How Do You Interpret Viewscan's B&W Histogram

[bwcolor]

I'm trying to decide what to do with my next roll and I have little experience to draw upon, so any help would be appreciated... no books yet.. but coming.

I am scanning 6x7 Arista EDU Ultra 100 @80/ Xtol 1:2 12 min @68 degrees. I agitated thirty seconds and then 15 sec of gentle agitation every two minutes. I locked film base exposure and color. The preview histograms in shot after shot have no information in the right-most 2 1/2 segments (stops?) and the data ends on the left with elevation. The subjects have a good range of brightness, extending beyond the film's ability to capture all values. I used my on camera meter (Mamiya 7) and limited the shots where there is no backlight and backgrounds are generally neutral, but some dark shadows. The grain structure, or lack thereof and tonality of these shots is very nice. Does this mean that I'm underexposing by 2 1/2 stops? If so, where do I go from here? I don't want to be shooting at such a low EI. Do I take the development time to 20 minutes, or so? I gather that this would capture more data on film, but images would show increased grain (no biggie in 6x7) and increased contrast. I could increase agitation and have slightly larger grain and capture more data, but I'm guessing that this would be a rather small change in the histogram??? Would increasing exposure and keeping development time the same decrease contrast?

 

[zenlibra]

From my experience using Vuescan I generally ignore the histogram. If you are shooting and developing your film "correctly" I wouldn't worry about it. I do the lock exposure routine and try to scan as flat as possible and use Lightroom with everything set to zero to work my images. That being said I am NOT an expert in film development so YMMV. You may know about this but if you do a search for "vuescan tutorial" there is a Flickr set that illustrates the best settings to use.

 

[bwcolor]

I just seems to me that I want data to fill the histogram when the subject contains a sufficient range of luminance. What I'm trying to do is get the correct exposure/development combination. That is what I don't have at this time.

 

[Bob Michaels]

I use the Vuescan histogram with every b&w neg I scan. I want to insure that neither end of the histogram is clipped and there is a smooth transition to either end.

If one end of the histogram is clipped significantly, and the thresholds are set at zero, that means there are either burnt out highlights or completely blocked up shadows on the negative. Given the large latitude of b&w film, that tells you that you have an exposure or development problem.

Other than that, I don't concern myself too much with the shape of the histogram or where it falls on the graph.

 

[bwcolor]

Yes, that makes sense to me, but when based on the above and based on a much greater experience in such things, where do I go from here?

 

[zenlibra]

I would trust the histogram in photoshop or other comparable software before vuescan. The histogram looks more spread out after I import into Lightroom. I'm not an expert so hopefully others will chime in.

 

[zenlibra]

Oops. I was a little slow to reply, thanks for the insight Bob. I'm interested in this topic also.

 

[bwcolor]

Yes, this is true, but it is also true that VueScan automatically selects the base of the curve on each side prior to scanning. I'm assuming that this is the only information captured.

 

[Bob Michaels]

I would trust the histogram in photoshop or other comparable software before vuescan. The histogram looks more spread out after I import into Lightroom. I'm not an expert so hopefully others will chime in.

the only thing the histogram in Vuescan tells me if that I have not clipped any data on the negative in the scan process. But I have found that is the only thing that matters in the scan process as I make all adjustments in my image editor.

Now the concept that the only thing that matters in the scan process is capturing all the data on the neg is a very broad statement but was meant to be exactly what it says.

 

[bwcolor]

If my curves in VueScan came to base on both sides, I would not be asking this question. I would rather throw data away intentionally in PS than lose it with exposure/development.

 

[Bob Michaels]

If my curves in VueScan came to base on both sides, I would not be asking this question. I would rather throw data away intentionally in PS than lose it with exposure/development.

If you have captured a scene with a five stop tonal range on film, it simply will not fill the histogram graph in Vuescan. In fact, I do not think it is possible to fill the Vuescan graph with any film. So long as the tonal range of the scene you are photographing fits along the flat part of the film's density / log exposure characteristic curve (see your film's data sheet) you are good.

I never worry if the Vuescan end points are set to leave large gaps on either side of the histogram. You subsequently adjust this via levels in your image editor anyway.

Personally, I tend to underdevelop slightly which gives a flatter negative. This allows me to place the contrast, via curves in Photoshop, in the midtones, highlights or shadows, where ever it works best to my eye.

 

[mfogiel]

bwcolor

OK, so let's get the things straight.

Do not lock film base or colour - select B&W negative, 16 bit grayscale scanning.

1 - under "colour" in the film type, select Kodak Tmax 400 , D76, CI 0.4
2- do the previev, then carefully select the image area, so that it DOES NOT contain anything else than the properly exposed film - so no clear film border, or even worse, filmless area
3- at this point the histogram should adjust itself to the proper brightness range, if it "overflows", redo the preview with the selected borders
4 - if the histogram still overflows, select instead of B&W negative the slide film


As to your negatives, if you look at them on a light table, you should be able to see through the darkest parts - if you do not, it means you have overdeveloped them - this is quite common as most manufacturers tend to give too long development times, particularly if you scan. If this is the case, next time develop 20-30% shorter.

 

[hlockwood]

I just seems to me that I want data to fill the histogram when the subject contains a sufficient range of luminance. What I'm trying to do is get the correct exposure/development combination. That is what I don't have at this time.

I can't comment on your film development, but that kind of histogram would send me a message. First, crop to be certain that only the exposed image is being sampled. If a histogram in NikonScan software shows a gap on the right, it can be nudged rightward by increasing the "Analog Gain" which is the exposure. Does Vuescan allow manual adjustment of exposure?

Harry

 

[bwcolor]

Locking the exposure and establishing a base contribution, which can then be subtracted, thus finding a true absolute black made a lot of sense to me and seemed like a good universal approach that can be used with any film. Color film, just requires establishing the white point in addition to the steps taken. I was using the presets prior to this.

That aside, I gather that you are saying that eyeballing and light box is the best bet for establishing a properly exposed negative. Upon doing this, I come to the same conclusion, but hard to quantify degree of underexposure. The darkest area of the negative is still thin and easily seen through.

I can adjust exposure, but did not based upon setting exposure through the exposure lock etc. in Vuescan based upon base density.

 

[Bob Michaels]

Locking the exposure and establishing a base contribution, which can then be subtracted, thus finding a true absolute black made a lot of sense to me and seemed like a good universal approach that can be used with any film. Color film, just requires establishing the white point in addition to the steps taken. I was using the presets prior to this.

That aside, I gather that you are saying that eyeballing and light box is the best bet for establishing a properly exposed negative. Upon doing this, I come to the same conclusion, but hard to quantify degree of underexposure. The darkest area of the negative is still thin and easily seen through.

I can adjust exposure, but did not based upon setting exposure through the exposure lock etc. in Vuescan based upon base density.

I have never done anything this involved in 8-9 years of using Vuescan on b&w negs. I tried early on and realized it got me no where. I just capture all of the data on the neg then adjust in Photoshop.

You have to remember that all those choices and adjustments in Vuescan are meant for someone who does not use an image editor and wants to make the file as good as possible straight from the scan. But those Vuescan adjustments are a poor comparison to what one can do with even the most basic image editor.

I guess I know what a properly exposed and developed neg looks like from experience. I don't think it is realistically possible to evaluate exposure and development from anything Vuescan tells you.

Just make sure you do not fall into the trap of believing you will get better results because your work flow is more complex.

 

[bwcolor]

Bob, we may be talking past each other. What I am discussing is a very very simple work flow. Establishing base density takes but a minute for the whole roll. What mfogiel suggested is also very simple, but gets more complicated as I try different emulsions. What I don't have is a properly exposed negative. I do believe that you are much more experienced than I at identifying good exposure. I will take mfogiel's advice on density and the lightbox and remember your warning regarding relying upon VueScan.

I setup my EI and development times based on reading here and through other sources on the Internet. My Xtol is a couple of months old, stored in a bottle with little oxygen, but perhaps it is partially oxidized and this is why my results vary from the expected. I think that I will try a roll with the same agitation, but extend development to twenty minutes and see what I end up with and go from there.

 

[Bob Michaels]

<snip>
I setup my EI and development times based on reading here and through other sources on the Internet. My Xtol is a couple of months old, stored in a bottle with little oxygen, but perhaps it is partially oxidized and this is why my results vary from the expected. I think that I will try a roll with the same agitation, but extend development to twenty minutes and see what I end up with and go from there.

I suggest you expose and develop a roll exactly as the film and developer manufacturer recommend. Download the data sheets from the manufacturers. It is rare that Kodak, Ilford, or Fuji is flat out wrong. It may turn out that you need some fine tuning down the road for optimum results, but make sure you are not re-inventing the wheel.

 

[bwcolor]

Bob:

I did start with Freestyle's PDF, but I didn't want to develop full strength in Xtol.

 

[Bob Michaels]

Bob:

I did start with Freestyle's PDF, but I didn't want to develop full strength in Xtol.

I hope I am not sounding harsh when I suggest you determine which film it really is (Arista Premium is Tri-X, Arista Legacy is Neopan 400) and then download that Tech Pub from Kodak or Fuji. Also download the Xtol tech pub from Kodak. There was a lot of scientific research that resulted in a lot of data included in those publications. They may be cookbook directions but they do work.

By all means, do what you would like to do. I just sense that you think Ed Hamrick has not figured out how Vuescan works or major film & developer manufacturers have not determined how to use their products. Certainly you may determine down the road that some tweaks to those old standards benefit you. But I suggest using those manufacturer standards as a baseline.

Have you made prints or JPGs, which ever is you final desired output? You should evaluate the interim processes such as exposure, development, scanning, or image editor adjustments as they impact the final result. Only when you evaluate the final result and determine if anything can be improved will you know if you need to change anything.

 

[bwcolor]

Bob:

Prior to doing anything, I downloaded Kodak, Arista and Fomapan Classic 100 pdfs. I looked at the Massive Development Chart and several other resources, including a number of photographic boards.

Here is one of the images..verticals through head and all, but it wasn't posed..she was playing with her chocolate: