street @ ISO 4000 (Fuji X-E1)

Petach

Hall of Famer
Location
UK, Essex
Name
Peter Tachauer
It was pretty grey, dismal and dim yesterday. I wanted to maintain 1/250s shutter with a reasonable f8 - f13 but without ramping up the ISO too much. The 1/250s at those f numbers and at iso 4000 produced underexposure by about 2 - 2.5 stops. Tweaking in LR4 with exposure seems to have done the trick and ISO looks quite useable to me.

8508873947_7697bbb529_b.jpg

street @ ISO 4000 XE1 by petach123 (Peter Tachauer), on Flickr

8509981274_25c41da3e8_b.jpg

street @ ISO 4000 XE1 by petach123 (Peter Tachauer), on Flickr

8509981052_96d2363a3b_b.jpg

street @ ISO 4000 XE1 by petach123 (Peter Tachauer), on Flickr

8508873073_0856af952b_b.jpg

street @ ISO 4000 XE1 by petach123 (Peter Tachauer), on Flickr

8509980506_7914253b46_b.jpg

street @ ISO 4000 XE1 by petach123 (Peter Tachauer), on Flickr

8508872541_6d80b227b7_b.jpg

street @ ISO 4000 XE1 by petach123 (Peter Tachauer), on Flickr
 
I find the noise distracting in a couple of these too, Pete. I shoot a lot at 6400 with the X-Pro and I think exposing properly at a higher ISO ends up with cleaner results (or more grain-like noise) than under-exposing at ISO 4000 and then pulling the shadows up. Here are a couple just to illustrate:

View attachment 65943
Philly Valentines Day-545-Edit by ramboorider1, on Flickr

View attachment 65944
Philly Valentines Day-552-Edit by ramboorider1, on Flickr

These sensors are capable of a LOT - don't be afraid to push them. That said, I'm not sure 6400 is a true 6400. Using the same EFL and same aperture on my OMD, I get the same exposure at about the same shutter speed at 3200 on my OMD that I do at 6400 on my X-Pro. So the ISO numbers look impressive but the actual low light capability of the two cameras isn't really very far apart at all...

-Ray
 
For an 18mm lens on an APS-C sensor you're being very conservative with your apertures with respect to depth-of-field. I'd be happy to use f/5.6 or even f/4 since the distance to subject doesn't vary greatly in that kind of situation.

You're asking a lot of that sensor in pushing an ISO 4000 file by 2-2.5 stops!
 
I agree with Ray. that a proper exposure, even at a higher ISO, delivers less noise than an underexposed image pulled-up in post.

GRAA0067-XL.jpg
Join to see EXIF info for this image (if available)

OM-D w/ O15 lens cap, ISO 12800

G
 
It was pretty grey, dismal and dim yesterday. I wanted to maintain 1/250s shutter with a reasonable f8 - f13 but without ramping up the ISO too much. The 1/250s at those f numbers and at iso 4000 produced underexposure by about 2 - 2.5 stops.

By conventional definition of ISO, an ISO 4000 file pushed 2+ stops is actually ISO 16000+ (many people say "effective ISO" it's just as legitimately "actual ISO"). So it's not really apples to apples to compare these to "properly exposed" ISO 6400 files. Pete's shots are properly exposed ISO 16000+ shots. On the other hand, Nic's point about being conservative with DOF (using a very deep DOF) holds.
 
By conventional definition of ISO, an ISO 4000 file pushed 2+ stops is actually ISO 16000+ (many people say "effective ISO" it's just as legitimately "actual ISO"). So it's not really apples to apples to compare these to "properly exposed" ISO 6400 files. Pete's shots are properly exposed ISO 16000+ shots. On the other hand, Nic's point about being conservative with DOF (using a very deep DOF) holds.

We are reading this very differently. For clarity sake and so not to confuse others ... I read that Pete calibrated the camera's lightmeter/sensor to ISO 4000 then underexposed the image 2-2.5 stops and added those stops back by tweaking the images in Light Room.

You're saying, essentially, that tweaking in LR, adding back the 2-2.5 stops by moving the slider, does not affect noise. Had Pete set the camera and shot at ISO 16000 - 24000, he would have the similar noise, per the adjustments after the fact, in LR. In summary, that underexposing and compensating in post does not effect noise quality or noise level.

I haven't seen that, but then I haven't looked for that either.

Gary

PS- I understand your point of apples to oranges, but our point is a proper exposure compared to an underexposure, is by its very nature, an apple to orange comparison.
G

PPS- I do agree with you and Nic that there seems to be a lot of wiggle room for ISO adjustments that shouldn't have a significant impact on the image capturing.
G
 
Nothing, sorry, I got confused by numbers.
I'll just say that this extreme use of the xtrans APSC seems to give worse results than the GRD. Perhaps it's the conservative aperture setting.
 
I thought only landscape shooters and macro heads ever went over f8. Those shots are definitely NOT indicative of what the Fuji files should look like at ISO4000. I use ISO6400 all the time and they are clean as a whistle.

Around Christmas time I shot a photo of our tree and the only light in the room was from the TV. Just for giggles, I turned it up to ISO25600 (!) It's not pretty, but it's useable in a pinch. This is a JPEG straight from camera with no additional noise reduction so it could obviously be improved as well.
8276836905_46e1b32e60_c.jpg

ISO 25600 by Lukinosity, on Flickr

Don't baby the Fuji.....PUSH it.
 
We are reading this very differently. For clarity sake and so not to confuse others ... I read that Pete calibrated the camera's lightmeter/sensor to ISO 4000 then underexposed the image 2-2.5 stops and added those stops back by tweaking the images in Light Room.

You're saying, essentially, that tweaking in LR, adding back the 2-2.5 stops by moving the slider, does not affect noise. Had Pete set the camera and shot at ISO 16000 - 24000, he would have the similar noise, per the adjustments after the fact, in LR. In summary, that underexposing and compensating in post does not effect noise quality or noise level.

What I am saying is that if Pete set his camera to ISO 4000, 1/250s, f/8 and needed to push his RAW file by 2 full stops (+2EV on the exposure slider in Lightroom) to get a resulting image of "normal brightness", then that resulting image is an ISO 16,000 image and will have about as much noise as you'd expect from an image taken under those same conditions with the camera set to ISO 16,000 if the camera had such a setting.

Here's a brief experiment to explain my point:

Shoot any image at ISO 3200 and write down the shutter speed, and f-number you used.

Repeat the same shot with the same shutter speed and f-number, but ISO 800 instead of 3200.

Open both RAW files in Lightroom and click on "Auto Tone" for each - they will look pretty much the same. The second one may have a hair more shadow noise, but there won't be much in it if you're using any of the recent cameras with low read noise sensors (E-M5, Fuji, etc).
 
Back
Top