What is causing this noise?

Stan in NH wrote:

What is the ISO? Did you apply a lot of sharpening in pp?

The ISO is 100. The image is a screen grap from RawTherapee. No PP had been applied. I just looked at the image.

jvc1 wrote:

Jeffry7 wrote:

Hi folks,

Not sure if this is the right place for my question, but the camera is a 550d, so here goes.

I took a daytime shot of the moon over the weekend with a manual lens and it turned out really noisy.

800mm, ISO 100, 1/60, 100% crop

Someone I showed this to suggested it was underexposure, but the histogram looks OK.

Full image via Rawtherapee

Any ideas?

The crop looks way over exposed. What lens? Looks like a cheap mirror lens. What's the full image look like? The area of deep space should be all black, the moon should be very bright. SS looks to slow, something's wrong with the whole thing. Can you attach the full image with exif?

Hi jvc1,

It is a cheap lens, but not a mirror. It is one of these:

https://www.celestron.com/products/travel-scope-70-portable-telescope

Which is to say a cheap telescope. Behind the telescope lens I have a t-mount 2x teleconverter.

This is a daytime shot. It was taken at 8am. The background is the correct color because the sky is blue during the day. I explained it was a daytime shot in the OP. Not sure how to make it plainer.

I will try to post the full with exif, but that is around a 100mb Tiff, and DPReview seems to be choking on it.

Hi folks,

So I want to thank you for taking a look at my picture and trying to sort me out. In going back to the picture to export a Tiff with the EXIF data as requested I noticed that the noise was gone. Or at least I did not see the noise when looking in Digital Photo Professional 4. So I looked again with Rawtherapee and the noise was back. Same RAW file, two very different results. I poked around in Rawtherapee and found the exposure compensation was not at a neutral value. When I reset it, the noise went away.

TL;DR, this is a case of operator error.

jvc1's insistance on seeing the unmessed around with file prompted my discovery of the error.

What is an apropriate way to mark the answer?

Jeffry7 wrote:

Stan in NH wrote:

What is the ISO? Did you apply a lot of sharpening in pp?

The ISO is 100. The image is a screen grap from RawTherapee. No PP had been applied. I just looked at the image.

A sensor makes colour noise, due to the CFA. Your image sample shows noise reduction gone bad. Can you show the image without any "processing" going on?  RAWTherapee clearly is doing something to the noise here.

A couple of people asked to see the full sized image.

Everything I have tried to include the EXIF data in the image has failed. The TIFF exported from Digital Photo Professional has the EXIF data, but I can't upload the 100+ meg file to DPReview (apparently. I left it going for hours and never got past the spinning please wait thingy.)

Every conversion I have found strips the EXIF, even when you ask them not to. One thing I did find is that IrfanView can copy the EXIF data to the clipboard. So here is that EXIF data from the original TIFF before conversion to GIF via RIOT. I have scrubbed some personnel information, but otherwise it is here.

Filename - IMG_7960.TIF
ImageWidth - 5184
ImageLength - 3456
BitsPerSample - 16 16 16
Compression - 1 (None)
PhotometricInterpretation - 2
Make - Canon
Model - Canon EOS 550D
StripOffset - 127678
Orientation - Top left
SamplesPerPixel - 3
RowsPerStrip - 8
StripByteCount - 129406
XResolution - 350
YResolution - 350
PlanarConfiguration - 1
ResolutionUnit - Inch
Software - Digital Photo Professional
DateTime - 2019:04:28 08:02:11
Artist -
Copyright -
ExifOffset - 131198
InterColorProfile - 135950
ExposureTime - 1/60 seconds
FNumber - 0
ExposureProgram - Manual control
ISOSpeedRatings - 100
ExifVersion - 0221
DateTimeOriginal - 2019:04:28 08:02:11
DateTimeDigitized - 2019:04:28 08:02:11
ComponentsConfiguration - YCbCr
ShutterSpeedValue - 1/64 seconds
ApertureValue - F 0.00
ExposureBiasValue - 0
MeteringMode - Multi-segment
Flash - Flash not fired, compulsory flash mode
FocalLength - 0 mm
UserComment -
SubsecTime - 03
SubsecTimeOriginal - 03
SubsecTimeDigitized - 03
FlashPixVersion - 0100
ColorSpace - sRGB
FocalPlaneXResolution - 5728.18
FocalPlaneYResolution - 5808.40
FocalPlaneResolutionUnit - Inch
CustomRendered - Normal process
ExposureMode - Manual
White Balance - Manual
SceneCaptureType - Standard

Maker Note (Vendor): -
Macro mode - Normal
Self timer - Off
Quality - Normal
Flash mode - Not fired
Sequence mode - Single or Timer
Focus mode - MF
Image size - Large
Easy shooting mode - Manual
Digital zoom - None
Contrast - Normal
Saturation - Normal
Sharpness - Low , -32769
ISO Value - 32767
Metering mode - Evaluative
Focus type - Auto
AF point selected - 0
Exposure mode - Manual
Focal length - 0.0 - 0.0 mm
Focal units - 1/mm
Flash activity - Not fired
Flash details -
Focus mode 2 - 65535
Auto ISO - 100
Base ISO - 100
White Balance - Sunny
Sequence number - 0
Flash bias - 0 EV
Subject Distance - 0
Image Type - Canon EOS 550D
Firmware Version - Firmware Version 1.0.9
Owner Name - XXXX
Camera Serial Number - XXXX
AF point selected - Single-point AF
Num AF Points - 1
Valid AF Points - 1
AF Image Width - 5184
AF Image Height - 3456
File number - 000 - 0000
Sharpness (A0) - 3

Thumbnail: -
ImageWidth - 252
ImageLength - 168
BitsPerSample - 8 8 8
Compression - 1 (None)
PhotometricInterpretation - 2
StripOffset - 188
SamplesPerPixel - 3
RowsPerStrip - 8
StripByteCount - 272
XResolution - 350
YResolution - 350
PlanarConfiguration - 1
ResolutionUnit - Inch

Not sure why you would want to post a TIFF. DPP JPEGs have EXIF.

The weird noise is gone now, RAW Therapy NR settings must have caused it, like I thought.

Do clean that sensor, though.

brightcolours wrote:

Not sure why you would want to post a TIFF. DPP JPEGs have EXIF.

Well I didn't want to post a lossy format when we were talking about noise. Also Tiff supports internally compressing the image data.

The weird noise is gone now, RAW Therapy NR settings must have caused it, like I thought.

I don't know about NR, but the exposure compensation in Rawtherapee was at +2 stops. I am going to have to see if there is a reset to defaults function in that program.

Do clean that sensor, though.

Yes.