[yusuf]

I was shooting today with my newly acquired M42 SMC Tak 105mm and noticed that histogram is consistently narrow compared to shooting with FA50/1.4 @f2 or Helios 44M-4 @f2.

When I took a few shot using FA50 or 44M @f2, the histogram is pretty rich, covering the entire spectrum. However, when I use SMC Tak 105mm, it's hardly covers 70% of the spectrum

This surprises me a bit because SMC Tak 105mm is pretty capable lens as I heard here. Also, I am shooting same subject under same environment.

None of the lenses uses hood. Also, it was a cloudy and overcast day so hood does not matter much.

Any ideas?

[bdery]

I believe since newer lenses pass along information to the camera, it's reasonable to assume they also pass along signal processing information. Or the camera knows how to handle specific lenses, maybe. What's sure is that the Tak does not pass anything along.

For the record, histograms with my 18-55 WR kit and my Sigma 17-70 are always different. The same is probably true with all my lenses, but I never checked toroughly. Pictures look different, too, straight from the camera.

[Lowell Goudge]

A cloudy day will result in a narrower histogram than a sunny day so please don't say it does not matter. Also, unless you are shooting the same scene with the same focal length the histogram will be different. Remember a histogram is only a graph of the distribution of intensity over the entire frame.

Aside from these points if the tak has a narrower histogram this indicates a higher contrast and suggests better overall lens coatings than a lens with lower contrast contrast is lost with internal lens reflections due to inferior coatings

[yusuf]

A cloudy day will result in a narrower histogram than a sunny day so please don't say it does not matter. Also, unless you are shooting the same scene with the same focal length the histogram will be different. Remember a histogram is only a graph of the distribution of intensity over the entire frame.

Aside from these points if the tak has a narrower histogram this indicates a higher contrast and suggests better overall lens coatings than a lens with lower contrast contrast is lost with internal lens reflections due to inferior coatings

Are you suggesting FA50mm has inferior coating than Tak? I beg to differ.

Also, my point was that all the lenses did not have hood. All, the pictures were shot in same environment and the same subject.

Please note, narrow histogram does not mean better contrast.

---

Last edited by yusuf; 08-25-2010 at 06:09 AM.

[Nesster]

Try them in manual exposure mode to remove the electronics variable.

It is true that different lenses will produce different histograms of the same scene, this is direct evidence of differing transfer functions. The widest possible histogram may seem desireable, but I'm not certain that's always a good thing... like Lowell says, ideally the histogram spreads out depending on EV range in the scene itself.

I've compared a 16-45 with a 43 ltd; initially the zoom looked to be contrastier when looking at the image itself, but checking the histogram the 43 actually had more overall contrast, i.e. the histogram was wider. Additionally, the detail resolution of the 43 was way better. I think this is one of the variables a lens designer has to play with - the contrast above the maximum resolution of the lens - to create an impression of sharpness and contrast. I.e. the 16-45 on its own produces spectacular photos... only in comparison with a truly superb lens such as the 43 does one notice its failings, in normal viewing. The 43 is quieter but when everything goes right, produces a superior image.

Someday when I have the chance, I'll check my SMC 105 vs a couple of others.

[yusuf]

Try them in manual exposure mode to remove the electronics variable.

It is true that different lenses will produce different histograms of the same scene, this is direct evidence of differing transfer functions. The widest possible histogram may seem desireable, but I'm not certain that's always a good thing... like Lowell says, ideally the histogram spreads out depending on EV range in the scene itself.

That's exactly the point, the scene is the same and hence the EV range. Somehow I feel that complete tonal range in not getting though 105mm Tak.

[Wheatfield]

Unless the pictures are exactly the same, expecting the histograms to be the same is not realistic. The 105 is a much longer lens than the 50, so things like sky values aren't going to be as prevalent, with the consequence that the histogram is narrower.
If the "tonal range" of the tak was blinded by 30% compared to your 50/1.4, the tak wouldn't work as a lens, so somehow I think your conclusion is wrong and you have missed some simple thing like what I mentioned above.

[Lowell Goudge]

Are you suggesting FA50mm has inferior coating than Tak? I beg to differ.

Also, my point was that all the lenses did not have hood. All, the pictures were shot in same environment and the same subject.

Please note, narrow histogram does not mean better contrast.

Actually yes. There was a thread here a while ago about the FA50 and how much it benefited in terms of contrast when used with a hood

As others have pointed out your test is not valid due to the different FOV of each lens. Try taking a series of shots of a block wall and then compare histograms to evaluate contrast

[yusuf]

Actually yes. There was a thread here a while ago about the FA50 and how much it benefited in terms of contrast when used with a hood

As others have pointed out your test is not valid due to the different FOV of each lens. Try taking a series of shots of a block wall and then compare histograms to evaluate contrast

I remember that discussion, that discussion was valid for f1.4 or f1.7. I am shooting at f2 and FA50 is pretty good at f2.

Moreover, you are leading to the case that narrow histogram is due to better contrast which IMO is not a correct assumption. Anyway, I will take some more shots and post it here.

[stevebrot]

This does not surprise me.

I have done comparisons between lenses of the same focal length of the same subject and gotten quite different histograms from each lens. This makes sense if you think about it a bit. How many times have you heard statements like:

• Lens A has much higher contrast than lens B
• Highlights really pop with lens A
• Lens B has excellent color rendition, particularly with <name color>

All of these "character" aspects of a particular lens will show up in the histogram. If you want to see a real eye-opening difference, put the camera in M mode, meter using a hand-held meter, and expose both lenses using the same settings!


Steve

[Nesster]

Moreover, you are leading to the case that narrow histogram is due to better contrast which IMO is not a correct assumption. Anyway, I will take some more shots and post it here.

I think of a narrow histogram as analogous to compression in music: it sounds louder though there actually is less dynamic range (defined as the difference in dB between the loudest and quietest parts of the music)

[Lowell Goudge]

Re histogram and contrast. I may have my logic backwards here between thinking of high contrast conditions and setting the jpeg values

In rethinking this, you set contrast low to compress the histogram with less vreyscale between stops.

Considering this a narrow histogram may actually be lower contrast

[Nesster]

An excellent article at luminous landscape talks about lens contrast:
Understanding Lens Contrast

But this topic goes to show how inadequate most published lens test data is, relative to real world performance. I mean, there is some correlation for sure re. perceived sharpness etc, but a lot more info can be gleaned.

[bimjo]

Unless the pictures are exactly the same, expecting the histograms to be the same is not realistic. The 105 is a much longer lens than the 50, so things like sky values aren't going to be as prevalent, with the consequence that the histogram is narrower.
If the "tonal range" of the tak was blinded by 30% compared to your 50/1.4, the tak wouldn't work as a lens, so somehow I think your conclusion is wrong and you have missed some simple thing like what I mentioned above.

Easily proved with a zoom lens. Longer focal length will have a compressed histogram, unless you're shooting just sky.

Focal length has the same effect on your light meter as well.

[georgweb]

Good read on this whole thread and a whole lot of things to consider.

Here's one more read about the subject, Manual lens on 5D -recommendations - FM Forums (and more stuff from same user on that forum)

Basically goes to show there's horses for courses - I am in the old-lens-on-digital camp and so I am soaking it all in .-)