Noisy wedding?

[Squidgy]

I've just been offered the chance to be second photographer to a pro at a wedding which will be my first time at such an event. I want to make an impression even though the pro (who has more faith than I) has told me that even if I don't get any good shots it will be a good experience for me so there is no pressure or high expectation but still, I want to give it my all. We will start at the brides house and do both indoor and outdoor shots. I have been practicing using the camera in AV mode with the flash set to high speed so I can use ambient light and the flash just fills in (I have no problem with setting white balance). However, I have been told that the church is very old and very dark which worries me, yes I know to take the ISO up but it's the noise this produces that makes me feel well out of my league when it comes to post processing. I've looked at lots of wedding photographers websites and their images which were obviously taken in poor light look wonderful and noise free. Has any one got any overall advice they would please share and is there any software to help reduce noise. Thanks for taking the time to read.

[The Blue Boy]

Hi Nigel,

What software have you already got?

[Davidd]

You're going to inevitably get noise in church photos because they're so shadowy, Nigel (more noise in shadows). Personally I find that the most recent ACR noise reduction feature does the job to my satisfaction. And realistically its reduction not removal. Its a trade off between noise and detail. I would reduce the noise only so much that it doesn't affect the quality of the image, if you see what I mean. There are many factors to consider in processing an image, and noise level is only one of them. Capture a nice well-composed image of the happy couple in the church using natural light and who's going to get fixated on noise? I'm sure that others here will give you good advice about other noise reduction options. By the way, there's another wedding thread running at the moment, check it out.

Good luck with this great opportunity to be second shooter!

[The Blue Boy]

David,

I find the reduction of noise in acr can only do so much. There was/is a thread here about the trade off between sharpening and noise reduction using the Bruce Fraser method. Damned if I can find it. The latest version of acr is using this method to a point, using the masking in the sharpening dialogue. You already know this, so I'll shut up.

And I've just seen your location! Fancy a pint?

[hoffstriker]

Hey Nigel,

What camera body you shooting? What PP software are you using? What lenses are you planning on using? Are you using on or off camera flash and either way make sure you use a diffuser. Just got into weddings this year (against Colin's advise) and have learned soooo much in a very short time. Fill in the answers and I think many people on here will be able to help.

[Colin Southern]

Just got into weddings this year (against Colin's advise) and have learned soooo much in a very short time.

As I'm sure anyone would Chris - the problem is though, our first duty of care has to be to the Bride & Groom. It's a bit like a patient who puts his life in the hands of a surgeon as the surgeon attempts his first operation. I don't think it would go down to well with the relatives if the patient died, but the surgeon learned a lot.

My attitude has always been that when it comes to wedding photography, failure isn't an option - and thus it needs to be treated as such (for the primary shooter anyway).

In my opinion - if (as primary shooters) they're not thoroughly familiar with their equipment - how to use it - it's limitations - all the techniques for poor light - bright light - off-camera flash - backup equipment - comfortable directing people etc then they really need to get up to speed in these areas before taking on responsibilities in what is probably THE single most demanding type of photography.

[hoffstriker]

As I'm sure anyone would Chris - the problem is though, our first duty of care has to be to the Bride & Groom. It's a bit like a patient who puts his life in the hands of a surgeon as the surgeon attempts his first operation. I don't think it would go down to well with the relatives if the patient died, but the surgeon learned a lot.

Hense, that is why my first wedding was basically a "cadaver" that I did not want to shoot, but was forced into. It was incredibly informal and I had an assistant to increase the photo stock available. There was no formal ceremony and it was on a beach during the worst lighting part of the day. The sky was overcast and there was a deep fog, but that is how everyone should start. There was a safety net (my asst) and I had understanding people that were not in a rush. It also helps that all they wanted was about a dozen photos to remember the day. I ended up providing them with 400.

What I am saying is that Nigel is doing this the right way. He's a second shooter and his shots really are not needed. My first wedding was basically the same, the couple would have been happy with anything just to show others it actually happened. But, there is always a first time, and we have to embrace these first times because they are when we learn so much with the leaving of our normal comfort zone. So long as we (and the couple to be married) realize our inadequacies considering our inexperience.

With all this being said I still think Colin is right 95% of the time...

[Colin Southern]

Hense, that is why my first wedding was basically a "cadaver" that I did not want to shoot, but was forced into. It was incredibly informal and I had an assistant to increase the photo stock available. There was no formal ceremony and it was on a beach during the worst lighting part of the day. The sky was overcast and there was a deep fog, but that is how everyone should start. There was a safety net (my asst) and I had understanding people that were not in a rush. It also helps that all they wanted was about a dozen photos to remember the day. I ended up providing them with 400.

And I think that's fine - just so long as everyone knows "what the deal is". But I still think that one needs to be very careful; in some ways it's a bit like driving without a seatbelt - most people would strongly recommend that we wear one whilst driving, and yes if we don't wear one - but don't have an accident - it still doesn't mean that it's not a good idea to wear one. Relating that to photography, few would want to pay a photographer several thousand dollars when they could get a friend or associate to do it for nothing (it becomes less of a "financial priority" when they have what they perceive to be "cheaper options"). The problem is that almost without exception, they have no idea how difficult and demanding wedding photography is (if a quality job is a pre-requisite anyway) - they've seen lots of "cool shots" (probably never seen a bad one) and basically just assume that "that's the kind of shot that modern cameras produce" -- and often the photographer isn't much wiser.

If they're happy with the result then great - the tog lives to shoot another day - but there have been many many many many instances where it turns into a disaster zone, and the tog only ends up discovering later that (a) it's harder than people realise, and (b) the B&G actually had much higher expectations than the tog realised. Big problem.

What I am saying is that Nigel is doing this the right way. He's a second shooter and his shots really are not needed.

Yep - so long as everyone understands the "rules of the game" then that's absolutely fine

With all this being said I still think Colin is right 95% of the time...

I think I'm right 100% of the time ...

... even when I'm wrong

[Black Pearl]

I've looked at lots of wedding photographers websites and their images which were obviously taken in poor light look wonderful and noise free.

Thats because noise is not the monster that many make it out to be also when these photographers reduced the size of their image to put on the web the noise ceased to be visible (if it was to start with) and the same goes for the finished print.

Try your camera at a higher than normal ISO in a low light location.
Make sure you get the exposure absolutely perfect as brightening the image with software WILL make noise an issue.
Go and get it professionally printed at the sort of sizes it will (might) appear in an album.

It will put your mind at rest I assure you.



If you are still worried and your camera has the function set it to shoot RAW and Jpeg with high ISO noise reduction switched to high in the menu. If you struggle with software getting the noise reduced when processing the RAW file just use the Jpeg as the camera will have done it for you and very often does a darn good job.

[Colin Southern]

I know to take the ISO up but it's the noise this produces that makes me feel well out of my league when it comes to post processing. I've looked at lots of wedding photographers websites and their images which were obviously taken in poor light look wonderful and noise free. Has any one got any overall advice they would please share and is there any software to help reduce noise. Thanks for taking the time to read.

Hi Nigel,

A quick lesson on ISO and noise ...

Most sensors can capture a dynamic range of about 12 stops at the camera's base (native) ISO - meaning the range of brightness between "any brighter and it overloads the sensor" and "any dimmer and the signal representing part of the image is so low it can't be distinguished from the noise". A reflective scene (eg bride in white dress and groom in black suit" only requires about 4 of these 12 stops - so assuming that the highlights ("white dress") were exposed right up to the sensors limit then you have essentially about an 8 stop safety margin (which is huge).

When one increases the camera ISO (say from 100 to 200), the physics of the sensor don't change - it's still only capable of recording the same 12 stop range - but what's really happening internally is that the camera is in-essence under-exposing the shot by 1 stop and then amplifying the signal to shift the levels back to where they should be. The problem is though that the noise gets amplified too - so the noise floor that was 12 stops down is now only 11 stops down - and although the dynamic range of the sensor hasn't changed, the dynamic range that the camera is able to capture HAS changed (from 12 to 11). For low ISO modes it's not a problem because even with a sloppy exposure there is still a big safety margin (if it was 8 stops at ISO 100 it might be 7 stops @ ISO 200 - 6 stops at ISO 400 - 5 stops @ ISO 800 - 4 stops @ ISO 1600 etc) (I'm sure you get the idea).

Put that aside for a sec while I introduce something else ...

... when we take a picture of a scene, the camera will expose that scene as if it was 18% grey (a "correct" exposure) - even if the scene IS 18% grey - and a "correct" exposure has been captured, it still doesn't mean that the whites will be exposed as highlights that push the sensor to the maximum; the camera includes a 1 to 2 stop "safety margin" (assuming RAW shooting throughout all of this) - so if for example you have the camera set to ISO 1600 (4 stop safety margin) and you shoot a white dress next to a black suit for a "normal" exposure, you're actually already under-exposing by 1 to 2 stops (lets call it 2 for the sake of the maths) - so the 4 stop safety margin you thought you had is probably a lot closer to only 2 stops -- and noise starts to appear in the shadows.

At 3200 ISO this noise is a lot worse, and at 6400 ISO it's getting pretty bad because the shadow areas are pretty much at the noise floor.

So - in summary - as one increases the cameras ISO, the margin of error for sloppy exposures gets less and less ... and most photographers normally shoot pretty sloppy exposures! So basically the "trick" is "don't be afraid to use high ISO modes, but if you want to minimise noise then you HAVE to expose-to-the-right and push the exposures as far as you can without blowing. In the case of a church wedding this means applying exposure compensation so that the brightest thing in the scene (probably the brides dress) is as close to blowing as possible.

The problem is though, how do you know when it's close to blowing? about all you can do is make sure that the highlight alert ("blinkies") on your camera is turned on, and watch the histogram - but - you also need to be aware that the highlight alert - histogram - and image on your review screen are ALL based on the camera's in-camera JPEG conversion of the RAW image, and they WILL "over-state" things (ie if they're showing the dress as slightly blown in reality it won't actually be).

So that's it in a nutshell really - use a high ISO mode, but push the exposure as far as you can. In that situation I'd be very happy if the dress was showing as slightly blown on the review screen, and the grooms black suit was looking grey rather than black (because I'd have the biggest separation of signal to noise).

How far can one push it? Well I'm afraid that that comes down to experience and testing -- something you can get a bit of and do a bit of before the big day - so get out there and test

So that's the first rule / "trick" of high ISO shooting: Don't under-expose (in this case with under-exposure being any remaining headroom safety margin between the highlights captured and the maximum that the camera is acapable of recording) (the image will probably look over-exposed on the screen).

In terms of noise - usually it's only visible at high magnifications; it's normally too small to see when looking at the entire image as a whole, unless one crops the image excessively - so "don't crop the image excessively" becomes rule #2.

In terms of noise reduction programs - personally - I find that they're a waste of money; for the most part, all they're doing is averaging the signal (with a few fancy algorithms thrown in for good measure), and all that THAT does is give you a soft picture. If you push your exposures - don't crop excessively - then I can almost guarantee that noise won't be a problem at any ISO; typically 97% of noise is sampled out when an image is down-sampled for internet / PC display, and physical prints can only reproduce a 4 stop range anyway and even noise that is visible on the screen ususally can't be seen an a physical print.

Hope this helps

[Colin Southern]

If you struggle with software getting the noise reduced when processing the RAW file just use the Jpeg as the camera will have done it for you and very often does a darn good job.

That's actually a good point that a lot of people are finding out the hard way ....

When a new camera comes out it typically has a sensor manufactured using more advanced technologies - and that technology gives a better signal to noise ratio - so higher ISO modes are possible - BUT - they also typically have higher pixel densities and that of course means more noise, so to a very real degree, the advances in technology that allow better signal to noise ratios are "countered" by the increased noise due to the higher pixel densities.

So if that's the case, then how are we able to get higher ISO modes without more noise I hear you all ask? The answer lies in what Robin mentions - in-camera noise reduction. Sounds like a good thing (and it is I guess), but what a lot of people don't realise is that it's ONLY applied to the on-camera generated JPEG and NOT to any RAW file.

So what happens is Canon / Nikon announce a new camera and show some wonderful noise free images shot inside a cave at ISO 3,000,000 - people think "wow - this is GREAT - gotta have one of those" - and then find that their RAW images (because all photo boffins shoot RAW) that they're getting at high ISO modes don't look anything like the "ones in the advertisements" (also made worse by that fact that they're not pushing their exposure to the absolute limit).

[MrB]

A quick lesson on ISO and noise ...
Hope this helps

Truly fascinating! Thank you for sharing your knowledge, Colin.

Philip

[Colin Southern]

Truly fascinating! Thank you for sharing your knowledge, Colin.

Philip

You're welcome Philip - not sure if I added more noise to the problem or not though!

[MrB]

When a new camera comes out it typically has a sensor manufactured using more advanced technologies - and that technology gives a better signal to noise ratio - so higher ISO modes are possible - BUT - they also typically have higher pixel densities and that of course means more noise, so to a very real degree, the advances in technology that allow better signal to noise ratios are "countered" by the increased noise due to the higher pixel densities.

A further question comes to mind -
I have often read the point about the increase in noise resulting from increasing pixel density, as it is often used to explain why compact cameras are more noisy than DSLRs. So my question is, for example, if the camera has a 12 megapixel sensor is there anything to be gained in terms of noise reduction by stepping down to 6 megapixels when shooting at high ISO values, assuming one is producing sensible sized prints (e.g 8x10 or A4)?

Philip

[Colin Southern]

A further question comes to mind -
I have often read the point about the increase in noise resulting from increasing pixel density, as it is often used to explain why compact cameras are more noisy than DSLRs. So my question is, for example, if the camera has a 12 megapixel sensor is there anything to be gained in terms of noise reduction by stepping down to 6 megapixels when shooting at high ISO values, assuming one is producing sensible sized prints (e.g 8x10 or A4)?

Philip

It's something you might like to test, but I doubt you'd see any improvement. I'm being a bit vague because it's hard to say how manufactures may choose to handle the smaller resolution; if they choose pixel binning wher the output from several photosites are averaged (which may happen at high ISOs) then you may effectively be shooting at a reduced resolution anyway (even if they fudge it back up afterwards) or they may just be sampling the extra pixels out. Either way, it still doesn't change the physics of that tiny sensor.

In a nutshell, camera sensors are kinda like a DSLR sensor, but with most of the safety margin removed. Try to recover a badly under-exposed shot to see what I mean - I had to work on one the other day, and it was not a thing of beauty

PS: Nigel - sorry to hijack the thread a bit here - I can split this off to a seperate thread if you prefer.

[MrB]

It's something you might like to test, but I doubt you'd see any improvement. I'm being a bit vague because it's hard to say how manufactures may choose to handle the smaller resolution; if they choose pixel binning wher the output from several photosites are averaged (which may happen at high ISOs) then you may effectively be shooting at a reduced resolution anyway (even if they fudge it back up afterwards) or they may just be sampling the extra pixels out. Either way, it still doesn't change the physics of that tiny sensor.

In a nutshell, camera sensors are kinda like a DSLR sensor, but with most of the safety margin removed. Try to recover a badly under-exposed shot to see what I mean - I had to work on one the other day, and it was not a thing of beauty

PS: Nigel - sorry to hijack the thread a bit here - I can split this off to a seperate thread if you prefer.

Colin, I don't think that these posts do "hijack the thread" and I think your replies are all relevant to Nigel's wedding photography. My fault for probably phrasing my last question in a way that was misleading to you. I mentioned the compact camera just as an example of the pixel density point, but I was actually wondering if it is possible for Nigel (or anyone) to reduce noise on a DSLR by reducing the resolution selected for high ISO shots.

(No need to answer this again. As you have already written - it is obviously something that needs investigating and might be different for different models of DSLR.)

Philip

[Davidd]

Well done Colin for taking the time and trouble to explain all that - I'm sure you've already done this several times before and then of course there are the tutorials etc. You make my response seem a bit thin! Must try harder... People caring enough to provide such high quality advice is what makes this forum special.

But where's Nigel? Bet you didn't expect all this - now you know, eh?

good wishes

[Squidgy]

WOW! What a response, I'm overwhelmed. Thanks to all who have commented or offered advice, I really do appreciate the time taken and I will carefully read and reply to each individual who has questions tomorrow afternoon (after the event) as I've unfortunately been hit with a rush job at work and wont get home until late tonight. I have taken a brief look at the posts just now while taking a short coffee break and appologise for not being able to reply to you all sooner. Watch this space for first time pics and replies to above. Many, Many thanks.

[Black Pearl]

I have often read the point about the increase in noise resulting from increasing pixel density, as it is often used to explain why compact cameras are more noisy than DSLRs. So my question is, for example, if the camera has a 12 megapixel sensor is there anything to be gained in terms of noise reduction by stepping down to 6 megapixels when shooting at high ISO values, assuming one is producing sensible sized prints (e.g 8x10 or A4)?

Nope.

It is all down to physical pixel size. If you pack more pixels into a given space they get smaller and their signals gets smaller. To get an image you need to amplify those signals more and the result is a random splatter (I'm sticking with that word) of odd colours and brightnesses - noise. You can electronically drop the pixel count on the camera but the same pixels are creating the original file and and the same properties will result.

Make sure your exposure is bang on - either use the in-camera noise reduction you get with a Jpeg file or careful and up to date software to reduce the noise later from a RAW file - don't pixel peep - try not to muck about with the resulting file too much - carefully sharpen your image and then view the results as a print.


I've said this above, I've said it in earlier posts and I'll keep on saying it - noise is not the monster it is made out to be.

For years and years and years............and years...........we shot with film that was bloody awful above about 400iso and the results were accepted as fine - because we printed our pictures and viewed them normally.
Now we have the capability of zooming into every shot to the degree where noise/grain is visible many have become obsessed with the need to have a image that is flawless at any iso. Digital sensors are way better than film ever was at high iso's and we just need to accept that at some point we will get imperfections but they are not the end of the world.

[Colin Southern]

Well done Colin for taking the time and trouble to explain all that - I'm sure you've already done this several times before and then of course there are the tutorials etc. You make my response seem a bit thin! Must try harder... People caring enough to provide such high quality advice is what makes this forum special.

But where's Nigel? Bet you didn't expect all this - now you know, eh?

good wishes

No worries David,

I must confess that I have a few "hobby horses" around here (high ISO "Noise" - HDR - Monitor "Calibration" - sharpening to name 4!) - often there is so much mis-information being repeated around the net that the only way to begin to counter it is to not only "state my truth", but to also take the time to explain why it's "the truth" (even if I do upset a few people along the way ).