Reading mode: Light Dark

lens noise vs sensor noise?

Go to last post
V

videonex

Member
Messages
16
Reaction score
1
I have a question about noise. I am hoping that some of you more technical or experienced people may shed some light. I have been under the assumption that noise comes from the sensor. In researching this issue online I have found that the vast majority of people insist that this is correct. I have found discussions where people go on for paragraphs about why noise is only a sensor issue and given two lenses with the same settings, you will get the same amount of noise--because you are increasing ISO, etc. This is sensible, but it does not match my recent experience.

I am shooting in a low light situation indoors with a Panasonic GH3 and the 25mm f1.4 getting great results. Out of curiosity for setting up a second camera, I did some simple tests comparing the 25mm with the old 14-42 kit lens. I shot some footage in Manual mode at f4 with both lenses. ISO at 1600, shutter speed 1/50th. Both shots are acceptable, but there is clearly some shimmering noise in the kit lens footage, most visible in the light colored wall. Now, I was expecting that the kit is not as sharp, clear, etc, but I did not expect noise, except noise due to the sensor. Since the 25mm gives no noise at the exact same settings, I have to assume that the difference is in the lens. Don't I?

Out of the several dozen opinions I have seen regarding this issue, only a few people mentioned types of noise that can come from the lens, while the vast majority insisted noise is a sensor issue only. I did find one forum (question asked by Phillip Bloom in 2005) where they were discussing noise and grain specific to particular lenses, which made me think that maybe I'm not crazy.

My question is: Is my experiment flawed? Am I missing something? Or are the majority of people simply uninformed about this issue because it only comes up in rare cases? Can you shed any light on how a cheaper lens makes this noise while a good lens at the same settings has none?

I appreciate any help in understanding this.
 
videonex said:
I have a question about noise. I am hoping that some of you more technical or experienced people may shed some light. I have been under the assumption that noise comes from the sensor. In researching this issue online I have found that the vast majority of people insist that this is correct. I have found discussions where people go on for paragraphs about why noise is only a sensor issue and given two lenses with the same settings, you will get the same amount of noise--because you are increasing ISO, etc. This is sensible, but it does not match my recent experience.

I am shooting in a low light situation indoors with a Panasonic GH3 and the 25mm f1.4 getting great results. Out of curiosity for setting up a second camera, I did some simple tests comparing the 25mm with the old 14-42 kit lens. I shot some footage in Manual mode at f4 with both lenses. ISO at 1600, shutter speed 1/50th. Both shots are acceptable, but there is clearly some shimmering noise in the kit lens footage, most visible in the light colored wall. Now, I was expecting that the kit is not as sharp, clear, etc, but I did not expect noise, except noise due to the sensor. Since the 25mm gives no noise at the exact same settings, I have to assume that the difference is in the lens. Don't I?
Click to expand...
Exact same settings. But the T-stop of these lenses are probably different. So for the same settings, different amount of light might be reaching the sensor.

IIRC I have read somewhere that some cameras, try to hide T-stop effects by manipulating ISO gain underneath. Otherwise, for the same setting, some lens will give brighter image, while some other lens will give darker image.

videonex said:
Out of the several dozen opinions I have seen regarding this issue, only a few people mentioned types of noise that can come from the lens, while the vast majority insisted noise is a sensor issue only. I did find one forum (question asked by Phillip Bloom in 2005) where they were discussing noise and grain specific to particular lenses, which made me think that maybe I'm not crazy.

My question is: Is my experiment flawed? Am I missing something? Or are the majority of people simply uninformed about this issue because it only comes up in rare cases? Can you shed any light on how a cheaper lens makes this noise while a good lens at the same settings has none?

I appreciate any help in understanding this.
Click to expand...
 
videonex said:
I have a question about noise. I am hoping that some of you more technical or experienced people may shed some light. I have been under the assumption that noise comes from the sensor. In researching this issue online I have found that the vast majority of people insist that this is correct. I have found discussions where people go on for paragraphs about why noise is only a sensor issue and given two lenses with the same settings, you will get the same amount of noise--because you are increasing ISO, etc. This is sensible, but it does not match my recent experience.

I am shooting in a low light situation indoors with a Panasonic GH3 and the 25mm f1.4 getting great results. Out of curiosity for setting up a second camera, I did some simple tests comparing the 25mm with the old 14-42 kit lens. I shot some footage in Manual mode at f4 with both lenses. ISO at 1600, shutter speed 1/50th. Both shots are acceptable, but there is clearly some shimmering noise in the kit lens footage, most visible in the light colored wall. Now, I was expecting that the kit is not as sharp, clear, etc, but I did not expect noise, except noise due to the sensor. Since the 25mm gives no noise at the exact same settings, I have to assume that the difference is in the lens. Don't I?

Out of the several dozen opinions I have seen regarding this issue, only a few people mentioned types of noise that can come from the lens, while the vast majority insisted noise is a sensor issue only. I did find one forum (question asked by Phillip Bloom in 2005) where they were discussing noise and grain specific to particular lenses, which made me think that maybe I'm not crazy.

My question is: Is my experiment flawed? Am I missing something? Or are the majority of people simply uninformed about this issue because it only comes up in rare cases? Can you shed any light on how a cheaper lens makes this noise while a good lens at the same settings has none?

I appreciate any help in understanding this.
Click to expand...
I suspect that you are agonizing over two entirely separate concepts, although it might just be the way you explain things.

When people talk about noise in digital photography, they mean the sum total of shot noise (the inherent randomness of photon noise), read noise (the noise created by the electronics that read the photon counts as individual charges and whatever other noise sources exist in the electronics chain. This all comes out as grain or mottling (chroma noise.)

You, on the other hand, are talking about something you call "shimmering noise" for which the only reference I can find is mosquito noise. That weird compression artifact that appears on edges. Is this what you mean?

If so, then one can speculate on all sorts of interesting causes ... for example, if the edge is a bit soft on one lens, then that lens might have more or less noise (I'm better on less for a softer edge.) And that would not shimmer at all. A lens that is biting sharp, on the other hand, might stimulate more compression artifacts. Of course, this might go the other way ... and that would support the more expensive lens having no shimmer.

Anyway .... perhaps that is the explanation.

The bottom line is that you have to decide which noise you are talking about. This shimmer is a compression artifact and has nothing to do with the sensor. If you were using a GH2 and patched it for much better video quality, you could probably get rid of shimmer entirely by reducing compression dramatically. The sensor noise is still there, but of course reducing compression improves the look of grain too ...
 
That's a good point ... but would T stops differ by more than 1/3 of a stop under normal circumstances? And would that be as visible as the OP is implying it is? From none to a visible shimmer?
 
You have us at a disadvantage here, as you have seen these two video clips, while everyone else has not.

Make the clips available for download, and folks will then be able to judge them and give you an opinion on what it is that you are seeing.

.
 
videonex said:
....

I am shooting in a low light situation indoors with a Panasonic GH3 and the 25mm f1.4 getting great results. Out of curiosity for setting up a second camera, I did some simple tests comparing the 25mm with the old 14-42 kit lens. I shot some footage in Manual mode at f4 with both lenses. ISO at 1600, shutter speed 1/50th. Both shots are acceptable, but there is clearly some shimmering noise in the kit lens footage, most visible in the light colored wall...
Click to expand...
There is a big difference between modern and old lenses in terms of their ability to transmit light. Modern good lenses lenses can usually transmit 85-90% of light. It means that if we assume that a lens is capable of 85% transmission then T2.8 will be around f2.6 (2.8*0.92 is about 2.6); T8 will be around f7.4.

The older lenses (and some of them are excellent in many respects) do not have such transmission capacity. Some of them, as I remember reading somewhere (so, take my words with a grain of salt) can transmit not more than 50%. This roughly equates to one stop. That's significant.

I do not know if this explains what you have observed in your case, but it's certainly something to look at.
 
Thank you all for your comments.

lancepring, I will try to post some examples soon.

I got an new 12-35mm f.28 lens yesterday and did the same test. I found it to be somewhere in between the other two, closer to the 25mm. Just a tiny bit of grainy movement. Sorry my description of the noise was not very thorough. I am not fluent with the various kinds of noise and what they are called. It looks like slight variations in color (chroma noise?) which would probably not be noticed at all except that they are slightly different in each frame, so appear to move or churn a bit. I guess grain might be a good description, but very fine grain.

Thanks again for your thoughts and I will try to get something to look at asap.
 
Here is a link to a comparison video.

Loading…

dl.dropboxusercontent.com

Since the original .mov files were huge, I brought them into Premiere into a 720 x 480 sequence, left them at 100% scale and chose a part of the frame where the noise was easiest to see--and saved it as an .mp4. I realize that a highly compressed mp4 file is not ideal for seeing subtle differences, but I think the difference between the two shots is fairly visible even so.

I am wondering if what we call chromatic abberation in a still photo looks like noise in video?
 
videonex said:
Here is a link to a comparison video.

https://dl.dropboxusercontent.com/u/3953932/noise.mp4

Since the original .mov files were huge, I brought them into Premiere into a 720 x 480 sequence, left them at 100% scale and chose a part of the frame where the noise was easiest to see--and saved it as an .mp4. I realize that a highly compressed mp4 file is not ideal for seeing subtle differences, but I think the difference between the two shots is fairly visible even so.

I am wondering if what we call chromatic abberation in a still photo looks like noise in video?
Click to expand...
Wow, I now see what you have been talking about. That is a dramatic difference.

I'm quite stumped. I don't know how just using a different lens could give such different results at the exact same settings. I would normally expect such a difference to be caused by using a much high ISO.
 
I think that there has to have been an error made somewhere here. I set my 14-42mm lens to 25mm, and the largest aperture that I can get is only 4.6, not 4.0

.
 
videonex said:
Here is a link to a comparison video.

https://dl.dropboxusercontent.com/u/3953932/noise.mp4
Click to expand...
The 14-42 must be being shot at a higher ISO and is therefore noiser. You must have made some mistake in shooting these and are not getting the ISO settings you think you are. The 25mm lens can shoot wider open and can therefore select a lower ISO. It's a bit tricky given the low resolution of the sample you posted, but it looks to me like there's less depth of field in the 25mm shot (the arm of the chair in the foreground looks blurrier to me and the camera is obviously hunting to shoot a focus point).
 
Yes, Sean, it would seem so. However, the reason I posted this question is because it is NOT so. That is why I did the experiment, and that is why I posted the video. The only way I can see that you could be right, is if I misunderstand how the GH3 works. Is it possible that in Manual mode, it could be automatically raising the ISO? I don't know, but that is the only way the ISO could be higher in one of the shots. The whole point of doing the experiment was to keep all things equal except for the change of lens, so it would be easy for anyone to duplicate this test and see if you get different results (assuming you have a couple of different lenses-one really good and one not as good). My reason for posting this question is because I assumed, just as many do, that it MUST be an ISO difference.
 
Don't worry too much about either kind of noise. There's a lot of tools which can help remove it completely.

I've seen DP's shy away from high ISOs due to the fear of noise.. only to underexpose the image and be shafted by a crappy histogram.

Noise is a solveable problem. Look here: http://www.adammyhill.com/color-grading

A
 
.... I would put substantial money on operator error. If I had any :-)
 
adamM said:
Don't worry too much about either kind of noise. There's a lot of tools which can help remove it completely.
Click to expand...
If it were actually "shimmering" as was originally named, it would be a crawling along edges that cannot be removed.

What it actually is (see the videos) is heavy chroma noise from very high ISO and yes, it can be removed ... but there is a cost that may not be worth paying ...
adamM said:
I've seen DP's shy away from high ISOs due to the fear of noise.. only to underexpose the image and be shafted by a crappy histogram.

Noise is a solveable problem. Look here: http://www.adammyhill.com/color-grading
Click to expand...
True ... but you missed the point of the thread. It's not about noise as an absolute ... it's about the relative noise between two identical settings on different lenses. Once you look at the example that was posted you will see that it has to have been some kind of operator error.
 
videonex said:
The only way I can see that you could be right, is if I misunderstand how the GH3 works. Is it possible that in Manual mode, it could be automatically raising the ISO?
Click to expand...
I'm not familiar enough with the GH3 to be able to answer your question. One thing I would check is to see if you can see the diaphragm of the 25mm lens stopped down while you're shooting to see if it's really at f/4. I don't know how the GH3 works in manual mode, but I can conceive that if the camera is set to f1.4 then it might automatically bump the ISO up when you mount a lens whose fastest aperture is over 3 stops slower than that.

One thing I can tell you is that what I see in that video is high-ISO noise. You'd have to work awfully hard to convince me that it was something else.
 
I agree it's not much of a problem. The obvious solution is use the better lens. I am just curious to understand a result that I didn't expect to see -- so unexpected that some don't believe it really happened. I wonder if anyone will be curious enough to try the experiment and see if they get different results.
 
[No message]
 
videonex said:
I agree it's not much of a problem. The obvious solution is use the better lens. I am just curious to understand a result that I didn't expect to see -- so unexpected that some don't believe it really happened. I wonder if anyone will be curious enough to try the experiment and see if they get different results.
Click to expand...
Sounds like a challenge ... well, ok then ... I will. I have the GH2 but that should only makes things worse :-)
 
You must log in or register to reply here.

Keyboard shortcuts

f
Forum
m
My threads
Mobile site
About
Editorial content
Cameras & Lenses
Community
All content, design, and layout are Copyright © 1998–2025 Digital Photography Review All Rights Reserved.
Reproduction in whole or part in any form or medium without specific written permission is prohibited.
When you use DPReview links to buy products, the site may earn a commission.
©GPS Media - Guides, Products, Services.
Back
Top