 50 image as you can.</p>

<p>However, things are much more complicated underneath. Not just for this particular problem, but for all similar problems.</p>

<p>I don't have much deep understanding of film chemistry, so the following is a bit different as far as photo development may be concerned.</p>

<p>The underlying cause of the look of Velvia 50 (or for any sensor, film or digital) is the spectral response of the image capturing device. You (the OP) can restate the problem as how to simulate Velvia 50 film's spectral response with a digital sensor.</p>

<p>One formal method to do so could be to assume that there are three kinds of spectral responses of the film and of the digital camera's sensor: Reg, Green and Blue. Next, take a set of images with Velvia 50 and somehow determine the RGB response of the film from that. Perhaps Fuji's data sheet for this film can do that for you. Next, you need to either reprogram the spectral response of the digital camera's sensor (so that it captures images just as Velvia 50 film does), or change it after the image has been captured. The looks of both the sensors' (film and digital camera) output will be similar if the both their spectral responses are similar. Now, the problem of matching a digital camera sensor's spectral response to Velvia 50's is the biggest problem. More like an engineering and optimization problem.</p>

<p>Hope this sheds more light in understanding the intricacies of this problem.</p>

<p> </p>

[sumo_kun]

<p>Do you guys think he has given up yet ;)<br>

As H S said, the 2 materials have different spectral responses. Mapping one to the other is not a linear process. I really don't think that you can emulate RVP on digital. This has been mentioned before but for one shot you could maybe sit there and persevere until they looked similar but then you would need to repeat the process for a different shot. Those colours you mentioned... why those specific colours? Do you think that once you map those colours it will all be fine?<br>

At work, we have a system for rendering CG with a simulated film look and guess what, RVP is one of the profiles. Looking into the code and theory behind it, it is much more complicated than you make it out to be and we are still working on it. For some background theory you should look at some tech papers which discuss film simulation for CG.<br>

If you want the look of Velvia you pretty much have to use Velvia. If you want your own personal look for "wow" images, then why are you asking us? Its a personal thing and you should tweak until you are happy.</p>

[boinkphoto]

<p>Everyone tries, but no one does, which is why I've gone back to shooting Velvia in addition to digital.</p>

<p>I think part of the problem is you literally cannot create data (or color) that does not exist. There is only so much you can tweak things.</p>

<p>That said, if you figure out a way that really works, there'll be plenty who want your recipe.</p>

[Nick D.]

<p>

<p>Try Fuji S3 PRO or S5, they have velvia settings in camera. They have unique Fuji sensor, resolution limited to around 8 mega pixels, but they have best colors in Jpegs and dynamic range still not reached by Nikon or Canon. </p>

<p>

 

</p>

</p>

<p> </p>

[h_s1]

<p>Chuk, I would rather that the OP not give up on his quest. The right way to see this is to realize what the fundamental problem is, what various factors involved are and what are the best options before the OP. Having this information, OP is not only in a better position to make a reasonable decision, but is also now more aware of all these underlying intricacies. In short, I think this thread is a very valuable informative thread for the OP and for everyone else.</p>

<p>In fact, I am glad and thankful that OP started this thread with his query.</p>

<p> </p>

[rickbehl]

<p>Here's my 2 cents :-)</p>

<p>Seems like there are roughly two groups emerging in this thread. The first group believes that the problem is a worthy one of solving and the second believes that Velvia (especially 50) is a somewhat magical film which is going to be hard to impossible to emulate.</p>

<p>Let me outline first that I also shoot film 90% of the time. And Velvia 50 is used in about 70% of those shots. I also believe to a certain extent that this film is exceptional in what it tries to achieve. And of course if you look at the passion with which other photographers past and present speak about it then you can see that I'm not alone here.</p>

<p>In a way this question really harks back to the old Film vs Digital debate. Here's my take on it. Many years ago film manufacturers (Kodak, Fuji, Ilford, etc) at sometimes great expense to themselves set up R&D departments filled with very smart scientists to try and create formulae for different types of film. These formulae would ultimately decide on how to take the 'input' from a camera (in the form of light waves) and convert it into 'output' (in the form of colour on negative/slide film). This 'layer' of the industry was an incredibly important part of the photographic world and hence competition inspired each iteration of film to improve upon the last. This period of time lasted a while and we can only theoretically imagine just how much effort or man days work and thought was put into this process over that period.</p>

<p>Move on to the present digital day and that 'layer' has all but dissappeared. Now we don't have rooms full of smart scientists doing that work. The onus of mapping input (light) to output (colour) has now been taken on by both camera manufacturers (Nikon, Canon, etc) and Software developers. Neither of these parties are yet or maybe never will be experts in the land of colour as those original 'rooms of scientists'. So is it really no surprise that when an image in RAW format pops out of digital camera these days it looks as 'dull as dishwater'? (especially when compared to viewing a well exposed Velvia 50 transparency on a light table). But work is being done in this area mostly by software developers and I am sure that over time we will get closer to simulating certain types of film. However we musn't forget that we also have the issue of comparing a backlit or projected transparency vs digital display image which is a whole other kettle of fish...</p>

<p>Anyway, I guess what I am saying here is, sure, go ahead and try and recreate the Velvia 50 look. And after all that is a scientific problem and so must have a scientific solution (assuming the gamut of one format can fit within the other). BUT, if you really want the Velvia look right now, then shoot film... It's not as hard or as expensive or as time consuming as it looks. And you will be surprised how much LESS time you spend in front of a computer after shooting the image.</p>

<p>Rick</p>

[h_s1]

<p>Les ... er, no. You are comparing apples to oranges. Velvia has no brains of its own, it does not feel anything. Programs are not creating a Velvia look by themselves. The central tenet in all this is the perception of humans. There are two mediums that are capturing light and transforming them before perception. At present, the two transformation are different, yet physical and quantifiable in non-obvious ways.</p>

<p>The other way to look at the problems is that the reference is the human color perception and the idea is to come up with a transformation for a digital cameras which gives the same perception as Velvia's output.</p>

<p>Velvia does not perceive anything (as a mouse does). A program is not doing anything by itself (as the machine in your analogy apparently is doing). There are no two independent references (as a machine's perception and the mice' in your analogy), only one (human's). Thus you are comparing apples to oranges.</p>

<p> </p>

[dan_south]

<p>D S:</p>

<p>If you are really interested in mimicking the look of Velvia 50, borrow someone's film camera and shoot a variety of scenes with the film side by side with your digital camera set to RAW capture. Get the film developed and put the chromes on a light box. Then open your digital files on the computer (you of course have calibrated your monitor, correct?) and try to get them to look like the chromes.</p>

<p>That's really all there is to it. If you want the result you need to get out there and do the work. Repeat the process as necessary if you need to make exposure adjustments for shooting a film with such a narrow latitude. That's probably going to be the toughest part for an all digital shooter to get used to.</p>

<p>I suspect that you'll get close to the look of Velvia in some cases but never an exact match. There are too many variables. I have tried some pre-packaged profiles that were supposed to make digital files look like Velvia, and they were an absolute joke. However, these were generic one-size-fits-all solutions. I'm sure that you'll have better success by working diligently on individual images.</p>

<p>Good luck!</p>

<p>Oh, and stop using my initials. ;-p</p>

<p> </p>

[mauro_franic]

<p>The only way to try to approximate the colors of Velvia is to use Velvia alongside your DSLR and then color match. Saving this as an action will NOT work for different shots.</p>

<p>Here is a test for those using DXO: Convert this Ektar shot to Velvia using software without color match to the Velvia shot and post the results.</p><div></div>

[mauro_franic]

<p>Since you have the answer, you can evaluate the quality of the results yourself.</p>

[h_s1]

<p>Les, you are right, I think I did. :-)</p>

[tim gray]

<p>Mauro: I'm a film lover and only shoot film. But you can get *close* to the the Velvia picture above without using color matching, at least on the basis of a shot-by-shot comparison. I'm not going to post my results because everyone's definition of 'close' is different, and furthermore, I don't have a calibrated monitor here, which actually doesn't make much of a difference in this kind of comparison, but I'm not going to get into that argument with the people who just fall back on the, 'You don't have a calibrated monitor, thus you are a moron and I win.'</p>

<p>That being said, try converting to LAB and add a curves layer. In curves, boost the contrast on L with a nice 'S' curve. In the a layer, shift the midpoint to the right by about 7 units, and in b, shift it to the left by 8. That's in the ball park. I actually didn't use those exact adjustments in a and b, since I thought it needed a yellower sky, but bluer water, so my a and b curves were more complex (particularly a, since there was a lot of magenta in strange places that needed to be taken care of), but the idea still stands.</p>

<p>It can also be done in RGB, but I think it's more straightforward in LAB. In RGB, I added two curves layers. The first, set to Luminosity, added contrast with an 'S' curve. The second, set to color, had the R and G midpoints shifted to the left by about 10 (give or take) and a more complex B curve with a boost in the shadows and a big cut in the highlights and a flattened midtone region. Again, this gave a yellowish sky and blueish water, which is what I see in the Velvia version.</p>

<p>To clarify, I don't think it's that difficult to 'get in the ball park' with a direct comparison between shots. The last 10-20% is much harder to get at, and I think it would probably be difficult to get in the right neighborhood without the Velvia shot to work with. Personally, if I wanted the Velvia look, I'd just shot Velvia, which is what I think you would say too.</p>

<p>Long live film.</p>

[boinkphoto]

<blockquote>

<p>"I'm not going to post my results because everyone's definition of 'close' is different"</p>

</blockquote>

<p>I think that's the crux of the problem - I've seen plenty of Velvia-like attempts, but I don't think I've seen anything where I would mistake the image for actual Velvia.</p>

<p>In some sense that doesn't matter - what matters is you're satisfied with the result. In fact if you had a guaranteed Velvia-o-matic filter, you might find that it trashes a lot of your photos that you think you want Velvia for. Instead what you want may really be only <strong>some </strong>of the characteristics.</p>

<p>The only reason I want one (a filter myself) is because then at least I could feel less guilty about manipulating the photo. Velvia did the manipulation for you and using it you had no choice, so you could guiltlessly feel like you were playing within the parameters you were handed. Whereas doing PP I feel like I'm creating a false image. I'm not sure there's really a difference there, but hey, I'm neurotic.</p>

<p>Finally, as noted before, I don't think a true Velvia filter is possible outside of some serious engineering, probably requiring significant sampling and actual computer code. Yes, "in the ballpark" or at least something you can convince yourself is similar, but not the real deal. There are just too many factors - reciprocity, sensor response (which varies by camera and may literally lack the requisite data), ISO differences, white balance effects, etc. Certainly I don't see any possible "one size fits all" solution.</p>

[timothy_gray]

<p>Why attempt to re-create the wheel? If you like the look of Velvia, by all means, SHOOT VELVIA.<br>

I'd like to understand the OP's rationale here. Why the need to reproduce a look digitally you can readily attain with film?<br>

It seems to me to be counter-intuitive.</p>

[boinkphoto]

<p>I understand the OP, as I think do many others - film is more work and less "instant gratification". It's also very easy to mess up with little hope of recovery. Being able to get all the advantages of digital in combination with the response characteristics of Velvia is very enticing.</p>

<p>That said, I'm shooting film again because I think the quest while valiant is hopeless and RZ67s are dirt cheap these days. Plus the work it takes, has some advantages in focusing you to the task.</p>

<p>In the mean time, I shoot Fuji S5 and S2s because they are the next best things (I have a D300 too - which is great, but different). Still, the Fujis are not Velvia...</p>