P4 versus P4P low light test

[Jase]

Hi guys

Was just testing the P4 against the P4P. In general I'm impressed with the P4P but I thought its real strength would be shooting in low light. I took these 2 images just before dark about a minute apart, one with the P4 and the other with the P4P. The 1st one is with the P4P which you can also see has a narrower FOV. Both were taken with an ISO of 1600 and an F-stop of 2.8 (although the P4 you cant control the aperture thats what it shows in the camera details) both with the exposure set at zero. For some reason the exposure time was 1/4 for the P4 and 1/15 for the P4P (Not sure if this exposure time difference has something to do with different focal lengths, the P4 being 20mm and the P4P being 24mm)

My question is why does the P4P have a lot more grain, if the professionals can please help me
Thanks
Jase

 

[Tomas Wangen]

When testing image quality it is important do set all parameters the same. The 1/15 shutter speed at ISO 1600 has same effect as ISO 6400 (assuming you are lifting exposure two stops in post) so you are essentially comparing apples to oranges here.

Try manual exposure for both with same ISO, same shutter speed and same aperture. P4p still has noise, but it should compare favorably to p4. Additionally, if light is changing very fast, taking the exposure at the same time is also helpful.

Edit: I do see that in this case the missing noise in the P4 capture is actually due to "smudging" where the details (and with that the noise) is essentially gone. Aggressive noise reduction on the p4p image will have some of the same effect, but will allow for more detail to be kept.

 

[Doctorcobani]

I see both pictures very blurry (just like a moved camera with a low shutter speed) and not in focus. I see both pictures at ISO 1600, but the JPG from the P4 has a very agressive noise reduction applied (you can see the strong lack of detail, regardless P4P has more detail because of the sensor).

Compare both pictures with manual exposure, same settings on both (remember tap to focus in P4P) and compare the DNG (RAW) to see both pictures processed in the same way...

 

[Spike_151]

 

[gr8pics]

The 1/15 shutter speed at ISO 1600 has same effect as ISO 6400

How did you come to that conclusion?
Normally a variation of the shutter speed is due to the lens brightness, more glass, more light, less glass, less light, thats why the P4P shutter speed is faster, cause the lens is brighter and therefore chooses a faster shutter speed to achieve the same exposure, in theory.

 

[Tomas Wangen]

How did you come to that conclusion?
Normally a variation of the shutter speed is due to the lens brightness, more glass, more light, less glass, less light, thats why the P4P shutter speed is faster, cause the lens is brighter and therefore chooses a faster shutter speed to achieve the same exposure, in theory.

Actually this is not correct. The reason we can use a external light meter and still get determinable results, is that the aperture along with the shutter speed and iso determines the exposure regardless of glass and camera as long as the calibration is reasonably good. This is true for large format cameras and mobile cameras and they certainly do not share any components.

I.e. f2.8 is f2.8

 

[gr8pics]

Actually this is not correct. The reason we can use a external light meter and still get determinable results, is that the aperture along with the shutter speed and iso determines the exposure regardless of glass and camera as long as the calibration is reasonably good. This is true for large format cameras and mobile cameras and they certainly do not share any components.

I.e. f2.8 is f2.8

I didnt understand your reply.
A f4 lens requires longer shutter speeds to achieve the same exposure as a f1.8 lens, thats the whole idea of brighter lenses. Unless you compensate with iso, and in dark environments a large aperture is your friend, not high iso, at least in most consumer cameras.
If aperture have been set to maximum, and iso is the same, then shutter speed is the only factor left you can compensate light with.

May i ask what background you have from photography? Are you a full time working professional, or just for fun and hobby? Im just curious, so i can better understand where all your knowledge comes from, cause i have problem understanding some of your statements and theories in here, that doesnt make any sense...
Im not trying to offend you, but understand you

 

[Tomas Wangen]

I didnt understand your reply.
A f4 lens requires longer shutter speeds to achieve the same exposure as a f1.8 lens, thats the whole idea of brighter lenses. Unless you compensate with iso, and in dark environments a large aperture is your friend, not high iso, at least in most consumer cameras.
If aperture have been set to maximum, and iso is the same, then shutter speed is the only factor left you can compensate light with.

May i ask what background you have from photography? Are you a full time working professional, or just for fun and hobby? Im just curious, so i can better understand where all your knowledge comes from, cause i have problem understanding some of your statements and theories in here, that doesnt make any sense...
Im not trying to offend you, but understand you

What are you talking about?

The OP said: "Both were taken with an ISO of 1600 and an F-stop of 2.8"

There was a two stop difference in shutterspeed which was what I commented.

If you have anything worthwhile to add, please do, but if not please leave it at that.

Btw, I see now that you removed my original qualifying text when you first quoted me, leaving a seemingly false statement. I will therefore repeat the text with qualifier present:

The 1/15 shutter speed at ISO 1600 has same effect as ISO 6400 (assuming you are lifting exposure two stops in post)

 

[gr8pics]

What are you talking about?

The OP said: "Both were taken with an ISO of 1600 and an F-stop of 2.8"

There was a two stop difference in shutterspeed which was what I commented.

If you have anything worthwhile to add, please do, but if not please leave it at that.

Btw, I see now that you removed my original qualifying text when you first quoted me, leaving a seemingly false statement. I will therefore repeat the text with qualifier present:

The 1/15 shutter speed at ISO 1600 has same effect as ISO 6400 (assuming you are lifting exposure two stops in post)

I asked you a simple question, cause your information in here sounds like a mixup of allot of different things you might only read theories about, but okay, if you dont want to share your background, thats fine.

The OP asked, why does P4P have more grain/noise than P4? He told us both pics where captured at f2.8 and iso 1600.
What has a theoretical iso compensation to do with the noise he asks about when both pics where capture with same iso and aperture?
I also assume his pics havent been retouched, but are straight out of the camera, maybe OP can clarify that?

So again, your theories about how much you need to raise the iso to compensate for 7 steps of shutter speed, is irrelevant to his question; why P4P have more noise than the P4, cause there will not be more noise from shutter 1/15 to 1/4

You also said:
"When testing image quality it is important do set all parameters the same, so you are essentially comparing apples to oranges here"

He already did, the difference in shutter speeds here will not make ANY difference in quality in terms of noise, so saying he compares apples to oranges is just nonsense.

Hes got a good point, noise in P4P should be less than the ones from P4, for obvious reasons.

My suggestion is, that he do a new test in raw format, as i assume this is jpeg since noise reduction have been applied, then we can determine the real differences, shutter speed does not introduce noise, unless you keep the shutter/sensor open for 30 secs or longer, then noise will appear due to heating of the sensor, the colder the sensor is, the less noise it will produce, hes also outside and its winter, so hot pixels will not be of any issue here ...

 

[Tomas Wangen]

cause your information in here sounds like a mixup of allot of different things you might only read theories about

Please clarify what I am mixing up. I challenge you to find any fault with reference to a reputable source.

If you insist on protesting everything I say for no apparent reason, you have to back it up somehow.

 

[gr8pics]

Please clarify what I am mixing up. I challenge you to find any fault with reference to a reputable source.

If you insist on protesting everything I say for no apparent reason, you have to back it up somehow.

I already did, see my previous post...
I only disagree when you post information that is misleading and/or confusing, and in this case, talking about how much iso equals 7 steps of shutter speed makes no sense in terms of OP concerns, if it does, then i dont understand your rhetoric.
He took two pictures with virtually the same settings, only thing that separates them is the focal length.

 

[Tomas Wangen]

I already did, see my previous post...
I only disagree when you post information that is misleading and/or confusing, and in this case, talking about how much iso equals 7 steps of shutter speed makes no sense in terms of OP concerns, if it does, then i dont understand your rhetoric.
He took two pictures with virtually the same settings, only thing that separates them is the focal length.

First, to OP, I am very sorry this went way off-topic.

I will not continue this thread now, but as I originally said - quote - "When testing image quality it is important do set all parameters the same" and a two-stop difference in exposure is not - quote - "virtually the same settings".

I did also say, again in the original comment, that after reviewing the images, it was apparent that noise-cancellation in the p4 image was the main reason for the difference in noise for this particular sample.

 

[Jase]

Thanks guys. Just the info I was hoping for.

I didnt realise that the P4 had noise reduction processed in the Jpg, that alone could be a big reason. I think I have not done a good enough job but was trying to get some quick information. I held hand both drones for the pics as I thought the stabilisation would be good enough and also because in the real world the drone will be hovering and I cant imagine that held holding would be much worse. But I guess things are in much closer proximity in these pictures than would be while hovering higher to do night photography which doesnt help. Regardless its probably not a good enough test. I did focus the P4P on an object about 2/3 away of the main focus area. I think the blurring is mainly from slow shutter and being hand held.

In terms of the replies, yes I was trying to see why at the same lighting, iso and f stop the noise seemed greater. As it might be that the P4 has had noise reduction I guess its not worth discussing further with these examples. As the sensor heat topic came up, where I stay, we are in the middle of our summer season in South Africa so its not cold at all

Would it be better if I did it again to rather keep the f-stop and exposure time the same and reduce the iso 2 stops as the sensor seems to be getting more light or more sensitive? This way I would be taking advantage of a lower iso in similar lighting conditions?

Cheers
Jase

 

[Tomas Wangen]

Would it be better if I did it again to rather keep the f-stop and exposure time the same and reduce the iso 2 stops as the sensor seems to be getting more light or more sensitive? This way I would be taking advantage of a lower iso in similar lighting conditions?
Cheers
Jase

Hi, to compare the performance of each sensor, it would be best (most fair) to have the exact same parameters for both cameras.

This means shooting in manual and (almost) at the same time and ofcourse the same scene.

To get a correct exposure, you can use one of the drones in a mode (at f2.8 and chosen iso) and then simply copying the suggested shutter value over to the (m)anual mode, for both drones.

For low-light shots it makes most sense to choose the 1600 iso you already did (as both drones support this).

(both will get the best performance at iso 100, but that is not very practical in low-light scenarios)

This will result in two equal exposures of for instance ISO 1600, f/2.8, 1/8 second which can then be directly compared.

To avoid the very different noise-reduction you saw in the last test, using raw mode will ensure that no initial noise-reduction is applied and therefore give directly comparable images.

Good luck!