I have two similar macro lenses. Mamiya 120/4 and Canon 100/2.8 IS. Both go from infinity to 1:1 and FL are very similar (and BTW, so is IQ). There are, of course differences. The Mamiya gets very long when you go from infinity (125mm) to 1:1 (205mm) while the Canon has a constant (125mm) length thanks to IF design. And of course, the Mamiya does not offer AF or IS.
I knew all that before but what surprised me is that there is a very big difference WRT light loss. I mounted my camera (7D) on a tripod and pointed it to an evenly lit (daylight) wall which was about 2m away. I also set it to spot metering from the center, just to be on the safe side. I used Av and kept ISO and all other parameters were not changed as I wanted that the only difference in shutter speed will be from the lenses themselves. I then mounted both lenses (the Canon was set to MF and IS off, just to be on the safe side) and set them to infinity and 1:1 and half-clicked the shutter release button in order to activate the light meter. The Canon was checked at f/4 (to equalize things as much as possible) and wide open.
To say that I was amazed would be an understatement.
Mamiya (wide open) at infinity - 1/250. Mamiya (wide open) at 1:1 - 1/20. A whopping 3.5 stops of light loss.
Canon (f/4) at infinity - 1/125. Canon (f/4) at 1:1 - 1/45. A much more moderate 1.5 stops of light loss.
Canon (wide open) at infinity - 1/250. Canon (wide open) at 1:1 - 1/90. Same light loss as above. Not very surprising.
Other conclusions: At 1:1 the Canon enters twice as much light to the camera. At infinity it enters half.
From http://www.cambridgeincolour.com/tutorials/macro-lenses.htm I know that light loss should be about two stops but none seem to be there. The Canon is slightly better but the Mamiya is far worse. The other differences also puzzles me.
Thus my question is simple: Why do they differ so much?
TIA.
Hello Yakim,
I have not had this situation before, neither am I a macro-photographer. Howerver I think there is a simple explanation for the Mamiya lens.
Since the FL changes from 120 to 205, it acts like it has, "a 205/120= 1,7 teleconverter" buildt in, on 1:1. ln((1,7)^2)/ln(2)=1,5 stops (just like a 1,4 TC has 1 stop light loss and a 2,0TC has 2 stops light loss). So 1,5 plus the 2 stops from the Cambridgeincolor.com for macro makes it 3,5 stops.
Why the Canon only has 1,5 stops loss and not 2 stops? I do not know. Maybe something with the IF mechnism. Does anybody know?
Greetings,
Jeroen
[quote name='Yakim' date='22 July 2010 - 02:25 PM' timestamp='1279808741' post='1232']
I have two similar macro lenses. Mamiya 120/4 and Canon 100/2.8 IS. Both go from infinity to 1:1 and FL are very similar (and BTW, so is IQ). There are, of course differences. The Mamiya gets very long when you go from infinity (125mm) to 1:1 (205mm) while the Canon has a constant (125mm) length thanks to IF design. And of course, the Mamiya does not offer AF or IS.
I knew all that before but what surprised me is that there is a very big difference WRT light loss. I mounted my camera (7D) on a tripod and pointed it to an evenly lit (daylight) wall which was about 2m away. I also set it to spot metering from the center, just to be on the safe side. I used Av and kept ISO and all other parameters were not changed as I wanted that the only difference in shutter speed will be from the lenses themselves. I then mounted both lenses (the Canon was set to MF and IS off, just to be on the safe side) and set them to infinity and 1:1 and half-clicked the shutter release button in order to activate the light meter. The Canon was checked at f/4 (to equalize things as much as possible) and wide open.
To say that I was amazed would be an understatement.
Mamiya (wide open) at infinity - 1/250. Mamiya (wide open) at 1:1 - 1/20. A whopping 3.5 stops of light loss.
Canon (f/4) at infinity - 1/125. Canon (f/4) at 1:1 - 1/45. A much more moderate 1.5 stops of light loss.
Canon (wide open) at infinity - 1/250. Canon (wide open) at 1:1 - 1/90. Same light loss as above. Not very surprising.
Other conclusions: At 1:1 the Canon enters twice as much light to the camera. At infinity it enters half.
From http://www.cambridgeincolour.com/tutorials/macro-lenses.htm I know that light loss should be about two stops but none seem to be there. The Canon is slightly better but the Mamiya is far worse. The other differences also puzzles me.
Thus my question is simple: Why do they differ so much?
TIA.
[/quote]
[quote name='Yakim' date='22 July 2010 - 04:25 PM' timestamp='1279808741' post='1232']
I have two similar macro lenses. Mamiya 120/4 and Canon 100/2.8 IS. Both go from infinity to 1:1 and FL are very similar (and BTW, so is IQ). There are, of course differences. The Mamiya gets very long when you go from infinity (125mm) to 1:1 (205mm) while the Canon has a constant (125mm) length thanks to IF design. And of course, the Mamiya does not offer AF or IS.
I knew all that before but what surprised me is that there is a very big difference WRT light loss. I mounted my camera (7D) on a tripod and pointed it to an evenly lit (daylight) wall which was about 2m away. I also set it to spot metering from the center, just to be on the safe side. I used Av and kept ISO and all other parameters were not changed as I wanted that the only difference in shutter speed will be from the lenses themselves. I then mounted both lenses (the Canon was set to MF and IS off, just to be on the safe side) and set them to infinity and 1:1 and half-clicked the shutter release button in order to activate the light meter. The Canon was checked at f/4 (to equalize things as much as possible) and wide open.
To say that I was amazed would be an understatement.
Mamiya (wide open) at infinity - 1/250. Mamiya (wide open) at 1:1 - 1/20. A whopping 3.5 stops of light loss.
Canon (f/4) at infinity - 1/125. Canon (f/4) at 1:1 - 1/45. A much more moderate 1.5 stops of light loss.
Canon (wide open) at infinity - 1/250. Canon (wide open) at 1:1 - 1/90. Same light loss as above. Not very surprising.
Other conclusions: At 1:1 the Canon enters twice as much light to the camera. At infinity it enters half.
From [url="http://www.cambridgeincolour.com/tutorials/macro-lenses.htm"]http://www.cambridge...acro-lenses.htm[/url] I know that light loss should be about two stops but none seem to be there. The Canon is slightly better but the Mamiya is far worse. The other differences also puzzles me.
Thus my question is simple: Why do they differ so much?
TIA.
[/quote]
Have you taken any photos to confirm that the metered brightness and actual brightness of the image have anything to do with each other?
enjoy
[quote name='jkeijmel' date='22 July 2010 - 04:13 PM' timestamp='1279811629' post='1236']
Since the FL changes from 120 to 205[/quote]
1. These are physical lengths, not FL.
2. The link states that light loss should always be 2 stops. It mentions nothing about lens design (IF or non-IF) so I conclude it is not a factor.
[quote name='joachim' date='22 July 2010 - 04:23 PM' timestamp='1279812227' post='1238']
Have you taken any photos to confirm that the metered brightness and actual brightness of the image have anything to do with each other?[/quote]
No, but I don't think it'll make any difference. In my previous experience I get the same exposure in the pic itself, regardless of the lens mounted.
Hi Jeroen,
[quote name='jkeijmel' date='22 July 2010 - 05:13 PM' timestamp='1279811629' post='1236']
Hello Yakim,
I have not had this situation before, neither am I a macro-photographer. Howerver I think there is a simple explanation for the Mamiya lens.
Since the FL changes from 120 to 205, it acts like it has, "a 205/120= 1,7 teleconverter" buildt in, on 1:1. ln((1,7)^2)/ln(2)=1,5 stops (just like a 1,4 TC has 1 stop light loss and a 2,0TC has 2 stops light loss). So 1,5 plus the 2 stops from the Cambridgeincolor.com for macro makes it 3,5 stops.
Why the Canon only has 1,5 stops loss and not 2 stops? I do not know. Maybe something with the IF mechnism. Does anybody know?
Greetings,
Jeroen[/quote]
The FL of the Mamiya doesn't change, the tube length and distance of the optical rear node to the sensor does. Furthermore, at 1:1 a 120 mm lens will have an extension of 240 mm to the rear optical node and 240 mm from the frontal node to the subject, under normal circumstances. You have to add the register distance of mount to sensor to the extension of the lens, IOW.
Neither does it have a teleconverter built in.
Any lens at 1:1 loses two stops of light, no matter what the FL is. That is a given, BTW. It is exactly that, no more, no less.
So how the differences between these two lenses?
First of all, the Canon has IF and shortens the FL at 1:1 to approximately 70 mm, no longer 100 mm IOW, and this is a contimous change from infinity to 1:1. This means it captures a lot more of the environmental light in the background than the Mamiya does, and that light may include a larger chunk of bright(er) areas. Spot metering with the 7D is still 4 percent or thereabouts.
Secondly, the Mamiya must be closed down manually for metering, while the Canon meters wide open. Here alone you will likely get a difference.
Thirdly metering with a manual lens, especially when stopping down, is not as accurate as with a EF lens. This may not be a large influence, if at all in case of using the lens wide open, however.
Then there is the issue of daylight and testign in an essentially non-controlled environemnt. This needs to be tested a bit more rigorously IMO, with a fixed amount of light rather than with day light, IOW, indoors with an evenly illuminated wall either with flash or some fixed lighting.
Lastly, I assume the Mamiya is an older lens, with coating not up to digital standards, and with more transmission losses due to lesser coating than the Canon. It may just have a rear element that is not coated at the back, the camera side, which will cause reflections. Furthermore, it may not be as transparant as the Canon lens due to lesser coatings, and suffer from some more reflections internally. These reflections may mislead the metering system. By focusing close, the reflections seen by the metering system may actually get considerably less, because the lens tube gets a lot longer, whiel at normal focusing distances this lens may well underexpose. I reckon the Canon macro and body combination is spot when metering.
All in all I think it is a combination of things at play.
Kind regards, Wim
Gear: Canon EOS R with 3 primes and 2 zooms, 4 EF-R adapters, Canon EOS 5 (analog), 9 Canon EF primes, a lone Canon EF zoom, 2 extenders, 2 converters, tubes; Olympus OM-D 1 Mk II & Pen F with 12 primes, 6 zooms, and 3 Metabones EF-MFT adapters ....
[quote name='Yakim' date='22 July 2010 - 05:28 PM' timestamp='1279812483' post='1240']
No, but I don't think it'll make any difference. In my previous experience I get the same exposure in the pic itself, regardless of the lens mounted.
[/quote]
Hi Yakim,
I really do think you need to. I honestly do reckon that wide open the Mamiya underexposes by about a stop, if the brightness levels were exactly the same when either lens was mounted.
I also think you need to test more rigorously to see what is really at play, which means you really have to create a setup indoors with controllable fixed lighting or flashes.
The Canon may be slightly optimistic when stopped down a little, and the Mamiya probably underexposes by a full stop, especially wide open, and likely suffers from both reflections and light loss due to lesser coatings. And that if the light levels realy stayed the same throughout yoru test.
If you take that first stop away, it is only 2.5 stops, and that is quite reasonable. Light measurement generally is no more acurate than about 1/3 of a stop anyway. I also think you need to compare all metering modes, not just spot metering; it is more easily misled than more integral metering modes.
Finally, you should also try this at different f-stops, not just F/4, i.e., wide open for the Mamiya. I think that is the only way to find out how the Mamiya behaves, and find possible causes. And you have to take pictures, of course, and check and compare those.
Kind regards, Wim
Gear: Canon EOS R with 3 primes and 2 zooms, 4 EF-R adapters, Canon EOS 5 (analog), 9 Canon EF primes, a lone Canon EF zoom, 2 extenders, 2 converters, tubes; Olympus OM-D 1 Mk II & Pen F with 12 primes, 6 zooms, and 3 Metabones EF-MFT adapters ....
[quote name='wim' date='22 July 2010 - 04:44 PM' timestamp='1279813441' post='1241']
Any lens at 1:1 loses two stops of light, no matter what the FL is. That is a given, BTW. It is exactly that, no more, no less.[/quote]
So why does my test not corroborating this?
[quote name='wim' date='22 July 2010 - 04:44 PM' timestamp='1279813441' post='1241']
First of all, the Canon has IF and shortens the FL at 1:1 to approximately 70 mm, no longer 100 mm IOW, and this is a contimous change from infinity to 1:1. This means it captures a lot more of the environmental light in the background than the Mamiya does, and that light may include a larger chunk of bright(er) areas. Spot metering with the 7D is still 4 percent or thereabouts.[/quote]
1. 7D spot metering is 2.3%.
2. The link mentions nothing about lens design (IF or non-IF) so I conclude it is not a factor.
[quote name='wim' date='22 July 2010 - 04:44 PM' timestamp='1279813441' post='1241']
Secondly, the Mamiya must be closed down manually for metering, while the Canon meters wide open. Here alone you will likely get a difference.
Thirdly metering with a manual lens, especially when stopping down, is not as accurate as with a EF lens. This may not be a large influence, if at all in case of using the lens wide open, however.[/quote]
Exactly. The Mamiya was metered wide open. I did not change the aperure in it.
[quote name='wim' date='22 July 2010 - 04:44 PM' timestamp='1279813441' post='1241']
Then there is the issue of daylight and testign in an essentially non-controlled environemnt. This needs to be tested a bit more rigorously IMO, with a fixed amount of light rather than with day light, IOW, indoors with an evenly illuminated wall either with flash or some fixed lighting.[/quote]
It took me a minute or so to change lenses and meter. Light levels did not change. I was so amazed that I did it again several minutes later. Same results.
[quote name='Yakim' date='22 July 2010 - 06:00 PM' timestamp='1279814405' post='1243']
So why does my test not corroborating this?
[/quote]
There may be a bunch of factors causing this. One of the reasons why I suggest you set up a completely controllable environment, and do take photographs as well.
Mathematically there si exactly 2 stops between infinity and 1:1, provided everything else stays the same. Since it doesn't, soemthing else must be causing this effect.
Quote:1. 7D spot metering is 2.3%.
2. The link mentions nothing about lens design (IF or non-IF) so I conclude it is not a factor.
Well, I am sorry but the AOV certainly can have an influence on this. If you have a dark object in the centre, and a bright background, that may make a difference. Or the other way around for that matter. The effect of perspective may play a role, because the images aren't exactly the same. Especially not at 1:1 (Canon: FL 70 mm, Mamiya: still 120 mm). And you are comparing two different lenses here, and one of those on a camera it doesn 't really belong on. There could be many factors IOW which could affect this, and it likely does.
The problem is I don't really know, because I don't know your setup at all, of the testing grounds so to speak.
Quote:Exactly. The Mamiya was metered wide open. I did not change the aperure in it.
It took me a minute or so to change lenses and meter. Light levels did not change. I was so amazed that I did it again several minutes later. Same results.
Ok, something is at play, and the only way to find out is make sure all variables are eliminated except one. Which means setting up a testing environment indoors, with controlled lighting, test all metering modes, too, and take photographs.
I have used many MF lenses, and they all metered differently, at different f-stops (closing the lens down manually). After a while you get to know the lenses, so I didn't really care too much about this. I do know. I also know that my C/Y 50 F/1.4 overexposes by more than 1 stop wide open on my Canon bodies, and is accurate from F/2.8, +/- 1/3 of an f-stop. What causes that I don't know. Personally, I took it as a given. Similarly, my Leitz-R 60 Macro. It underexposes at any f-stop, from about a 1/2 wide open to 1/3 at the others.
Kind regards, Wim
Gear: Canon EOS R with 3 primes and 2 zooms, 4 EF-R adapters, Canon EOS 5 (analog), 9 Canon EF primes, a lone Canon EF zoom, 2 extenders, 2 converters, tubes; Olympus OM-D 1 Mk II & Pen F with 12 primes, 6 zooms, and 3 Metabones EF-MFT adapters ....
[quote name='wim' date='22 July 2010 - 05:34 PM' timestamp='1279816442' post='1245']
Ok, something is at play, and the only way to find out is make sure all variables are eliminated except one. Which means setting up a testing environment indoors, with controlled lighting, test all metering modes, too, and take photographs.[/quote]
I do not have the facilities and resources to do that.
[quote name='wim' date='22 July 2010 - 05:34 PM' timestamp='1279816442' post='1245']
I have used many MF lenses, and they all metered differently, at different f-stops (closing the lens down manually). After a while you get to know the lenses, so I didn't really care too much about this. I do know. I also know that my C/Y 50 F/1.4 overexposes by more than 1 stop wide open on my Canon bodies, and is accurate from F/2.8, +/- 1/3 of an f-stop. What causes that I don't know. Personally, I took it as a given. Similarly, my Leitz-R 60 Macro. It underexposes at any f-stop, from about a 1/2 wide open to 1/3 at the others.[/quote]
I must admit this is the most amazing thing I have ever heard. I never thought things like this exist with such magnitude. Thank you for your insight. Nevertheless, when shooting with the Mamiya (i.e. actually taking pictures, not testing) I never had to apply any EC. I always assumed it's because the camera meters TTL and does not care which lens is attached.
|