Win a Pentax Limited Lens: Raffle Now Live!

TGregory · 03-12-2009, 02:54 PM

This may have been discussed before but I could not find it. I seem to be having focus issues with the Sigma 30mm 1.4 lens. It seems to be a bigger problem closer to wide open in lower light.

Has anyone had similar experience and if so, what was your solution.

Thanks

RT1 · 03-12-2009, 03:13 PM

I had this problem with my Sigma 24-60 f2.8, but my 30mm 1.4 has been spot on from day one. It's easily my favorite lens.

I sent the 24-60 in to Sigma USA and they re calibrated it and it's been fantastic ever since.

Is your 30mm searching a lot, or are you getting blurred shots? If it's the later, you do have to remember that this lens has a very thin depth of field when wide open.

dlacouture · 03-13-2009, 12:59 AM

When you say "in lower light", do you mean indoor, by tungsten?
If this is the case, you'll have to know that tungsten lighting offsets the AF system, and results in front-focusing. This has to do with the light spectrum shifting toward the reds, and as the AF module is calibrated for green, oups!

If you are familiar with lenses, you'll see that some have a little red mark under the distance scale, slightly off the main mark. See below, right between 4 and 11.

This is the focus indicator for infrared, and shows quite prettily the impact of wavelength on focus.

Anyway, a body focusing perfectly outdoor will be a little off indoor.

Quicksand · 03-13-2009, 10:35 AM

Originally posted by dlacouture

When you say "in lower light", do you mean indoor, by tungsten?
If this is the case, you'll have to know that tungsten lighting offsets the AF system, and results in front-focusing. This has to do with the light spectrum shifting toward the reds, and as the AF module is calibrated for green, oups!

My FA50/1.4 shows exactly the same behavior. Outdoors, I can autofocus and shoot wide-open with no problems. Indoors, I have to dial in -9 in focus compensation on my K20D. Stopped down to f/2.8+, there's enough DOF that it doesn't make much difference either way.

I have mostly gotten used to it at this point, though I wish there were an easier way to switch among focus compensation values without having to drill into the "Custom" menu every time. I'd be especially happy if it automatically compensated based on white balance, but anyway . . .

**StephenMerola** · 03-13-2009, 10:37 AM

Originally posted by dlacouture

When you say "in lower light", do you mean indoor, by tungsten?
If this is the case, you'll have to know that tungsten lighting offsets the AF system, and results in front-focusing. This has to do with the light spectrum shifting toward the reds, and as the AF module is calibrated for green, oups!

If you are familiar with lenses, you'll see that some have a little red mark under the distance scale, slightly off the main mark. See below, right between 4 and 11.

This is the focus indicator for infrared, and shows quite prettily the impact of wavelength on focus.

Anyway, a body focusing perfectly outdoor will be a little off indoor.

So what is the proper lighting for testing focus indoors?

Gooshin · 03-13-2009, 10:38 AM

get a daylight balanced bulb.

TGregory · 03-13-2009, 10:41 AM

I did not know that about tungsten light but I did some test with a tripod to make sure there was no camera movement and you are correct.

I went back to some other shots I had taken on an airplane and they are perfectly focused. I do not know what type of light is on an airplane but it seems to make a difference on the AF.

Thanks for the educations.

G

**Digitalis** · 03-13-2009, 06:57 PM

"If this is the case, you'll have to know that tungsten lighting offsets the AF system, and results in front-focusing. This has to do with the light spectrum shifting toward the reds, and as the AF module is calibrated for green"

I thought AF sensors only saw in monochrome, interesting hypothesis.

**Digitalis** · 03-13-2009, 07:26 PM

well dlacouture, I tested your hypothesis. and it appears that you are wrong. I tested it using my Pentax 16-45mm f/4 ED AL ( I removed the polariser I had on it, they mess with images displayed on LCD's) using this PSD file which was displayed on my 28" LCD screen. The PSD file has three layers, one high contrast pattern that any lens can lock on to, and a colour pattern that uses colours that a monochrome sensor wouldn't be able to see(it would just see uniform grey)...if the sensor actually saw colour then it would AF on the colour pattern no problems. But mine couldn't achieve a focus lock...the last pattern is what a monochrome sensor would see...it's simply a desaturated version of the color patern which I used as a control to see if the lens was focusing on the pixels on my screen.

the PSD file I used for this test is in this zip file.

stevebrot · 03-13-2009, 07:42 PM

Originally posted by Digitalis

"If this is the case, you'll have to know that tungsten lighting offsets the AF system, and results in front-focusing. This has to do with the light spectrum shifting toward the reds, and as the AF module is calibrated for green"

I thought AF sensors only saw in monochrome, interesting hypothesis.

Ummm...is that red monochrome or green monochrome...

Think about it.

Steve

**Digitalis** · 03-13-2009, 07:50 PM

Originally posted by stevebrot

Ummm...is that red monochrome or green monochrome...

Think about it.

Steve

well if you looked at how I tested it...the sensor is unfiltered.

stevebrot · 03-13-2009, 07:52 PM

Originally posted by stevebrot

Ummm...is that red monochrome or green monochrome...

Think about it.

Steve

Sorry about the quick quip. In reality, I think the spectral balance vs. focus accuracy issue is probably a little more complex that a simple red/green scenario. One that most of us are unlikely to have good comprehension of unless we were an optical engineer of some sort.

Steve

stevebrot · 03-13-2009, 08:04 PM

Originally posted by Digitalis

well if you looked at how I tested it...the sensor is unfiltered.

I don't have Photoshop, though I think I understand what kind of images are on your layers.

There are two issues here that are unrelated. The first is the plane of focus for different wavelengths of light. Most modern lenses are pretty well corrected such that the visible spectra converge pretty much on the same plane. The second is the spectral sensitivity of the AF sensor. My expectation is that if the sensor was most sensitive to green light and the subject illumination was primarily on the red end of the spectrum, precision would suffer rather than skewing the accuracy. (The scene would "appear" more "dim" with inherit lower contrast.)

Now since I truly don't know what I am talking about, I am going to shut up now!

Steve

**Digitalis** · 03-13-2009, 08:20 PM

To the best of my knowledge there aren't any lenses below 100mm that are truly Apochromatic (which means that optically all three wavelengths of light are brought into focus at exactly the same plane) But the colour shifts with non-APO lenses aren't as obvious at normal focus distances..It's only when you're focusing closely at wide apertures when those colour shifts can inhibit maximum sharpness. Generally, lenses with Large apertures seem to suffer from it more than lenses with smaller apertures.

first image is of the Voigtlander Lanthar 125mm f/2.5 APO, second image is from the sigma 50mm f/1.4 - to demonstrate the difference from APO lens designs against non-APO lenses- note the red/cyan fringing in the out of focus areas on the sigma.

dantekgeek · 03-13-2009, 09:53 PM

I clicked through this thread expecting a discussion of my very favorite lens, and wound up learning something about how light wavelength can effect autofocus! Much love PF!

03-12-2009, 02:54 PM	#1
TGregory Inactive Account Join Date: Oct 2008 Location: Austin, TX Photos: Gallery Posts: 42	Sigma 30mm 1.4 This may have been discussed before but I could not find it. I seem to be having focus issues with the Sigma 30mm 1.4 lens. It seems to be a bigger problem closer to wide open in lower light. Has anyone had similar experience and if so, what was your solution. Thanks

03-13-2009, 07:26 PM	#9
Digitalis Site Supporter Join Date: Mar 2009 Location: Melbourne, Victoria Photos: Gallery Posts: 11,695	well dlacouture, I tested your hypothesis. and it appears that you are wrong. I tested it using my Pentax 16-45mm f/4 ED AL ( I removed the polariser I had on it, they mess with images displayed on LCD's) using this PSD file which was displayed on my 28" LCD screen. The PSD file has three layers, one high contrast pattern that any lens can lock on to, and a colour pattern that uses colours that a monochrome sensor wouldn't be able to see(it would just see uniform grey)...if the sensor actually saw colour then it would AF on the colour pattern no problems. But mine couldn't achieve a focus lock...the last pattern is what a monochrome sensor would see...it's simply a desaturated version of the color patern which I used as a control to see if the lens was focusing on the pixels on my screen. the PSD file I used for this test is in this zip file. Last edited by Digitalis; 04-07-2009 at 06:05 PM.

03-13-2009, 08:20 PM	#14
Digitalis Site Supporter Join Date: Mar 2009 Location: Melbourne, Victoria Photos: Gallery Posts: 11,695	To the best of my knowledge there aren't any lenses below 100mm that are truly Apochromatic (which means that optically all three wavelengths of light are brought into focus at exactly the same plane) But the colour shifts with non-APO lenses aren't as obvious at normal focus distances..It's only when you're focusing closely at wide apertures when those colour shifts can inhibit maximum sharpness. Generally, lenses with Large apertures seem to suffer from it more than lenses with smaller apertures. first image is of the Voigtlander Lanthar 125mm f/2.5 APO, second image is from the sigma 50mm f/1.4 - to demonstrate the difference from APO lens designs against non-APO lenses- note the red/cyan fringing in the out of focus areas on the sigma. Last edited by Digitalis; 05-08-2009 at 12:31 AM.

Similar Threads
Thread	Thread Starter	Forum	Replies	Last Post
People A few with the Sigma 30mm f/1.4	palmor	Post Your Photos!	4	07-26-2010 12:23 AM
Sigma 28mm f/1.8 and Sigma 30mm f/1.4: 3 Questions	Biro	Pentax SLR Lens Discussion	0	03-06-2010 03:12 PM
sigma 30mm vs. pentax 30mm	jay legere	Pentax SLR Lens Discussion	6	02-01-2008 04:15 PM

03-12-2009, 03:13 PM	#2
RT1 Forum Member Join Date: Sep 2008 Location: Kansas City Photos: Gallery Posts: 85	I had this problem with my Sigma 24-60 f2.8, but my 30mm 1.4 has been spot on from day one. It's easily my favorite lens. I sent the 24-60 in to Sigma USA and they re calibrated it and it's been fantastic ever since. Is your 30mm searching a lot, or are you getting blurred shots? If it's the later, you do have to remember that this lens has a very thin depth of field when wide open.

03-13-2009, 12:59 AM	#3
dlacouture Veteran Member Join Date: Jul 2008 Location: Var, South of France Photos: Albums Posts: 1,074	When you say "in lower light", do you mean indoor, by tungsten? If this is the case, you'll have to know that tungsten lighting offsets the AF system, and results in front-focusing. This has to do with the light spectrum shifting toward the reds, and as the AF module is calibrated for green, oups! If you are familiar with lenses, you'll see that some have a little red mark under the distance scale, slightly off the main mark. See below, right between 4 and 11. This is the focus indicator for infrared, and shows quite prettily the impact of wavelength on focus. Anyway, a body focusing perfectly outdoor will be a little off indoor.

03-13-2009, 10:35 AM	#4
Quicksand Pentaxian Join Date: Apr 2008 Location: Central Ohio (formerly SF Bay Area) Posts: 1,519	Originally posted by dlacouture When you say "in lower light", do you mean indoor, by tungsten? If this is the case, you'll have to know that tungsten lighting offsets the AF system, and results in front-focusing. This has to do with the light spectrum shifting toward the reds, and as the AF module is calibrated for green, oups! My FA50/1.4 shows exactly the same behavior. Outdoors, I can autofocus and shoot wide-open with no problems. Indoors, I have to dial in -9 in focus compensation on my K20D. Stopped down to f/2.8+, there's enough DOF that it doesn't make much difference either way. I have mostly gotten used to it at this point, though I wish there were an easier way to switch among focus compensation values without having to drill into the "Custom" menu every time. I'd be especially happy if it automatically compensated based on white balance, but anyway . . .

03-13-2009, 10:37 AM	#5
StephenMerola Site Supporter Join Date: Feb 2009 Location: Wisconsin Photos: Albums Posts: 423	Originally posted by dlacouture When you say "in lower light", do you mean indoor, by tungsten? If this is the case, you'll have to know that tungsten lighting offsets the AF system, and results in front-focusing. This has to do with the light spectrum shifting toward the reds, and as the AF module is calibrated for green, oups! If you are familiar with lenses, you'll see that some have a little red mark under the distance scale, slightly off the main mark. See below, right between 4 and 11. This is the focus indicator for infrared, and shows quite prettily the impact of wavelength on focus. Anyway, a body focusing perfectly outdoor will be a little off indoor. So what is the proper lighting for testing focus indoors?

03-13-2009, 10:38 AM	#6
Gooshin Veteran Member Join Date: Sep 2007 Location: Toronto, the one in Canada. Posts: 5,610	get a daylight balanced bulb.

03-13-2009, 10:41 AM	#7
TGregory Inactive Account Join Date: Oct 2008 Location: Austin, TX Photos: Gallery Posts: 42 Original Poster	I did not know that about tungsten light but I did some test with a tripod to make sure there was no camera movement and you are correct. I went back to some other shots I had taken on an airplane and they are perfectly focused. I do not know what type of light is on an airplane but it seems to make a difference on the AF. Thanks for the educations. G

03-13-2009, 06:57 PM	#8
Digitalis Site Supporter Join Date: Mar 2009 Location: Melbourne, Victoria Photos: Gallery Posts: 11,695	"If this is the case, you'll have to know that tungsten lighting offsets the AF system, and results in front-focusing. This has to do with the light spectrum shifting toward the reds, and as the AF module is calibrated for green" I thought AF sensors only saw in monochrome, interesting hypothesis.

03-13-2009, 07:42 PM	#10
stevebrot Otis Memorial Pentaxian Join Date: Mar 2007 Location: Vancouver (USA) Photos: Gallery \| Albums Posts: 42,007	Originally posted by Digitalis "If this is the case, you'll have to know that tungsten lighting offsets the AF system, and results in front-focusing. This has to do with the light spectrum shifting toward the reds, and as the AF module is calibrated for green" I thought AF sensors only saw in monochrome, interesting hypothesis. Ummm...is that red monochrome or green monochrome... Think about it. Steve

03-13-2009, 07:50 PM	#11
Digitalis Site Supporter Join Date: Mar 2009 Location: Melbourne, Victoria Photos: Gallery Posts: 11,695	Originally posted by stevebrot Ummm...is that red monochrome or green monochrome... Think about it. Steve well if you looked at how I tested it...the sensor is unfiltered.

03-13-2009, 07:52 PM	#12
stevebrot Otis Memorial Pentaxian Join Date: Mar 2007 Location: Vancouver (USA) Photos: Gallery \| Albums Posts: 42,007	Originally posted by stevebrot Ummm...is that red monochrome or green monochrome... Think about it. Steve Sorry about the quick quip. In reality, I think the spectral balance vs. focus accuracy issue is probably a little more complex that a simple red/green scenario. One that most of us are unlikely to have good comprehension of unless we were an optical engineer of some sort. Steve

03-13-2009, 08:04 PM	#13
stevebrot Otis Memorial Pentaxian Join Date: Mar 2007 Location: Vancouver (USA) Photos: Gallery \| Albums Posts: 42,007	Originally posted by Digitalis well if you looked at how I tested it...the sensor is unfiltered. I don't have Photoshop, though I think I understand what kind of images are on your layers. There are two issues here that are unrelated. The first is the plane of focus for different wavelengths of light. Most modern lenses are pretty well corrected such that the visible spectra converge pretty much on the same plane. The second is the spectral sensitivity of the AF sensor. My expectation is that if the sensor was most sensitive to green light and the subject illumination was primarily on the red end of the spectrum, precision would suffer rather than skewing the accuracy. (The scene would "appear" more "dim" with inherit lower contrast.) Now since I truly don't know what I am talking about, I am going to shut up now! Steve

03-13-2009, 09:53 PM	#15
dantekgeek Senior Member Join Date: Oct 2008 Location: NYC Posts: 100	I clicked through this thread expecting a discussion of my very favorite lens, and wound up learning something about how light wavelength can effect autofocus! Much love PF!